Revert "TEMP: Try python 3.8."

This reverts commit 8def4d5177.

.github/workflows/main.yml

@@ -0,0 +1,206 @@
+name: ci
+on: ["push", "pull_request", "workflow_dispatch"]
+
+jobs:
+
+  lint:
+    name: lint
+    runs-on: ubuntu-20.04
+    steps:
+      - uses: actions/checkout@v3
+      - uses: actions/setup-python@v4
+        with:
+          python-version: '3.9'
+      - name: extract pip cache
+        uses: actions/cache@v3
+        with:
+          path: ~/.cache/pip
+          key: ${{ runner.os }}-pip-${{ hashFiles('setup.py') }}
+          restore-keys: ${{ runner.os }}-pip-
+      - run: pip install --user --upgrade pip wheel
+      - run: pip install -e .[lint]
+      - run: make lint
+
+  tests-unit:
+    name: "tests / unit"
+    strategy:
+      matrix:
+        os:
+          - ubuntu-20.04
+          - macos-latest
+          - windows-latest
+    runs-on: ${{ matrix.os }}
+    steps:
+      - uses: actions/checkout@v3
+      - uses: actions/setup-python@v4
+        with:
+          python-version: '3.9'
+      - name: set pip cache dir
+        shell: bash
+        run: echo "PIP_CACHE_DIR=$(pip cache dir)" >> $GITHUB_ENV
+      - name: extract pip cache
+        uses: actions/cache@v3
+        with:
+          path: ${{ env.PIP_CACHE_DIR }}
+          key: ${{ runner.os }}-pip-${{ hashFiles('setup.py') }}
+          restore-keys: ${{ runner.os }}-pip-
+      - id: os-name
+        uses: ASzc/change-string-case-action@v5
+        with:
+          string: ${{ runner.os }}
+      - run: python -m pip install --user --upgrade pip wheel
+      - if: startsWith(runner.os, 'linux')
+        run: pip install -e .[test]
+      - if: startsWith(runner.os, 'linux')
+        env:
+          HOME: /tmp
+        run: make test-unit-coverage
+      - if: startsWith(runner.os, 'linux') != true
+        run: pip install -e .[test]
+      - if: startsWith(runner.os, 'linux') != true
+        env:
+          HOME: /tmp
+        run: coverage run --source=lbry -m unittest tests/unit/test_conf.py
+      - name: submit coverage report
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          COVERALLS_FLAG_NAME: tests-unit-${{ steps.os-name.outputs.lowercase }}
+          COVERALLS_PARALLEL: true
+        run: |
+          pip install coveralls
+          coveralls --service=github
+
+  tests-integration:
+    name: "tests / integration"
+    runs-on: ubuntu-20.04
+    strategy:
+      matrix:
+        test:
+          - datanetwork
+          - blockchain
+          - claims
+          - takeovers
+          - transactions
+          - other
+    steps:
+      - name: Configure sysctl limits
+        run: |
+          sudo swapoff -a
+          sudo sysctl -w vm.swappiness=1
+          sudo sysctl -w fs.file-max=262144
+          sudo sysctl -w vm.max_map_count=262144
+      - name: Runs Elasticsearch
+        uses: elastic/elastic-github-actions/elasticsearch@master
+        with:
+          stack-version: 7.12.1
+      - uses: actions/checkout@v3
+      - uses: actions/setup-python@v4
+        with:
+          python-version: '3.9'
+      - if: matrix.test == 'other'
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y --no-install-recommends ffmpeg
+      - name: extract pip cache
+        uses: actions/cache@v3
+        with:
+          path: ./.tox
+          key: tox-integration-${{ matrix.test }}-${{ hashFiles('setup.py') }}
+          restore-keys: txo-integration-${{ matrix.test }}-
+      - run: pip install tox coverage coveralls
+      - if: matrix.test == 'claims'
+        run: rm -rf .tox
+      - run: tox -e ${{ matrix.test }}
+      - name: submit coverage report
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          COVERALLS_FLAG_NAME: tests-integration-${{ matrix.test }}
+          COVERALLS_PARALLEL: true
+        run: |
+          coverage combine tests
+          coveralls --service=github
+
+
+  coverage:
+    needs: ["tests-unit", "tests-integration"]
+    runs-on: ubuntu-20.04
+    steps:
+      - name: finalize coverage report submission
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          pip install coveralls
+          coveralls --service=github --finish
+
+  build:
+    needs: ["lint", "tests-unit", "tests-integration"]
+    name: "build / binary"
+    strategy:
+      matrix:
+        os:
+          - ubuntu-20.04
+          - macos-latest
+          - windows-latest
+    runs-on: ${{ matrix.os }}
+    steps:
+      - uses: actions/checkout@v3
+      - uses: actions/setup-python@v4
+        with:
+          python-version: '3.9'
+      - id: os-name
+        uses: ASzc/change-string-case-action@v5
+        with:
+          string: ${{ runner.os }}
+      - name: set pip cache dir
+        shell: bash
+        run: echo "PIP_CACHE_DIR=$(pip cache dir)" >> $GITHUB_ENV
+      - name: extract pip cache
+        uses: actions/cache@v3
+        with:
+          path: ${{ env.PIP_CACHE_DIR }}
+          key: ${{ runner.os }}-pip-${{ hashFiles('setup.py') }}
+          restore-keys: ${{ runner.os }}-pip-
+      - run: pip install pyinstaller==4.6
+      - run: pip install -e .
+      - if: startsWith(github.ref, 'refs/tags/v')
+        run: python docker/set_build.py
+      - if: startsWith(runner.os, 'linux') || startsWith(runner.os, 'mac')
+        name: Build & Run (Unix)
+        run: |
+          pyinstaller --onefile --name lbrynet lbry/extras/cli.py
+          dist/lbrynet --version
+      - if: startsWith(runner.os, 'windows')
+        name: Build & Run (Windows)
+        run: |
+          pip install pywin32==301
+          pyinstaller --additional-hooks-dir=scripts/. --icon=icons/lbry256.ico --onefile --name lbrynet lbry/extras/cli.py
+          dist/lbrynet.exe --version
+      - uses: actions/upload-artifact@v3
+        with:
+          name: lbrynet-${{ steps.os-name.outputs.lowercase }}
+          path: dist/
+
+  release:
+    name: "release"
+    if: startsWith(github.ref, 'refs/tags/v')
+    needs: ["build"]
+    runs-on: ubuntu-20.04
+    steps:
+      - uses: actions/checkout@v1
+      - uses: actions/download-artifact@v2
+      - name: upload binaries
+        env:
+          GITHUB_TOKEN: ${{ secrets.RELEASE_API_TOKEN }}
+        run: |
+          pip install githubrelease
+          chmod +x lbrynet-macos/lbrynet
+          chmod +x lbrynet-linux/lbrynet
+          zip --junk-paths lbrynet-mac.zip lbrynet-macos/lbrynet
+          zip --junk-paths lbrynet-linux.zip lbrynet-linux/lbrynet
+          zip --junk-paths lbrynet-windows.zip lbrynet-windows/lbrynet.exe
+          ls -lh
+          githubrelease release lbryio/lbry-sdk info ${GITHUB_REF#refs/tags/}
+          githubrelease asset lbryio/lbry-sdk upload ${GITHUB_REF#refs/tags/} \
+            lbrynet-mac.zip lbrynet-linux.zip lbrynet-windows.zip
+          githubrelease release lbryio/lbry-sdk publish ${GITHUB_REF#refs/tags/}
+

.github/workflows/release.yml

@@ -0,0 +1,22 @@
+name: slack
+
+on:
+  release:
+    types: [published]
+
+jobs:
+  release:
+    name: "slack notification"
+    runs-on: ubuntu-20.04
+    steps:
+      - uses: LoveToKnow/slackify-markdown-action@v1.0.0
+        id: markdown
+        with:
+          text: "There is a new SDK release: ${{github.event.release.html_url}}\n${{ github.event.release.body }}"
+      - uses: slackapi/slack-github-action@v1.14.0
+        env:
+          CHANGELOG: '<!channel> ${{ steps.markdown.outputs.text }}'
+          SLACK_WEBHOOK_URL: ${{ secrets.SLACK_RELEASE_BOT_WEBHOOK }}
+        with:
+          payload: '{"type": "mrkdwn", "text": ${{ toJSON(env.CHANGELOG) }} }'
+

.gitignore

@@ -1,2 +1,22 @@
 /.idea
 /.DS_Store
+/build
+/dist
+/.tox
+/.coverage*
+/lbry-venv
+/venv
+/lbry/blockchain
+
+lbry.egg-info
+__pycache__
+_trial_temp/
+trending*.log
+
+/tests/integration/claims/files
+/tests/.coverage.*
+
+/lbry/wallet/bin
+
+/.vscode
+/.gitignore

.travis.yml

@@ -1,146 +0,0 @@
-dist: xenial
-language: python
-python: "3.7"
-
-jobs:
-  include:
-
-    - stage: code quality
-      name: "pylint & mypy"
-      install:
-        - make install
-      script: make lint
-
-    - stage: test
-      name: "LBRY Unit Tests"
-      install:
-        - make install
-      script:
-        - cd lbry && HOME=/tmp coverage run -p --source=lbry -m unittest discover -vv tests.unit
-      after_success:
-        - coverage combine lbry/
-        - bash <(curl -s https://codecov.io/bash)
-
-    - name: "LBRY Integration Tests"
-      install:
-        - pip install coverage tox-travis
-        - sudo mount -o mode=1777,nosuid,nodev -t tmpfs tmpfs /tmp
-      script: cd lbry && tox
-      after_success:
-        - coverage combine lbry
-        - bash <(curl -s https://codecov.io/bash)
-
-    - &torba-tests
-      name: "Torba Unit Tests"
-      env: TESTTYPE=unit
-      install:
-        - pip install coverage tox-travis
-      script: cd torba && tox
-      after_success:
-        - coverage combine torba/tests
-        - bash <(curl -s https://codecov.io/bash)
-
-    - <<: *torba-tests
-      name: "Torba Integration Tests"
-      env: TESTTYPE=integration
-
-    - name: "Run Examples"
-      install:
-        - make install
-      script:
-        - cd lbry && HOME=/tmp coverage run -p --source=lbry scripts/generate_json_api.py
-      after_success:
-        - coverage combine lbry
-        - bash <(curl -s https://codecov.io/bash)
-
-    - &build
-      stage: build
-      name: "Linux"
-      env: OS=linux
-      install:
-        - pip install pyinstaller awscli
-        - cd torba && pip install -e . && cd ..
-        - cd lbry
-        - python scripts/set_build.py
-        - pip install -e .
-      script:
-        - pyinstaller -F -n lbrynet lbry/extras/cli.py
-        - cd dist
-        - chmod +x lbrynet
-        - zip -j lbrynet-${OS}.zip lbrynet
-        - shasum -a 256 -b lbrynet-${OS}.zip
-        - ./lbrynet --version
-      after_success:
-        - aws configure set aws_access_key_id $ARTIFACTS_KEY
-        - aws configure set aws_secret_access_key $ARTIFACTS_SECRET
-        - aws configure set region us-east-1
-        - export S3_PATH="daemon/build-${TRAVIS_BUILD_NUMBER}_commit-${TRAVIS_COMMIT:0:7}_branch-${TRAVIS_BRANCH}$([ ! -z ${TRAVIS_TAG} ] && echo _tag-${TRAVIS_TAG})"
-        - aws s3 cp lbrynet-${OS}.zip s3://build.lbry.io/${S3_PATH}/lbrynet-${OS}.zip
-      deploy:
-        provider: releases
-        api_key:
-          secure: "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"
-        file: lbrynet-${OS}.zip
-        skip_cleanup: true
-        overwrite: true
-        draft: true
-        on:
-          tags: true
-
-    - <<: *build
-      name: "Mac"
-      os: osx
-      osx_image: xcode8.3
-      language: shell
-      env: OS=mac
-      before_install:
-        - brew uninstall mercurial
-        - brew upgrade python || true
-        - pip3 install --user --upgrade pip virtualenv
-        - /Users/travis/Library/Python/3.7/bin/virtualenv --clear $HOME/venv
-        - source $HOME/venv/bin/activate
-      before_cache:
-        - brew cleanup
-
-    - <<: *build
-      name: "Windows"
-      os: windows
-      language: shell
-      env:
-        - OS=windows
-        - PATH=/c/Python37:/c/Python37/Scripts:/C/Windows/System32/downlevel:$PATH
-      before_install:
-        - choco install python --version=3.7.4 --x86
-        - python -m pip install --upgrade pip
-        - pip install pywin32
-      script:
-        - pyinstaller --additional-hooks-dir=scripts/. --icon=icons/lbry256.ico -F -n lbrynet lbry/extras/cli.py
-        - cd dist
-        - 7z a -tzip lbrynet-windows.zip lbrynet.exe
-        - sha256sum -b lbrynet-windows.zip
-        - ./lbrynet.exe --version
-    - if: tag IS present
-      stage: build
-      name: "Wallet Server Docker Image - Tagged Release"
-      script:
-          - set -e
-          - echo "$DOCKER_PASSWORD" | docker login --username "$DOCKER_USERNAME" --password-stdin
-          - travis_retry docker build -t lbry/wallet-server:$TRAVIS_TAG -f lbry/scripts/Dockerfile.wallet_server .
-          - docker push lbry/wallet-server:$TRAVIS_TAG
-
-    - if: tag IS blank AND branch = master AND NOT type IN (pull_request)
-      stage: build
-      name: "Wallet Server Docker Image - Master"
-      script:
-          - set -e
-          - echo "$DOCKER_PASSWORD" | docker login --username "$DOCKER_USERNAME" --password-stdin
-          - travis_retry docker build -t lbry/wallet-server:master -f lbry/scripts/Dockerfile.wallet_server .
-          - docker push lbry/wallet-server:master
-
-cache:
-  directories:
-    - $HOME/venv
-    - $HOME/.cache/pip
-    - $HOME/Library/Caches/pip
-    - $HOME/Library/Caches/Homebrew
-    - $TRAVIS_BUILD_DIR/.tox

INSTALL.md

@@ -9,20 +9,29 @@ Here's a video walkthrough of this setup, which is itself hosted by the LBRY net
 
 ## Prerequisites
 
-Running `lbrynet` from source requires Python 3.7 or higher. Get the installer for your OS [here](https://www.python.org/downloads/release/python-370/).
+Running `lbrynet` from source requires Python 3.7. Get the installer for your OS [here](https://www.python.org/downloads/release/python-370/).
 
-After installing python 3, you'll need to install some additional libraries depending on your operating system.
+After installing Python 3.7, you'll need to install some additional libraries depending on your operating system.
+
+Because of [issue #2769](https://github.com/lbryio/lbry-sdk/issues/2769)
+at the moment the `lbrynet` daemon will only work correctly with Python 3.7.
+If Python 3.8+ is used, the daemon will start but the RPC server
+may not accept messages, returning the following:
+```
+Could not connect to daemon. Are you sure it's running?
+```
 
 ### macOS
 
 macOS users will need to install [xcode command line tools](https://developer.xamarin.com/guides/testcloud/calabash/configuring/osx/install-xcode-command-line-tools/) and [homebrew](http://brew.sh/).
 
 These environment variables also need to be set:
-1. PYTHONUNBUFFERED=1
-2. EVENT_NOKQUEUE=1
+```
+PYTHONUNBUFFERED=1
+EVENT_NOKQUEUE=1
+```
 
 Remaining dependencies can then be installed by running:
-
 ```
 brew install python protobuf
 ```
@@ -31,14 +40,17 @@ Assistance installing Python3: https://docs.python-guide.org/starting/install3/o
 
 ### Linux
 
-On Ubuntu (16.04 minimum, we recommend 18.04), install the following:
-
+On Ubuntu (we recommend 18.04 or 20.04), install the following:
 ```
 sudo add-apt-repository ppa:deadsnakes/ppa
 sudo apt-get update
 sudo apt-get install build-essential python3.7 python3.7-dev git python3.7-venv libssl-dev python-protobuf
 ```
 
+The [deadsnakes PPA](https://launchpad.net/~deadsnakes/+archive/ubuntu/ppa) provides Python 3.7
+for those Ubuntu distributions that no longer have it in their
+official repositories.
+
 On Raspbian, you will also need to install `python-pyparsing`.
 
 If you're running another Linux distro, install the equivalent of the above packages for your system.
@@ -47,65 +59,119 @@ If you're running another Linux distro, install the equivalent of the above pack
 
 ### Linux/Mac
 
-To install on Linux/Mac:
+Clone the repository:
+```bash
+git clone https://github.com/lbryio/lbry-sdk.git
+cd lbry-sdk
+```
 
- ```
- Clone the repository:
- $ git clone https://github.com/lbryio/lbry-sdk.git
- $ cd lbry-sdk
+Create a Python virtual environment for lbry-sdk:
+```bash
+python3.7 -m venv lbry-venv
+```
 
- Create a Python virtual environment for lbry-sdk:
- $ python3.7 -m venv lbry-venv
+Activate virtual environment:
+```bash
+source lbry-venv/bin/activate
+```
 
- Activating lbry-sdk virtual environment:
- $ source lbry-venv/bin/activate
+Make sure you're on Python 3.7+ as default in the virtual environment:
+```bash
+python --version
+```
 
- Make sure you're on Python 3.7+ (as the default Python in virtual environment):
- $ python --version
+Install packages:
+```bash
+make install
+```
 
- Install packages:
- $ make install
+If you are on Linux and using PyCharm, generates initial configs:
+```bash
+make idea
+```
 
- If you are on Linux and using PyCharm, generates initial configs:
- $ make idea
- ```
+To verify your installation, `which lbrynet` should return a path inside
+of the `lbry-venv` folder.
+```bash
+(lbry-venv) $ which lbrynet
+/opt/lbry-sdk/lbry-venv/bin/lbrynet
+```
 
-To verify your installation, `which lbrynet` should return a path inside of the `lbry-venv` folder created by the `python3.7 -m venv lbry-venv` command.
+To exit the virtual environment simply use the command `deactivate`.
 
 ### Windows
 
-To install on Windows:
+Clone the repository:
+```bash
+git clone https://github.com/lbryio/lbry-sdk.git
+cd lbry-sdk
+```
 
- ```
- Clone the repository:
- > git clone https://github.com/lbryio/lbry-sdk.git
- > cd lbry-sdk
+Create a Python virtual environment for lbry-sdk:
+```bash
+python -m venv lbry-venv
+```
 
- Create a Python virtual environment for lbry-sdk:
- > python -m venv lbry-venv
+Activate virtual environment:
+```bash
+lbry-venv\Scripts\activate
+```
 
- Activating lbry-sdk virtual environment:
- > lbry-venv\Scripts\activate
-
- Install packages:
- > cd torba
- > pip install -e .
- > cd ../lbry
- > pip install -e .
- ```
+Install packages:
+```bash
+pip install -e .
+```
 
 ## Run the tests
+### Elasticsearch
 
-To run the unit tests from the repo directory:
+For running integration tests, Elasticsearch is required to be available at localhost:9200/
 
- ```
- python -m unittest discover -s lbry tests.unit
- ```
+The easiest way to start it is using docker with:
+```bash
+make elastic-docker
+```
+
+Alternative installation methods are available [at Elasticsearch website](https://www.elastic.co/guide/en/elasticsearch/reference/current/install-elasticsearch.html).
+
+To run the unit and integration tests from the repo directory:
+```
+python -m unittest discover tests.unit
+python -m unittest discover tests.integration
+```
 
 ## Usage
 
 To start the API server:
-    `lbrynet start`
+```
+lbrynet start
+```
 
+Whenever the code inside [lbry-sdk/lbry](./lbry)
+is modified we should run `make install` to recompile the `lbrynet`
+executable with the newest code.
+
+## Development
+
+When developing, remember to enter the environment,
+and if you wish start the server interactively.
+```bash
+$ source lbry-venv/bin/activate
+
+(lbry-venv) $ python lbry/extras/cli.py start
+```
+
+Parameters can be passed in the same way.
+```bash
+(lbry-venv) $ python lbry/extras/cli.py wallet balance
+```
+
+If a Python debugger (`pdb` or `ipdb`) is installed we can also start it
+in this way, set up break points, and step through the code.
+```bash
+(lbry-venv) $ pip install ipdb
+
+(lbry-venv) $ ipdb lbry/extras/cli.py
+```
 
 Happy hacking!

LICENSE

@@ -1,6 +1,6 @@
 The MIT License (MIT)
 
-Copyright (c) 2015-2019 LBRY Inc
+Copyright (c) 2015-2022 LBRY Inc
 
 Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the
 "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish,

MANIFEST.in

@@ -1,4 +1,4 @@
 include README.md
 include CHANGELOG.md
 include LICENSE
-recursive-include torba *.txt *.py
+recursive-include lbry *.txt *.py

Makefile

@@ -1,18 +1,26 @@
+.PHONY: install tools lint test test-unit test-unit-coverage test-integration idea
+
 install:
-	cd torba && pip install -e .
-	cd lbry && pip install -e .
-	pip install mypy==0.701
-	pip install coverage astroid pylint
+	pip install -e .
 
 lint:
-	cd lbry && pylint lbry
-	cd torba && pylint --rcfile=setup.cfg torba
-	cd torba && mypy --ignore-missing-imports torba
+	pylint --rcfile=setup.cfg lbry
+	#mypy --ignore-missing-imports lbry
 
-test:
-	cd lbry && tox
-	cd torba && tox
+test: test-unit test-integration
+
+test-unit:
+	python -m unittest discover tests.unit
+
+test-unit-coverage:
+	coverage run --source=lbry -m unittest discover -vv tests.unit
+
+test-integration:
+	tox
 
 idea:
 	mkdir -p .idea
-	cp -r lbry/scripts/idea/* .idea
+	cp -r scripts/idea/* .idea
+
+elastic-docker:
+	docker run -d -v lbryhub:/usr/share/elasticsearch/data -p 9200:9200 -p 9300:9300 -e"ES_JAVA_OPTS=-Xms512m -Xmx512m" -e "discovery.type=single-node" docker.elastic.co/elasticsearch/elasticsearch:7.12.1

README.md

@@ -1,15 +1,15 @@
-# <img src="https://raw.githubusercontent.com/lbryio/lbry-sdk/master/lbry/lbry.png" alt="LBRY" width="48" height="36" /> LBRY SDK [![Build Status](https://travis-ci.org/lbryio/lbry-sdk.svg?branch=master)](https://travis-ci.org/lbryio/lbry-sdk) [![Test Coverage](https://codecov.io/gh/lbryio/lbry-sdk/branch/master/graph/badge.svg)](https://codecov.io/gh/lbryio/lbry-sdk)
+# <img src="https://raw.githubusercontent.com/lbryio/lbry-sdk/master/lbry.png" alt="LBRY" width="48" height="36" /> LBRY SDK [![build](https://github.com/lbryio/lbry-sdk/actions/workflows/main.yml/badge.svg)](https://github.com/lbryio/lbry-sdk/actions/workflows/main.yml) [![coverage](https://coveralls.io/repos/github/lbryio/lbry-sdk/badge.svg)](https://coveralls.io/github/lbryio/lbry-sdk)
 
 LBRY is a decentralized peer-to-peer protocol for publishing and accessing digital content. It utilizes the [LBRY blockchain](https://github.com/lbryio/lbrycrd) as a global namespace and database of digital content. Blockchain entries contain searchable content metadata, identities, rights and access rules. LBRY also provides a data network that consists of peers (seeders) uploading and downloading data from other peers, possibly in exchange for payments, as well as a distributed hash table used by peers to discover other peers.
 
 LBRY SDK for Python is currently the most fully featured implementation of the LBRY Network protocols and includes many useful components and tools for building decentralized applications. Primary features and components include:
 
- * Built on Python 3.7+ and `asyncio`.
- * Kademlia DHT (Distributed Hash Table) implementation for finding peers to download from and announcing to peers what we have to host ([lbry.dht](https://github.com/lbryio/lbry-sdk/tree/master/lbry/lbry/dht)).
- * Blob exchange protocol for transferring encrypted blobs of content and negotiating payments ([lbry.blob_exchange](https://github.com/lbryio/lbry-sdk/tree/master/lbry/lbry/blob_exchange)).
- * Protobuf schema for encoding and decoding metadata stored on the blockchain ([lbry.schema](https://github.com/lbryio/lbry-sdk/tree/master/lbry/lbry/schema)).
- * Wallet implementation for the LBRY blockchain ([lbry.wallet](https://github.com/lbryio/lbry-sdk/tree/master/lbry/lbry/wallet)).
- * Daemon with a JSON-RPC API to ease building end user applications in any language and for automating various tasks ([lbry.extras.daemon](https://github.com/lbryio/lbry-sdk/tree/master/lbry/lbry/extras/daemon)). 
+ * Built on Python 3.7 and `asyncio`.
+ * Kademlia DHT (Distributed Hash Table) implementation for finding peers to download from and announcing to peers what we have to host ([lbry.dht](https://github.com/lbryio/lbry-sdk/tree/master/lbry/dht)).
+ * Blob exchange protocol for transferring encrypted blobs of content and negotiating payments ([lbry.blob_exchange](https://github.com/lbryio/lbry-sdk/tree/master/lbry/blob_exchange)).
+ * Protobuf schema for encoding and decoding metadata stored on the blockchain ([lbry.schema](https://github.com/lbryio/lbry-sdk/tree/master/lbry/schema)).
+ * Wallet implementation for the LBRY blockchain ([lbry.wallet](https://github.com/lbryio/lbry-sdk/tree/master/lbry/wallet)).
+ * Daemon with a JSON-RPC API to ease building end user applications in any language and for automating various tasks ([lbry.extras.daemon](https://github.com/lbryio/lbry-sdk/tree/master/lbry/extras/daemon)). 
 
 ## Installation
 
@@ -41,7 +41,7 @@ This project is MIT licensed. For the full license, see [LICENSE](LICENSE).
 
 ## Security
 
-We take security seriously. Please contact security@lbry.com regarding any security issues. [Our GPG key is here](https://lbry.com/faq/gpg-key) if you need it.
+We take security seriously. Please contact security@lbry.com regarding any security issues. [Our PGP key is here](https://lbry.com/faq/pgp-key) if you need it.
 
 ## Contact
 
@@ -53,4 +53,4 @@ The documentation for the API can be found [here](https://lbry.tech/api/sdk).
 
 Daemon defaults, ports, and other settings are documented [here](https://lbry.tech/resources/daemon-settings).
 
-Settings can be configured using a daemon-settings.yml file. An example can be found [here](https://github.com/lbryio/lbry-sdk/blob/master/lbry/example_daemon_settings.yml).
+Settings can be configured using a daemon-settings.yml file. An example can be found [here](https://github.com/lbryio/lbry-sdk/blob/master/example_daemon_settings.yml).

SECURITY.md

@@ -0,0 +1,9 @@
+# Security Policy
+
+## Supported Versions
+
+While we are not at v1.0 yet, only the latest release will be supported.
+
+## Reporting a Vulnerability
+
+See https://lbry.com/faq/security

docker/Dockerfile.dht_node

@@ -0,0 +1,43 @@
+FROM debian:10-slim
+
+ARG user=lbry
+ARG projects_dir=/home/$user
+ARG db_dir=/database
+
+ARG DOCKER_TAG
+ARG DOCKER_COMMIT=docker
+ENV DOCKER_TAG=$DOCKER_TAG DOCKER_COMMIT=$DOCKER_COMMIT
+
+RUN apt-get update && \
+    apt-get -y --no-install-recommends install \
+      wget \
+      automake libtool \
+      tar unzip \
+      build-essential \
+      pkg-config \
+      libleveldb-dev \
+      python3.7 \
+      python3-dev \
+      python3-pip \
+      python3-wheel \
+      python3-setuptools && \
+    update-alternatives --install /usr/bin/pip pip /usr/bin/pip3 1 && \
+    rm -rf /var/lib/apt/lists/*
+
+RUN groupadd -g 999 $user && useradd -m -u 999 -g $user $user
+
+COPY . $projects_dir
+RUN chown -R $user:$user $projects_dir
+RUN mkdir -p $db_dir
+RUN chown -R $user:$user $db_dir
+
+USER $user
+WORKDIR $projects_dir
+
+RUN python3 -m pip install -U setuptools pip
+RUN make install
+RUN python3 docker/set_build.py
+RUN rm ~/.cache -rf
+VOLUME $db_dir
+ENTRYPOINT ["python3", "scripts/dht_node.py"]
+

docker/Dockerfile.wallet_server

@@ -0,0 +1,56 @@
+FROM debian:10-slim
+
+ARG user=lbry
+ARG db_dir=/database
+ARG projects_dir=/home/$user
+
+ARG DOCKER_TAG
+ARG DOCKER_COMMIT=docker
+ENV DOCKER_TAG=$DOCKER_TAG DOCKER_COMMIT=$DOCKER_COMMIT
+
+RUN apt-get update && \
+    apt-get -y --no-install-recommends install \
+      wget \
+      tar unzip \
+      build-essential \
+      automake libtool \
+      pkg-config \
+      libleveldb-dev \
+      python3.7 \
+      python3-dev \
+      python3-pip \
+      python3-wheel \
+      python3-cffi \
+      python3-setuptools && \
+    update-alternatives --install /usr/bin/pip pip /usr/bin/pip3 1 && \
+    rm -rf /var/lib/apt/lists/*
+
+RUN groupadd -g 999 $user && useradd -m -u 999 -g $user $user
+RUN mkdir -p $db_dir
+RUN chown -R $user:$user $db_dir
+
+COPY . $projects_dir
+RUN chown -R $user:$user $projects_dir
+
+USER $user
+WORKDIR $projects_dir
+
+RUN pip install uvloop
+RUN make install
+RUN python3 docker/set_build.py
+RUN rm ~/.cache -rf
+
+# entry point
+ARG host=0.0.0.0
+ARG tcp_port=50001
+ARG daemon_url=http://lbry:lbry@localhost:9245/
+VOLUME $db_dir
+ENV TCP_PORT=$tcp_port
+ENV HOST=$host
+ENV DAEMON_URL=$daemon_url
+ENV DB_DIRECTORY=$db_dir
+ENV MAX_SESSIONS=1000000000
+ENV MAX_SEND=1000000000000000000
+ENV EVENT_LOOP_POLICY=uvloop
+COPY ./docker/wallet_server_entrypoint.sh /entrypoint.sh
+ENTRYPOINT ["/entrypoint.sh"]

docker/Dockerfile.web

@@ -0,0 +1,45 @@
+FROM debian:10-slim
+
+ARG user=lbry
+ARG downloads_dir=/database
+ARG projects_dir=/home/$user
+
+ARG DOCKER_TAG
+ARG DOCKER_COMMIT=docker
+ENV DOCKER_TAG=$DOCKER_TAG DOCKER_COMMIT=$DOCKER_COMMIT
+
+RUN apt-get update && \
+    apt-get -y --no-install-recommends install \
+      wget \
+      automake libtool \
+      tar unzip \
+      build-essential \
+      pkg-config \
+      libleveldb-dev \
+      python3.7 \
+      python3-dev \
+      python3-pip \
+      python3-wheel \
+      python3-setuptools && \
+    update-alternatives --install /usr/bin/pip pip /usr/bin/pip3 1 && \
+    rm -rf /var/lib/apt/lists/*
+
+RUN groupadd -g 999 $user && useradd -m -u 999 -g $user $user
+RUN mkdir -p $downloads_dir
+RUN chown -R $user:$user $downloads_dir
+
+COPY . $projects_dir
+RUN chown -R $user:$user $projects_dir
+
+USER $user
+WORKDIR $projects_dir
+
+RUN pip install uvloop
+RUN make install
+RUN python3 docker/set_build.py
+RUN rm ~/.cache -rf
+
+# entry point
+VOLUME $downloads_dir
+COPY ./docker/webconf.yaml /webconf.yaml
+ENTRYPOINT ["/home/lbry/.local/bin/lbrynet", "start", "--config=/webconf.yaml"]

docker/README.md

@@ -0,0 +1,9 @@
+### How to run with docker-compose
+1. Edit config file and after that fix permissions with
+```
+sudo chown -R 999:999 webconf.yaml
+```
+2. Start SDK with
+```
+docker-compose up -d
+```

docker/docker-compose-wallet-server.yml

@@ -0,0 +1,49 @@
+version: "3"
+
+volumes:
+  wallet_server:
+  es01:
+
+services:
+  wallet_server:
+    depends_on: 
+      - es01
+    image: lbry/wallet-server:${WALLET_SERVER_TAG:-latest-release}
+    restart: always
+    network_mode: host
+    ports:
+      - "50001:50001" # rpc port
+      - "2112:2112"   # uncomment to enable prometheus
+    volumes:
+      - "wallet_server:/database"
+    environment:
+      - DAEMON_URL=http://lbry:lbry@127.0.0.1:9245
+      - MAX_QUERY_WORKERS=4
+      - CACHE_MB=1024
+      - CACHE_ALL_TX_HASHES=
+      - CACHE_ALL_CLAIM_TXOS=
+      - MAX_SEND=1000000000000000000
+      - MAX_RECEIVE=1000000000000000000
+      - MAX_SESSIONS=100000
+      - HOST=0.0.0.0
+      - TCP_PORT=50001
+      - PROMETHEUS_PORT=2112
+      - FILTERING_CHANNEL_IDS=770bd7ecba84fd2f7607fb15aedd2b172c2e153f 95e5db68a3101df19763f3a5182e4b12ba393ee8
+      - BLOCKING_CHANNEL_IDS=dd687b357950f6f271999971f43c785e8067c3a9 06871aa438032244202840ec59a469b303257cad b4a2528f436eca1bf3bf3e10ff3f98c57bd6c4c6
+  es01:
+    image: docker.elastic.co/elasticsearch/elasticsearch:7.11.0
+    container_name: es01
+    environment:
+      - node.name=es01
+      - discovery.type=single-node
+      - indices.query.bool.max_clause_count=8192
+      - bootstrap.memory_lock=true
+      - "ES_JAVA_OPTS=-Xms4g -Xmx4g"  # no more than 32, remember to disable swap
+    ulimits:
+      memlock:
+        soft: -1
+        hard: -1
+    volumes:
+      - es01:/usr/share/elasticsearch/data
+    ports:
+      - 127.0.0.1:9200:9200

docker/docker-compose.yml

@@ -0,0 +1,9 @@
+version: '3'
+services:
+    websdk:
+        image: vshyba/websdk
+        ports:
+            - '5279:5279'
+            - '5280:5280'
+        volumes:
+            - ./webconf.yaml:/webconf.yaml

docker/hooks/build

@@ -0,0 +1,7 @@
+#!/bin/bash
+
+DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" >/dev/null 2>&1 && pwd )"
+cd "$DIR/../.." ## make sure we're in the right place. Docker Hub screws this up sometimes
+echo "docker build dir: $(pwd)"
+
+docker build --build-arg DOCKER_TAG=$DOCKER_TAG --build-arg DOCKER_COMMIT=$SOURCE_COMMIT -f $DOCKERFILE_PATH -t $IMAGE_NAME .

docker/install_choco.ps1

@@ -0,0 +1,11 @@
+# requires powershell and .NET 4+. see https://chocolatey.org/install for more info.
+
+$chocoVersion = powershell choco -v
+if(-not($chocoVersion)){
+    Write-Output "Chocolatey is not installed, installing now"
+    Write-Output "IF YOU KEEP GETTING THIS MESSAGE ON EVERY BUILD, TRY RESTARTING THE GITLAB RUNNER SO IT GETS CHOCO INTO IT'S ENV"
+    Set-ExecutionPolicy Bypass -Scope Process -Force; iex ((New-Object System.Net.WebClient).DownloadString('https://chocolatey.org/install.ps1'))
+}
+else{
+    Write-Output "Chocolatey version $chocoVersion is already installed"
+}

docker/set_build.py

@@ -0,0 +1,44 @@
+import sys
+import os
+import re
+import logging
+import lbry.build_info as build_info_mod
+
+log = logging.getLogger()
+log.addHandler(logging.StreamHandler())
+log.setLevel(logging.DEBUG)
+
+
+def _check_and_set(d: dict, key: str, value: str):
+    try:
+        d[key]
+    except KeyError:
+        raise Exception(f"{key} var does not exist in {build_info_mod.__file__}")
+    d[key] = value
+
+
+def main():
+    build_info = {item: build_info_mod.__dict__[item] for item in dir(build_info_mod) if not item.startswith("__")}
+
+    commit_hash = os.getenv('DOCKER_COMMIT', os.getenv('GITHUB_SHA'))
+    if commit_hash is None:
+        raise ValueError("Commit hash not found in env vars")
+    _check_and_set(build_info, "COMMIT_HASH", commit_hash[:6])
+
+    docker_tag = os.getenv('DOCKER_TAG')
+    if docker_tag:
+        _check_and_set(build_info, "DOCKER_TAG", docker_tag)
+        _check_and_set(build_info, "BUILD", "docker")
+    else:
+        if re.match(r'refs/tags/v\d+\.\d+\.\d+$', str(os.getenv('GITHUB_REF'))):
+            _check_and_set(build_info, "BUILD", "release")
+        else:
+            _check_and_set(build_info, "BUILD", "qa")
+
+    log.debug("build info: %s", ", ".join([f"{k}={v}" for k, v in build_info.items()]))
+    with open(build_info_mod.__file__, 'w') as f:
+        f.write("\n".join([f"{k} = \"{v}\"" for k, v in build_info.items()]) + "\n")
+
+
+if __name__ == '__main__':
+    sys.exit(main())

docker/wallet_server_entrypoint.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+
+# entrypoint for wallet server Docker image
+
+set -euo pipefail
+
+SNAPSHOT_URL="${SNAPSHOT_URL:-}" #off by default. latest snapshot at https://lbry.com/snapshot/wallet
+
+if [[ -n "$SNAPSHOT_URL" ]] && [[ ! -f /database/lbry-leveldb ]]; then
+  files="$(ls)"
+  echo "Downloading wallet snapshot from $SNAPSHOT_URL"
+  wget --no-verbose --trust-server-names --content-disposition "$SNAPSHOT_URL"
+  echo "Extracting snapshot..."
+  filename="$(grep -vf <(echo "$files") <(ls))" # finds the file that was not there before
+  case "$filename" in
+    *.tgz|*.tar.gz|*.tar.bz2 )  tar xvf "$filename" --directory /database ;;
+    *.zip ) unzip "$filename" -d /database ;;
+    * ) echo "Don't know how to extract ${filename}. SNAPSHOT COULD NOT BE LOADED" && exit 1 ;;
+  esac
+  rm "$filename"
+fi
+
+/home/lbry/.local/bin/lbry-hub-elastic-sync
+echo 'starting server'
+/home/lbry/.local/bin/lbry-hub "$@"

docker/webconf.yaml

@@ -0,0 +1,9 @@
+allowed_origin: "*"
+max_key_fee: "0.0 USD"
+save_files: false
+save_blobs: false
+streaming_server: "0.0.0.0:5280"
+api: "0.0.0.0:5279"
+data_dir: /tmp
+download_dir: /tmp
+wallet_dir: /tmp

lbry/.gitattributes

@@ -1 +0,0 @@
-/CHANGELOG.md merge=union

lbry/.gitignore

@@ -1,13 +0,0 @@
-.DS_Store
-/build
-/dist
-/.tox
-/.idea
-/.coverage
-/lbry-venv
-
-lbry.egg-info
-__pycache__
-_trial_temp/
-
-/tests/integration/files

lbry/.pylintrc

@@ -1,441 +0,0 @@
-[MASTER]
-
-# Specify a configuration file.
-#rcfile=
-
-# Python code to execute, usually for sys.path manipulation such as
-# pygtk.require().
-#init-hook=
-
-# Add files or directories to the blacklist. They should be base names, not
-# paths.
-ignore=CVS,schema
-
-# Add files or directories matching the regex patterns to the
-# blacklist. The regex matches against base names, not paths.
-# `\.#.*` - add emacs tmp files to the blacklist
-ignore-patterns=\.#.*
-
-# Pickle collected data for later comparisons.
-persistent=yes
-
-# List of plugins (as comma separated values of python modules names) to load,
-# usually to register additional checkers.
-load-plugins=
-
-# Use multiple processes to speed up Pylint.
-jobs=4
-
-# Allow loading of arbitrary C extensions. Extensions are imported into the
-# active Python interpreter and may run arbitrary code.
-unsafe-load-any-extension=no
-
-# A comma-separated list of package or module names from where C extensions may
-# be loaded. Extensions are loading into the active Python interpreter and may
-# run arbitrary code
-# extension-pkg-whitelist=
-
-# Allow optimization of some AST trees. This will activate a peephole AST
-# optimizer, which will apply various small optimizations. For instance, it can
-# be used to obtain the result of joining multiple strings with the addition
-# operator. Joining a lot of strings can lead to a maximum recursion error in
-# Pylint and this flag can prevent that. It has one side effect, the resulting
-# AST will be different than the one from reality.
-optimize-ast=no
-
-
-[MESSAGES CONTROL]
-
-# Only show warnings with the listed confidence levels. Leave empty to show
-# all. Valid levels: HIGH, INFERENCE, INFERENCE_FAILURE, UNDEFINED
-confidence=
-
-# Enable the message, report, category or checker with the given id(s). You can
-# either give multiple identifier separated by comma (,) or put this option
-# multiple time (only on the command line, not in the configuration file where
-# it should appear only once). See also the "--disable" option for examples.
-#enable=
-
-# Disable the message, report, category or checker with the given id(s). You
-# can either give multiple identifiers separated by comma (,) or put this
-# option multiple times (only on the command line, not in the configuration
-# file where it should appear only once).You can also use "--disable=all" to
-# disable everything first and then re-enable specific checks. For example, if
-# you want to run only the similarities checker, you can use "--disable=all
-# --enable=similarities". If you want to run only the classes checker, but have
-# no Warning level messages displayed, use"--disable=all --enable=classes
-# --disable=W"
-disable=
-  anomalous-backslash-in-string,
-  arguments-differ,
-  attribute-defined-outside-init,
-  bad-continuation,
-  bare-except,
-  broad-except,
-  cell-var-from-loop,
-  consider-iterating-dictionary,
-  dangerous-default-value,
-  duplicate-code,
-  fixme,
-  invalid-name,
-  len-as-condition,
-  locally-disabled,
-  logging-not-lazy,
-  missing-docstring,
-  no-else-return,
-  no-init,
-  no-member,
-  no-self-use,
-  protected-access,
-  redefined-builtin,
-  redefined-outer-name,
-  redefined-variable-type,
-  relative-import,
-  signature-differs,
-  super-init-not-called,
-  too-few-public-methods,
-  too-many-arguments,
-  too-many-branches,
-  too-many-instance-attributes,
-  too-many-lines,
-  too-many-locals,
-  too-many-nested-blocks,
-  too-many-public-methods,
-  too-many-return-statements,
-  too-many-statements,
-  trailing-newlines,
-  undefined-loop-variable,
-  ungrouped-imports,
-  unnecessary-lambda,
-  unused-argument,
-  unused-variable,
-  wrong-import-order,
-  wrong-import-position,
-  deprecated-lambda,
-  simplifiable-if-statement,
-  unidiomatic-typecheck,
-  inconsistent-return-statements,
-  keyword-arg-before-vararg,
-  assignment-from-no-return,
-  useless-return,
-  assignment-from-none,
-  stop-iteration-return
-
-
-[REPORTS]
-
-# Set the output format. Available formats are text, parseable, colorized, msvs
-# (visual studio) and html. You can also give a reporter class, eg
-# mypackage.mymodule.MyReporterClass.
-output-format=text
-
-# Put messages in a separate file for each module / package specified on the
-# command line instead of printing them on stdout. Reports (if any) will be
-# written in a file name "pylint_global.[txt|html]".
-files-output=no
-
-# Tells whether to display a full report or only the messages
-reports=no
-
-# Python expression which should return a note less than 10 (10 is the highest
-# note). You have access to the variables errors warning, statement which
-# respectively contain the number of errors / warnings messages and the total
-# number of statements analyzed. This is used by the global evaluation report
-# (RP0004).
-evaluation=10.0 - ((float(5 * error + warning + refactor + convention) / statement) * 10)
-
-# Template used to display messages. This is a python new-style format string
-# used to format the message information. See doc for all details
-#msg-template=
-
-
-[VARIABLES]
-
-# Tells whether we should check for unused import in __init__ files.
-init-import=no
-
-# A regular expression matching the name of dummy variables (i.e. expectedly
-# not used).
-dummy-variables-rgx=_$|dummy
-
-# List of additional names supposed to be defined in builtins. Remember that
-# you should avoid to define new builtins when possible.
-additional-builtins=
-
-# List of strings which can identify a callback function by name. A callback
-# name must start or end with one of those strings.
-callbacks=cb_,_cb
-
-
-[LOGGING]
-
-# Logging modules to check that the string format arguments are in logging
-# function parameter format
-logging-modules=logging
-
-
-[BASIC]
-
-# List of builtins function names that should not be used, separated by a comma
-bad-functions=map,filter,input
-
-# Good variable names which should always be accepted, separated by a comma
-# allow `d` as its used frequently for deferred callback chains
-good-names=i,j,k,ex,Run,_,d
-
-# Bad variable names which should always be refused, separated by a comma
-bad-names=foo,bar,baz,toto,tutu,tata
-
-# Colon-delimited sets of names that determine each other's naming style when
-# the name regexes allow several styles.
-name-group=
-
-# Include a hint for the correct naming format with invalid-name
-include-naming-hint=no
-
-# Regular expression matching correct function names
-function-rgx=[a-z_][a-z0-9_]{2,30}$
-
-# Naming hint for function names
-function-name-hint=[a-z_][a-z0-9_]{2,30}$
-
-# Regular expression matching correct variable names
-variable-rgx=[a-z_][a-z0-9_]{2,30}$
-
-# Naming hint for variable names
-variable-name-hint=[a-z_][a-z0-9_]{2,30}$
-
-# Regular expression matching correct constant names
-const-rgx=(([A-Z_][A-Z0-9_]*)|(__.*__))$
-
-# Naming hint for constant names
-const-name-hint=(([A-Z_][A-Z0-9_]*)|(__.*__))$
-
-# Regular expression matching correct attribute names
-attr-rgx=[a-z_][a-z0-9_]{2,30}$
-
-# Naming hint for attribute names
-attr-name-hint=[a-z_][a-z0-9_]{2,30}$
-
-# Regular expression matching correct argument names
-argument-rgx=[a-z_][a-z0-9_]{2,30}$
-
-# Naming hint for argument names
-argument-name-hint=[a-z_][a-z0-9_]{2,30}$
-
-# Regular expression matching correct class attribute names
-class-attribute-rgx=([A-Za-z_][A-Za-z0-9_]{2,30}|(__.*__))$
-
-# Naming hint for class attribute names
-class-attribute-name-hint=([A-Za-z_][A-Za-z0-9_]{2,30}|(__.*__))$
-
-# Regular expression matching correct inline iteration names
-inlinevar-rgx=[A-Za-z_][A-Za-z0-9_]*$
-
-# Naming hint for inline iteration names
-inlinevar-name-hint=[A-Za-z_][A-Za-z0-9_]*$
-
-# Regular expression matching correct class names
-class-rgx=[A-Z_][a-zA-Z0-9]+$
-
-# Naming hint for class names
-class-name-hint=[A-Z_][a-zA-Z0-9]+$
-
-# Regular expression matching correct module names
-module-rgx=(([a-z_][a-z0-9_]*)|([A-Z][a-zA-Z0-9]+))$
-
-# Naming hint for module names
-module-name-hint=(([a-z_][a-z0-9_]*)|([A-Z][a-zA-Z0-9]+))$
-
-# Regular expression matching correct method names
-method-rgx=[a-z_][a-z0-9_]{2,30}$
-
-# Naming hint for method names
-method-name-hint=[a-z_][a-z0-9_]{2,30}$
-
-# Regular expression which should only match function or class names that do
-# not require a docstring.
-no-docstring-rgx=^_
-
-# Minimum line length for functions/classes that require docstrings, shorter
-# ones are exempt.
-docstring-min-length=-1
-
-
-[ELIF]
-
-# Maximum number of nested blocks for function / method body
-max-nested-blocks=5
-
-
-[SPELLING]
-
-# Spelling dictionary name. Available dictionaries: none. To make it working
-# install python-enchant package.
-spelling-dict=
-
-# List of comma separated words that should not be checked.
-spelling-ignore-words=
-
-# A path to a file that contains private dictionary; one word per line.
-spelling-private-dict-file=
-
-# Tells whether to store unknown words to indicated private dictionary in
-# --spelling-private-dict-file option instead of raising a message.
-spelling-store-unknown-words=no
-
-
-[FORMAT]
-
-# Maximum number of characters on a single line.
-max-line-length=120
-
-# Regexp for a line that is allowed to be longer than the limit.
-ignore-long-lines=^\s*(# )?<?https?://\S+>?$
-
-# Allow the body of an if to be on the same line as the test if there is no
-# else.
-single-line-if-stmt=no
-
-# List of optional constructs for which whitespace checking is disabled. `dict-
-# separator` is used to allow tabulation in dicts, etc.: {1  : 1,\n222: 2}.
-# `trailing-comma` allows a space between comma and closing bracket: (a, ).
-# `empty-line` allows space-only lines.
-no-space-check=trailing-comma,dict-separator
-
-# Maximum number of lines in a module
-max-module-lines=1000
-
-# String used as indentation unit. This is usually "    " (4 spaces) or "\t" (1
-# tab).
-indent-string='    '
-
-# Number of spaces of indent required inside a hanging  or continued line.
-indent-after-paren=4
-
-# Expected format of line ending, e.g. empty (any line ending), LF or CRLF.
-expected-line-ending-format=
-
-
-[MISCELLANEOUS]
-
-# List of note tags to take in consideration, separated by a comma.
-notes=FIXME,XXX,TODO
-
-
-[SIMILARITIES]
-
-# Minimum lines number of a similarity.
-min-similarity-lines=4
-
-# Ignore comments when computing similarities.
-ignore-comments=yes
-
-# Ignore docstrings when computing similarities.
-ignore-docstrings=yes
-
-# Ignore imports when computing similarities.
-ignore-imports=no
-
-
-[TYPECHECK]
-
-# Tells whether missing members accessed in mixin class should be ignored. A
-# mixin class is detected if its name ends with "mixin" (case insensitive).
-ignore-mixin-members=yes
-
-# List of module names for which member attributes should not be checked
-# (useful for modules/projects where namespaces are manipulated during runtime
-# and thus existing member attributes cannot be deduced by static analysis. It
-# supports qualified module names, as well as Unix pattern matching.
-ignored-modules=leveldb,distutils
-# Ignoring distutils because: https://github.com/PyCQA/pylint/issues/73
-
-# List of classes names for which member attributes should not be checked
-# (useful for classes with attributes dynamically set). This supports can work
-# with qualified names.
-# ignored-classes=
-
-
-# List of members which are set dynamically and missed by pylint inference
-# system, and so shouldn't trigger E1101 when accessed. Python regular
-# expressions are accepted.
-generated-members=lbrynet.lbrynet_daemon.LBRYDaemon.Parameters
-
-
-[IMPORTS]
-
-# Deprecated modules which should not be used, separated by a comma
-deprecated-modules=regsub,TERMIOS,Bastion,rexec
-
-# Create a graph of every (i.e. internal and external) dependencies in the
-# given file (report RP0402 must not be disabled)
-import-graph=
-
-# Create a graph of external dependencies in the given file (report RP0402 must
-# not be disabled)
-ext-import-graph=
-
-# Create a graph of internal dependencies in the given file (report RP0402 must
-# not be disabled)
-int-import-graph=
-
-
-[DESIGN]
-
-# Maximum number of arguments for function / method
-max-args=10
-
-# Argument names that match this expression will be ignored. Default to name
-# with leading underscore
-ignored-argument-names=_.*
-
-# Maximum number of locals for function / method body
-max-locals=15
-
-# Maximum number of return / yield for function / method body
-max-returns=6
-
-# Maximum number of branch for function / method body
-max-branches=12
-
-# Maximum number of statements in function / method body
-max-statements=50
-
-# Maximum number of parents for a class (see R0901).
-max-parents=8
-
-# Maximum number of attributes for a class (see R0902).
-max-attributes=7
-
-# Minimum number of public methods for a class (see R0903).
-min-public-methods=2
-
-# Maximum number of public methods for a class (see R0904).
-max-public-methods=20
-
-# Maximum number of boolean expressions in a if statement
-max-bool-expr=5
-
-
-[CLASSES]
-
-# List of method names used to declare (i.e. assign) instance attributes.
-defining-attr-methods=__init__,__new__,setUp
-
-# List of valid names for the first argument in a class method.
-valid-classmethod-first-arg=cls
-
-# List of valid names for the first argument in a metaclass class method.
-valid-metaclass-classmethod-first-arg=mcs
-
-# List of member names, which should be excluded from the protected access
-# warning.
-exclude-protected=_asdict,_fields,_replace,_source,_make
-
-
-[EXCEPTIONS]
-
-# Exceptions that will emit a warning when being caught. Defaults to
-# "Exception"
-overgeneral-exceptions=Exception

lbry/README.md

@@ -1,56 +0,0 @@
-# <img src="https://raw.githubusercontent.com/lbryio/lbry-sdk/master/lbry/lbry.png" alt="LBRY" width="48" height="36" /> LBRY SDK [![Build Status](https://travis-ci.org/lbryio/lbry-sdk.svg?branch=master)](https://travis-ci.org/lbryio/lbry-sdk) [![Test Coverage](https://codecov.io/gh/lbryio/lbry-sdk/branch/master/graph/badge.svg)](https://codecov.io/gh/lbryio/lbry-sdk)
-
-LBRY is a decentralized peer-to-peer protocol for publishing and accessing digital content. It utilizes the [LBRY blockchain](https://github.com/lbryio/lbrycrd) as a global namespace and database of digital content. Blockchain entries contain searchable content metadata, identities, rights and access rules. LBRY also provides a data network that consists of peers (seeders) uploading and downloading data from other peers, possibly in exchange for payments, as well as a distributed hash table used by peers to discover other peers.
-
-LBRY SDK for Python is currently the most full featured implementation of the LBRY Network protocols and includes many useful components and tools for building decentralized applications. Primary features and components:
-
- * Built on Python 3.7+ and `asyncio`.
- * Kademlia DHT (Distributed Hash Table) implementation for finding peers to download from and announcing to peers what we have to host ([lbry.dht](https://github.com/lbryio/lbry-sdk/tree/master/lbry/lbry/dht)).
- * Blob exchange protocol for transferring encrypted blobs of content and negotiating payments ([lbry.blob_exchange](https://github.com/lbryio/lbry-sdk/tree/master/lbry/lbry/blob_exchange)).
- * Protobuf schema for encoding and decoding metadata stored on the blockchain ([lbry.schema](https://github.com/lbryio/lbry-sdk/tree/master/lbry/lbry/schema)).
- * Wallet implementation for the LBRY blockchain ([lbry.wallet](https://github.com/lbryio/lbry-sdk/tree/master/lbry/lbry/wallet)).
- * Daemon with a JSON-RPC API to ease building end user applications in any language and for automating various tasks ([lbry.extras.daemon](https://github.com/lbryio/lbry-sdk/tree/master/lbry/lbry/extras/daemon)). 
-
-## Installation
-
-Our [releases page](https://github.com/lbryio/lbry-sdk/releases) contains pre-built binaries of the latest release, pre-releases, and past releases for macOS, Debian-based Linux, and Windows. [Automated travis builds](http://build.lbry.io/daemon/) are also available for testing.
-
-## Usage
-
-Run `lbrynet start` to launch the API server.
-
-By default, `lbrynet` will provide a JSON-RPC server at `http://localhost:5279`. It is easy to interact with via cURL or sane programming languages.
-
-Our [quickstart guide](https://lbry.tech/playground) provides a simple walkthrough and examples for learning.
-
-With the daemon running, `lbrynet commands` will show you a list of commands.
-
-The full API is documented [here](https://lbry.tech/api/sdk).
-
-## Running from source
-
-Installing from source is also relatively painless. Full instructions are in [INSTALL.md](INSTALL.md)
-
-## Contributing
-
-Contributions to this project are welcome, encouraged, and compensated. For more details, please check [this](https://lbry.tech/contribute) link.
-
-## License
-
-This project is MIT licensed. For the full license, see [LICENSE](LICENSE).
-
-## Security
-
-We take security seriously. Please contact security@lbry.com regarding any security issues. [Our GPG key is here](https://lbry.com/faq/gpg-key) if you need it.
-
-## Contact
-
-The primary contact for this project is [@eukreign](mailto:lex@lbry.com).
-
-## Additional information and links
-
-The documentation for the API can be found [here](https://lbry.tech/api/sdk).
-
-Daemon defaults, ports, and other settings are documented [here](https://lbry.tech/resources/daemon-settings).
-
-Settings can be configured using a daemon-settings.yml file. An example can be found [here](https://github.com/lbryio/lbry-sdk/blob/master/lbry/example_daemon_settings.yml).

lbry/__init__.py

@@ -0,0 +1,2 @@
+__version__ = "0.113.0"
+version = tuple(map(int, __version__.split('.')))  # pylint: disable=invalid-name

lbry/blob/__init__.py

@@ -0,0 +1,6 @@
+from lbry.utils import get_lbry_hash_obj
+
+MAX_BLOB_SIZE = 2 * 2 ** 20
+
+# digest_size is in bytes, and blob hashes are hex encoded
+BLOBHASH_LENGTH = get_lbry_hash_obj().digest_size * 2

lbry/blob/blob_file.py

@@ -1,5 +1,6 @@
 import os
 import re
+import time
 import asyncio
 import binascii
 import logging
@@ -9,18 +10,20 @@ from io import BytesIO
 from cryptography.hazmat.primitives.ciphers import Cipher, modes
 from cryptography.hazmat.primitives.ciphers.algorithms import AES
 from cryptography.hazmat.primitives.padding import PKCS7
+from cryptography.hazmat.backends import default_backend
 
-from lbry.cryptoutils import backend, get_lbry_hash_obj
+from lbry.utils import get_lbry_hash_obj
 from lbry.error import DownloadCancelledError, InvalidBlobHashError, InvalidDataError
 
-from lbry.blob import MAX_BLOB_SIZE, blobhash_length
+from lbry.blob import MAX_BLOB_SIZE, BLOBHASH_LENGTH
 from lbry.blob.blob_info import BlobInfo
 from lbry.blob.writer import HashBlobWriter
 
 log = logging.getLogger(__name__)
 
 
-_hexmatch = re.compile("^[a-f,0-9]+$")
+HEXMATCH = re.compile("^[a-f,0-9]+$")
+BACKEND = default_backend()
 
 
 def is_valid_blobhash(blobhash: str) -> bool:
@@ -31,11 +34,11 @@ def is_valid_blobhash(blobhash: str) -> bool:
 
     @return: True/False
     """
-    return len(blobhash) == blobhash_length and _hexmatch.match(blobhash)
+    return len(blobhash) == BLOBHASH_LENGTH and HEXMATCH.match(blobhash)
 
 
 def encrypt_blob_bytes(key: bytes, iv: bytes, unencrypted: bytes) -> typing.Tuple[bytes, str]:
-    cipher = Cipher(AES(key), modes.CBC(iv), backend=backend)
+    cipher = Cipher(AES(key), modes.CBC(iv), backend=BACKEND)
     padder = PKCS7(AES.block_size).padder()
     encryptor = cipher.encryptor()
     encrypted = encryptor.update(padder.update(unencrypted) + padder.finalize()) + encryptor.finalize()
@@ -47,7 +50,7 @@ def encrypt_blob_bytes(key: bytes, iv: bytes, unencrypted: bytes) -> typing.Tupl
 def decrypt_blob_bytes(data: bytes, length: int, key: bytes, iv: bytes) -> bytes:
     if len(data) != length:
         raise ValueError("unexpected length")
-    cipher = Cipher(AES(key), modes.CBC(iv), backend=backend)
+    cipher = Cipher(AES(key), modes.CBC(iv), backend=BACKEND)
     unpadder = PKCS7(AES.block_size).unpadder()
     decryptor = cipher.decryptor()
     return unpadder.update(decryptor.update(data) + decryptor.finalize()) + unpadder.finalize()
@@ -68,21 +71,27 @@ class AbstractBlob:
         'writers',
         'verified',
         'writing',
-        'readers'
+        'readers',
+        'added_on',
+        'is_mine',
     ]
 
-    def __init__(self, loop: asyncio.AbstractEventLoop, blob_hash: str, length: typing.Optional[int] = None,
-                 blob_completed_callback: typing.Optional[typing.Callable[['AbstractBlob'], asyncio.Task]] = None,
-                 blob_directory: typing.Optional[str] = None):
+    def __init__(
+        self, loop: asyncio.AbstractEventLoop, blob_hash: str, length: typing.Optional[int] = None,
+        blob_completed_callback: typing.Optional[typing.Callable[['AbstractBlob'], asyncio.Task]] = None,
+        blob_directory: typing.Optional[str] = None, added_on: typing.Optional[int] = None, is_mine: bool = False,
+    ):
         self.loop = loop
         self.blob_hash = blob_hash
         self.length = length
         self.blob_completed_callback = blob_completed_callback
         self.blob_directory = blob_directory
         self.writers: typing.Dict[typing.Tuple[typing.Optional[str], typing.Optional[int]], HashBlobWriter] = {}
-        self.verified: asyncio.Event = asyncio.Event(loop=self.loop)
-        self.writing: asyncio.Event = asyncio.Event(loop=self.loop)
+        self.verified: asyncio.Event = asyncio.Event()
+        self.writing: asyncio.Event = asyncio.Event()
         self.readers: typing.List[typing.BinaryIO] = []
+        self.added_on = added_on or time.time()
+        self.is_mine = is_mine
 
         if not is_valid_blobhash(blob_hash):
             raise InvalidBlobHashError(blob_hash)
@@ -108,7 +117,7 @@ class AbstractBlob:
                 if reader in self.readers:
                     self.readers.remove(reader)
 
-    def _write_blob(self, blob_bytes: bytes):
+    def _write_blob(self, blob_bytes: bytes) -> asyncio.Task:
         raise NotImplementedError()
 
     def set_length(self, length) -> None:
@@ -142,7 +151,7 @@ class AbstractBlob:
 
     def close(self):
         while self.writers:
-            peer, writer = self.writers.popitem()
+            _, writer = self.writers.popitem()
             if writer and writer.finished and not writer.finished.done() and not self.loop.is_closed():
                 writer.finished.cancel()
         while self.readers:
@@ -163,7 +172,10 @@ class AbstractBlob:
         if not self.is_readable():
             raise OSError('blob files cannot be read')
         with self.reader_context() as handle:
-            return await self.loop.sendfile(writer.transport, handle, count=self.get_length())
+            try:
+                return await self.loop.sendfile(writer.transport, handle, count=self.get_length())
+            except (ConnectionError, BrokenPipeError, RuntimeError, OSError, AttributeError):
+                return -1
 
     def decrypt(self, key: bytes, iv: bytes) -> bytes:
         """
@@ -175,35 +187,42 @@ class AbstractBlob:
 
     @classmethod
     async def create_from_unencrypted(
-            cls, loop: asyncio.AbstractEventLoop, blob_dir: typing.Optional[str], key: bytes, iv: bytes,
-            unencrypted: bytes, blob_num: int,
-            blob_completed_callback: typing.Optional[typing.Callable[['AbstractBlob'], None]] = None) -> BlobInfo:
+        cls, loop: asyncio.AbstractEventLoop, blob_dir: typing.Optional[str], key: bytes, iv: bytes,
+        unencrypted: bytes, blob_num: int, added_on: int, is_mine: bool,
+        blob_completed_callback: typing.Optional[typing.Callable[['AbstractBlob'], None]] = None,
+    ) -> BlobInfo:
         """
         Create an encrypted BlobFile from plaintext bytes
         """
 
         blob_bytes, blob_hash = encrypt_blob_bytes(key, iv, unencrypted)
         length = len(blob_bytes)
-        blob = cls(loop, blob_hash, length, blob_completed_callback, blob_dir)
+        blob = cls(loop, blob_hash, length, blob_completed_callback, blob_dir, added_on, is_mine)
         writer = blob.get_blob_writer()
         writer.write(blob_bytes)
         await blob.verified.wait()
-        return BlobInfo(blob_num, length, binascii.hexlify(iv).decode(), blob_hash)
+        return BlobInfo(blob_num, length, binascii.hexlify(iv).decode(), added_on, blob_hash, is_mine)
 
     def save_verified_blob(self, verified_bytes: bytes):
         if self.verified.is_set():
             return
-        if self.is_writeable():
-            self._write_blob(verified_bytes)
+
+        def update_events(_):
             self.verified.set()
+            self.writing.clear()
+
+        if self.is_writeable():
+            self.writing.set()
+            task = self._write_blob(verified_bytes)
+            task.add_done_callback(update_events)
             if self.blob_completed_callback:
-                self.blob_completed_callback(self)
+                task.add_done_callback(lambda _: self.blob_completed_callback(self))
 
     def get_blob_writer(self, peer_address: typing.Optional[str] = None,
                         peer_port: typing.Optional[int] = None) -> HashBlobWriter:
         if (peer_address, peer_port) in self.writers and not self.writers[(peer_address, peer_port)].closed():
             raise OSError(f"attempted to download blob twice from {peer_address}:{peer_port}")
-        fut = asyncio.Future(loop=self.loop)
+        fut = asyncio.Future()
         writer = HashBlobWriter(self.blob_hash, self.get_length, fut)
         self.writers[(peer_address, peer_port)] = writer
 
@@ -237,11 +256,13 @@ class BlobBuffer(AbstractBlob):
     """
     An in-memory only blob
     """
-    def __init__(self, loop: asyncio.AbstractEventLoop, blob_hash: str, length: typing.Optional[int] = None,
-                 blob_completed_callback: typing.Optional[typing.Callable[['AbstractBlob'], asyncio.Task]] = None,
-                 blob_directory: typing.Optional[str] = None):
+    def __init__(
+        self, loop: asyncio.AbstractEventLoop, blob_hash: str, length: typing.Optional[int] = None,
+        blob_completed_callback: typing.Optional[typing.Callable[['AbstractBlob'], asyncio.Task]] = None,
+        blob_directory: typing.Optional[str] = None, added_on: typing.Optional[int] = None, is_mine: bool = False
+    ):
         self._verified_bytes: typing.Optional[BytesIO] = None
-        super().__init__(loop, blob_hash, length, blob_completed_callback, blob_directory)
+        super().__init__(loop, blob_hash, length, blob_completed_callback, blob_directory, added_on, is_mine)
 
     @contextlib.contextmanager
     def _reader_context(self) -> typing.ContextManager[typing.BinaryIO]:
@@ -256,9 +277,11 @@ class BlobBuffer(AbstractBlob):
             self.verified.clear()
 
     def _write_blob(self, blob_bytes: bytes):
-        if self._verified_bytes:
-            raise OSError("already have bytes for blob")
-        self._verified_bytes = BytesIO(blob_bytes)
+        async def write():
+            if self._verified_bytes:
+                raise OSError("already have bytes for blob")
+            self._verified_bytes = BytesIO(blob_bytes)
+        return self.loop.create_task(write())
 
     def delete(self):
         if self._verified_bytes:
@@ -276,10 +299,12 @@ class BlobFile(AbstractBlob):
     """
     A blob existing on the local file system
     """
-    def __init__(self, loop: asyncio.AbstractEventLoop, blob_hash: str, length: typing.Optional[int] = None,
-                 blob_completed_callback: typing.Optional[typing.Callable[['AbstractBlob'], asyncio.Task]] = None,
-                 blob_directory: typing.Optional[str] = None):
-        super().__init__(loop, blob_hash, length, blob_completed_callback, blob_directory)
+    def __init__(
+        self, loop: asyncio.AbstractEventLoop, blob_hash: str, length: typing.Optional[int] = None,
+        blob_completed_callback: typing.Optional[typing.Callable[['AbstractBlob'], asyncio.Task]] = None,
+        blob_directory: typing.Optional[str] = None, added_on: typing.Optional[int] = None, is_mine: bool = False
+    ):
+        super().__init__(loop, blob_hash, length, blob_completed_callback, blob_directory, added_on, is_mine)
         if not blob_directory or not os.path.isdir(blob_directory):
             raise OSError(f"invalid blob directory '{blob_directory}'")
         self.file_path = os.path.join(self.blob_directory, self.blob_hash)
@@ -314,22 +339,28 @@ class BlobFile(AbstractBlob):
             handle.close()
 
     def _write_blob(self, blob_bytes: bytes):
-        with open(self.file_path, 'wb') as f:
-            f.write(blob_bytes)
+        def _write_blob():
+            with open(self.file_path, 'wb') as f:
+                f.write(blob_bytes)
+
+        async def write_blob():
+            await self.loop.run_in_executor(None, _write_blob)
+
+        return self.loop.create_task(write_blob())
 
     def delete(self):
+        super().delete()
         if os.path.isfile(self.file_path):
             os.remove(self.file_path)
-        return super().delete()
 
     @classmethod
     async def create_from_unencrypted(
-            cls, loop: asyncio.AbstractEventLoop, blob_dir: typing.Optional[str], key: bytes, iv: bytes,
-            unencrypted: bytes, blob_num: int,
-            blob_completed_callback: typing.Optional[typing.Callable[['AbstractBlob'],
-                                                                     asyncio.Task]] = None) -> BlobInfo:
+        cls, loop: asyncio.AbstractEventLoop, blob_dir: typing.Optional[str], key: bytes, iv: bytes,
+        unencrypted: bytes, blob_num: int, added_on: float, is_mine: bool,
+        blob_completed_callback: typing.Optional[typing.Callable[['AbstractBlob'], asyncio.Task]] = None
+    ) -> BlobInfo:
         if not blob_dir or not os.path.isdir(blob_dir):
             raise OSError(f"cannot create blob in directory: '{blob_dir}'")
         return await super().create_from_unencrypted(
-            loop, blob_dir, key, iv, unencrypted, blob_num, blob_completed_callback
+            loop, blob_dir, key, iv, unencrypted, blob_num, added_on, is_mine, blob_completed_callback
         )

lbry/blob/blob_info.py

@@ -7,13 +7,19 @@ class BlobInfo:
         'blob_num',
         'length',
         'iv',
+        'added_on',
+        'is_mine'
     ]
 
-    def __init__(self, blob_num: int, length: int, iv: str, blob_hash: typing.Optional[str] = None):
+    def __init__(
+            self, blob_num: int, length: int, iv: str, added_on,
+             blob_hash: typing.Optional[str] = None, is_mine=False):
         self.blob_hash = blob_hash
         self.blob_num = blob_num
         self.length = length
         self.iv = iv
+        self.added_on = added_on
+        self.is_mine = is_mine
 
     def as_dict(self) -> typing.Dict:
         d = {

lbry/blob/blob_manager.py

@@ -2,7 +2,7 @@ import os
 import typing
 import asyncio
 import logging
-from lbry.utils import LRUCache
+from lbry.utils import LRUCacheWithMetrics
 from lbry.blob.blob_file import is_valid_blobhash, BlobFile, BlobBuffer, AbstractBlob
 from lbry.stream.descriptor import StreamDescriptor
 from lbry.connection_manager import ConnectionManager
@@ -32,34 +32,34 @@ class BlobManager:
             else self._node_data_store.completed_blobs
         self.blobs: typing.Dict[str, AbstractBlob] = {}
         self.config = config
-        self.decrypted_blob_lru_cache = None if not self.config.blob_lru_cache_size else LRUCache(
+        self.decrypted_blob_lru_cache = None if not self.config.blob_lru_cache_size else LRUCacheWithMetrics(
             self.config.blob_lru_cache_size)
         self.connection_manager = ConnectionManager(loop)
 
-    def _get_blob(self, blob_hash: str, length: typing.Optional[int] = None):
+    def _get_blob(self, blob_hash: str, length: typing.Optional[int] = None, is_mine: bool = False):
         if self.config.save_blobs or (
                 is_valid_blobhash(blob_hash) and os.path.isfile(os.path.join(self.blob_dir, blob_hash))):
             return BlobFile(
-                self.loop, blob_hash, length, self.blob_completed, self.blob_dir
+                self.loop, blob_hash, length, self.blob_completed, self.blob_dir, is_mine=is_mine
             )
         return BlobBuffer(
-            self.loop, blob_hash, length, self.blob_completed, self.blob_dir
+            self.loop, blob_hash, length, self.blob_completed, self.blob_dir, is_mine=is_mine
         )
 
-    def get_blob(self, blob_hash, length: typing.Optional[int] = None):
+    def get_blob(self, blob_hash, length: typing.Optional[int] = None, is_mine: bool = False):
         if blob_hash in self.blobs:
             if self.config.save_blobs and isinstance(self.blobs[blob_hash], BlobBuffer):
                 buffer = self.blobs.pop(blob_hash)
                 if blob_hash in self.completed_blob_hashes:
                     self.completed_blob_hashes.remove(blob_hash)
-                self.blobs[blob_hash] = self._get_blob(blob_hash, length)
+                self.blobs[blob_hash] = self._get_blob(blob_hash, length, is_mine)
                 if buffer.is_readable():
                     with buffer.reader_context() as reader:
                         self.blobs[blob_hash].write_blob(reader.read())
             if length and self.blobs[blob_hash].length is None:
                 self.blobs[blob_hash].set_length(length)
         else:
-            self.blobs[blob_hash] = self._get_blob(blob_hash, length)
+            self.blobs[blob_hash] = self._get_blob(blob_hash, length, is_mine)
         return self.blobs[blob_hash]
 
     def is_blob_verified(self, blob_hash: str, length: typing.Optional[int] = None) -> bool:
@@ -83,6 +83,8 @@ class BlobManager:
         to_add = await self.storage.sync_missing_blobs(in_blobfiles_dir)
         if to_add:
             self.completed_blob_hashes.update(to_add)
+        # check blobs that aren't set as finished but were seen on disk
+        await self.ensure_completed_blobs_status(in_blobfiles_dir - to_add)
         if self.config.track_bandwidth:
             self.connection_manager.start()
         return True
@@ -105,13 +107,26 @@ class BlobManager:
         if isinstance(blob, BlobFile):
             if blob.blob_hash not in self.completed_blob_hashes:
                 self.completed_blob_hashes.add(blob.blob_hash)
-            return self.loop.create_task(self.storage.add_blobs((blob.blob_hash, blob.length), finished=True))
+            return self.loop.create_task(self.storage.add_blobs(
+                (blob.blob_hash, blob.length, blob.added_on, blob.is_mine), finished=True)
+            )
         else:
-            return self.loop.create_task(self.storage.add_blobs((blob.blob_hash, blob.length), finished=False))
+            return self.loop.create_task(self.storage.add_blobs(
+                (blob.blob_hash, blob.length, blob.added_on, blob.is_mine), finished=False)
+            )
 
-    def check_completed_blobs(self, blob_hashes: typing.List[str]) -> typing.List[str]:
-        """Returns of the blobhashes_to_check, which are valid"""
-        return [blob_hash for blob_hash in blob_hashes if self.is_blob_verified(blob_hash)]
+    async def ensure_completed_blobs_status(self, blob_hashes: typing.Iterable[str]):
+        """Ensures that completed blobs from a given list of blob hashes are set as 'finished' in the database."""
+        to_add = []
+        for blob_hash in blob_hashes:
+            if not self.is_blob_verified(blob_hash):
+                continue
+            blob = self.get_blob(blob_hash)
+            to_add.append((blob.blob_hash, blob.length, blob.added_on, blob.is_mine))
+            if len(to_add) > 500:
+                await self.storage.add_blobs(*to_add, finished=True)
+                to_add.clear()
+        return await self.storage.add_blobs(*to_add, finished=True)
 
     def delete_blob(self, blob_hash: str):
         if not is_valid_blobhash(blob_hash):

lbry/blob/disk_space_manager.py

@@ -0,0 +1,77 @@
+import asyncio
+import logging
+
+log = logging.getLogger(__name__)
+
+
+class DiskSpaceManager:
+
+    def __init__(self, config, db, blob_manager, cleaning_interval=30 * 60, analytics=None):
+        self.config = config
+        self.db = db
+        self.blob_manager = blob_manager
+        self.cleaning_interval = cleaning_interval
+        self.running = False
+        self.task = None
+        self.analytics = analytics
+        self._used_space_bytes = None
+
+    async def get_free_space_mb(self, is_network_blob=False):
+        limit_mb = self.config.network_storage_limit if is_network_blob else self.config.blob_storage_limit
+        space_used_mb = await self.get_space_used_mb()
+        space_used_mb = space_used_mb['network_storage'] if is_network_blob else space_used_mb['content_storage']
+        return max(0, limit_mb - space_used_mb)
+
+    async def get_space_used_bytes(self):
+        self._used_space_bytes = await self.db.get_stored_blob_disk_usage()
+        return self._used_space_bytes
+
+    async def get_space_used_mb(self, cached=True):
+        cached = cached and self._used_space_bytes is not None
+        space_used_bytes = self._used_space_bytes if cached else await self.get_space_used_bytes()
+        return {key: int(value/1024.0/1024.0) for key, value in space_used_bytes.items()}
+
+    async def clean(self):
+        await self._clean(False)
+        await self._clean(True)
+
+    async def _clean(self, is_network_blob=False):
+        space_used_mb = await self.get_space_used_mb(cached=False)
+        if is_network_blob:
+            space_used_mb = space_used_mb['network_storage']
+        else:
+            space_used_mb = space_used_mb['content_storage'] + space_used_mb['private_storage']
+        storage_limit_mb = self.config.network_storage_limit if is_network_blob else self.config.blob_storage_limit
+        if self.analytics:
+            asyncio.create_task(
+                self.analytics.send_disk_space_used(space_used_mb, storage_limit_mb, is_network_blob)
+            )
+        delete = []
+        available = storage_limit_mb - space_used_mb
+        if storage_limit_mb == 0 if not is_network_blob else available >= 0:
+            return 0
+        for blob_hash, file_size, _ in await self.db.get_stored_blobs(is_mine=False, is_network_blob=is_network_blob):
+            delete.append(blob_hash)
+            available += int(file_size/1024.0/1024.0)
+            if available >= 0:
+                break
+        if delete:
+            await self.db.stop_all_files()
+            await self.blob_manager.delete_blobs(delete, delete_from_db=True)
+        self._used_space_bytes = None
+        return len(delete)
+
+    async def cleaning_loop(self):
+        while self.running:
+            await asyncio.sleep(self.cleaning_interval)
+            await self.clean()
+
+    async def start(self):
+        self.running = True
+        self.task = asyncio.create_task(self.cleaning_loop())
+        self.task.add_done_callback(lambda _: log.info("Stopping blob cleanup service."))
+
+    async def stop(self):
+        if self.running:
+            self.running = False
+            self.task.cancel()

lbry/blob/writer.py

@@ -3,7 +3,7 @@ import logging
 import asyncio
 from io import BytesIO
 from lbry.error import InvalidBlobHashError, InvalidDataError
-from lbry.cryptoutils import get_lbry_hash_obj
+from lbry.utils import get_lbry_hash_obj
 
 log = logging.getLogger(__name__)
 

lbry/blob_exchange/client.py

@@ -32,7 +32,7 @@ class BlobExchangeClientProtocol(asyncio.Protocol):
         self.buf = b''
 
         # this is here to handle the race when the downloader is closed right as response_fut gets a result
-        self.closed = asyncio.Event(loop=self.loop)
+        self.closed = asyncio.Event()
 
     def data_received(self, data: bytes):
         if self.connection_manager:
@@ -95,7 +95,7 @@ class BlobExchangeClientProtocol(asyncio.Protocol):
             if self._response_fut and not self._response_fut.done():
                 self._response_fut.set_exception(err)
 
-    async def _download_blob(self) -> typing.Tuple[int, Optional['BlobExchangeClientProtocol']]:
+    async def _download_blob(self) -> typing.Tuple[int, Optional['BlobExchangeClientProtocol']]:  # pylint: disable=too-many-return-statements
         """
         :return: download success (bool), connected protocol (BlobExchangeClientProtocol)
         """
@@ -111,7 +111,7 @@ class BlobExchangeClientProtocol(asyncio.Protocol):
             self.transport.write(msg)
             if self.connection_manager:
                 self.connection_manager.sent_data(f"{self.peer_address}:{self.peer_port}", len(msg))
-            response: BlobResponse = await asyncio.wait_for(self._response_fut, self.peer_timeout, loop=self.loop)
+            response: BlobResponse = await asyncio.wait_for(self._response_fut, self.peer_timeout)
             availability_response = response.get_availability_response()
             price_response = response.get_price_response()
             blob_response = response.get_blob_response()
@@ -151,7 +151,9 @@ class BlobExchangeClientProtocol(asyncio.Protocol):
                 f" timeout in {self.peer_timeout}"
             log.debug(msg)
             msg = f"downloaded {self.blob.blob_hash[:8]} from {self.peer_address}:{self.peer_port}"
-            await asyncio.wait_for(self.writer.finished, self.peer_timeout, loop=self.loop)
+            await asyncio.wait_for(self.writer.finished, self.peer_timeout)
+            # wait for the io to finish
+            await self.blob.verified.wait()
             log.info("%s at %fMB/s", msg,
                      round((float(self._blob_bytes_received) /
                             float(time.perf_counter() - start_time)) / 1000000.0, 2))
@@ -185,7 +187,7 @@ class BlobExchangeClientProtocol(asyncio.Protocol):
         try:
             self._blob_bytes_received = 0
             self.blob, self.writer = blob, blob.get_blob_writer(self.peer_address, self.peer_port)
-            self._response_fut = asyncio.Future(loop=self.loop)
+            self._response_fut = asyncio.Future()
             return await self._download_blob()
         except OSError:
             # i'm not sure how to fix this race condition - jack
@@ -213,11 +215,11 @@ class BlobExchangeClientProtocol(asyncio.Protocol):
             self.connection_manager.connection_made(f"{self.peer_address}:{self.peer_port}")
         log.debug("connection made to %s:%i", self.peer_address, self.peer_port)
 
-    def connection_lost(self, reason):
+    def connection_lost(self, exc):
         if self.connection_manager:
             self.connection_manager.outgoing_connection_lost(f"{self.peer_address}:{self.peer_port}")
-        log.debug("connection lost to %s:%i (reason: %s, %s)", self.peer_address, self.peer_port, str(reason),
-                  str(type(reason)))
+        log.debug("connection lost to %s:%i (reason: %s, %s)", self.peer_address, self.peer_port, str(exc),
+                  str(type(exc)))
         self.close()
 
 
@@ -242,7 +244,7 @@ async def request_blob(loop: asyncio.AbstractEventLoop, blob: Optional['Abstract
     try:
         if not connected_protocol:
             await asyncio.wait_for(loop.create_connection(lambda: protocol, address, tcp_port),
-                                   peer_connect_timeout, loop=loop)
+                                   peer_connect_timeout)
             connected_protocol = protocol
         if blob is None or blob.get_is_verified() or not blob.is_writeable():
             # blob is None happens when we are just opening a connection

lbry/blob_exchange/downloader.py

@@ -3,6 +3,7 @@ import typing
 import logging
 from lbry.utils import cache_concurrent
 from lbry.blob_exchange.client import request_blob
+from lbry.dht.node import get_kademlia_peers_from_hosts
 if typing.TYPE_CHECKING:
     from lbry.conf import Config
     from lbry.dht.node import Node
@@ -27,9 +28,9 @@ class BlobDownloader:
         self.ignored: typing.Dict['KademliaPeer', int] = {}
         self.scores: typing.Dict['KademliaPeer', int] = {}
         self.failures: typing.Dict['KademliaPeer', int] = {}
-        self.connection_failures: typing.List['KademliaPeer'] = []
+        self.connection_failures: typing.Set['KademliaPeer'] = set()
         self.connections: typing.Dict['KademliaPeer', 'BlobExchangeClientProtocol'] = {}
-        self.is_running = asyncio.Event(loop=self.loop)
+        self.is_running = asyncio.Event()
 
     def should_race_continue(self, blob: 'AbstractBlob'):
         max_probes = self.config.max_connections_per_download * (1 if self.connections else 10)
@@ -48,7 +49,7 @@ class BlobDownloader:
             connection_id=connection_id, connection_manager=self.blob_manager.connection_manager
         )
         if not bytes_received and not protocol and peer not in self.connection_failures:
-            self.connection_failures.append(peer)
+            self.connection_failures.add(peer)
         if not protocol and peer not in self.ignored:
             self.ignored[peer] = self.loop.time()
             log.debug("drop peer %s:%i", peer.address, peer.tcp_port)
@@ -63,8 +64,8 @@ class BlobDownloader:
             self.scores[peer] = bytes_received / elapsed if bytes_received and elapsed else 1
 
     async def new_peer_or_finished(self):
-        active_tasks = list(self.active_connections.values()) + [asyncio.sleep(1)]
-        await asyncio.wait(active_tasks, loop=self.loop, return_when='FIRST_COMPLETED')
+        active_tasks = list(self.active_connections.values()) + [asyncio.create_task(asyncio.sleep(1))]
+        await asyncio.wait(active_tasks, return_when='FIRST_COMPLETED')
 
     def cleanup_active(self):
         if not self.active_connections and not self.connections:
@@ -89,30 +90,30 @@ class BlobDownloader:
         self.is_running.set()
         try:
             while not blob.get_is_verified() and self.is_running.is_set():
-                batch: typing.Set['KademliaPeer'] = set()
+                batch: typing.Set['KademliaPeer'] = set(self.connections.keys())
                 while not self.peer_queue.empty():
                     batch.update(self.peer_queue.get_nowait())
-                if batch:
-                    self.peer_queue.put_nowait(list(batch))
                 log.debug(
-                    "running, %d peers, %d ignored, %d active, %s connections",
+                    "%s running, %d peers, %d ignored, %d active, %s connections", blob_hash[:6],
                     len(batch), len(self.ignored), len(self.active_connections), len(self.connections)
                 )
                 for peer in sorted(batch, key=lambda peer: self.scores.get(peer, 0), reverse=True):
-                    if peer in self.ignored or peer in self.active_connections:
+                    if peer in self.ignored:
+                        continue
+                    if peer in self.active_connections or not self.should_race_continue(blob):
                         continue
-                    if not self.should_race_continue(blob):
-                        break
                     log.debug("request %s from %s:%i", blob_hash[:8], peer.address, peer.tcp_port)
-                    just_probe = len(self.connections) == 0
-                    t = self.loop.create_task(self.request_blob_from_peer(blob, peer, connection_id, just_probe))
+                    t = self.loop.create_task(self.request_blob_from_peer(blob, peer, connection_id))
                     self.active_connections[peer] = t
+                self.peer_queue.put_nowait(list(batch))
                 await self.new_peer_or_finished()
                 self.cleanup_active()
             log.debug("downloaded %s", blob_hash[:8])
             return blob
         finally:
             blob.close()
+            if self.loop.is_running():
+                self.loop.call_soon(self.cleanup_active)
 
     def close(self):
         self.connection_failures.clear()
@@ -123,11 +124,14 @@ class BlobDownloader:
             protocol.close()
 
 
-async def download_blob(loop, config: 'Config', blob_manager: 'BlobManager', node: 'Node',
+async def download_blob(loop, config: 'Config', blob_manager: 'BlobManager', dht_node: 'Node',
                         blob_hash: str) -> 'AbstractBlob':
-    search_queue = asyncio.Queue(loop=loop, maxsize=config.max_connections_per_download)
+    search_queue = asyncio.Queue(maxsize=config.max_connections_per_download)
     search_queue.put_nowait(blob_hash)
-    peer_queue, accumulate_task = node.accumulate_peers(search_queue)
+    peer_queue, accumulate_task = dht_node.accumulate_peers(search_queue)
+    fixed_peers = None if not config.fixed_peers else await get_kademlia_peers_from_hosts(config.fixed_peers)
+    if fixed_peers:
+        loop.call_later(config.fixed_peer_delay, peer_queue.put_nowait, fixed_peers)
     downloader = BlobDownloader(loop, config, blob_manager, peer_queue)
     try:
         return await downloader.download_blob(blob_hash)

lbry/blob_exchange/serialization.py

@@ -46,7 +46,7 @@ class BlobAvailabilityRequest(BlobMessage):
 
     def __init__(self, requested_blobs: typing.List[str], lbrycrd_address: typing.Optional[bool] = True,
                  **kwargs) -> None:
-        assert len(requested_blobs)
+        assert len(requested_blobs) > 0
         self.requested_blobs = requested_blobs
         self.lbrycrd_address = lbrycrd_address
 
@@ -134,9 +134,9 @@ class BlobErrorResponse(BlobMessage):
         }
 
 
-blob_request_types = typing.Union[BlobPriceRequest, BlobAvailabilityRequest, BlobDownloadRequest,
+blob_request_types = typing.Union[BlobPriceRequest, BlobAvailabilityRequest, BlobDownloadRequest,  # pylint: disable=invalid-name
                                   BlobPaymentAddressRequest]
-blob_response_types = typing.Union[BlobPriceResponse, BlobAvailabilityResponse, BlobDownloadResponse,
+blob_response_types = typing.Union[BlobPriceResponse, BlobAvailabilityResponse, BlobDownloadResponse,  # pylint: disable=invalid-name
                                    BlobErrorResponse, BlobPaymentAddressResponse]
 
 
@@ -157,10 +157,10 @@ def _parse_blob_response(response_msg: bytes) -> typing.Tuple[typing.Optional[ty
         except ValueError:
             continue
         possible_response_keys = {
-                    BlobPaymentAddressResponse.key,
-                    BlobAvailabilityResponse.key,
-                    BlobPriceResponse.key,
-                    BlobDownloadResponse.key
+            BlobPaymentAddressResponse.key,
+            BlobAvailabilityResponse.key,
+            BlobPriceResponse.key,
+            BlobDownloadResponse.key
         }
         if isinstance(response, dict) and response.keys():
             if set(response.keys()).issubset(possible_response_keys):
@@ -179,7 +179,7 @@ class BlobRequest:
         return d
 
     def _get_request(self, request_type: blob_request_types):
-        request = tuple(filter(lambda r: type(r) == request_type, self.requests))
+        request = tuple(filter(lambda r: type(r) == request_type, self.requests))  # pylint: disable=unidiomatic-typecheck
         if request:
             return request[0]
 
@@ -235,7 +235,7 @@ class BlobResponse:
         return d
 
     def _get_response(self, response_type: blob_response_types):
-        response = tuple(filter(lambda r: type(r) == response_type, self.responses))
+        response = tuple(filter(lambda r: type(r) == response_type, self.responses))  # pylint: disable=unidiomatic-typecheck
         if response:
             return response[0]
 
@@ -280,4 +280,3 @@ class BlobResponse:
                 if response_type.key in response
             ])
         return cls(requests, extra)
-

lbry/blob_exchange/server.py

@@ -1,6 +1,7 @@
 import asyncio
 import binascii
 import logging
+import socket
 import typing
 from json.decoder import JSONDecodeError
 from lbry.blob_exchange.serialization import BlobResponse, BlobRequest, blob_response_types
@@ -24,19 +25,19 @@ class BlobServerProtocol(asyncio.Protocol):
         self.idle_timeout = idle_timeout
         self.transfer_timeout = transfer_timeout
         self.server_task: typing.Optional[asyncio.Task] = None
-        self.started_listening = asyncio.Event(loop=self.loop)
+        self.started_listening = asyncio.Event()
         self.buf = b''
         self.transport: typing.Optional[asyncio.Transport] = None
         self.lbrycrd_address = lbrycrd_address
         self.peer_address_and_port: typing.Optional[str] = None
-        self.started_transfer = asyncio.Event(loop=self.loop)
-        self.transfer_finished = asyncio.Event(loop=self.loop)
+        self.started_transfer = asyncio.Event()
+        self.transfer_finished = asyncio.Event()
         self.close_on_idle_task: typing.Optional[asyncio.Task] = None
 
     async def close_on_idle(self):
         while self.transport:
             try:
-                await asyncio.wait_for(self.started_transfer.wait(), self.idle_timeout, loop=self.loop)
+                await asyncio.wait_for(self.started_transfer.wait(), self.idle_timeout)
             except asyncio.TimeoutError:
                 log.debug("closing idle connection from %s", self.peer_address_and_port)
                 return self.close()
@@ -96,22 +97,30 @@ class BlobServerProtocol(asyncio.Protocol):
                 incoming_blob = {'blob_hash': blob.blob_hash, 'length': blob.length}
                 responses.append(BlobDownloadResponse(incoming_blob=incoming_blob))
                 self.send_response(responses)
-                bh = blob.blob_hash[:8]
-                log.debug("send %s to %s:%i", bh, peer_address, peer_port)
+                blob_hash = blob.blob_hash[:8]
+                log.debug("send %s to %s:%i", blob_hash, peer_address, peer_port)
                 self.started_transfer.set()
                 try:
-                    sent = await asyncio.wait_for(blob.sendfile(self), self.transfer_timeout, loop=self.loop)
-                    self.blob_manager.connection_manager.sent_data(self.peer_address_and_port, sent)
-                    log.info("sent %s (%i bytes) to %s:%i", bh, sent, peer_address, peer_port)
-                except (ConnectionResetError, BrokenPipeError, RuntimeError, OSError, asyncio.TimeoutError) as err:
-                    if isinstance(err, asyncio.TimeoutError):
-                        log.debug("timed out sending blob %s to %s", bh, peer_address)
+                    sent = await asyncio.wait_for(blob.sendfile(self), self.transfer_timeout)
+                    if sent and sent > 0:
+                        self.blob_manager.connection_manager.sent_data(self.peer_address_and_port, sent)
+                        log.info("sent %s (%i bytes) to %s:%i", blob_hash, sent, peer_address, peer_port)
                     else:
-                        log.debug("stopped sending %s to %s:%i", bh, peer_address, peer_port)
+                        self.close()
+                        log.debug("stopped sending %s to %s:%i", blob_hash, peer_address, peer_port)
+                        return
+                except (OSError, ValueError, asyncio.TimeoutError) as err:
+                    if isinstance(err, asyncio.TimeoutError):
+                        log.debug("timed out sending blob %s to %s", blob_hash, peer_address)
+                    else:
+                        log.warning("could not read blob %s to send %s:%i", blob_hash, peer_address, peer_port)
                     self.close()
+                    return
                 finally:
                     self.transfer_finished.set()
-        if responses:
+            else:
+                log.info("don't have %s to send %s:%i", blob.blob_hash[:8], peer_address, peer_port)
+        if responses and not self.transport.is_closing():
             self.send_response(responses)
 
     def data_received(self, data):
@@ -122,14 +131,14 @@ class BlobServerProtocol(asyncio.Protocol):
             return
         if data:
             self.blob_manager.connection_manager.received_data(self.peer_address_and_port, len(data))
-            message, separator, remainder = data.rpartition(b'}')
+            _, separator, remainder = data.rpartition(b'}')
             if not separator:
                 self.buf += data
                 return
             try:
                 request = BlobRequest.deserialize(self.buf + data)
                 self.buf = remainder
-            except JSONDecodeError:
+            except (UnicodeDecodeError, JSONDecodeError):
                 log.error("request from %s is not valid json (%i bytes): %s", self.peer_address_and_port,
                           len(self.buf + data), '' if not data else binascii.hexlify(self.buf + data).decode())
                 self.close()
@@ -148,7 +157,7 @@ class BlobServer:
         self.loop = loop
         self.blob_manager = blob_manager
         self.server_task: typing.Optional[asyncio.Task] = None
-        self.started_listening = asyncio.Event(loop=self.loop)
+        self.started_listening = asyncio.Event()
         self.lbrycrd_address = lbrycrd_address
         self.idle_timeout = idle_timeout
         self.transfer_timeout = transfer_timeout
@@ -159,6 +168,13 @@ class BlobServer:
             raise Exception("already running")
 
         async def _start_server():
+            # checking if the port is in use
+            # thx https://stackoverflow.com/a/52872579
+            with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as s:
+                if s.connect_ex(('localhost', port)) == 0:
+                    # the port is already in use!
+                    log.error("Failed to bind TCP %s:%d", interface, port)
+
             server = await self.loop.create_server(
                 lambda: self.server_protocol_class(self.loop, self.blob_manager, self.lbrycrd_address,
                                                    self.idle_timeout, self.transfer_timeout),

lbry/build_info.py

@@ -1,3 +1,4 @@
 # don't touch this. CI server changes this during build/deployment
 BUILD = "dev"
-BUILD_COMMIT = "source installation"
+COMMIT_HASH = "none"
+DOCKER_TAG = "none"

lbry/conf.py

@@ -1,21 +1,21 @@
 import os
 import re
 import sys
-import typing
 import logging
-import yaml
+from typing import List, Dict, Tuple, Union, TypeVar, Generic, Optional
 from argparse import ArgumentParser
 from contextlib import contextmanager
 from appdirs import user_data_dir, user_config_dir
+import yaml
 from lbry.error import InvalidCurrencyError
 from lbry.dht import constants
-from torba.client.coinselection import STRATEGIES
+from lbry.wallet.coinselection import STRATEGIES
 
 log = logging.getLogger(__name__)
 
 
-NOT_SET = type('NOT_SET', (object,), {})
-T = typing.TypeVar('T')
+NOT_SET = type('NOT_SET', (object,), {})  # pylint: disable=invalid-name
+T = TypeVar('T')
 
 CURRENCIES = {
     'BTC': {'type': 'crypto'},
@@ -24,18 +24,18 @@ CURRENCIES = {
 }
 
 
-class Setting(typing.Generic[T]):
+class Setting(Generic[T]):
 
-    def __init__(self, doc: str, default: typing.Optional[T] = None,
-                 previous_names: typing.Optional[typing.List[str]] = None,
-                 metavar: typing.Optional[str] = None):
+    def __init__(self, doc: str, default: Optional[T] = None,
+                 previous_names: Optional[List[str]] = None,
+                 metavar: Optional[str] = None):
         self.doc = doc
         self.default = default
         self.previous_names = previous_names or []
         self.metavar = metavar
 
     def __set_name__(self, owner, name):
-        self.name = name
+        self.name = name  # pylint: disable=attribute-defined-outside-init
 
     @property
     def cli_name(self):
@@ -45,7 +45,7 @@ class Setting(typing.Generic[T]):
     def no_cli_name(self):
         return f"--no-{self.name.replace('_', '-')}"
 
-    def __get__(self, obj: typing.Optional['BaseConfig'], owner) -> T:
+    def __get__(self, obj: Optional['BaseConfig'], owner) -> T:
         if obj is None:
             return self
         for location in obj.search_order:
@@ -53,7 +53,7 @@ class Setting(typing.Generic[T]):
                 return location[self.name]
         return self.default
 
-    def __set__(self, obj: 'BaseConfig', val: typing.Union[T, NOT_SET]):
+    def __set__(self, obj: 'BaseConfig', val: Union[T, NOT_SET]):
         if val == NOT_SET:
             for location in obj.modify_order:
                 if self.name in location:
@@ -63,13 +63,25 @@ class Setting(typing.Generic[T]):
             for location in obj.modify_order:
                 location[self.name] = val
 
-    def validate(self, val):
+    def is_set(self, obj: 'BaseConfig') -> bool:
+        for location in obj.search_order:
+            if self.name in location:
+                return True
+        return False
+
+    def is_set_to_default(self, obj: 'BaseConfig') -> bool:
+        for location in obj.search_order:
+            if self.name in location:
+                return location[self.name] == self.default
+        return False
+
+    def validate(self, value):
         raise NotImplementedError()
 
-    def deserialize(self, value):
+    def deserialize(self, value):  # pylint: disable=no-self-use
         return value
 
-    def serialize(self, value):
+    def serialize(self, value):  # pylint: disable=no-self-use
         return value
 
     def contribute_to_argparse(self, parser: ArgumentParser):
@@ -82,14 +94,18 @@ class Setting(typing.Generic[T]):
 
 
 class String(Setting[str]):
-    def validate(self, val):
-        assert isinstance(val, str), \
+    def validate(self, value):
+        assert isinstance(value, str), \
             f"Setting '{self.name}' must be a string."
 
+    # TODO: removes this after pylint starts to understand generics
+    def __get__(self, obj: Optional['BaseConfig'], owner) -> str:  # pylint: disable=useless-super-delegation
+        return super().__get__(obj, owner)
+
 
 class Integer(Setting[int]):
-    def validate(self, val):
-        assert isinstance(val, int), \
+    def validate(self, value):
+        assert isinstance(value, int), \
             f"Setting '{self.name}' must be an integer."
 
     def deserialize(self, value):
@@ -97,8 +113,8 @@ class Integer(Setting[int]):
 
 
 class Float(Setting[float]):
-    def validate(self, val):
-        assert isinstance(val, float), \
+    def validate(self, value):
+        assert isinstance(value, float), \
             f"Setting '{self.name}' must be a decimal."
 
     def deserialize(self, value):
@@ -106,8 +122,8 @@ class Float(Setting[float]):
 
 
 class Toggle(Setting[bool]):
-    def validate(self, val):
-        assert isinstance(val, bool), \
+    def validate(self, value):
+        assert isinstance(value, bool), \
             f"Setting '{self.name}' must be a true/false value."
 
     def contribute_to_argparse(self, parser: ArgumentParser):
@@ -127,10 +143,10 @@ class Toggle(Setting[bool]):
 
 
 class Path(String):
-    def __init__(self, doc: str, default: str = '', *args, **kwargs):
+    def __init__(self, doc: str, *args, default: str = '', **kwargs):
         super().__init__(doc, default, *args, **kwargs)
 
-    def __get__(self, obj, owner):
+    def __get__(self, obj, owner) -> str:
         value = super().__get__(obj, owner)
         if isinstance(value, str):
             return os.path.expanduser(os.path.expandvars(value))
@@ -187,7 +203,7 @@ class MaxKeyFee(Setting[dict]):
         )
         parser.add_argument(
             self.no_cli_name,
-            help=f"Disable maximum key fee check.",
+            help="Disable maximum key fee check.",
             dest=self.name,
             const=None,
             action="store_const",
@@ -196,7 +212,7 @@ class MaxKeyFee(Setting[dict]):
 
 
 class StringChoice(String):
-    def __init__(self, doc: str, valid_values: typing.List[str], default: str, *args, **kwargs):
+    def __init__(self, doc: str, valid_values: List[str], default: str, *args, **kwargs):
         super().__init__(doc, default, *args, **kwargs)
         if not valid_values:
             raise ValueError("No valid values provided")
@@ -204,16 +220,16 @@ class StringChoice(String):
             raise ValueError(f"Default value must be one of: {', '.join(valid_values)}")
         self.valid_values = valid_values
 
-    def validate(self, val):
-        super().validate(val)
-        if val not in self.valid_values:
+    def validate(self, value):
+        super().validate(value)
+        if value not in self.valid_values:
             raise ValueError(f"Setting '{self.name}' value must be one of: {', '.join(self.valid_values)}")
 
 
 class ListSetting(Setting[list]):
 
-    def validate(self, val):
-        assert isinstance(val, (tuple, list)), \
+    def validate(self, value):
+        assert isinstance(value, (tuple, list)), \
             f"Setting '{self.name}' must be a tuple or list."
 
     def contribute_to_argparse(self, parser: ArgumentParser):
@@ -226,10 +242,10 @@ class ListSetting(Setting[list]):
 
 class Servers(ListSetting):
 
-    def validate(self, val):
-        assert isinstance(val, (tuple, list)), \
+    def validate(self, value):
+        assert isinstance(value, (tuple, list)), \
             f"Setting '{self.name}' must be a tuple or list of servers."
-        for idx, server in enumerate(val):
+        for idx, server in enumerate(value):
             assert isinstance(server, (tuple, list)) and len(server) == 2, \
                 f"Server defined '{server}' at index {idx} in setting " \
                 f"'{self.name}' must be a tuple or list of two items."
@@ -260,26 +276,104 @@ class Servers(ListSetting):
 
 class Strings(ListSetting):
 
-    def validate(self, val):
-        assert isinstance(val, (tuple, list)), \
+    def validate(self, value):
+        assert isinstance(value, (tuple, list)), \
             f"Setting '{self.name}' must be a tuple or list of strings."
-        for idx, string in enumerate(val):
+        for idx, string in enumerate(value):
             assert isinstance(string, str), \
                 f"Value of '{string}' at index {idx} in setting " \
                 f"'{self.name}' must be a string."
 
 
+class KnownHubsList:
+
+    def __init__(self, config: 'Config' = None, file_name: str = 'known_hubs.yml'):
+        self.file_name = file_name
+        self.path = os.path.join(config.wallet_dir, self.file_name) if config else None
+        self.hubs: Dict[Tuple[str, int], Dict] = {}
+        if self.exists:
+            self.load()
+
+    @property
+    def exists(self):
+        return self.path and os.path.exists(self.path)
+
+    @property
+    def serialized(self) -> Dict[str, Dict]:
+        return {f"{host}:{port}": details for (host, port), details in self.hubs.items()}
+
+    def filter(self, match_none=False, **kwargs):
+        if not kwargs:
+            return self.hubs
+        result = {}
+        for hub, details in self.hubs.items():
+            for key, constraint in kwargs.items():
+                value = details.get(key)
+                if value == constraint or (match_none and value is None):
+                    result[hub] = details
+                    break
+        return result
+
+    def load(self):
+        if self.path:
+            with open(self.path, 'r') as known_hubs_file:
+                raw = known_hubs_file.read()
+                for hub, details in yaml.safe_load(raw).items():
+                    self.set(hub, details)
+
+    def save(self):
+        if self.path:
+            with open(self.path, 'w') as known_hubs_file:
+                known_hubs_file.write(yaml.safe_dump(self.serialized, default_flow_style=False))
+
+    def set(self, hub: str, details: Dict):
+        if hub and hub.count(':') == 1:
+            host, port = hub.split(':')
+            hub_parts = (host, int(port))
+            if hub_parts not in self.hubs:
+                self.hubs[hub_parts] = details
+                return hub
+
+    def add_hubs(self, hubs: List[str]):
+        added = False
+        for hub in hubs:
+            if self.set(hub, {}) is not None:
+                added = True
+        return added
+
+    def items(self):
+        return self.hubs.items()
+
+    def __bool__(self):
+        return len(self) > 0
+
+    def __len__(self):
+        return self.hubs.__len__()
+
+    def __iter__(self):
+        return iter(self.hubs)
+
+
 class EnvironmentAccess:
     PREFIX = 'LBRY_'
 
-    def __init__(self, environ: dict):
-        self.environ = environ
+    def __init__(self, config: 'BaseConfig', environ: dict):
+        self.configuration = config
+        self.data = {}
+        if environ:
+            self.load(environ)
+
+    def load(self, environ):
+        for setting in self.configuration.get_settings():
+            value = environ.get(f'{self.PREFIX}{setting.name.upper()}', NOT_SET)
+            if value != NOT_SET and not (isinstance(setting, ListSetting) and value is None):
+                self.data[setting.name] = setting.deserialize(value)
 
     def __contains__(self, item: str):
-        return f'{self.PREFIX}{item.upper()}' in self.environ
+        return item in self.data
 
     def __getitem__(self, item: str):
-        return self.environ[f'{self.PREFIX}{item.upper()}']
+        return self.data[item]
 
 
 class ArgumentAccess:
@@ -320,7 +414,7 @@ class ConfigFileAccess:
         cls = type(self.configuration)
         with open(self.path, 'r') as config_file:
             raw = config_file.read()
-        serialized = yaml.load(raw) or {}
+        serialized = yaml.safe_load(raw) or {}
         for key, value in serialized.items():
             attr = getattr(cls, key, None)
             if attr is None:
@@ -364,7 +458,7 @@ class ConfigFileAccess:
         del self.data[key]
 
 
-TBC = typing.TypeVar('TBC', bound='BaseConfig')
+TBC = TypeVar('TBC', bound='BaseConfig')
 
 
 class BaseConfig:
@@ -438,7 +532,7 @@ class BaseConfig:
         self.arguments = ArgumentAccess(self, args)
 
     def set_environment(self, environ=None):
-        self.environment = EnvironmentAccess(environ or os.environ)
+        self.environment = EnvironmentAccess(self, environ or os.environ)
 
     def set_persisted(self, config_file_path=None):
         if config_file_path is None:
@@ -457,7 +551,27 @@ class BaseConfig:
             self.persisted.save()
 
 
-class CLIConfig(BaseConfig):
+class TranscodeConfig(BaseConfig):
+
+    ffmpeg_path = String('A list of places to check for ffmpeg and ffprobe. '
+                         f'$data_dir/ffmpeg/bin and $PATH are checked afterward. Separator: {os.pathsep}',
+                         '', previous_names=['ffmpeg_folder'])
+    video_encoder = String('FFmpeg codec and parameters for the video encoding. '
+                           'Example: libaom-av1 -crf 25 -b:v 0 -strict experimental',
+                           'libx264 -crf 24 -preset faster -pix_fmt yuv420p')
+    video_bitrate_maximum = Integer('Maximum bits per second allowed for video streams (0 to disable).', 5_000_000)
+    video_scaler = String('FFmpeg scaling parameters for reducing bitrate. '
+                          'Example: -vf "scale=-2:720,fps=24" -maxrate 5M -bufsize 3M',
+                          r'-vf "scale=if(gte(iw\,ih)\,min(1920\,iw)\,-2):if(lt(iw\,ih)\,min(1920\,ih)\,-2)" '
+                          r'-maxrate 5500K -bufsize 5000K')
+    audio_encoder = String('FFmpeg codec and parameters for the audio encoding. '
+                           'Example: libopus -b:a 128k',
+                           'aac -b:a 160k')
+    volume_filter = String('FFmpeg filter for audio normalization. Exmple: -af loudnorm', '')
+    volume_analysis_time = Integer('Maximum seconds into the file that we examine audio volume (0 to disable).', 240)
+
+
+class CLIConfig(TranscodeConfig):
 
     api = String('Host name and port for lbrynet daemon API.', 'localhost:5279', metavar='HOST:PORT')
 
@@ -475,6 +589,9 @@ class CLIConfig(BaseConfig):
 
 
 class Config(CLIConfig):
+
+    jurisdiction = String("Limit interactions to wallet server in this jurisdiction.")
+
     # directories
     data_dir = Path("Directory path to store blobs.", metavar='DIR')
     download_dir = Path(
@@ -496,7 +613,8 @@ class Config(CLIConfig):
         "ports or have firewall rules you likely want to disable this.", True
     )
     udp_port = Integer("UDP port for communicating on the LBRY DHT", 4444, previous_names=['dht_node_port'])
-    tcp_port = Integer("TCP port to listen for incoming blob requests", 3333, previous_names=['peer_port'])
+    tcp_port = Integer("TCP port to listen for incoming blob requests", 4444, previous_names=['peer_port'])
+    prometheus_port = Integer("Port to expose prometheus metrics (off by default)", 0)
     network_interface = String("Interface to use for the DHT and blob exchange", '0.0.0.0')
 
     # routing table
@@ -504,17 +622,24 @@ class Config(CLIConfig):
         "Routing table bucket index below which we always split the bucket if given a new key to add to it and "
         "the bucket is full. As this value is raised the depth of the routing table (and number of peers in it) "
         "will increase. This setting is used by seed nodes, you probably don't want to change it during normal "
-        "use.", 1
+        "use.", 2
+    )
+    is_bootstrap_node = Toggle(
+        "When running as a bootstrap node, disable all logic related to balancing the routing table, so we can "
+        "add as many peers as possible and better help first-runs.", False
     )
 
     # protocol timeouts
     download_timeout = Float("Cumulative timeout for a stream to begin downloading before giving up", 30.0)
     blob_download_timeout = Float("Timeout to download a blob from a peer", 30.0)
+    hub_timeout = Float("Timeout when making a hub request", 30.0)
     peer_connect_timeout = Float("Timeout to establish a TCP connection to a peer", 3.0)
-    node_rpc_timeout = Float("Timeout when making a DHT request", constants.rpc_timeout)
+    node_rpc_timeout = Float("Timeout when making a DHT request", constants.RPC_TIMEOUT)
 
     # blob announcement and download
     save_blobs = Toggle("Save encrypted blob files for hosting, otherwise download blobs to memory only.", True)
+    network_storage_limit = Integer("Disk space in MB to be allocated for helping the P2P network. 0 = disable", 0)
+    blob_storage_limit = Integer("Disk space in MB to be allocated for blob storage. 0 = no limit", 0)
     blob_lru_cache_size = Integer(
         "LRU cache size for decrypted downloaded blobs used to minimize re-downloading the same blobs when "
         "replying to a range request. Set to 0 to disable.", 32
@@ -531,6 +656,7 @@ class Config(CLIConfig):
         "Maximum number of peers to connect to while downloading a blob", 4,
         previous_names=['max_connections_per_stream']
     )
+    concurrent_hub_requests = Integer("Maximum number of concurrent hub requests", 32)
     fixed_peer_delay = Float(
         "Amount of seconds before adding the reflector servers as potential peers to download from in case dht"
         "peers are not found or are slow", 2.0
@@ -539,6 +665,7 @@ class Config(CLIConfig):
         "Don't download streams with fees exceeding this amount. When set to "
         "null, the amount is unbounded.", {'currency': 'USD', 'amount': 50.0}
     )
+    max_wallet_server_fee = String("Maximum daily LBC amount allowed as payment for wallet servers.", "0.0")
 
     # reflector settings
     reflect_streams = Toggle(
@@ -550,42 +677,62 @@ class Config(CLIConfig):
     )
 
     # servers
-    reflector_servers = Servers("Reflector re-hosting servers", [
+    reflector_servers = Servers("Reflector re-hosting servers for mirroring publishes", [
         ('reflector.lbry.com', 5566)
     ])
+
+    fixed_peers = Servers("Fixed peers to fall back to if none are found on P2P for a blob", [
+        ('cdn.reflector.lbry.com', 5567)
+    ])
+
+    tracker_servers = Servers("BitTorrent-compatible (BEP15) UDP trackers for helping P2P discovery", [
+        ('tracker.lbry.com', 9252),
+        ('tracker.lbry.grin.io', 9252),
+        ('tracker.lbry.pigg.es', 9252),
+        ('tracker.lizard.technology', 9252),
+        ('s1.lbry.network', 9252),
+    ])
+
     lbryum_servers = Servers("SPV wallet servers", [
-        ('spv1.lbry.com', 50001),
-        ('spv2.lbry.com', 50001),
-        ('spv3.lbry.com', 50001),
-        ('spv4.lbry.com', 50001),
-        ('spv5.lbry.com', 50001),
-        ('spv6.lbry.com', 50001),
-        ('spv7.lbry.com', 50001),
-        ('spv8.lbry.com', 50001),
-        ('spv9.lbry.com', 50001),
+        ('spv11.lbry.com', 50001),
+        ('spv12.lbry.com', 50001),
+        ('spv13.lbry.com', 50001),
+        ('spv14.lbry.com', 50001),
+        ('spv15.lbry.com', 50001),
+        ('spv16.lbry.com', 50001),
+        ('spv17.lbry.com', 50001),
+        ('spv18.lbry.com', 50001),
+        ('spv19.lbry.com', 50001),
+        ('hub.lbry.grin.io', 50001),
+        ('hub.lizard.technology', 50001),
+        ('s1.lbry.network', 50001),
     ])
     known_dht_nodes = Servers("Known nodes for bootstrapping connection to the DHT", [
+        ('dht.lbry.grin.io', 4444),  # Grin
+        ('dht.lbry.madiator.com', 4444),  # Madiator
+        ('dht.lbry.pigg.es', 4444), # Pigges
         ('lbrynet1.lbry.com', 4444),  # US EAST
         ('lbrynet2.lbry.com', 4444),  # US WEST
         ('lbrynet3.lbry.com', 4444),  # EU
-        ('lbrynet4.lbry.com', 4444)  # ASIA
+        ('lbrynet4.lbry.com', 4444),  # ASIA
+        ('dht.lizard.technology', 4444),  # Jack
+        ('s2.lbry.network', 4444),
     ])
 
-    comment_server = String("Comment server API URL", "https://comments.lbry.com/api")
-
     # blockchain
     blockchain_name = String("Blockchain name - lbrycrd_main, lbrycrd_regtest, or lbrycrd_testnet", 'lbrycrd_main')
-    s3_headers_depth = Integer("download headers from s3 when the local height is more than 10 chunks behind", 96 * 10)
-    cache_time = Integer("Time to cache resolved claims", 150)  # TODO: use this
 
     # daemon
-    save_files = Toggle("Save downloaded files when calling `get` by default", True)
+    save_files = Toggle("Save downloaded files when calling `get` by default", False)
     components_to_skip = Strings("components which will be skipped during start-up of daemon", [])
     share_usage_data = Toggle(
-        "Whether to share usage stats and diagnostic info with LBRY.", True,
+        "Whether to share usage stats and diagnostic info with LBRY.", False,
         previous_names=['upload_log', 'upload_log', 'share_debug_info']
     )
     track_bandwidth = Toggle("Track bandwidth usage", True)
+    allowed_origin = String(
+        "Allowed `Origin` header value for API request (sent by browser), use * to allow "
+        "all hosts; default is to only allow API requests with no `Origin` value.", "")
 
     # media server
     streaming_server = String('Host name and port to serve streaming media over range requests',
@@ -595,7 +742,14 @@ class Config(CLIConfig):
 
     coin_selection_strategy = StringChoice(
         "Strategy to use when selecting UTXOs for a transaction",
-        STRATEGIES, "standard")
+        STRATEGIES, "prefer_confirmed"
+    )
+
+    transaction_cache_size = Integer("Transaction cache size", 2 ** 17)
+    save_resolved_claims = Toggle(
+        "Save content claims to the database when they are resolved to keep file_list up to date, "
+        "only disable this if file_x commands are not needed", True
+    )
 
     @property
     def streaming_host(self):
@@ -608,6 +762,7 @@ class Config(CLIConfig):
     def __init__(self, **kwargs):
         super().__init__(**kwargs)
         self.set_default_paths()
+        self.known_hubs = KnownHubsList(self)
 
     def set_default_paths(self):
         if 'darwin' in sys.platform.lower():
@@ -629,10 +784,14 @@ class Config(CLIConfig):
         return os.path.join(self.data_dir, 'lbrynet.log')
 
 
-def get_windows_directories() -> typing.Tuple[str, str, str]:
-    from lbry.winpaths import get_path, FOLDERID, UserHandle  # pylint: disable=import-outside-toplevel
+def get_windows_directories() -> Tuple[str, str, str]:
+    from lbry.winpaths import get_path, FOLDERID, UserHandle, \
+        PathNotFoundException  # pylint: disable=import-outside-toplevel
 
-    download_dir = get_path(FOLDERID.Downloads, UserHandle.current)
+    try:
+        download_dir = get_path(FOLDERID.Downloads, UserHandle.current)
+    except PathNotFoundException:
+        download_dir = os.getcwd()
 
     # old
     appdata = get_path(FOLDERID.RoamingAppData, UserHandle.current)
@@ -644,23 +803,23 @@ def get_windows_directories() -> typing.Tuple[str, str, str]:
     # new
     data_dir = user_data_dir('lbrynet', 'lbry')
     lbryum_dir = user_data_dir('lbryum', 'lbry')
-    download_dir = get_path(FOLDERID.Downloads, UserHandle.current)
     return data_dir, lbryum_dir, download_dir
 
 
-def get_darwin_directories() -> typing.Tuple[str, str, str]:
+def get_darwin_directories() -> Tuple[str, str, str]:
     data_dir = user_data_dir('LBRY')
     lbryum_dir = os.path.expanduser('~/.lbryum')
     download_dir = os.path.expanduser('~/Downloads')
     return data_dir, lbryum_dir, download_dir
 
 
-def get_linux_directories() -> typing.Tuple[str, str, str]:
+def get_linux_directories() -> Tuple[str, str, str]:
     try:
         with open(os.path.join(user_config_dir(), 'user-dirs.dirs'), 'r') as xdg:
-            down_dir = re.search(r'XDG_DOWNLOAD_DIR=(.+)', xdg.read()).group(1)
-        down_dir = re.sub(r'\$HOME', os.getenv('HOME') or os.path.expanduser("~/"), down_dir)
-        download_dir = re.sub('\"', '', down_dir)
+            down_dir = re.search(r'XDG_DOWNLOAD_DIR=(.+)', xdg.read())
+        if down_dir:
+            down_dir = re.sub(r'\$HOME', os.getenv('HOME') or os.path.expanduser("~/"), down_dir.group(1))
+            download_dir = re.sub('\"', '', down_dir)
     except OSError:
         download_dir = os.getenv('XDG_DOWNLOAD_DIR')
     if not download_dir:

lbry/connection_manager.py

@@ -67,18 +67,18 @@ class ConnectionManager:
 
         while True:
             last = time.perf_counter()
-            await asyncio.sleep(0.1, loop=self.loop)
+            await asyncio.sleep(0.1)
             self._status['incoming_bps'].clear()
             self._status['outgoing_bps'].clear()
             now = time.perf_counter()
             while self.outgoing:
-                k, v = self.outgoing.popitem()
-                self._status['total_sent'] += v
-                self._status['outgoing_bps'][k] = v / (now - last)
+                k, sent = self.outgoing.popitem()
+                self._status['total_sent'] += sent
+                self._status['outgoing_bps'][k] = sent / (now - last)
             while self.incoming:
-                k, v = self.incoming.popitem()
-                self._status['total_received'] += v
-                self._status['incoming_bps'][k] = v / (now - last)
+                k, received = self.incoming.popitem()
+                self._status['total_received'] += received
+                self._status['incoming_bps'][k] = received / (now - last)
             self._status['total_outgoing_mbs'] = int(sum(list(self._status['outgoing_bps'].values())
                                                          )) / 1000000.0
             self._status['total_incoming_mbs'] = int(sum(list(self._status['incoming_bps'].values())

lbry/constants.py

@@ -0,0 +1,2 @@
+CENT = 1000000
+COIN = 100*CENT

lbry/crypto/base58.py

@@ -0,0 +1,86 @@
+from lbry.crypto.hash import double_sha256
+from lbry.crypto.util import bytes_to_int, int_to_bytes
+
+
+class Base58Error(Exception):
+    """ Exception used for Base58 errors. """
+
+
+class Base58:
+    """ Class providing base 58 functionality. """
+
+    chars = '123456789ABCDEFGHJKLMNPQRSTUVWXYZabcdefghijkmnopqrstuvwxyz'
+    assert len(chars) == 58
+    char_map = {c: n for n, c in enumerate(chars)}
+
+    @classmethod
+    def char_value(cls, c):
+        val = cls.char_map.get(c)
+        if val is None:
+            raise Base58Error(f'invalid base 58 character "{c}"')
+        return val
+
+    @classmethod
+    def decode(cls, txt):
+        """ Decodes txt into a big-endian bytearray. """
+        if isinstance(txt, memoryview):
+            txt = str(txt)
+
+        if isinstance(txt, bytes):
+            txt = txt.decode()
+
+        if not isinstance(txt, str):
+            raise TypeError('a string is required')
+
+        if not txt:
+            raise Base58Error('string cannot be empty')
+
+        value = 0
+        for c in txt:
+            value = value * 58 + cls.char_value(c)
+
+        result = int_to_bytes(value)
+
+        # Prepend leading zero bytes if necessary
+        count = 0
+        for c in txt:
+            if c != '1':
+                break
+            count += 1
+        if count:
+            result = bytes((0,)) * count + result
+
+        return result
+
+    @classmethod
+    def encode(cls, be_bytes):
+        """Converts a big-endian bytearray into a base58 string."""
+        value = bytes_to_int(be_bytes)
+
+        txt = ''
+        while value:
+            value, mod = divmod(value, 58)
+            txt += cls.chars[mod]
+
+        for byte in be_bytes:
+            if byte != 0:
+                break
+            txt += '1'
+
+        return txt[::-1]
+
+    @classmethod
+    def decode_check(cls, txt, hash_fn=double_sha256):
+        """ Decodes a Base58Check-encoded string to a payload. The version prefixes it. """
+        be_bytes = cls.decode(txt)
+        result, check = be_bytes[:-4], be_bytes[-4:]
+        if check != hash_fn(result)[:4]:
+            raise Base58Error(f'invalid base 58 checksum for {txt}')
+        return result
+
+    @classmethod
+    def encode_check(cls, payload, hash_fn=double_sha256):
+        """ Encodes a payload bytearray (which includes the version byte(s))
+            into a Base58Check string."""
+        be_bytes = payload + hash_fn(payload)[:4]
+        return cls.encode(be_bytes)

lbry/crypto/crypt.py

@@ -0,0 +1,71 @@
+import os
+import base64
+import typing
+from cryptography.hazmat.primitives.kdf.scrypt import Scrypt
+from cryptography.hazmat.primitives.ciphers import Cipher, modes
+from cryptography.hazmat.primitives.ciphers.algorithms import AES
+from cryptography.hazmat.primitives.padding import PKCS7
+from cryptography.hazmat.backends import default_backend
+
+from lbry.error import InvalidPasswordError
+from lbry.crypto.hash import double_sha256
+
+
+def aes_encrypt(secret: str, value: str, init_vector: bytes = None) -> str:
+    if init_vector is not None:
+        assert len(init_vector) == 16
+    else:
+        init_vector = os.urandom(16)
+    key = double_sha256(secret.encode())
+    encryptor = Cipher(AES(key), modes.CBC(init_vector), default_backend()).encryptor()
+    padder = PKCS7(AES.block_size).padder()
+    padded_data = padder.update(value.encode()) + padder.finalize()
+    encrypted_data = encryptor.update(padded_data) + encryptor.finalize()
+    return base64.b64encode(init_vector + encrypted_data).decode()
+
+
+def aes_decrypt(secret: str, value: str) -> typing.Tuple[str, bytes]:
+    try:
+        data = base64.b64decode(value.encode())
+        key = double_sha256(secret.encode())
+        init_vector, data = data[:16], data[16:]
+        decryptor = Cipher(AES(key), modes.CBC(init_vector), default_backend()).decryptor()
+        unpadder = PKCS7(AES.block_size).unpadder()
+        result = unpadder.update(decryptor.update(data)) + unpadder.finalize()
+        return result.decode(), init_vector
+    except UnicodeDecodeError:
+        raise InvalidPasswordError()
+    except ValueError as e:
+        if e.args[0] == 'Invalid padding bytes.':
+            raise InvalidPasswordError()
+        raise
+
+
+def better_aes_encrypt(secret: str, value: bytes) -> bytes:
+    init_vector = os.urandom(16)
+    key = scrypt(secret.encode(), salt=init_vector)
+    encryptor = Cipher(AES(key), modes.CBC(init_vector), default_backend()).encryptor()
+    padder = PKCS7(AES.block_size).padder()
+    padded_data = padder.update(value) + padder.finalize()
+    encrypted_data = encryptor.update(padded_data) + encryptor.finalize()
+    return base64.b64encode(b's:8192:16:1:' + init_vector + encrypted_data)
+
+
+def better_aes_decrypt(secret: str, value: bytes) -> bytes:
+    try:
+        data = base64.b64decode(value)
+        _, scryp_n, scrypt_r, scrypt_p, data = data.split(b':', maxsplit=4)
+        init_vector, data = data[:16], data[16:]
+        key = scrypt(secret.encode(), init_vector, int(scryp_n), int(scrypt_r), int(scrypt_p))
+        decryptor = Cipher(AES(key), modes.CBC(init_vector), default_backend()).decryptor()
+        unpadder = PKCS7(AES.block_size).unpadder()
+        return unpadder.update(decryptor.update(data)) + unpadder.finalize()
+    except ValueError as e:
+        if e.args[0] == 'Invalid padding bytes.':
+            raise InvalidPasswordError()
+        raise
+
+
+def scrypt(passphrase, salt, scrypt_n=1<<13, scrypt_r=16, scrypt_p=1):
+    kdf = Scrypt(salt, length=32, n=scrypt_n, r=scrypt_r, p=scrypt_p, backend=default_backend())
+    return kdf.derive(passphrase)

lbry/crypto/hash.py

@@ -0,0 +1,47 @@
+import hashlib
+import hmac
+from binascii import hexlify, unhexlify
+
+
+def sha256(x):
+    """ Simple wrapper of hashlib sha256. """
+    return hashlib.sha256(x).digest()
+
+
+def sha512(x):
+    """ Simple wrapper of hashlib sha512. """
+    return hashlib.sha512(x).digest()
+
+
+def ripemd160(x):
+    """ Simple wrapper of hashlib ripemd160. """
+    h = hashlib.new('ripemd160')
+    h.update(x)
+    return h.digest()
+
+
+def double_sha256(x):
+    """ SHA-256 of SHA-256, as used extensively in bitcoin. """
+    return sha256(sha256(x))
+
+
+def hmac_sha512(key, msg):
+    """ Use SHA-512 to provide an HMAC. """
+    return hmac.new(key, msg, hashlib.sha512).digest()
+
+
+def hash160(x):
+    """ RIPEMD-160 of SHA-256.
+        Used to make bitcoin addresses from pubkeys. """
+    return ripemd160(sha256(x))
+
+
+def hash_to_hex_str(x):
+    """ Convert a big-endian binary hash to displayed hex string.
+        Display form of a binary hash is reversed and converted to hex. """
+    return hexlify(reversed(x))
+
+
+def hex_str_to_hash(x):
+    """ Convert a displayed hex string to a binary hash. """
+    return reversed(unhexlify(x))

lbry/crypto/util.py

@@ -0,0 +1,13 @@
+from binascii import unhexlify, hexlify
+
+
+def bytes_to_int(be_bytes):
+    """ Interprets a big-endian sequence of bytes as an integer. """
+    return int(hexlify(be_bytes), 16)
+
+
+def int_to_bytes(value):
+    """ Converts an integer to a big-endian sequence of bytes. """
+    length = (value.bit_length() + 7) // 8
+    s = '%x' % value
+    return unhexlify(('0' * (len(s) % 2) + s).zfill(length * 2))

lbry/dht/blob_announcer.py

@@ -0,0 +1,78 @@
+import asyncio
+import typing
+import logging
+
+from prometheus_client import Counter, Gauge
+
+if typing.TYPE_CHECKING:
+    from lbry.dht.node import Node
+    from lbry.extras.daemon.storage import SQLiteStorage
+
+log = logging.getLogger(__name__)
+
+
+class BlobAnnouncer:
+    announcements_sent_metric = Counter(
+        "announcements_sent", "Number of announcements sent and their respective status.", namespace="dht_node",
+        labelnames=("peers", "error"),
+    )
+    announcement_queue_size_metric = Gauge(
+        "announcement_queue_size", "Number of hashes waiting to be announced.", namespace="dht_node",
+        labelnames=("scope",)
+    )
+
+    def __init__(self, loop: asyncio.AbstractEventLoop, node: 'Node', storage: 'SQLiteStorage'):
+        self.loop = loop
+        self.node = node
+        self.storage = storage
+        self.announce_task: asyncio.Task = None
+        self.announce_queue: typing.List[str] = []
+        self._done = asyncio.Event()
+        self.announced = set()
+
+    async def _run_consumer(self):
+        while self.announce_queue:
+            try:
+                blob_hash = self.announce_queue.pop()
+                peers = len(await self.node.announce_blob(blob_hash))
+                self.announcements_sent_metric.labels(peers=peers, error=False).inc()
+                if peers > 4:
+                    self.announced.add(blob_hash)
+                else:
+                    log.debug("failed to announce %s, could only find %d peers, retrying soon.", blob_hash[:8], peers)
+            except Exception as err:
+                self.announcements_sent_metric.labels(peers=0, error=True).inc()
+                log.warning("error announcing %s: %s", blob_hash[:8], str(err))
+
+    async def _announce(self, batch_size: typing.Optional[int] = 10):
+        while batch_size:
+            if not self.node.joined.is_set():
+                await self.node.joined.wait()
+            await asyncio.sleep(60)
+            if not self.node.protocol.routing_table.get_peers():
+                log.warning("No peers in DHT, announce round skipped")
+                continue
+            self.announce_queue.extend(await self.storage.get_blobs_to_announce())
+            self.announcement_queue_size_metric.labels(scope="global").set(len(self.announce_queue))
+            log.debug("announcer task wake up, %d blobs to announce", len(self.announce_queue))
+            while len(self.announce_queue) > 0:
+                log.info("%i blobs to announce", len(self.announce_queue))
+                await asyncio.gather(*[self._run_consumer() for _ in range(batch_size)])
+                announced = list(filter(None, self.announced))
+                if announced:
+                    await self.storage.update_last_announced_blobs(announced)
+                    log.info("announced %i blobs", len(announced))
+                    self.announced.clear()
+            self._done.set()
+            self._done.clear()
+
+    def start(self, batch_size: typing.Optional[int] = 10):
+        assert not self.announce_task or self.announce_task.done(), "already running"
+        self.announce_task = self.loop.create_task(self._announce(batch_size))
+
+    def stop(self):
+        if self.announce_task and not self.announce_task.done():
+            self.announce_task.cancel()
+
+    def wait(self):
+        return self._done.wait()

lbry/dht/constants.py

@@ -0,0 +1,40 @@
+import hashlib
+import os
+
+HASH_CLASS = hashlib.sha384  # pylint: disable=invalid-name
+HASH_LENGTH = HASH_CLASS().digest_size
+HASH_BITS = HASH_LENGTH * 8
+ALPHA = 5
+K = 8
+SPLIT_BUCKETS_UNDER_INDEX = 1
+REPLACEMENT_CACHE_SIZE = 8
+RPC_TIMEOUT = 5.0
+RPC_ATTEMPTS = 5
+RPC_ATTEMPTS_PRUNING_WINDOW = 600
+ITERATIVE_LOOKUP_DELAY = RPC_TIMEOUT / 2.0  # TODO: use config val / 2 if rpc timeout is provided
+REFRESH_INTERVAL = 3600  # 1 hour
+REPLICATE_INTERVAL = REFRESH_INTERVAL
+DATA_EXPIRATION = 86400  # 24 hours
+TOKEN_SECRET_REFRESH_INTERVAL = 300  # 5 minutes
+MAYBE_PING_DELAY = 300  # 5 minutes
+CHECK_REFRESH_INTERVAL = REFRESH_INTERVAL / 5
+RPC_ID_LENGTH = 20
+PROTOCOL_VERSION = 1
+MSG_SIZE_LIMIT = 1400
+
+
+def digest(data: bytes) -> bytes:
+    h = HASH_CLASS()
+    h.update(data)
+    return h.digest()
+
+
+def generate_id(num=None) -> bytes:
+    if num is not None:
+        return digest(str(num).encode())
+    else:
+        return digest(os.urandom(32))
+
+
+def generate_rpc_id(num=None) -> bytes:
+    return generate_id(num)[:RPC_ID_LENGTH]

lbry/dht/node.py

@@ -0,0 +1,282 @@
+import logging
+import asyncio
+import typing
+import socket
+
+from prometheus_client import Gauge
+
+from lbry.utils import aclosing, resolve_host
+from lbry.dht import constants
+from lbry.dht.peer import make_kademlia_peer
+from lbry.dht.protocol.distance import Distance
+from lbry.dht.protocol.iterative_find import IterativeNodeFinder, IterativeValueFinder
+from lbry.dht.protocol.protocol import KademliaProtocol
+
+if typing.TYPE_CHECKING:
+    from lbry.dht.peer import PeerManager
+    from lbry.dht.peer import KademliaPeer
+
+log = logging.getLogger(__name__)
+
+
+class Node:
+    storing_peers_metric = Gauge(
+        "storing_peers", "Number of peers storing blobs announced to this node", namespace="dht_node",
+        labelnames=("scope",),
+    )
+    stored_blob_with_x_bytes_colliding = Gauge(
+        "stored_blobs_x_bytes_colliding", "Number of blobs with at least X bytes colliding with this node id prefix",
+        namespace="dht_node", labelnames=("amount",)
+    )
+    def __init__(self, loop: asyncio.AbstractEventLoop, peer_manager: 'PeerManager', node_id: bytes, udp_port: int,
+                 internal_udp_port: int, peer_port: int, external_ip: str, rpc_timeout: float = constants.RPC_TIMEOUT,
+                 split_buckets_under_index: int = constants.SPLIT_BUCKETS_UNDER_INDEX, is_bootstrap_node: bool = False,
+                 storage: typing.Optional['SQLiteStorage'] = None):
+        self.loop = loop
+        self.internal_udp_port = internal_udp_port
+        self.protocol = KademliaProtocol(loop, peer_manager, node_id, external_ip, udp_port, peer_port, rpc_timeout,
+                                         split_buckets_under_index, is_bootstrap_node)
+        self.listening_port: asyncio.DatagramTransport = None
+        self.joined = asyncio.Event()
+        self._join_task: asyncio.Task = None
+        self._refresh_task: asyncio.Task = None
+        self._storage = storage
+
+    @property
+    def stored_blob_hashes(self):
+        return self.protocol.data_store.keys()
+
+    async def refresh_node(self, force_once=False):
+        while True:
+            # remove peers with expired blob announcements from the datastore
+            self.protocol.data_store.removed_expired_peers()
+
+            total_peers: typing.List['KademliaPeer'] = []
+            # add all peers in the routing table
+            total_peers.extend(self.protocol.routing_table.get_peers())
+            # add all the peers who have announced blobs to us
+            storing_peers = self.protocol.data_store.get_storing_contacts()
+            self.storing_peers_metric.labels("global").set(len(storing_peers))
+            total_peers.extend(storing_peers)
+
+            counts = {0: 0, 1: 0, 2: 0}
+            node_id = self.protocol.node_id
+            for blob_hash in self.protocol.data_store.keys():
+                bytes_colliding = 0 if blob_hash[0] != node_id[0] else 2 if blob_hash[1] == node_id[1] else 1
+                counts[bytes_colliding] += 1
+            self.stored_blob_with_x_bytes_colliding.labels(amount=0).set(counts[0])
+            self.stored_blob_with_x_bytes_colliding.labels(amount=1).set(counts[1])
+            self.stored_blob_with_x_bytes_colliding.labels(amount=2).set(counts[2])
+
+            # get ids falling in the midpoint of each bucket that hasn't been recently updated
+            node_ids = self.protocol.routing_table.get_refresh_list(0, True)
+
+            if self.protocol.routing_table.get_peers():
+                # if we have node ids to look up, perform the iterative search until we have k results
+                while node_ids:
+                    peers = await self.peer_search(node_ids.pop())
+                    total_peers.extend(peers)
+            else:
+                if force_once:
+                    break
+                fut = asyncio.Future()
+                self.loop.call_later(constants.REFRESH_INTERVAL // 4, fut.set_result, None)
+                await fut
+                continue
+
+            # ping the set of peers; upon success/failure the routing able and last replied/failed time will be updated
+            to_ping = [peer for peer in set(total_peers) if self.protocol.peer_manager.peer_is_good(peer) is not True]
+            if to_ping:
+                self.protocol.ping_queue.enqueue_maybe_ping(*to_ping, delay=0)
+            if self._storage:
+                await self._storage.save_kademlia_peers(self.protocol.routing_table.get_peers())
+            if force_once:
+                break
+
+            fut = asyncio.Future()
+            self.loop.call_later(constants.REFRESH_INTERVAL, fut.set_result, None)
+            await fut
+
+    async def announce_blob(self, blob_hash: str) -> typing.List[bytes]:
+        hash_value = bytes.fromhex(blob_hash)
+        assert len(hash_value) == constants.HASH_LENGTH
+        peers = await self.peer_search(hash_value)
+
+        if not self.protocol.external_ip:
+            raise Exception("Cannot determine external IP")
+        log.debug("Store to %i peers", len(peers))
+        for peer in peers:
+            log.debug("store to %s %s %s", peer.address, peer.udp_port, peer.tcp_port)
+        stored_to_tup = await asyncio.gather(
+            *(self.protocol.store_to_peer(hash_value, peer) for peer in peers)
+        )
+        stored_to = [node_id for node_id, contacted in stored_to_tup if contacted]
+        if stored_to:
+            log.debug(
+                "Stored %s to %i of %i attempted peers", hash_value.hex()[:8],
+                len(stored_to), len(peers)
+            )
+        else:
+            log.debug("Failed announcing %s, stored to 0 peers", blob_hash[:8])
+        return stored_to
+
+    def stop(self) -> None:
+        if self.joined.is_set():
+            self.joined.clear()
+        if self._join_task:
+            self._join_task.cancel()
+        if self._refresh_task and not (self._refresh_task.done() or self._refresh_task.cancelled()):
+            self._refresh_task.cancel()
+        if self.protocol and self.protocol.ping_queue.running:
+            self.protocol.ping_queue.stop()
+            self.protocol.stop()
+        if self.listening_port is not None:
+            self.listening_port.close()
+        self._join_task = None
+        self.listening_port = None
+        log.info("Stopped DHT node")
+
+    async def start_listening(self, interface: str = '0.0.0.0') -> None:
+        if not self.listening_port:
+            self.listening_port, _ = await self.loop.create_datagram_endpoint(
+                lambda: self.protocol, (interface, self.internal_udp_port)
+            )
+            log.info("DHT node listening on UDP %s:%i", interface, self.internal_udp_port)
+            self.protocol.start()
+        else:
+            log.warning("Already bound to port %s", self.listening_port)
+
+    async def join_network(self, interface: str = '0.0.0.0',
+                           known_node_urls: typing.Optional[typing.List[typing.Tuple[str, int]]] = None):
+        def peers_from_urls(urls: typing.Optional[typing.List[typing.Tuple[bytes, str, int, int]]]):
+            peer_addresses = []
+            for node_id, address, udp_port, tcp_port in urls:
+                if (node_id, address, udp_port, tcp_port) not in peer_addresses and \
+                        (address, udp_port) != (self.protocol.external_ip, self.protocol.udp_port):
+                    peer_addresses.append((node_id, address, udp_port, tcp_port))
+            return [make_kademlia_peer(*peer_address) for peer_address in peer_addresses]
+
+        if not self.listening_port:
+            await self.start_listening(interface)
+        self.protocol.ping_queue.start()
+        self._refresh_task = self.loop.create_task(self.refresh_node())
+
+        while True:
+            if self.protocol.routing_table.get_peers():
+                if not self.joined.is_set():
+                    self.joined.set()
+                    log.info(
+                        "joined dht, %i peers known in %i buckets", len(self.protocol.routing_table.get_peers()),
+                        self.protocol.routing_table.buckets_with_contacts()
+                    )
+            else:
+                if self.joined.is_set():
+                    self.joined.clear()
+                seed_peers = peers_from_urls(
+                    await self._storage.get_persisted_kademlia_peers()
+                ) if self._storage else []
+                if not seed_peers:
+                    try:
+                        seed_peers.extend(peers_from_urls([
+                            (None, await resolve_host(address, udp_port, 'udp'), udp_port, None)
+                            for address, udp_port in known_node_urls or []
+                        ]))
+                    except socket.gaierror:
+                        await asyncio.sleep(30)
+                        continue
+
+                self.protocol.peer_manager.reset()
+                self.protocol.ping_queue.enqueue_maybe_ping(*seed_peers, delay=0.0)
+                await self.peer_search(self.protocol.node_id, shortlist=seed_peers, count=32)
+
+            await asyncio.sleep(1)
+
+    def start(self, interface: str, known_node_urls: typing.Optional[typing.List[typing.Tuple[str, int]]] = None):
+        self._join_task = self.loop.create_task(self.join_network(interface, known_node_urls))
+
+    def get_iterative_node_finder(self, key: bytes, shortlist: typing.Optional[typing.List['KademliaPeer']] = None,
+                                  max_results: int = constants.K) -> IterativeNodeFinder:
+        shortlist = shortlist or self.protocol.routing_table.find_close_peers(key)
+        return IterativeNodeFinder(self.loop, self.protocol, key, max_results, shortlist)
+
+    def get_iterative_value_finder(self, key: bytes, shortlist: typing.Optional[typing.List['KademliaPeer']] = None,
+                                   max_results: int = -1) -> IterativeValueFinder:
+        shortlist = shortlist or self.protocol.routing_table.find_close_peers(key)
+        return IterativeValueFinder(self.loop, self.protocol, key, max_results, shortlist)
+
+    async def peer_search(self, node_id: bytes, count=constants.K, max_results=constants.K * 2,
+                          shortlist: typing.Optional[typing.List['KademliaPeer']] = None
+                          ) -> typing.List['KademliaPeer']:
+        peers = []
+        async with aclosing(self.get_iterative_node_finder(
+                node_id, shortlist=shortlist, max_results=max_results)) as node_finder:
+            async for iteration_peers in node_finder:
+                peers.extend(iteration_peers)
+        distance = Distance(node_id)
+        peers.sort(key=lambda peer: distance(peer.node_id))
+        return peers[:count]
+
+    async def _accumulate_peers_for_value(self, search_queue: asyncio.Queue, result_queue: asyncio.Queue):
+        tasks = []
+        try:
+            while True:
+                blob_hash = await search_queue.get()
+                tasks.append(self.loop.create_task(self._peers_for_value_producer(blob_hash, result_queue)))
+        finally:
+            for task in tasks:
+                task.cancel()
+
+    async def _peers_for_value_producer(self, blob_hash: str, result_queue: asyncio.Queue):
+        async def put_into_result_queue_after_pong(_peer):
+            try:
+                await self.protocol.get_rpc_peer(_peer).ping()
+                result_queue.put_nowait([_peer])
+                log.debug("pong from %s:%i for %s", _peer.address, _peer.udp_port, blob_hash)
+            except asyncio.TimeoutError:
+                pass
+
+        # prioritize peers who reply to a dht ping first
+        # this minimizes attempting to make tcp connections that won't work later to dead or unreachable peers
+        async with aclosing(self.get_iterative_value_finder(bytes.fromhex(blob_hash))) as value_finder:
+            async for results in value_finder:
+                to_put = []
+                for peer in results:
+                    if peer.address == self.protocol.external_ip and self.protocol.peer_port == peer.tcp_port:
+                        continue
+                    is_good = self.protocol.peer_manager.peer_is_good(peer)
+                    if is_good:
+                        # the peer has replied recently over UDP, it can probably be reached on the TCP port
+                        to_put.append(peer)
+                    elif is_good is None:
+                        if not peer.udp_port:
+                            # TODO: use the same port for TCP and UDP
+                            # the udp port must be guessed
+                            # default to the ports being the same. if the TCP port appears to be <=0.48.0 default,
+                            # including on a network with several nodes, then assume the udp port is proportionately
+                            # based on a starting port of 4444
+                            udp_port_to_try = peer.tcp_port
+                            if 3400 > peer.tcp_port > 3332:
+                                udp_port_to_try = (peer.tcp_port - 3333) + 4444
+                            self.loop.create_task(put_into_result_queue_after_pong(
+                                make_kademlia_peer(peer.node_id, peer.address, udp_port_to_try, peer.tcp_port)
+                            ))
+                        else:
+                            self.loop.create_task(put_into_result_queue_after_pong(peer))
+                    else:
+                        # the peer is known to be bad/unreachable, skip trying to connect to it over TCP
+                        log.debug("skip bad peer %s:%i for %s", peer.address, peer.tcp_port, blob_hash)
+                if to_put:
+                    result_queue.put_nowait(to_put)
+
+    def accumulate_peers(self, search_queue: asyncio.Queue,
+                         peer_queue: typing.Optional[asyncio.Queue] = None
+                         ) -> typing.Tuple[asyncio.Queue, asyncio.Task]:
+        queue = peer_queue or asyncio.Queue()
+        return queue, self.loop.create_task(self._accumulate_peers_for_value(search_queue, queue))
+
+
+async def get_kademlia_peers_from_hosts(peer_list: typing.List[typing.Tuple[str, int]]) -> typing.List['KademliaPeer']:
+    peer_address_list = [(await resolve_host(url, port, proto='tcp'), port) for url, port in peer_list]
+    kademlia_peer_list = [make_kademlia_peer(None, address, None, tcp_port=port, allow_localhost=True)
+                          for address, port in peer_address_list]
+    return kademlia_peer_list

lbry/dht/peer.py

@@ -1,44 +1,54 @@
 import typing
 import asyncio
 import logging
-import ipaddress
-from binascii import hexlify
 from dataclasses import dataclass, field
 from functools import lru_cache
 
+from prometheus_client import Gauge
+
+from lbry.utils import is_valid_public_ipv4 as _is_valid_public_ipv4, LRUCache
 from lbry.dht import constants
 from lbry.dht.serialization.datagram import make_compact_address, make_compact_ip, decode_compact_address
 
+ALLOW_LOCALHOST = False
+CACHE_SIZE = 16384
 log = logging.getLogger(__name__)
 
 
-@lru_cache(1024)
+@lru_cache(CACHE_SIZE)
 def make_kademlia_peer(node_id: typing.Optional[bytes], address: typing.Optional[str],
                        udp_port: typing.Optional[int] = None,
-                       tcp_port: typing.Optional[int] = None) -> 'KademliaPeer':
-    return KademliaPeer(address, node_id, udp_port, tcp_port=tcp_port)
+                       tcp_port: typing.Optional[int] = None,
+                       allow_localhost: bool = False) -> 'KademliaPeer':
+    return KademliaPeer(address, node_id, udp_port, tcp_port=tcp_port, allow_localhost=allow_localhost)
 
 
-def is_valid_ipv4(address):
-    try:
-        ip = ipaddress.ip_address(address)
-        return ip.version == 4
-    except ipaddress.AddressValueError:
-        return False
+def is_valid_public_ipv4(address, allow_localhost: bool = False):
+    allow_localhost = bool(allow_localhost or ALLOW_LOCALHOST)
+    return _is_valid_public_ipv4(address, allow_localhost)
 
 
 class PeerManager:
+    peer_manager_keys_metric = Gauge(
+        "peer_manager_keys", "Number of keys tracked by PeerManager dicts (sum)", namespace="dht_node",
+        labelnames=("scope",)
+    )
     def __init__(self, loop: asyncio.AbstractEventLoop):
         self._loop = loop
         self._rpc_failures: typing.Dict[
             typing.Tuple[str, int], typing.Tuple[typing.Optional[float], typing.Optional[float]]
-        ] = {}
-        self._last_replied: typing.Dict[typing.Tuple[str, int], float] = {}
-        self._last_sent: typing.Dict[typing.Tuple[str, int], float] = {}
-        self._last_requested: typing.Dict[typing.Tuple[str, int], float] = {}
-        self._node_id_mapping: typing.Dict[typing.Tuple[str, int], bytes] = {}
-        self._node_id_reverse_mapping: typing.Dict[bytes, typing.Tuple[str, int]] = {}
-        self._node_tokens: typing.Dict[bytes, (float, bytes)] = {}
+        ] = LRUCache(CACHE_SIZE)
+        self._last_replied: typing.Dict[typing.Tuple[str, int], float] = LRUCache(CACHE_SIZE)
+        self._last_sent: typing.Dict[typing.Tuple[str, int], float] = LRUCache(CACHE_SIZE)
+        self._last_requested: typing.Dict[typing.Tuple[str, int], float] = LRUCache(CACHE_SIZE)
+        self._node_id_mapping: typing.Dict[typing.Tuple[str, int], bytes] = LRUCache(CACHE_SIZE)
+        self._node_id_reverse_mapping: typing.Dict[bytes, typing.Tuple[str, int]] = LRUCache(CACHE_SIZE)
+        self._node_tokens: typing.Dict[bytes, (float, bytes)] = LRUCache(CACHE_SIZE)
+
+    def count_cache_keys(self):
+        return len(self._rpc_failures) + len(self._last_replied) + len(self._last_sent) + len(
+            self._last_requested) + len(self._node_id_mapping) + len(self._node_id_reverse_mapping) + len(
+            self._node_tokens)
 
     def reset(self):
         for statistic in (self._rpc_failures, self._last_replied, self._last_sent, self._last_requested):
@@ -70,7 +80,7 @@ class PeerManager:
 
     def get_node_token(self, node_id: bytes) -> typing.Optional[bytes]:
         ts, token = self._node_tokens.get(node_id, (0, None))
-        if ts and ts > self._loop.time() - constants.token_secret_refresh_interval:
+        if ts and ts > self._loop.time() - constants.TOKEN_SECRET_REFRESH_INTERVAL:
             return token
 
     def get_last_replied(self, address: str, udp_port: int) -> typing.Optional[float]:
@@ -88,28 +98,32 @@ class PeerManager:
             self._node_id_mapping.pop(self._node_id_reverse_mapping.pop(node_id))
         self._node_id_mapping[(address, udp_port)] = node_id
         self._node_id_reverse_mapping[node_id] = (address, udp_port)
+        self.peer_manager_keys_metric.labels("global").set(self.count_cache_keys())
+
+    def get_node_id_for_endpoint(self, address, port):
+        return self._node_id_mapping.get((address, port))
 
     def prune(self):  # TODO: periodically call this
         now = self._loop.time()
         to_pop = []
         for (address, udp_port), (_, last_failure) in self._rpc_failures.items():
-            if last_failure and last_failure < now - constants.rpc_attempts_pruning_window:
+            if last_failure and last_failure < now - constants.RPC_ATTEMPTS_PRUNING_WINDOW:
                 to_pop.append((address, udp_port))
         while to_pop:
             del self._rpc_failures[to_pop.pop()]
         to_pop = []
-        for node_id, (age, token) in self._node_tokens.items():
-            if age < now - constants.token_secret_refresh_interval:
+        for node_id, (age, token) in self._node_tokens.items():  # pylint: disable=unused-variable
+            if age < now - constants.TOKEN_SECRET_REFRESH_INTERVAL:
                 to_pop.append(node_id)
         while to_pop:
             del self._node_tokens[to_pop.pop()]
 
-    def contact_triple_is_good(self, node_id: bytes, address: str, udp_port: int):
+    def contact_triple_is_good(self, node_id: bytes, address: str, udp_port: int):  # pylint: disable=too-many-return-statements
         """
         :return: False if peer is bad, None if peer is unknown, or True if peer is good
         """
 
-        delay = self._loop.time() - constants.check_refresh_interval
+        delay = self._loop.time() - constants.CHECK_REFRESH_INTERVAL
 
         # fixme: find a way to re-enable that without breaking other parts
         # if node_id not in self._node_id_reverse_mapping or (address, udp_port) not in self._node_id_mapping:
@@ -139,9 +153,10 @@ class PeerManager:
     def peer_is_good(self, peer: 'KademliaPeer'):
         return self.contact_triple_is_good(peer.node_id, peer.address, peer.udp_port)
 
-    def decode_tcp_peer_from_compact_address(self, compact_address: bytes) -> 'KademliaPeer':
-        node_id, address, tcp_port = decode_compact_address(compact_address)
-        return make_kademlia_peer(node_id, address, udp_port=None, tcp_port=tcp_port)
+
+def decode_tcp_peer_from_compact_address(compact_address: bytes) -> 'KademliaPeer':  # pylint: disable=no-self-use
+    node_id, address, tcp_port = decode_compact_address(compact_address)
+    return make_kademlia_peer(node_id, address, udp_port=None, tcp_port=tcp_port)
 
 
 @dataclass(unsafe_hash=True)
@@ -151,16 +166,17 @@ class KademliaPeer:
     udp_port: typing.Optional[int] = field(hash=True)
     tcp_port: typing.Optional[int] = field(compare=False, hash=False)
     protocol_version: typing.Optional[int] = field(default=1, compare=False, hash=False)
+    allow_localhost: bool = field(default=False, compare=False, hash=False)
 
     def __post_init__(self):
         if self._node_id is not None:
-            if not len(self._node_id) == constants.hash_length:
-                raise ValueError("invalid node_id: {}".format(hexlify(self._node_id).decode()))
-        if self.udp_port is not None and not 1 <= self.udp_port <= 65535:
-            raise ValueError("invalid udp port")
-        if self.tcp_port is not None and not 1 <= self.tcp_port <= 65535:
-            raise ValueError("invalid tcp port")
-        if not is_valid_ipv4(self.address):
+            if not len(self._node_id) == constants.HASH_LENGTH:
+                raise ValueError("invalid node_id: {}".format(self._node_id.hex()))
+        if self.udp_port is not None and not 1024 <= self.udp_port <= 65535:
+            raise ValueError(f"invalid udp port: {self.address}:{self.udp_port}")
+        if self.tcp_port is not None and not 1024 <= self.tcp_port <= 65535:
+            raise ValueError(f"invalid tcp port: {self.address}:{self.tcp_port}")
+        if not is_valid_public_ipv4(self.address, self.allow_localhost):
             raise ValueError(f"invalid ip address: '{self.address}'")
 
     def update_tcp_port(self, tcp_port: int):
@@ -178,3 +194,6 @@ class KademliaPeer:
 
     def compact_ip(self):
         return make_compact_ip(self.address)
+
+    def __str__(self):
+        return f"{self.__class__.__name__}({self.node_id.hex()[:8]}@{self.address}:{self.udp_port}-{self.tcp_port})"

lbry/dht/protocol/data_store.py

@@ -16,13 +16,19 @@ class DictDataStore:
         self._peer_manager = peer_manager
         self.completed_blobs: typing.Set[str] = set()
 
+    def keys(self):
+        return self._data_store.keys()
+
+    def __len__(self):
+        return self._data_store.__len__()
+
     def removed_expired_peers(self):
         now = self.loop.time()
         keys = list(self._data_store.keys())
         for key in keys:
             to_remove = []
             for (peer, ts) in self._data_store[key]:
-                if ts + constants.data_expiration < now or self._peer_manager.peer_is_good(peer) is False:
+                if ts + constants.DATA_EXPIRATION < now or self._peer_manager.peer_is_good(peer) is False:
                     to_remove.append((peer, ts))
             for item in to_remove:
                 self._data_store[key].remove(item)
@@ -43,7 +49,7 @@ class DictDataStore:
         """
         now = self.loop.time()
         for (peer, ts) in self._data_store.get(key, []):
-            if ts + constants.data_expiration > now:
+            if ts + constants.DATA_EXPIRATION > now:
                 yield peer
 
     def has_peers_for_blob(self, key: bytes) -> bool:
@@ -53,7 +59,7 @@ class DictDataStore:
         now = self.loop.time()
         if key in self._data_store:
             current = list(filter(lambda x: x[0] == contact, self._data_store[key]))
-            if len(current):
+            if len(current) > 0:
                 self._data_store[key][self._data_store[key].index(current[0])] = contact, now
             else:
                 self._data_store[key].append((contact, now))
@@ -65,6 +71,6 @@ class DictDataStore:
 
     def get_storing_contacts(self) -> typing.List['KademliaPeer']:
         peers = set()
-        for key, stored in self._data_store.items():
+        for _, stored in self._data_store.items():
             peers.update(set(map(lambda tup: tup[0], stored)))
         return list(peers)

lbry/dht/protocol/distance.py

@@ -9,17 +9,17 @@ class Distance:
     """
 
     def __init__(self, key: bytes):
-        if len(key) != constants.hash_length:
+        if len(key) != constants.HASH_LENGTH:
             raise ValueError(f"invalid key length: {len(key)}")
         self.key = key
         self.val_key_one = int.from_bytes(key, 'big')
 
     def __call__(self, key_two: bytes) -> int:
-        if len(key_two) != constants.hash_length:
+        if len(key_two) != constants.HASH_LENGTH:
             raise ValueError(f"invalid length of key to compare: {len(key_two)}")
         val_key_two = int.from_bytes(key_two, 'big')
         return self.val_key_one ^ val_key_two
 
-    def is_closer(self, a: bytes, b: bytes) -> bool:
-        """Returns true is `a` is closer to `key` than `b` is"""
-        return self(a) < self(b)
+    def is_closer(self, key_a: bytes, key_b: bytes) -> bool:
+        """Returns true is `key_a` is closer to `key` than `key_b` is"""
+        return self(key_a) < self(key_b)

lbry/dht/protocol/iterative_find.py

@@ -0,0 +1,361 @@
+import asyncio
+from itertools import chain
+from collections import defaultdict, OrderedDict
+from collections.abc import AsyncIterator
+import typing
+import logging
+from typing import TYPE_CHECKING
+from lbry.dht import constants
+from lbry.dht.error import RemoteException, TransportNotConnected
+from lbry.dht.protocol.distance import Distance
+from lbry.dht.peer import make_kademlia_peer, decode_tcp_peer_from_compact_address
+from lbry.dht.serialization.datagram import PAGE_KEY
+
+if TYPE_CHECKING:
+    from lbry.dht.protocol.protocol import KademliaProtocol
+    from lbry.dht.peer import PeerManager, KademliaPeer
+
+log = logging.getLogger(__name__)
+
+
+class FindResponse:
+    @property
+    def found(self) -> bool:
+        raise NotImplementedError()
+
+    def get_close_triples(self) -> typing.List[typing.Tuple[bytes, str, int]]:
+        raise NotImplementedError()
+
+    def get_close_kademlia_peers(self, peer_info) -> typing.Generator[typing.Iterator['KademliaPeer'], None, None]:
+        for contact_triple in self.get_close_triples():
+            node_id, address, udp_port = contact_triple
+            try:
+                yield make_kademlia_peer(node_id, address, udp_port)
+            except ValueError:
+                log.warning("misbehaving peer %s:%i returned peer with reserved ip %s:%i", peer_info.address,
+                            peer_info.udp_port, address, udp_port)
+
+
+class FindNodeResponse(FindResponse):
+    def __init__(self, key: bytes, close_triples: typing.List[typing.Tuple[bytes, str, int]]):
+        self.key = key
+        self.close_triples = close_triples
+
+    @property
+    def found(self) -> bool:
+        return self.key in [triple[0] for triple in self.close_triples]
+
+    def get_close_triples(self) -> typing.List[typing.Tuple[bytes, str, int]]:
+        return self.close_triples
+
+
+class FindValueResponse(FindResponse):
+    def __init__(self, key: bytes, result_dict: typing.Dict):
+        self.key = key
+        self.token = result_dict[b'token']
+        self.close_triples: typing.List[typing.Tuple[bytes, bytes, int]] = result_dict.get(b'contacts', [])
+        self.found_compact_addresses = result_dict.get(key, [])
+        self.pages = int(result_dict.get(PAGE_KEY, 0))
+
+    @property
+    def found(self) -> bool:
+        return len(self.found_compact_addresses) > 0
+
+    def get_close_triples(self) -> typing.List[typing.Tuple[bytes, str, int]]:
+        return [(node_id, address.decode(), port) for node_id, address, port in self.close_triples]
+
+
+class IterativeFinder(AsyncIterator):
+    def __init__(self, loop: asyncio.AbstractEventLoop,
+                 protocol: 'KademliaProtocol', key: bytes,
+                 max_results: typing.Optional[int] = constants.K,
+                 shortlist: typing.Optional[typing.List['KademliaPeer']] = None):
+        if len(key) != constants.HASH_LENGTH:
+            raise ValueError("invalid key length: %i" % len(key))
+        self.loop = loop
+        self.peer_manager = protocol.peer_manager
+        self.protocol = protocol
+
+        self.key = key
+        self.max_results = max(constants.K, max_results)
+
+        self.active: typing.Dict['KademliaPeer', int] = OrderedDict()  # peer: distance, sorted
+        self.contacted: typing.Set['KademliaPeer'] = set()
+        self.distance = Distance(key)
+
+        self.iteration_queue = asyncio.Queue()
+
+        self.running_probes: typing.Dict['KademliaPeer', asyncio.Task] = {}
+        self.iteration_count = 0
+        self.running = False
+        self.tasks: typing.List[asyncio.Task] = []
+        for peer in shortlist:
+            if peer.node_id:
+                self._add_active(peer, force=True)
+            else:
+                # seed nodes
+                self._schedule_probe(peer)
+
+    async def send_probe(self, peer: 'KademliaPeer') -> FindResponse:
+        """
+        Send the rpc request to the peer and return an object with the FindResponse interface
+        """
+        raise NotImplementedError()
+
+    def search_exhausted(self):
+        """
+        This method ends the iterator due no more peers to contact.
+        Override to provide last time results.
+        """
+        self.iteration_queue.put_nowait(None)
+
+    def check_result_ready(self, response: FindResponse):
+        """
+        Called after adding peers from an rpc result to the shortlist.
+        This method is responsible for putting a result for the generator into the Queue
+        """
+        raise NotImplementedError()
+
+    def get_initial_result(self) -> typing.List['KademliaPeer']:  #pylint: disable=no-self-use
+        """
+        Get an initial or cached result to be put into the Queue. Used for findValue requests where the blob
+        has peers in the local data store of blobs announced to us
+        """
+        return []
+
+    def _add_active(self, peer, force=False):
+        if not force and self.peer_manager.peer_is_good(peer) is False:
+            return
+        if peer in self.contacted:
+            return
+        if peer not in self.active and peer.node_id and peer.node_id != self.protocol.node_id:
+            self.active[peer] = self.distance(peer.node_id)
+            self.active = OrderedDict(sorted(self.active.items(), key=lambda item: item[1]))
+
+    async def _handle_probe_result(self, peer: 'KademliaPeer', response: FindResponse):
+        self._add_active(peer)
+        for new_peer in response.get_close_kademlia_peers(peer):
+            self._add_active(new_peer)
+        self.check_result_ready(response)
+        self._log_state(reason="check result")
+
+    def _reset_closest(self, peer):
+        if peer in self.active:
+            del self.active[peer]
+
+    async def _send_probe(self, peer: 'KademliaPeer'):
+        try:
+            response = await self.send_probe(peer)
+        except asyncio.TimeoutError:
+            self._reset_closest(peer)
+            return
+        except asyncio.CancelledError:
+            log.debug("%s[%x] cancelled probe",
+                      type(self).__name__, id(self))
+            raise
+        except ValueError as err:
+            log.warning(str(err))
+            self._reset_closest(peer)
+            return
+        except TransportNotConnected:
+            await self._aclose(reason="not connected")
+            return
+        except RemoteException:
+            self._reset_closest(peer)
+            return
+        return await self._handle_probe_result(peer, response)
+
+    def _search_round(self):
+        """
+        Send up to constants.alpha (5) probes to closest active peers
+        """
+
+        added = 0
+        for index, peer in enumerate(self.active.keys()):
+            if index == 0:
+                log.debug("%s[%x] closest to probe: %s",
+                          type(self).__name__, id(self),
+                          peer.node_id.hex()[:8])
+            if peer in self.contacted:
+                continue
+            if len(self.running_probes) >= constants.ALPHA:
+                break
+            if index > (constants.K + len(self.running_probes)):
+                break
+            origin_address = (peer.address, peer.udp_port)
+            if peer.node_id == self.protocol.node_id:
+                continue
+            if origin_address == (self.protocol.external_ip, self.protocol.udp_port):
+                continue
+            self._schedule_probe(peer)
+            added += 1
+        log.debug("%s[%x] running %d probes for key %s",
+                  type(self).__name__, id(self),
+                  len(self.running_probes), self.key.hex()[:8])
+        if not added and not self.running_probes:
+            log.debug("%s[%x] search for %s exhausted",
+                      type(self).__name__, id(self),
+                      self.key.hex()[:8])
+            self.search_exhausted()
+
+    def _schedule_probe(self, peer: 'KademliaPeer'):
+        self.contacted.add(peer)
+
+        t = self.loop.create_task(self._send_probe(peer))
+
+        def callback(_):
+            self.running_probes.pop(peer, None)
+            if self.running:
+                self._search_round()
+
+        t.add_done_callback(callback)
+        self.running_probes[peer] = t
+
+    def _log_state(self, reason="?"):
+        log.debug("%s[%x] [%s] %s: %i active nodes %i contacted %i produced %i queued",
+                  type(self).__name__, id(self), self.key.hex()[:8],
+                  reason, len(self.active), len(self.contacted),
+                  self.iteration_count, self.iteration_queue.qsize())
+
+    def __aiter__(self):
+        if self.running:
+            raise Exception("already running")
+        self.running = True
+        self.loop.call_soon(self._search_round)
+        return self
+
+    async def __anext__(self) -> typing.List['KademliaPeer']:
+        try:
+            if self.iteration_count == 0:
+                result = self.get_initial_result() or await self.iteration_queue.get()
+            else:
+                result = await self.iteration_queue.get()
+            if not result:
+                raise StopAsyncIteration
+            self.iteration_count += 1
+            return result
+        except asyncio.CancelledError:
+            await self._aclose(reason="cancelled")
+            raise
+        except StopAsyncIteration:
+            await self._aclose(reason="no more results")
+            raise
+
+    async def _aclose(self, reason="?"):
+        log.debug("%s[%x] [%s] shutdown because %s: %i active nodes %i contacted %i produced %i queued",
+                  type(self).__name__, id(self), self.key.hex()[:8],
+                  reason, len(self.active), len(self.contacted),
+                  self.iteration_count, self.iteration_queue.qsize())
+        self.running = False
+        self.iteration_queue.put_nowait(None)
+        for task in chain(self.tasks, self.running_probes.values()):
+            task.cancel()
+        self.tasks.clear()
+        self.running_probes.clear()
+
+    async def aclose(self):
+        if self.running:
+            await self._aclose(reason="aclose")
+        log.debug("%s[%x] [%s] async close completed",
+                  type(self).__name__, id(self), self.key.hex()[:8])
+
+class IterativeNodeFinder(IterativeFinder):
+    def __init__(self, loop: asyncio.AbstractEventLoop,
+                 protocol: 'KademliaProtocol', key: bytes,
+                 max_results: typing.Optional[int] = constants.K,
+                 shortlist: typing.Optional[typing.List['KademliaPeer']] = None):
+        super().__init__(loop, protocol, key, max_results, shortlist)
+        self.yielded_peers: typing.Set['KademliaPeer'] = set()
+
+    async def send_probe(self, peer: 'KademliaPeer') -> FindNodeResponse:
+        log.debug("probe %s:%d (%s) for NODE %s",
+                  peer.address, peer.udp_port, peer.node_id.hex()[:8] if peer.node_id else '', self.key.hex()[:8])
+        response = await self.protocol.get_rpc_peer(peer).find_node(self.key)
+        return FindNodeResponse(self.key, response)
+
+    def search_exhausted(self):
+        self.put_result(self.active.keys(), finish=True)
+
+    def put_result(self, from_iter: typing.Iterable['KademliaPeer'], finish=False):
+        not_yet_yielded = [
+            peer for peer in from_iter
+            if peer not in self.yielded_peers
+            and peer.node_id != self.protocol.node_id
+            and self.peer_manager.peer_is_good(peer) is True  # return only peers who answered
+        ]
+        not_yet_yielded.sort(key=lambda peer: self.distance(peer.node_id))
+        to_yield = not_yet_yielded[:max(constants.K, self.max_results)]
+        if to_yield:
+            self.yielded_peers.update(to_yield)
+            self.iteration_queue.put_nowait(to_yield)
+        if finish:
+            self.iteration_queue.put_nowait(None)
+
+    def check_result_ready(self, response: FindNodeResponse):
+        found = response.found and self.key != self.protocol.node_id
+
+        if found:
+            log.debug("found")
+            return self.put_result(self.active.keys(), finish=True)
+
+
+class IterativeValueFinder(IterativeFinder):
+    def __init__(self, loop: asyncio.AbstractEventLoop,
+                 protocol: 'KademliaProtocol', key: bytes,
+                 max_results: typing.Optional[int] = constants.K,
+                 shortlist: typing.Optional[typing.List['KademliaPeer']] = None):
+        super().__init__(loop, protocol, key, max_results, shortlist)
+        self.blob_peers: typing.Set['KademliaPeer'] = set()
+        # this tracks the index of the most recent page we requested from each peer
+        self.peer_pages: typing.DefaultDict['KademliaPeer', int] = defaultdict(int)
+        # this tracks the set of blob peers returned by each peer
+        self.discovered_peers: typing.Dict['KademliaPeer', typing.Set['KademliaPeer']] = defaultdict(set)
+
+    async def send_probe(self, peer: 'KademliaPeer') -> FindValueResponse:
+        log.debug("probe %s:%d (%s) for VALUE %s",
+                  peer.address, peer.udp_port, peer.node_id.hex()[:8], self.key.hex()[:8])
+        page = self.peer_pages[peer]
+        response = await self.protocol.get_rpc_peer(peer).find_value(self.key, page=page)
+        parsed = FindValueResponse(self.key, response)
+        if not parsed.found:
+            return parsed
+        already_known = len(self.discovered_peers[peer])
+        decoded_peers = set()
+        for compact_addr in parsed.found_compact_addresses:
+            try:
+                decoded_peers.add(decode_tcp_peer_from_compact_address(compact_addr))
+            except ValueError:
+                log.warning("misbehaving peer %s:%i returned invalid peer for blob",
+                            peer.address, peer.udp_port)
+                self.peer_manager.report_failure(peer.address, peer.udp_port)
+                parsed.found_compact_addresses.clear()
+                return parsed
+        self.discovered_peers[peer].update(decoded_peers)
+        log.debug("probed %s:%i page %i, %i known", peer.address, peer.udp_port, page,
+                  already_known + len(parsed.found_compact_addresses))
+        if len(self.discovered_peers[peer]) != already_known + len(parsed.found_compact_addresses):
+            log.warning("misbehaving peer %s:%i returned duplicate peers for blob", peer.address, peer.udp_port)
+        elif len(parsed.found_compact_addresses) >= constants.K and self.peer_pages[peer] < parsed.pages:
+            # the peer returned a full page and indicates it has more
+            self.peer_pages[peer] += 1
+            if peer in self.contacted:
+                # the peer must be removed from self.contacted so that it will be probed for the next page
+                self.contacted.remove(peer)
+        return parsed
+
+    def check_result_ready(self, response: FindValueResponse):
+        if response.found:
+            blob_peers = [decode_tcp_peer_from_compact_address(compact_addr)
+                          for compact_addr in response.found_compact_addresses]
+            to_yield = []
+            for blob_peer in blob_peers:
+                if blob_peer not in self.blob_peers:
+                    self.blob_peers.add(blob_peer)
+                    to_yield.append(blob_peer)
+            if to_yield:
+                self.iteration_queue.put_nowait(to_yield)
+
+    def get_initial_result(self) -> typing.List['KademliaPeer']:
+        if self.protocol.data_store.has_peers_for_blob(self.key):
+            return self.protocol.data_store.get_peers_for_blob(self.key)
+        return []

lbry/dht/protocol/protocol.py

@@ -3,13 +3,16 @@ import socket
 import functools
 import hashlib
 import asyncio
+import time
 import typing
-import binascii
 import random
 from asyncio.protocols import DatagramProtocol
 from asyncio.transports import DatagramTransport
 
+from prometheus_client import Gauge, Counter, Histogram
+
 from lbry.dht import constants
+from lbry.dht.serialization.bencoding import DecodeError
 from lbry.dht.serialization.datagram import decode_datagram, ErrorDatagram, ResponseDatagram, RequestDatagram
 from lbry.dht.serialization.datagram import RESPONSE_TYPE, ERROR_TYPE, PAGE_KEY
 from lbry.dht.error import RemoteException, TransportNotConnected
@@ -23,13 +26,18 @@ if typing.TYPE_CHECKING:
 log = logging.getLogger(__name__)
 
 
-old_protocol_errors = {
+OLD_PROTOCOL_ERRORS = {
     "findNode() takes exactly 2 arguments (5 given)": "0.19.1",
     "findValue() takes exactly 2 arguments (5 given)": "0.19.1"
 }
 
 
 class KademliaRPC:
+    stored_blob_metric = Gauge(
+        "stored_blobs", "Number of blobs announced by other peers", namespace="dht_node",
+        labelnames=("scope",),
+    )
+
     def __init__(self, protocol: 'KademliaProtocol', loop: asyncio.AbstractEventLoop, peer_port: int = 3333):
         self.protocol = protocol
         self.loop = loop
@@ -48,35 +56,36 @@ class KademliaRPC:
         return b'pong'
 
     def store(self, rpc_contact: 'KademliaPeer', blob_hash: bytes, token: bytes, port: int) -> bytes:
-        if len(blob_hash) != constants.hash_bits // 8:
+        if len(blob_hash) != constants.HASH_BITS // 8:
             raise ValueError(f"invalid length of blob hash: {len(blob_hash)}")
         if not 0 < port < 65535:
             raise ValueError(f"invalid tcp port: {port}")
         rpc_contact.update_tcp_port(port)
         if not self.verify_token(token, rpc_contact.compact_ip()):
-            if self.loop.time() - self.protocol.started_listening_time < constants.token_secret_refresh_interval:
+            if self.loop.time() - self.protocol.started_listening_time < constants.TOKEN_SECRET_REFRESH_INTERVAL:
                 pass
             else:
                 raise ValueError("Invalid token")
         self.protocol.data_store.add_peer_to_blob(
             rpc_contact, blob_hash
         )
+        self.stored_blob_metric.labels("global").set(len(self.protocol.data_store))
         return b'OK'
 
     def find_node(self, rpc_contact: 'KademliaPeer', key: bytes) -> typing.List[typing.Tuple[bytes, str, int]]:
-        if len(key) != constants.hash_length:
+        if len(key) != constants.HASH_LENGTH:
             raise ValueError("invalid contact node_id length: %i" % len(key))
 
         contacts = self.protocol.routing_table.find_close_peers(key, sender_node_id=rpc_contact.node_id)
         contact_triples = []
-        for contact in contacts[:constants.k * 2]:
+        for contact in contacts[:constants.K * 2]:
             contact_triples.append((contact.node_id, contact.address, contact.udp_port))
         return contact_triples
 
     def find_value(self, rpc_contact: 'KademliaPeer', key: bytes, page: int = 0):
         page = page if page > 0 else 0
 
-        if len(key) != constants.hash_length:
+        if len(key) != constants.HASH_LENGTH:
             raise ValueError("invalid blob_exchange hash length: %i" % len(key))
 
         response = {
@@ -84,7 +93,7 @@ class KademliaRPC:
         }
 
         if not page:
-            response[b'contacts'] = self.find_node(rpc_contact, key)[:constants.k]
+            response[b'contacts'] = self.find_node(rpc_contact, key)[:constants.K]
 
         if self.protocol.protocol_version:
             response[b'protocolVersion'] = self.protocol.protocol_version
@@ -96,16 +105,16 @@ class KademliaRPC:
             if not rpc_contact.tcp_port or peer.compact_address_tcp() != rpc_contact.compact_address_tcp()
         ]
         # if we don't have k storing peers to return and we have this hash locally, include our contact information
-        if len(peers) < constants.k and binascii.hexlify(key).decode() in self.protocol.data_store.completed_blobs:
+        if len(peers) < constants.K and key.hex() in self.protocol.data_store.completed_blobs:
             peers.append(self.compact_address())
         if not peers:
             response[PAGE_KEY] = 0
         else:
-            response[PAGE_KEY] = (len(peers) // (constants.k + 1)) + 1  # how many pages of peers we have for the blob
-        if len(peers) > constants.k:
+            response[PAGE_KEY] = (len(peers) // (constants.K + 1)) + 1  # how many pages of peers we have for the blob
+        if len(peers) > constants.K:
             random.Random(self.protocol.node_id).shuffle(peers)
-        if page * constants.k < len(peers):
-            response[key] = peers[page * constants.k:page * constants.k + constants.k]
+        if page * constants.K < len(peers):
+            response[key] = peers[page * constants.K:page * constants.K + constants.K]
         return response
 
     def refresh_token(self):  # TODO: this needs to be called periodically
@@ -154,7 +163,7 @@ class RemoteKademliaRPC:
         :param blob_hash: blob hash as bytes
         :return: b'OK'
         """
-        if len(blob_hash) != constants.hash_bits // 8:
+        if len(blob_hash) != constants.HASH_BITS // 8:
             raise ValueError(f"invalid length of blob hash: {len(blob_hash)}")
         if not self.protocol.peer_port or not 0 < self.protocol.peer_port < 65535:
             raise ValueError(f"invalid tcp port: {self.protocol.peer_port}")
@@ -171,7 +180,7 @@ class RemoteKademliaRPC:
         """
         :return: [(node_id, address, udp_port), ...]
         """
-        if len(key) != constants.hash_bits // 8:
+        if len(key) != constants.HASH_BITS // 8:
             raise ValueError(f"invalid length of find node key: {len(key)}")
         response = await self.protocol.send_request(
             self.peer, RequestDatagram.make_find_node(self.protocol.node_id, key)
@@ -186,7 +195,7 @@ class RemoteKademliaRPC:
             <key bytes>: [<blob_peer_compact_address, ...]
         }
         """
-        if len(key) != constants.hash_bits // 8:
+        if len(key) != constants.HASH_BITS // 8:
             raise ValueError(f"invalid length of find value key: {len(key)}")
         response = await self.protocol.send_request(
             self.peer, RequestDatagram.make_find_value(self.protocol.node_id, key, page=page)
@@ -203,12 +212,16 @@ class PingQueue:
         self._process_task: asyncio.Task = None
         self._running = False
         self._running_pings: typing.Set[asyncio.Task] = set()
-        self._default_delay = constants.maybe_ping_delay
+        self._default_delay = constants.MAYBE_PING_DELAY
 
     @property
     def running(self):
         return self._running
 
+    @property
+    def busy(self):
+        return self._running and (any(self._running_pings) or any(self._pending_contacts))
+
     def enqueue_maybe_ping(self, *peers: 'KademliaPeer', delay: typing.Optional[float] = None):
         delay = delay if delay is not None else self._default_delay
         now = self._loop.time()
@@ -220,11 +233,11 @@ class PingQueue:
         async def ping_task():
             try:
                 if self._protocol.peer_manager.peer_is_good(peer):
-                    if peer not in self._protocol.routing_table.get_peers():
+                    if not self._protocol.routing_table.get_peer(peer.node_id):
                         self._protocol.add_peer(peer)
                     return
                 await self._protocol.get_rpc_peer(peer).ping()
-            except asyncio.TimeoutError:
+            except (asyncio.TimeoutError, RemoteException):
                 pass
 
         task = self._loop.create_task(ping_task())
@@ -240,7 +253,7 @@ class PingQueue:
                     del self._pending_contacts[peer]
                     self.maybe_ping(peer)
                     break
-            await asyncio.sleep(1, loop=self._loop)
+            await asyncio.sleep(1)
 
     def start(self):
         assert not self._running
@@ -259,9 +272,33 @@ class PingQueue:
 
 
 class KademliaProtocol(DatagramProtocol):
+    request_sent_metric = Counter(
+        "request_sent", "Number of requests send from DHT RPC protocol", namespace="dht_node",
+        labelnames=("method",),
+    )
+    request_success_metric = Counter(
+        "request_success", "Number of successful requests", namespace="dht_node",
+        labelnames=("method",),
+    )
+    request_error_metric = Counter(
+        "request_error", "Number of errors returned from request to other peers", namespace="dht_node",
+        labelnames=("method",),
+    )
+    HISTOGRAM_BUCKETS = (
+        .005, .01, .025, .05, .075, .1, .25, .5, .75, 1.0, 2.5, 3.0, 3.5, 4.0, 4.50, 5.0, 5.50, 6.0, float('inf')
+    )
+    response_time_metric = Histogram(
+        "response_time", "Response times of DHT RPC requests", namespace="dht_node", buckets=HISTOGRAM_BUCKETS,
+        labelnames=("method",)
+    )
+    received_request_metric = Counter(
+        "received_request", "Number of received DHT RPC requests", namespace="dht_node",
+        labelnames=("method",),
+    )
+
     def __init__(self, loop: asyncio.AbstractEventLoop, peer_manager: 'PeerManager', node_id: bytes, external_ip: str,
-                 udp_port: int, peer_port: int, rpc_timeout: float = constants.rpc_timeout,
-                 split_buckets_under_index: int = constants.split_buckets_under_index):
+                 udp_port: int, peer_port: int, rpc_timeout: float = constants.RPC_TIMEOUT,
+                 split_buckets_under_index: int = constants.SPLIT_BUCKETS_UNDER_INDEX, is_boostrap_node: bool = False):
         self.peer_manager = peer_manager
         self.loop = loop
         self.node_id = node_id
@@ -271,20 +308,21 @@ class KademliaProtocol(DatagramProtocol):
         self.is_seed_node = False
         self.partial_messages: typing.Dict[bytes, typing.Dict[bytes, bytes]] = {}
         self.sent_messages: typing.Dict[bytes, typing.Tuple['KademliaPeer', asyncio.Future, RequestDatagram]] = {}
-        self.protocol_version = constants.protocol_version
+        self.protocol_version = constants.PROTOCOL_VERSION
         self.started_listening_time = 0
         self.transport: DatagramTransport = None
         self.old_token_secret = constants.generate_id()
         self.token_secret = constants.generate_id()
-        self.routing_table = TreeRoutingTable(self.loop, self.peer_manager, self.node_id, split_buckets_under_index)
+        self.routing_table = TreeRoutingTable(
+            self.loop, self.peer_manager, self.node_id, split_buckets_under_index, is_bootstrap_node=is_boostrap_node)
         self.data_store = DictDataStore(self.loop, self.peer_manager)
         self.ping_queue = PingQueue(self.loop, self)
         self.node_rpc = KademliaRPC(self, self.loop, self.peer_port)
         self.rpc_timeout = rpc_timeout
-        self._split_lock = asyncio.Lock(loop=self.loop)
+        self._split_lock = asyncio.Lock()
         self._to_remove: typing.Set['KademliaPeer'] = set()
         self._to_add: typing.Set['KademliaPeer'] = set()
-        self._wakeup_routing_task = asyncio.Event(loop=self.loop)
+        self._wakeup_routing_task = asyncio.Event()
         self.maintaing_routing_task: typing.Optional[asyncio.Task] = None
 
     @functools.lru_cache(128)
@@ -323,72 +361,10 @@ class KademliaProtocol(DatagramProtocol):
         return args, {}
 
     async def _add_peer(self, peer: 'KademliaPeer'):
-        if not peer.node_id:
-            log.warning("Tried adding a peer with no node id!")
-            return False
-        for p in self.routing_table.get_peers():
-            if (p.address, p.udp_port) == (peer.address, peer.udp_port) and p.node_id != peer.node_id:
-                self.routing_table.remove_peer(p)
-                self.routing_table.join_buckets()
-        bucket_index = self.routing_table.kbucket_index(peer.node_id)
-        if self.routing_table.buckets[bucket_index].add_peer(peer):
-            return True
-
-        # The bucket is full; see if it can be split (by checking if its range includes the host node's node_id)
-        if self.routing_table.should_split(bucket_index, peer.node_id):
-            self.routing_table.split_bucket(bucket_index)
-            # Retry the insertion attempt
-            result = await self._add_peer(peer)
-            self.routing_table.join_buckets()
-            return result
-        else:
-            # We can't split the k-bucket
-            #
-            # The 13 page kademlia paper specifies that the least recently contacted node in the bucket
-            # shall be pinged. If it fails to reply it is replaced with the new contact. If the ping is successful
-            # the new contact is ignored and not added to the bucket (sections 2.2 and 2.4).
-            #
-            # A reasonable extension to this is BEP 0005, which extends the above:
-            #
-            #    Not all nodes that we learn about are equal. Some are "good" and some are not.
-            #    Many nodes using the DHT are able to send queries and receive responses,
-            #    but are not able to respond to queries from other nodes. It is important that
-            #    each node's routing table must contain only known good nodes. A good node is
-            #    a node has responded to one of our queries within the last 15 minutes. A node
-            #    is also good if it has ever responded to one of our queries and has sent us a
-            #    query within the last 15 minutes. After 15 minutes of inactivity, a node becomes
-            #    questionable. Nodes become bad when they fail to respond to multiple queries
-            #    in a row. Nodes that we know are good are given priority over nodes with unknown status.
-            #
-            # When there are bad or questionable nodes in the bucket, the least recent is selected for
-            # potential replacement (BEP 0005). When all nodes in the bucket are fresh, the head (least recent)
-            # contact is selected as described in section 2.2 of the kademlia paper. In both cases the new contact
-            # is ignored if the pinged node replies.
-
-            not_good_contacts = self.routing_table.buckets[bucket_index].get_bad_or_unknown_peers()
-            not_recently_replied = []
-            for p in not_good_contacts:
-                last_replied = self.peer_manager.get_last_replied(p.address, p.udp_port)
-                if not last_replied or last_replied + 60 < self.loop.time():
-                    not_recently_replied.append(p)
-            if not_recently_replied:
-                to_replace = not_recently_replied[0]
-            else:
-                to_replace = self.routing_table.buckets[bucket_index].peers[0]
-                last_replied = self.peer_manager.get_last_replied(to_replace.address, to_replace.udp_port)
-                if last_replied and last_replied + 60 > self.loop.time():
-                    return False
-            log.debug("pinging %s:%s", to_replace.address, to_replace.udp_port)
-            try:
-                to_replace_rpc = self.get_rpc_peer(to_replace)
-                await to_replace_rpc.ping()
-                return False
-            except asyncio.TimeoutError:
-                log.debug("Replacing dead contact in bucket %i: %s:%i with %s:%i ", bucket_index,
-                          to_replace.address, to_replace.udp_port, peer.address, peer.udp_port)
-                if to_replace in self.routing_table.buckets[bucket_index]:
-                    self.routing_table.buckets[bucket_index].remove_peer(to_replace)
-                return await self._add_peer(peer)
+        async def probe(some_peer: 'KademliaPeer'):
+            rpc_peer = self.get_rpc_peer(some_peer)
+            await rpc_peer.ping()
+        return await self.routing_table.add_peer(peer, probe)
 
     def add_peer(self, peer: 'KademliaPeer'):
         if peer.node_id == self.node_id:
@@ -406,35 +382,34 @@ class KademliaProtocol(DatagramProtocol):
                 async with self._split_lock:
                     peer = self._to_remove.pop()
                     self.routing_table.remove_peer(peer)
-                    self.routing_table.join_buckets()
             while self._to_add:
                 async with self._split_lock:
                     await self._add_peer(self._to_add.pop())
-            await asyncio.gather(self._wakeup_routing_task.wait(), asyncio.sleep(.1, loop=self.loop), loop=self.loop)
+            await asyncio.gather(self._wakeup_routing_task.wait(), asyncio.sleep(.1))
             self._wakeup_routing_task.clear()
 
     def _handle_rpc(self, sender_contact: 'KademliaPeer', message: RequestDatagram):
-        assert sender_contact.node_id != self.node_id, (binascii.hexlify(sender_contact.node_id)[:8].decode(),
-                                                        binascii.hexlify(self.node_id)[:8].decode())
+        assert sender_contact.node_id != self.node_id, (sender_contact.node_id.hex()[:8],
+                                                        self.node_id.hex()[:8])
         method = message.method
         if method not in [b'ping', b'store', b'findNode', b'findValue']:
             raise AttributeError('Invalid method: %s' % message.method.decode())
         if message.args and isinstance(message.args[-1], dict) and b'protocolVersion' in message.args[-1]:
             # args don't need reformatting
-            args, kw = tuple(message.args[:-1]), message.args[-1]
+            args, kwargs = tuple(message.args[:-1]), message.args[-1]
         else:
-            args, kw = self._migrate_incoming_rpc_args(sender_contact, message.method, *message.args)
+            args, kwargs = self._migrate_incoming_rpc_args(sender_contact, message.method, *message.args)
         log.debug("%s:%i RECV CALL %s %s:%i", self.external_ip, self.udp_port, message.method.decode(),
                   sender_contact.address, sender_contact.udp_port)
 
         if method == b'ping':
             result = self.node_rpc.ping()
         elif method == b'store':
-            blob_hash, token, port, original_publisher_id, age = args[:5]
+            blob_hash, token, port, original_publisher_id, age = args[:5]  # pylint: disable=unused-variable
             result = self.node_rpc.store(sender_contact, blob_hash, token, port)
         else:
             key = args[0]
-            page = kw.get(PAGE_KEY, 0)
+            page = kwargs.get(PAGE_KEY, 0)
             if method == b'findNode':
                 result = self.node_rpc.find_node(sender_contact, key)
             else:
@@ -447,11 +422,15 @@ class KademliaProtocol(DatagramProtocol):
 
     def handle_request_datagram(self, address: typing.Tuple[str, int], request_datagram: RequestDatagram):
         # This is an RPC method request
+        self.received_request_metric.labels(method=request_datagram.method).inc()
         self.peer_manager.report_last_requested(address[0], address[1])
-        try:
-            peer = self.routing_table.get_peer(request_datagram.node_id)
-        except IndexError:
-            peer = make_kademlia_peer(request_datagram.node_id, address[0], address[1])
+        peer = self.routing_table.get_peer(request_datagram.node_id)
+        if not peer:
+            try:
+                peer = make_kademlia_peer(request_datagram.node_id, address[0], address[1])
+            except ValueError as err:
+                log.warning("error replying to %s: %s", address[0], str(err))
+                return
         try:
             self._handle_rpc(peer, request_datagram)
             # if the contact is not known to be bad (yet) and we haven't yet queried it, send it a ping so that it
@@ -484,25 +463,25 @@ class KademliaProtocol(DatagramProtocol):
     def handle_response_datagram(self, address: typing.Tuple[str, int], response_datagram: ResponseDatagram):
         # Find the message that triggered this response
         if response_datagram.rpc_id in self.sent_messages:
-            peer, df, request = self.sent_messages[response_datagram.rpc_id]
+            peer, future, _ = self.sent_messages[response_datagram.rpc_id]
             if peer.address != address[0]:
-                df.set_exception(RemoteException(
-                    f"response from {address[0]}, expected {peer.address}")
+                future.set_exception(
+                    RemoteException(f"response from {address[0]}, expected {peer.address}")
                 )
                 return
 
             # We got a result from the RPC
             if peer.node_id == self.node_id:
-                df.set_exception(RemoteException("node has our node id"))
+                future.set_exception(RemoteException("node has our node id"))
                 return
             elif response_datagram.node_id == self.node_id:
-                df.set_exception(RemoteException("incoming message is from our node id"))
+                future.set_exception(RemoteException("incoming message is from our node id"))
                 return
             peer = make_kademlia_peer(response_datagram.node_id, address[0], address[1])
             self.peer_manager.report_last_replied(address[0], address[1])
             self.peer_manager.update_contact_triple(peer.node_id, address[0], address[1])
-            if not df.cancelled():
-                df.set_result(response_datagram)
+            if not future.cancelled():
+                future.set_result(response_datagram)
                 self.add_peer(peer)
             else:
                 log.warning("%s:%i replied, but after we cancelled the request attempt",
@@ -516,11 +495,13 @@ class KademliaProtocol(DatagramProtocol):
         # The RPC request raised a remote exception; raise it locally
         remote_exception = RemoteException(f"{error_datagram.exception_type}({error_datagram.response})")
         if error_datagram.rpc_id in self.sent_messages:
-            peer, df, request = self.sent_messages.pop(error_datagram.rpc_id)
+            peer, future, request = self.sent_messages.pop(error_datagram.rpc_id)
             if (peer.address, peer.udp_port) != address:
-                df.set_exception(RemoteException(
-                    f"response from {address[0]}:{address[1]}, "
-                    f"expected {peer.address}:{peer.udp_port}")
+                future.set_exception(
+                    RemoteException(
+                        f"response from {address[0]}:{address[1]}, "
+                        f"expected {peer.address}:{peer.udp_port}"
+                    )
                 )
                 return
             error_msg = f"" \
@@ -529,28 +510,32 @@ class KademliaProtocol(DatagramProtocol):
                 f"Raised: {str(remote_exception)}"
             if 'Invalid token' in error_msg:
                 log.debug(error_msg)
-            elif error_datagram.response not in old_protocol_errors:
+            elif error_datagram.response not in OLD_PROTOCOL_ERRORS:
                 log.warning(error_msg)
             else:
-                log.debug("known dht protocol backwards compatibility error with %s:%i (lbrynet v%s)",
-                            peer.address, peer.udp_port, old_protocol_errors[error_datagram.response])
-            df.set_exception(remote_exception)
+                log.debug(
+                    "known dht protocol backwards compatibility error with %s:%i (lbrynet v%s)",
+                    peer.address, peer.udp_port, OLD_PROTOCOL_ERRORS[error_datagram.response]
+                )
+            future.set_exception(remote_exception)
             return
         else:
-            if error_datagram.response not in old_protocol_errors:
+            if error_datagram.response not in OLD_PROTOCOL_ERRORS:
                 msg = f"Received error from {address[0]}:{address[1]}, but it isn't in response to a " \
                     f"pending request: {str(remote_exception)}"
                 log.warning(msg)
             else:
-                log.debug("known dht protocol backwards compatibility error with %s:%i (lbrynet v%s)",
-                            address[0], address[1], old_protocol_errors[error_datagram.response])
+                log.debug(
+                    "known dht protocol backwards compatibility error with %s:%i (lbrynet v%s)",
+                    address[0], address[1], OLD_PROTOCOL_ERRORS[error_datagram.response]
+                )
 
-    def datagram_received(self, datagram: bytes, address: typing.Tuple[str, int]) -> None:
+    def datagram_received(self, datagram: bytes, address: typing.Tuple[str, int]) -> None:  # pylint: disable=arguments-renamed
         try:
             message = decode_datagram(datagram)
-        except (ValueError, TypeError):
+        except (ValueError, TypeError, DecodeError):
             self.peer_manager.report_failure(address[0], address[1])
-            log.warning("Couldn't decode dht datagram from %s: %s", address, binascii.hexlify(datagram).decode())
+            log.warning("Couldn't decode dht datagram from %s: %s", address, datagram.hex())
             return
 
         if isinstance(message, RequestDatagram):
@@ -565,14 +550,19 @@ class KademliaProtocol(DatagramProtocol):
         self._send(peer, request)
         response_fut = self.sent_messages[request.rpc_id][1]
         try:
+            self.request_sent_metric.labels(method=request.method).inc()
+            start = time.perf_counter()
             response = await asyncio.wait_for(response_fut, self.rpc_timeout)
+            self.response_time_metric.labels(method=request.method).observe(time.perf_counter() - start)
             self.peer_manager.report_last_replied(peer.address, peer.udp_port)
+            self.request_success_metric.labels(method=request.method).inc()
             return response
         except asyncio.CancelledError:
             if not response_fut.done():
                 response_fut.cancel()
             raise
         except (asyncio.TimeoutError, RemoteException):
+            self.request_error_metric.labels(method=request.method).inc()
             self.peer_manager.report_failure(peer.address, peer.udp_port)
             if self.peer_manager.peer_is_good(peer) is False:
                 self.remove_peer(peer)
@@ -589,12 +579,12 @@ class KademliaProtocol(DatagramProtocol):
             raise TransportNotConnected()
 
         data = message.bencode()
-        if len(data) > constants.msg_size_limit:
+        if len(data) > constants.MSG_SIZE_LIMIT:
             log.warning("cannot send datagram larger than %i bytes (packet is %i bytes)",
-                        constants.msg_size_limit, len(data))
-            log.debug("Packet is too large to send: %s", binascii.hexlify(data[:3500]).decode())
+                        constants.MSG_SIZE_LIMIT, len(data))
+            log.debug("Packet is too large to send: %s", data[:3500].hex())
             raise ValueError(
-                f"cannot send datagram larger than {constants.msg_size_limit} bytes (packet is {len(data)} bytes)"
+                f"cannot send datagram larger than {constants.MSG_SIZE_LIMIT} bytes (packet is {len(data)} bytes)"
             )
         if isinstance(message, (RequestDatagram, ResponseDatagram)):
             assert message.node_id == self.node_id, message
@@ -637,35 +627,38 @@ class KademliaProtocol(DatagramProtocol):
         return constants.digest(self.token_secret + compact_ip)
 
     def verify_token(self, token, compact_ip):
-        h = constants.hash_class()
+        h = constants.HASH_CLASS()
         h.update(self.token_secret + compact_ip)
         if self.old_token_secret and not token == h.digest():  # TODO: why should we be accepting the previous token?
-            h = constants.hash_class()
+            h = constants.HASH_CLASS()
             h.update(self.old_token_secret + compact_ip)
             if not token == h.digest():
                 return False
         return True
 
-    async def store_to_peer(self, hash_value: bytes, peer: 'KademliaPeer',
+    async def store_to_peer(self, hash_value: bytes, peer: 'KademliaPeer',  # pylint: disable=too-many-return-statements
                             retry: bool = True) -> typing.Tuple[bytes, bool]:
         async def __store():
             res = await self.get_rpc_peer(peer).store(hash_value)
             if res != b"OK":
                 raise ValueError(res)
-            log.debug("Stored %s to %s", binascii.hexlify(hash_value).decode()[:8], peer)
+            log.debug("Stored %s to %s", hash_value.hex()[:8], peer)
             return peer.node_id, True
 
         try:
             return await __store()
         except asyncio.TimeoutError:
-            log.debug("Timeout while storing blob_hash %s at %s", binascii.hexlify(hash_value).decode()[:8], peer)
+            log.debug("Timeout while storing blob_hash %s at %s", hash_value.hex()[:8], peer)
             return peer.node_id, False
         except ValueError as err:
-            log.error("Unexpected response: %s" % err)
+            log.error("Unexpected response: %s", err)
             return peer.node_id, False
         except RemoteException as err:
+            if 'findValue() takes exactly 2 arguments (5 given)' in str(err):
+                log.debug("peer %s:%i is running an incompatible version of lbrynet", peer.address, peer.udp_port)
+                return peer.node_id, False
             if 'Invalid token' not in str(err):
-                log.exception("Unexpected error while storing blob_hash")
+                log.warning("Unexpected error while storing blob_hash: %s", err)
                 return peer.node_id, False
         self.peer_manager.clear_token(peer.node_id)
         if not retry:

lbry/dht/protocol/routing_table.py

@@ -4,7 +4,11 @@ import logging
 import typing
 import itertools
 
+from prometheus_client import Gauge
+
+from lbry import utils
 from lbry.dht import constants
+from lbry.dht.error import RemoteException
 from lbry.dht.protocol.distance import Distance
 if typing.TYPE_CHECKING:
     from lbry.dht.peer import KademliaPeer, PeerManager
@@ -13,10 +17,20 @@ log = logging.getLogger(__name__)
 
 
 class KBucket:
-    """ Description - later
     """
+    Kademlia K-bucket implementation.
+    """
+    peer_in_routing_table_metric = Gauge(
+        "peers_in_routing_table", "Number of peers on routing table", namespace="dht_node",
+        labelnames=("scope",)
+    )
+    peer_with_x_bit_colliding_metric = Gauge(
+        "peer_x_bit_colliding", "Number of peers with at least X bits colliding with this node id",
+        namespace="dht_node", labelnames=("amount",)
+    )
 
-    def __init__(self, peer_manager: 'PeerManager', range_min: int, range_max: int, node_id: bytes):
+    def __init__(self, peer_manager: 'PeerManager', range_min: int, range_max: int,
+                 node_id: bytes, capacity: int = constants.K):
         """
         @param range_min: The lower boundary for the range in the n-bit ID
                          space covered by this k-bucket
@@ -24,12 +38,12 @@ class KBucket:
                          covered by this k-bucket
         """
         self._peer_manager = peer_manager
-        self.last_accessed = 0
         self.range_min = range_min
         self.range_max = range_max
         self.peers: typing.List['KademliaPeer'] = []
         self._node_id = node_id
         self._distance_to_self = Distance(node_id)
+        self.capacity = capacity
 
     def add_peer(self, peer: 'KademliaPeer') -> bool:
         """ Add contact to _contact list in the right order. This will move the
@@ -50,24 +64,25 @@ class KBucket:
             self.peers.append(peer)
             return True
         else:
-            for i in range(len(self.peers)):
-                p = self.peers[i]
-                if p.node_id == peer.node_id:
-                    self.peers.remove(p)
+            for i, _ in enumerate(self.peers):
+                local_peer = self.peers[i]
+                if local_peer.node_id == peer.node_id:
+                    self.peers.remove(local_peer)
                     self.peers.append(peer)
                     return True
-        if len(self.peers) < constants.k:
+        if len(self.peers) < self.capacity:
             self.peers.append(peer)
+            self.peer_in_routing_table_metric.labels("global").inc()
+            bits_colliding = utils.get_colliding_prefix_bits(peer.node_id, self._node_id)
+            self.peer_with_x_bit_colliding_metric.labels(amount=bits_colliding).inc()
             return True
         else:
             return False
-            # raise BucketFull("No space in bucket to insert contact")
 
     def get_peer(self, node_id: bytes) -> 'KademliaPeer':
         for peer in self.peers:
             if peer.node_id == node_id:
                 return peer
-        raise IndexError(node_id)
 
     def get_peers(self, count=-1, exclude_contact=None, sort_distance_to=None) -> typing.List['KademliaPeer']:
         """ Returns a list containing up to the first count number of contacts
@@ -101,8 +116,8 @@ class KBucket:
         current_len = len(peers)
 
         # If count greater than k - return only k contacts
-        if count > constants.k:
-            count = constants.k
+        if count > constants.K:
+            count = constants.K
 
         if not current_len:
             return peers
@@ -124,6 +139,9 @@ class KBucket:
 
     def remove_peer(self, peer: 'KademliaPeer') -> None:
         self.peers.remove(peer)
+        self.peer_in_routing_table_metric.labels("global").dec()
+        bits_colliding = utils.get_colliding_prefix_bits(peer.node_id, self._node_id)
+        self.peer_with_x_bit_colliding_metric.labels(amount=bits_colliding).dec()
 
     def key_in_range(self, key: bytes) -> bool:
         """ Tests whether the specified key (i.e. node ID) is in the range
@@ -161,31 +179,43 @@ class TreeRoutingTable:
     version of the Kademlia paper, in section 2.4. It does, however, use the
     ping RPC-based k-bucket eviction algorithm described in section 2.2 of
     that paper.
+
+    BOOTSTRAP MODE: if set to True, we always add all peers. This is so a
+    bootstrap node does not get a bias towards its own node id and replies are
+    the best it can provide (joining peer knows its neighbors immediately).
+    Over time, this will need to be optimized so we use the disk as holding
+    everything in memory won't be feasible anymore.
+    See: https://github.com/bittorrent/bootstrap-dht
     """
+    bucket_in_routing_table_metric = Gauge(
+        "buckets_in_routing_table", "Number of buckets on routing table", namespace="dht_node",
+        labelnames=("scope",)
+    )
 
     def __init__(self, loop: asyncio.AbstractEventLoop, peer_manager: 'PeerManager', parent_node_id: bytes,
-                 split_buckets_under_index: int = constants.split_buckets_under_index):
+                 split_buckets_under_index: int = constants.SPLIT_BUCKETS_UNDER_INDEX, is_bootstrap_node: bool = False):
         self._loop = loop
         self._peer_manager = peer_manager
         self._parent_node_id = parent_node_id
         self._split_buckets_under_index = split_buckets_under_index
         self.buckets: typing.List[KBucket] = [
             KBucket(
-                self._peer_manager, range_min=0, range_max=2 ** constants.hash_bits, node_id=self._parent_node_id
+                self._peer_manager, range_min=0, range_max=2 ** constants.HASH_BITS, node_id=self._parent_node_id,
+                capacity=1 << 32 if is_bootstrap_node else constants.K
             )
         ]
 
     def get_peers(self) -> typing.List['KademliaPeer']:
         return list(itertools.chain.from_iterable(map(lambda bucket: bucket.peers, self.buckets)))
 
-    def should_split(self, bucket_index: int, to_add: bytes) -> bool:
+    def _should_split(self, bucket_index: int, to_add: bytes) -> bool:
         #  https://stackoverflow.com/questions/32129978/highly-unbalanced-kademlia-routing-table/32187456#32187456
         if bucket_index < self._split_buckets_under_index:
             return True
         contacts = self.get_peers()
         distance = Distance(self._parent_node_id)
         contacts.sort(key=lambda c: distance(c.node_id))
-        kth_contact = contacts[-1] if len(contacts) < constants.k else contacts[constants.k - 1]
+        kth_contact = contacts[-1] if len(contacts) < constants.K else contacts[constants.K - 1]
         return distance(to_add) < distance(kth_contact.node_id)
 
     def find_close_peers(self, key: bytes, count: typing.Optional[int] = None,
@@ -193,7 +223,7 @@ class TreeRoutingTable:
         exclude = [self._parent_node_id]
         if sender_node_id:
             exclude.append(sender_node_id)
-        count = count or constants.k
+        count = count or constants.K
         distance = Distance(key)
         contacts = self.get_peers()
         contacts = [c for c in contacts if c.node_id not in exclude]
@@ -203,39 +233,32 @@ class TreeRoutingTable:
         return []
 
     def get_peer(self, contact_id: bytes) -> 'KademliaPeer':
-        """
-        @raise IndexError: No contact with the specified contact ID is known
-                           by this node
-        """
-        return self.buckets[self.kbucket_index(contact_id)].get_peer(contact_id)
+        return self.buckets[self._kbucket_index(contact_id)].get_peer(contact_id)
 
     def get_refresh_list(self, start_index: int = 0, force: bool = False) -> typing.List[bytes]:
-        bucket_index = start_index
         refresh_ids = []
-        now = int(self._loop.time())
-        for bucket in self.buckets[start_index:]:
-            if force or now - bucket.last_accessed >= constants.refresh_interval:
-                to_search = self.midpoint_id_in_bucket_range(bucket_index)
-                refresh_ids.append(to_search)
-            bucket_index += 1
+        for offset, _ in enumerate(self.buckets[start_index:]):
+            refresh_ids.append(self._midpoint_id_in_bucket_range(start_index + offset))
+        # if we have 3 or fewer populated buckets get two random ids in the range of each to try and
+        # populate/split the buckets further
+        buckets_with_contacts = self.buckets_with_contacts()
+        if buckets_with_contacts <= 3:
+            for i in range(buckets_with_contacts):
+                refresh_ids.append(self._random_id_in_bucket_range(i))
+                refresh_ids.append(self._random_id_in_bucket_range(i))
         return refresh_ids
 
     def remove_peer(self, peer: 'KademliaPeer') -> None:
         if not peer.node_id:
             return
-        bucket_index = self.kbucket_index(peer.node_id)
+        bucket_index = self._kbucket_index(peer.node_id)
         try:
             self.buckets[bucket_index].remove_peer(peer)
+            self._join_buckets()
         except ValueError:
             return
 
-    def touch_kbucket(self, key: bytes) -> None:
-        self.touch_kbucket_by_index(self.kbucket_index(key))
-
-    def touch_kbucket_by_index(self, bucket_index: int):
-        self.buckets[bucket_index].last_accessed = int(self._loop.time())
-
-    def kbucket_index(self, key: bytes) -> int:
+    def _kbucket_index(self, key: bytes) -> int:
         i = 0
         for bucket in self.buckets:
             if bucket.key_in_range(key):
@@ -244,19 +267,19 @@ class TreeRoutingTable:
                 i += 1
         return i
 
-    def random_id_in_bucket_range(self, bucket_index: int) -> bytes:
+    def _random_id_in_bucket_range(self, bucket_index: int) -> bytes:
         random_id = int(random.randrange(self.buckets[bucket_index].range_min, self.buckets[bucket_index].range_max))
         return Distance(
             self._parent_node_id
-        )(random_id.to_bytes(constants.hash_length, 'big')).to_bytes(constants.hash_length, 'big')
+        )(random_id.to_bytes(constants.HASH_LENGTH, 'big')).to_bytes(constants.HASH_LENGTH, 'big')
 
-    def midpoint_id_in_bucket_range(self, bucket_index: int) -> bytes:
+    def _midpoint_id_in_bucket_range(self, bucket_index: int) -> bytes:
         half = int((self.buckets[bucket_index].range_max - self.buckets[bucket_index].range_min) // 2)
         return Distance(self._parent_node_id)(
-            int(self.buckets[bucket_index].range_min + half).to_bytes(constants.hash_length, 'big')
-        ).to_bytes(constants.hash_length, 'big')
+            int(self.buckets[bucket_index].range_min + half).to_bytes(constants.HASH_LENGTH, 'big')
+        ).to_bytes(constants.HASH_LENGTH, 'big')
 
-    def split_bucket(self, old_bucket_index: int) -> None:
+    def _split_bucket(self, old_bucket_index: int) -> None:
         """ Splits the specified k-bucket into two new buckets which together
         cover the same range in the key/ID space
 
@@ -279,11 +302,12 @@ class TreeRoutingTable:
         # ...and remove them from the old bucket
         for contact in new_bucket.peers:
             old_bucket.remove_peer(contact)
+        self.bucket_in_routing_table_metric.labels("global").set(len(self.buckets))
 
-    def join_buckets(self):
+    def _join_buckets(self):
         if len(self.buckets) == 1:
             return
-        to_pop = [i for i, bucket in enumerate(self.buckets) if not len(bucket)]
+        to_pop = [i for i, bucket in enumerate(self.buckets) if len(bucket) == 0]
         if not to_pop:
             return
         log.info("join buckets %i", len(to_pop))
@@ -302,18 +326,79 @@ class TreeRoutingTable:
         elif can_go_higher:
             self.buckets[bucket_index_to_pop + 1].range_min = bucket.range_min
         self.buckets.remove(bucket)
-        return self.join_buckets()
-
-    def contact_in_routing_table(self, address_tuple: typing.Tuple[str, int]) -> bool:
-        for bucket in self.buckets:
-            for contact in bucket.get_peers(sort_distance_to=False):
-                if address_tuple[0] == contact.address and address_tuple[1] == contact.udp_port:
-                    return True
-        return False
+        self.bucket_in_routing_table_metric.labels("global").set(len(self.buckets))
+        return self._join_buckets()
 
     def buckets_with_contacts(self) -> int:
         count = 0
         for bucket in self.buckets:
-            if len(bucket):
+            if len(bucket) > 0:
                 count += 1
         return count
+
+    async def add_peer(self, peer: 'KademliaPeer', probe: typing.Callable[['KademliaPeer'], typing.Awaitable]):
+        if not peer.node_id:
+            log.warning("Tried adding a peer with no node id!")
+            return False
+        for my_peer in self.get_peers():
+            if (my_peer.address, my_peer.udp_port) == (peer.address, peer.udp_port) and my_peer.node_id != peer.node_id:
+                self.remove_peer(my_peer)
+                self._join_buckets()
+        bucket_index = self._kbucket_index(peer.node_id)
+        if self.buckets[bucket_index].add_peer(peer):
+            return True
+
+        # The bucket is full; see if it can be split (by checking if its range includes the host node's node_id)
+        if self._should_split(bucket_index, peer.node_id):
+            self._split_bucket(bucket_index)
+            # Retry the insertion attempt
+            result = await self.add_peer(peer, probe)
+            self._join_buckets()
+            return result
+        else:
+            # We can't split the k-bucket
+            #
+            # The 13 page kademlia paper specifies that the least recently contacted node in the bucket
+            # shall be pinged. If it fails to reply it is replaced with the new contact. If the ping is successful
+            # the new contact is ignored and not added to the bucket (sections 2.2 and 2.4).
+            #
+            # A reasonable extension to this is BEP 0005, which extends the above:
+            #
+            #    Not all nodes that we learn about are equal. Some are "good" and some are not.
+            #    Many nodes using the DHT are able to send queries and receive responses,
+            #    but are not able to respond to queries from other nodes. It is important that
+            #    each node's routing table must contain only known good nodes. A good node is
+            #    a node has responded to one of our queries within the last 15 minutes. A node
+            #    is also good if it has ever responded to one of our queries and has sent us a
+            #    query within the last 15 minutes. After 15 minutes of inactivity, a node becomes
+            #    questionable. Nodes become bad when they fail to respond to multiple queries
+            #    in a row. Nodes that we know are good are given priority over nodes with unknown status.
+            #
+            # When there are bad or questionable nodes in the bucket, the least recent is selected for
+            # potential replacement (BEP 0005). When all nodes in the bucket are fresh, the head (least recent)
+            # contact is selected as described in section 2.2 of the kademlia paper. In both cases the new contact
+            # is ignored if the pinged node replies.
+
+            not_good_contacts = self.buckets[bucket_index].get_bad_or_unknown_peers()
+            not_recently_replied = []
+            for my_peer in not_good_contacts:
+                last_replied = self._peer_manager.get_last_replied(my_peer.address, my_peer.udp_port)
+                if not last_replied or last_replied + 60 < self._loop.time():
+                    not_recently_replied.append(my_peer)
+            if not_recently_replied:
+                to_replace = not_recently_replied[0]
+            else:
+                to_replace = self.buckets[bucket_index].peers[0]
+                last_replied = self._peer_manager.get_last_replied(to_replace.address, to_replace.udp_port)
+                if last_replied and last_replied + 60 > self._loop.time():
+                    return False
+            log.debug("pinging %s:%s", to_replace.address, to_replace.udp_port)
+            try:
+                await probe(to_replace)
+                return False
+            except (asyncio.TimeoutError, RemoteException):
+                log.debug("Replacing dead contact in bucket %i: %s:%i with %s:%i ", bucket_index,
+                          to_replace.address, to_replace.udp_port, peer.address, peer.udp_port)
+                if to_replace in self.buckets[bucket_index]:
+                    self.buckets[bucket_index].remove_peer(to_replace)
+                return await self.add_peer(peer, probe)

lbry/dht/serialization/bencoding.py

@@ -52,8 +52,7 @@ def _bdecode(data: bytes, start_index: int = 0) -> typing.Tuple[typing.Union[int
             raise DecodeError(err)
         start_index = split_pos + 1
         end_pos = start_index + length
-        b = data[start_index:end_pos]
-        return b, end_pos
+        return data[start_index:end_pos], end_pos
 
 
 def bencode(data: typing.Dict) -> bytes:

lbry/dht/serialization/datagram.py

@@ -7,9 +7,13 @@ REQUEST_TYPE = 0
 RESPONSE_TYPE = 1
 ERROR_TYPE = 2
 
+OPTIONAL_ARG_OFFSET = 100
+
 # bencode representation of argument keys
 PAGE_KEY = b'p'
 
+OPTIONAL_FIELDS = ()
+
 
 class KademliaDatagramBase:
     """
@@ -18,7 +22,7 @@ class KademliaDatagramBase:
     these correspond to the packet_type, rpc_id, and node_id args
     """
 
-    fields = [
+    required_fields = [
         'packet_type',
         'rpc_id',
         'node_id'
@@ -30,21 +34,26 @@ class KademliaDatagramBase:
         self.packet_type = packet_type
         if self.expected_packet_type != packet_type:
             raise ValueError(f"invalid packet type: {packet_type}, expected {self.expected_packet_type}")
-        if len(rpc_id) != constants.rpc_id_length:
+        if len(rpc_id) != constants.RPC_ID_LENGTH:
             raise ValueError(f"invalid rpc node_id: {len(rpc_id)} bytes (expected 20)")
-        if not len(node_id) == constants.hash_length:
+        if not len(node_id) == constants.HASH_LENGTH:
             raise ValueError(f"invalid node node_id: {len(node_id)} bytes (expected 48)")
         self.rpc_id = rpc_id
         self.node_id = node_id
 
     def bencode(self) -> bytes:
-        return bencode({
-           i: getattr(self, k) for i, k in enumerate(self.fields)
-        })
+        datagram = {
+            i: getattr(self, k) for i, k in enumerate(self.required_fields)
+        }
+        for i, k in enumerate(OPTIONAL_FIELDS):
+            value = getattr(self, k, None)
+            if value is not None:
+                datagram[i + OPTIONAL_ARG_OFFSET] = value
+        return bencode(datagram)
 
 
 class RequestDatagram(KademliaDatagramBase):
-    fields = [
+    required_fields = [
         'packet_type',
         'rpc_id',
         'node_id',
@@ -68,18 +77,18 @@ class RequestDatagram(KademliaDatagramBase):
 
     @classmethod
     def make_ping(cls, from_node_id: bytes, rpc_id: typing.Optional[bytes] = None) -> 'RequestDatagram':
-        rpc_id = rpc_id or constants.generate_id()[:constants.rpc_id_length]
+        rpc_id = rpc_id or constants.generate_id()[:constants.RPC_ID_LENGTH]
         return cls(REQUEST_TYPE, rpc_id, from_node_id, b'ping')
 
     @classmethod
     def make_store(cls, from_node_id: bytes, blob_hash: bytes, token: bytes, port: int,
                    rpc_id: typing.Optional[bytes] = None) -> 'RequestDatagram':
-        rpc_id = rpc_id or constants.generate_id()[:constants.rpc_id_length]
-        if len(blob_hash) != constants.hash_bits // 8:
+        rpc_id = rpc_id or constants.generate_id()[:constants.RPC_ID_LENGTH]
+        if len(blob_hash) != constants.HASH_BITS // 8:
             raise ValueError(f"invalid blob hash length: {len(blob_hash)}")
         if not 0 < port < 65536:
             raise ValueError(f"invalid port: {port}")
-        if len(token) != constants.hash_bits // 8:
+        if len(token) != constants.HASH_BITS // 8:
             raise ValueError(f"invalid token length: {len(token)}")
         store_args = [blob_hash, token, port, from_node_id, 0]
         return cls(REQUEST_TYPE, rpc_id, from_node_id, b'store', store_args)
@@ -87,16 +96,16 @@ class RequestDatagram(KademliaDatagramBase):
     @classmethod
     def make_find_node(cls, from_node_id: bytes, key: bytes,
                        rpc_id: typing.Optional[bytes] = None) -> 'RequestDatagram':
-        rpc_id = rpc_id or constants.generate_id()[:constants.rpc_id_length]
-        if len(key) != constants.hash_bits // 8:
+        rpc_id = rpc_id or constants.generate_id()[:constants.RPC_ID_LENGTH]
+        if len(key) != constants.HASH_BITS // 8:
             raise ValueError(f"invalid key length: {len(key)}")
         return cls(REQUEST_TYPE, rpc_id, from_node_id, b'findNode', [key])
 
     @classmethod
     def make_find_value(cls, from_node_id: bytes, key: bytes,
                         rpc_id: typing.Optional[bytes] = None, page: int = 0) -> 'RequestDatagram':
-        rpc_id = rpc_id or constants.generate_id()[:constants.rpc_id_length]
-        if len(key) != constants.hash_bits // 8:
+        rpc_id = rpc_id or constants.generate_id()[:constants.RPC_ID_LENGTH]
+        if len(key) != constants.HASH_BITS // 8:
             raise ValueError(f"invalid key length: {len(key)}")
         if page < 0:
             raise ValueError(f"cannot request a negative page ({page})")
@@ -104,7 +113,7 @@ class RequestDatagram(KademliaDatagramBase):
 
 
 class ResponseDatagram(KademliaDatagramBase):
-    fields = [
+    required_fields = [
         'packet_type',
         'rpc_id',
         'node_id',
@@ -119,7 +128,7 @@ class ResponseDatagram(KademliaDatagramBase):
 
 
 class ErrorDatagram(KademliaDatagramBase):
-    fields = [
+    required_fields = [
         'packet_type',
         'rpc_id',
         'node_id',
@@ -135,7 +144,7 @@ class ErrorDatagram(KademliaDatagramBase):
         self.response = response.decode()
 
 
-def decode_datagram(datagram: bytes) -> typing.Union[RequestDatagram, ResponseDatagram, ErrorDatagram]:
+def _decode_datagram(datagram: bytes):
     msg_types = {
         REQUEST_TYPE: RequestDatagram,
         RESPONSE_TYPE: ResponseDatagram,
@@ -143,23 +152,36 @@ def decode_datagram(datagram: bytes) -> typing.Union[RequestDatagram, ResponseDa
     }
 
     primitive: typing.Dict = bdecode(datagram)
-    if primitive[0] in [REQUEST_TYPE, ERROR_TYPE, RESPONSE_TYPE]:  # pylint: disable=unsubscriptable-object
-        datagram_type = primitive[0]  # pylint: disable=unsubscriptable-object
+
+    converted = {
+        str(k).encode() if not isinstance(k, bytes) else k: v for k, v in primitive.items()
+    }
+
+    if converted[b'0'] in [REQUEST_TYPE, ERROR_TYPE, RESPONSE_TYPE]:  # pylint: disable=unsubscriptable-object
+        datagram_type = converted[b'0']  # pylint: disable=unsubscriptable-object
     else:
         raise ValueError("invalid datagram type")
     datagram_class = msg_types[datagram_type]
-    return datagram_class(**{
-            k: primitive[i]  # pylint: disable=unsubscriptable-object
-            for i, k in enumerate(datagram_class.fields)
-            if i in primitive  # pylint: disable=unsupported-membership-test
-        }
-    )
+    decoded = {
+        k: converted[str(i).encode()]  # pylint: disable=unsubscriptable-object
+        for i, k in enumerate(datagram_class.required_fields)
+        if str(i).encode() in converted  # pylint: disable=unsupported-membership-test
+    }
+    for i, _ in enumerate(OPTIONAL_FIELDS):
+        if str(i + OPTIONAL_ARG_OFFSET).encode() in converted:
+            decoded[i + OPTIONAL_ARG_OFFSET] = converted[str(i + OPTIONAL_ARG_OFFSET).encode()]
+    return decoded, datagram_class
+
+
+def decode_datagram(datagram: bytes) -> typing.Union[RequestDatagram, ResponseDatagram, ErrorDatagram]:
+    decoded, datagram_class = _decode_datagram(datagram)
+    return datagram_class(**decoded)
 
 
 def make_compact_ip(address: str) -> bytearray:
     compact_ip = reduce(lambda buff, x: buff + bytearray([int(x)]), address.split('.'), bytearray())
     if len(compact_ip) != 4:
-        raise ValueError(f"invalid IPv4 length")
+        raise ValueError("invalid IPv4 length")
     return compact_ip
 
 
@@ -167,8 +189,8 @@ def make_compact_address(node_id: bytes, address: str, port: int) -> bytearray:
     compact_ip = make_compact_ip(address)
     if not 0 < port < 65536:
         raise ValueError(f'Invalid port: {port}')
-    if len(node_id) != constants.hash_bits // 8:
-        raise ValueError(f"invalid node node_id length")
+    if len(node_id) != constants.HASH_BITS // 8:
+        raise ValueError("invalid node node_id length")
     return compact_ip + port.to_bytes(2, 'big') + node_id
 
 
@@ -178,6 +200,6 @@ def decode_compact_address(compact_address: bytes) -> typing.Tuple[bytes, str, i
     node_id = compact_address[6:]
     if not 0 < port < 65536:
         raise ValueError(f'Invalid port: {port}')
-    if len(node_id) != constants.hash_bits // 8:
-        raise ValueError(f"invalid node node_id length")
+    if len(node_id) != constants.HASH_BITS // 8:
+        raise ValueError("invalid node node_id length")
     return node_id, address, port

lbry/error/Makefile

@@ -0,0 +1,5 @@
+generate:
+	python generate.py generate > __init__.py
+
+analyze:
+	python generate.py analyze

lbry/error/README.md

@@ -0,0 +1,95 @@
+# Exceptions
+
+Exceptions in LBRY are defined and generated from the Markdown table at the end of this README.
+
+## Guidelines
+
+When possible, use [built-in Python exceptions](https://docs.python.org/3/library/exceptions.html) or `aiohttp` [general client](https://docs.aiohttp.org/en/latest/client_reference.html#client-exceptions) / [HTTP](https://docs.aiohttp.org/en/latest/web_exceptions.html) exceptions, unless:
+1. You want to provide a better error message (extend the closest built-in/`aiohttp` exception in this case).
+2. You need to represent a new situation.
+
+When defining your own exceptions, consider:
+1. Extending a built-in Python or `aiohttp` exception.
+2. Using contextual variables in the error message.
+
+## Table Column Definitions
+
+Column | Meaning
+---|---
+Code | Codes are used only to define the hierarchy of exceptions and do not end up in the generated output, it is okay to re-number things as necessary at anytime to achieve the desired hierarchy.
+Name | Becomes the class name of the exception with "Error" appended to the end. Changing names of existing exceptions makes the API backwards incompatible. When extending other exceptions you must specify the full class name, manually adding "Error" as necessary (if extending another SDK exception).
+Message | User friendly error message explaining the exceptional event. Supports Python formatted strings: any variables used in the string will be generated as arguments in the `__init__` method. Use `--` to provide a doc string after the error message to be added to the class definition.
+
+## Exceptions Table
+
+Code | Name | Message
+---:|---|---
+**1xx** | UserInput | User input errors.
+**10x** | Command | Errors preparing to execute commands.
+101 | CommandDoesNotExist | Command '{command}' does not exist.
+102 | CommandDeprecated | Command '{command}' is deprecated.
+103 | CommandInvalidArgument | Invalid argument '{argument}' to command '{command}'.
+104 | CommandTemporarilyUnavailable | Command '{command}' is temporarily unavailable. -- Such as waiting for required components to start.
+105 | CommandPermanentlyUnavailable | Command '{command}' is permanently unavailable. -- such as when required component was intentionally configured not to start.
+**11x** | InputValue(ValueError) | Invalid argument value provided to command.
+111 | GenericInputValue | The value '{value}' for argument '{argument}' is not valid.
+112 | InputValueIsNone | None or null is not valid value for argument '{argument}'.
+113 | ConflictingInputValue | Only '{first_argument}' or '{second_argument}' is allowed, not both.
+114 | InputStringIsBlank | {argument} cannot be blank.
+115 | EmptyPublishedFile | Cannot publish empty file: {file_path}
+116 | MissingPublishedFile | File does not exist: {file_path}
+117 | InvalidStreamURL | Invalid LBRY stream URL: '{url}' -- When an URL cannot be downloaded, such as '@Channel/' or a collection
+**2xx** | Configuration | Configuration errors.
+201 | ConfigWrite | Cannot write configuration file '{path}'. -- When writing the default config fails on startup, such as due to permission issues.
+202 | ConfigRead | Cannot find provided configuration file '{path}'. -- Can't open the config file user provided via command line args.
+203 | ConfigParse | Failed to parse the configuration file '{path}'. -- Includes the syntax error / line number to help user fix it.
+204 | ConfigMissing | Configuration file '{path}' is missing setting that has no default / fallback.
+205 | ConfigInvalid | Configuration file '{path}' has setting with invalid value.
+**3xx** | Network | **Networking**
+301 | NoInternet | No internet connection.
+302 | NoUPnPSupport | Router does not support UPnP.
+**4xx** | Wallet | **Wallet Errors**
+401 | TransactionRejected | Transaction rejected, unknown reason.
+402 | TransactionFeeTooLow | Fee too low.
+403 | TransactionInvalidSignature | Invalid signature.
+404 | InsufficientFunds |  Not enough funds to cover this transaction. -- determined by wallet prior to attempting to broadcast a tx; this is different for example from a TX being created and sent but then rejected by lbrycrd for unspendable utxos.
+405 | ChannelKeyNotFound | Channel signing key not found.
+406 | ChannelKeyInvalid | Channel signing key is out of date. -- For example, channel was updated but you don't have the updated key.
+407 | DataDownload | Failed to download blob. *generic*
+408 | PrivateKeyNotFound | Couldn't find private key for {key} '{value}'.
+410 | Resolve | Failed to resolve '{url}'.
+411 | ResolveTimeout | Failed to resolve '{url}' within the timeout.
+411 | ResolveCensored | Resolve of '{url}' was censored by channel with claim id '{censor_id}'.
+420 | KeyFeeAboveMaxAllowed | {message}
+421 | InvalidPassword | Password is invalid.
+422 | IncompatibleWalletServer | '{server}:{port}' has an incompatibly old version.
+423 | TooManyClaimSearchParameters | {key} cant have more than {limit} items.
+424 | AlreadyPurchased | You already have a purchase for claim_id '{claim_id_hex}'. Use --allow-duplicate-purchase flag to override.
+431 | ServerPaymentInvalidAddress | Invalid address from wallet server: '{address}' - skipping payment round.
+432 | ServerPaymentWalletLocked | Cannot spend funds with locked wallet, skipping payment round.
+433 | ServerPaymentFeeAboveMaxAllowed | Daily server fee of {daily_fee} exceeds maximum configured of {max_fee} LBC.
+434 | WalletNotLoaded | Wallet {wallet_id} is not loaded.
+435 | WalletAlreadyLoaded | Wallet {wallet_path} is already loaded.
+436 | WalletNotFound | Wallet not found at {wallet_path}.
+437 | WalletAlreadyExists | Wallet {wallet_path} already exists, use `wallet_add` to load it.
+**5xx** | Blob | **Blobs**
+500 | BlobNotFound | Blob not found.
+501 | BlobPermissionDenied | Permission denied to read blob.
+502 | BlobTooBig | Blob is too big.
+503 | BlobEmpty | Blob is empty.
+510 | BlobFailedDecryption | Failed to decrypt blob.
+511 | CorruptBlob | Blobs is corrupted.
+520 | BlobFailedEncryption | Failed to encrypt blob.
+531 | DownloadCancelled | Download was canceled.
+532 | DownloadSDTimeout | Failed to download sd blob {download} within timeout.
+533 | DownloadDataTimeout | Failed to download data blobs for sd hash {download} within timeout.
+534 | InvalidStreamDescriptor | {message}
+535 | InvalidData | {message}
+536 | InvalidBlobHash | {message}
+**6xx** | Component | **Components**
+601 | ComponentStartConditionNotMet | Unresolved dependencies for: {components}
+602 | ComponentsNotStarted | {message}
+**7xx** | CurrencyExchange | **Currency Exchange**
+701 | InvalidExchangeRateResponse | Failed to get exchange rate from {source}: {reason}
+702 | CurrencyConversion | {message}
+703 | InvalidCurrency | Invalid currency: {currency} is not a supported currency.

lbry/error/__init__.py

@@ -0,0 +1,494 @@
+from .base import BaseError, claim_id
+
+
+class UserInputError(BaseError):
+    """
+    User input errors.
+    """
+
+
+class CommandError(UserInputError):
+    """
+    Errors preparing to execute commands.
+    """
+
+
+class CommandDoesNotExistError(CommandError):
+
+    def __init__(self, command):
+        self.command = command
+        super().__init__(f"Command '{command}' does not exist.")
+
+
+class CommandDeprecatedError(CommandError):
+
+    def __init__(self, command):
+        self.command = command
+        super().__init__(f"Command '{command}' is deprecated.")
+
+
+class CommandInvalidArgumentError(CommandError):
+
+    def __init__(self, argument, command):
+        self.argument = argument
+        self.command = command
+        super().__init__(f"Invalid argument '{argument}' to command '{command}'.")
+
+
+class CommandTemporarilyUnavailableError(CommandError):
+    """
+    Such as waiting for required components to start.
+    """
+
+    def __init__(self, command):
+        self.command = command
+        super().__init__(f"Command '{command}' is temporarily unavailable.")
+
+
+class CommandPermanentlyUnavailableError(CommandError):
+    """
+    such as when required component was intentionally configured not to start.
+    """
+
+    def __init__(self, command):
+        self.command = command
+        super().__init__(f"Command '{command}' is permanently unavailable.")
+
+
+class InputValueError(UserInputError, ValueError):
+    """
+    Invalid argument value provided to command.
+    """
+
+
+class GenericInputValueError(InputValueError):
+
+    def __init__(self, value, argument):
+        self.value = value
+        self.argument = argument
+        super().__init__(f"The value '{value}' for argument '{argument}' is not valid.")
+
+
+class InputValueIsNoneError(InputValueError):
+
+    def __init__(self, argument):
+        self.argument = argument
+        super().__init__(f"None or null is not valid value for argument '{argument}'.")
+
+
+class ConflictingInputValueError(InputValueError):
+
+    def __init__(self, first_argument, second_argument):
+        self.first_argument = first_argument
+        self.second_argument = second_argument
+        super().__init__(f"Only '{first_argument}' or '{second_argument}' is allowed, not both.")
+
+
+class InputStringIsBlankError(InputValueError):
+
+    def __init__(self, argument):
+        self.argument = argument
+        super().__init__(f"{argument} cannot be blank.")
+
+
+class EmptyPublishedFileError(InputValueError):
+
+    def __init__(self, file_path):
+        self.file_path = file_path
+        super().__init__(f"Cannot publish empty file: {file_path}")
+
+
+class MissingPublishedFileError(InputValueError):
+
+    def __init__(self, file_path):
+        self.file_path = file_path
+        super().__init__(f"File does not exist: {file_path}")
+
+
+class InvalidStreamURLError(InputValueError):
+    """
+    When an URL cannot be downloaded, such as '@Channel/' or a collection
+    """
+
+    def __init__(self, url):
+        self.url = url
+        super().__init__(f"Invalid LBRY stream URL: '{url}'")
+
+
+class ConfigurationError(BaseError):
+    """
+    Configuration errors.
+    """
+
+
+class ConfigWriteError(ConfigurationError):
+    """
+    When writing the default config fails on startup, such as due to permission issues.
+    """
+
+    def __init__(self, path):
+        self.path = path
+        super().__init__(f"Cannot write configuration file '{path}'.")
+
+
+class ConfigReadError(ConfigurationError):
+    """
+    Can't open the config file user provided via command line args.
+    """
+
+    def __init__(self, path):
+        self.path = path
+        super().__init__(f"Cannot find provided configuration file '{path}'.")
+
+
+class ConfigParseError(ConfigurationError):
+    """
+    Includes the syntax error / line number to help user fix it.
+    """
+
+    def __init__(self, path):
+        self.path = path
+        super().__init__(f"Failed to parse the configuration file '{path}'.")
+
+
+class ConfigMissingError(ConfigurationError):
+
+    def __init__(self, path):
+        self.path = path
+        super().__init__(f"Configuration file '{path}' is missing setting that has no default / fallback.")
+
+
+class ConfigInvalidError(ConfigurationError):
+
+    def __init__(self, path):
+        self.path = path
+        super().__init__(f"Configuration file '{path}' has setting with invalid value.")
+
+
+class NetworkError(BaseError):
+    """
+    **Networking**
+    """
+
+
+class NoInternetError(NetworkError):
+
+    def __init__(self):
+        super().__init__("No internet connection.")
+
+
+class NoUPnPSupportError(NetworkError):
+
+    def __init__(self):
+        super().__init__("Router does not support UPnP.")
+
+
+class WalletError(BaseError):
+    """
+    **Wallet Errors**
+    """
+
+
+class TransactionRejectedError(WalletError):
+
+    def __init__(self):
+        super().__init__("Transaction rejected, unknown reason.")
+
+
+class TransactionFeeTooLowError(WalletError):
+
+    def __init__(self):
+        super().__init__("Fee too low.")
+
+
+class TransactionInvalidSignatureError(WalletError):
+
+    def __init__(self):
+        super().__init__("Invalid signature.")
+
+
+class InsufficientFundsError(WalletError):
+    """
+    determined by wallet prior to attempting to broadcast a tx; this is different for example from a TX
+    being created and sent but then rejected by lbrycrd for unspendable utxos.
+    """
+
+    def __init__(self):
+        super().__init__("Not enough funds to cover this transaction.")
+
+
+class ChannelKeyNotFoundError(WalletError):
+
+    def __init__(self):
+        super().__init__("Channel signing key not found.")
+
+
+class ChannelKeyInvalidError(WalletError):
+    """
+    For example, channel was updated but you don't have the updated key.
+    """
+
+    def __init__(self):
+        super().__init__("Channel signing key is out of date.")
+
+
+class DataDownloadError(WalletError):
+
+    def __init__(self):
+        super().__init__("Failed to download blob. *generic*")
+
+
+class PrivateKeyNotFoundError(WalletError):
+
+    def __init__(self, key, value):
+        self.key = key
+        self.value = value
+        super().__init__(f"Couldn't find private key for {key} '{value}'.")
+
+
+class ResolveError(WalletError):
+
+    def __init__(self, url):
+        self.url = url
+        super().__init__(f"Failed to resolve '{url}'.")
+
+
+class ResolveTimeoutError(WalletError):
+
+    def __init__(self, url):
+        self.url = url
+        super().__init__(f"Failed to resolve '{url}' within the timeout.")
+
+
+class ResolveCensoredError(WalletError):
+
+    def __init__(self, url, censor_id, censor_row):
+        self.url = url
+        self.censor_id = censor_id
+        self.censor_row = censor_row
+        super().__init__(f"Resolve of '{url}' was censored by channel with claim id '{censor_id}'.")
+
+
+class KeyFeeAboveMaxAllowedError(WalletError):
+
+    def __init__(self, message):
+        self.message = message
+        super().__init__(f"{message}")
+
+
+class InvalidPasswordError(WalletError):
+
+    def __init__(self):
+        super().__init__("Password is invalid.")
+
+
+class IncompatibleWalletServerError(WalletError):
+
+    def __init__(self, server, port):
+        self.server = server
+        self.port = port
+        super().__init__(f"'{server}:{port}' has an incompatibly old version.")
+
+
+class TooManyClaimSearchParametersError(WalletError):
+
+    def __init__(self, key, limit):
+        self.key = key
+        self.limit = limit
+        super().__init__(f"{key} cant have more than {limit} items.")
+
+
+class AlreadyPurchasedError(WalletError):
+    """
+    allow-duplicate-purchase flag to override.
+    """
+
+    def __init__(self, claim_id_hex):
+        self.claim_id_hex = claim_id_hex
+        super().__init__(f"You already have a purchase for claim_id '{claim_id_hex}'. Use")
+
+
+class ServerPaymentInvalidAddressError(WalletError):
+
+    def __init__(self, address):
+        self.address = address
+        super().__init__(f"Invalid address from wallet server: '{address}' - skipping payment round.")
+
+
+class ServerPaymentWalletLockedError(WalletError):
+
+    def __init__(self):
+        super().__init__("Cannot spend funds with locked wallet, skipping payment round.")
+
+
+class ServerPaymentFeeAboveMaxAllowedError(WalletError):
+
+    def __init__(self, daily_fee, max_fee):
+        self.daily_fee = daily_fee
+        self.max_fee = max_fee
+        super().__init__(f"Daily server fee of {daily_fee} exceeds maximum configured of {max_fee} LBC.")
+
+
+class WalletNotLoadedError(WalletError):
+
+    def __init__(self, wallet_id):
+        self.wallet_id = wallet_id
+        super().__init__(f"Wallet {wallet_id} is not loaded.")
+
+
+class WalletAlreadyLoadedError(WalletError):
+
+    def __init__(self, wallet_path):
+        self.wallet_path = wallet_path
+        super().__init__(f"Wallet {wallet_path} is already loaded.")
+
+
+class WalletNotFoundError(WalletError):
+
+    def __init__(self, wallet_path):
+        self.wallet_path = wallet_path
+        super().__init__(f"Wallet not found at {wallet_path}.")
+
+
+class WalletAlreadyExistsError(WalletError):
+
+    def __init__(self, wallet_path):
+        self.wallet_path = wallet_path
+        super().__init__(f"Wallet {wallet_path} already exists, use `wallet_add` to load it.")
+
+
+class BlobError(BaseError):
+    """
+    **Blobs**
+    """
+
+
+class BlobNotFoundError(BlobError):
+
+    def __init__(self):
+        super().__init__("Blob not found.")
+
+
+class BlobPermissionDeniedError(BlobError):
+
+    def __init__(self):
+        super().__init__("Permission denied to read blob.")
+
+
+class BlobTooBigError(BlobError):
+
+    def __init__(self):
+        super().__init__("Blob is too big.")
+
+
+class BlobEmptyError(BlobError):
+
+    def __init__(self):
+        super().__init__("Blob is empty.")
+
+
+class BlobFailedDecryptionError(BlobError):
+
+    def __init__(self):
+        super().__init__("Failed to decrypt blob.")
+
+
+class CorruptBlobError(BlobError):
+
+    def __init__(self):
+        super().__init__("Blobs is corrupted.")
+
+
+class BlobFailedEncryptionError(BlobError):
+
+    def __init__(self):
+        super().__init__("Failed to encrypt blob.")
+
+
+class DownloadCancelledError(BlobError):
+
+    def __init__(self):
+        super().__init__("Download was canceled.")
+
+
+class DownloadSDTimeoutError(BlobError):
+
+    def __init__(self, download):
+        self.download = download
+        super().__init__(f"Failed to download sd blob {download} within timeout.")
+
+
+class DownloadDataTimeoutError(BlobError):
+
+    def __init__(self, download):
+        self.download = download
+        super().__init__(f"Failed to download data blobs for sd hash {download} within timeout.")
+
+
+class InvalidStreamDescriptorError(BlobError):
+
+    def __init__(self, message):
+        self.message = message
+        super().__init__(f"{message}")
+
+
+class InvalidDataError(BlobError):
+
+    def __init__(self, message):
+        self.message = message
+        super().__init__(f"{message}")
+
+
+class InvalidBlobHashError(BlobError):
+
+    def __init__(self, message):
+        self.message = message
+        super().__init__(f"{message}")
+
+
+class ComponentError(BaseError):
+    """
+    **Components**
+    """
+
+
+class ComponentStartConditionNotMetError(ComponentError):
+
+    def __init__(self, components):
+        self.components = components
+        super().__init__(f"Unresolved dependencies for: {components}")
+
+
+class ComponentsNotStartedError(ComponentError):
+
+    def __init__(self, message):
+        self.message = message
+        super().__init__(f"{message}")
+
+
+class CurrencyExchangeError(BaseError):
+    """
+    **Currency Exchange**
+    """
+
+
+class InvalidExchangeRateResponseError(CurrencyExchangeError):
+
+    def __init__(self, source, reason):
+        self.source = source
+        self.reason = reason
+        super().__init__(f"Failed to get exchange rate from {source}: {reason}")
+
+
+class CurrencyConversionError(CurrencyExchangeError):
+
+    def __init__(self, message):
+        self.message = message
+        super().__init__(f"{message}")
+
+
+class InvalidCurrencyError(CurrencyExchangeError):
+
+    def __init__(self, currency):
+        self.currency = currency
+        super().__init__(f"Invalid currency: {currency} is not a supported currency.")

lbry/error/base.py

@@ -0,0 +1,9 @@
+from binascii import hexlify
+
+
+def claim_id(claim_hash):
+    return hexlify(claim_hash[::-1]).decode()
+
+
+class BaseError(Exception):
+    pass

lbry/error/generate.py

@@ -0,0 +1,167 @@
+import re
+import sys
+import argparse
+from pathlib import Path
+from textwrap import fill, indent
+
+
+INDENT = ' ' * 4
+
+CLASS = """
+
+class {name}({parents}):{doc}
+"""
+
+INIT = """
+    def __init__({args}):{fields}
+        super().__init__({format}"{message}")
+"""
+
+FUNCTIONS = ['claim_id']
+
+
+class ErrorClass:
+
+    def __init__(self, hierarchy, name, message):
+        self.hierarchy = hierarchy.replace('**', '')
+        self.other_parents = []
+        if '(' in name:
+            assert ')' in name, f"Missing closing parenthesis in '{name}'."
+            self.other_parents = name[name.find('(')+1:name.find(')')].split(',')
+            name = name[:name.find('(')]
+        self.name = name
+        self.class_name = name+'Error'
+        self.message = message
+        self.comment = ""
+        if '--' in message:
+            self.message, self.comment = message.split('--')
+        self.message = self.message.strip()
+        self.comment = self.comment.strip()
+
+    @property
+    def is_leaf(self):
+        return 'x' not in self.hierarchy
+
+    @property
+    def code(self):
+        return self.hierarchy.replace('x', '')
+
+    @property
+    def parent_codes(self):
+        return self.hierarchy[0:2], self.hierarchy[0]
+
+    def get_arguments(self):
+        args = ['self']
+        for arg in re.findall('{([a-z0-1_()]+)}', self.message):
+            for func in FUNCTIONS:
+                if arg.startswith(f'{func}('):
+                    arg = arg[len(f'{func}('):-1]
+                    break
+            args.append(arg)
+        return args
+
+    @staticmethod
+    def get_fields(args):
+        if len(args) > 1:
+            return ''.join(f'\n{INDENT*2}self.{field} = {field}' for field in args[1:])
+        return ''
+
+    @staticmethod
+    def get_doc_string(doc):
+        if doc:
+            return f'\n{INDENT}"""\n{indent(fill(doc, 100), INDENT)}\n{INDENT}"""'
+        return ""
+
+    def render(self, out, parent):
+        if not parent:
+            parents = ['BaseError']
+        else:
+            parents = [parent.class_name]
+        parents += self.other_parents
+        args = self.get_arguments()
+        if self.is_leaf:
+            out.write((CLASS + INIT).format(
+                name=self.class_name, parents=', '.join(parents),
+                args=', '.join(args), fields=self.get_fields(args),
+                message=self.message, doc=self.get_doc_string(self.comment), format='f' if len(args) > 1 else ''
+            ))
+        else:
+            out.write(CLASS.format(
+                name=self.class_name, parents=', '.join(parents),
+                doc=self.get_doc_string(self.comment or self.message)
+            ))
+
+
+def get_errors():
+    with open('README.md', 'r') as readme:
+        lines = iter(readme.readlines())
+        for line in lines:
+            if line.startswith('## Exceptions Table'):
+                break
+        for line in lines:
+            if line.startswith('---:|'):
+                break
+        for line in lines:
+            if not line:
+                break
+            yield ErrorClass(*[c.strip() for c in line.split('|')])
+
+
+def find_parent(stack, child):
+    for parent_code in child.parent_codes:
+        parent = stack.get(parent_code)
+        if parent:
+            return parent
+
+
+def generate(out):
+    out.write(f"from .base import BaseError, {', '.join(FUNCTIONS)}\n")
+    stack = {}
+    for error in get_errors():
+        error.render(out, find_parent(stack, error))
+        if not error.is_leaf:
+            assert error.code not in stack, f"Duplicate code: {error.code}"
+            stack[error.code] = error
+
+
+def analyze():
+    errors = {e.class_name: [] for e in get_errors() if e.is_leaf}
+    here = Path(__file__).absolute().parents[0]
+    module = here.parent
+    for file_path in module.glob('**/*.py'):
+        if here in file_path.parents:
+            continue
+        with open(file_path) as src_file:
+            src = src_file.read()
+            for error in errors.keys():
+                found = src.count(error)
+                if found > 0:
+                    errors[error].append((file_path, found))
+
+    print('Unused Errors:\n')
+    for error, used in errors.items():
+        if used:
+            print(f' - {error}')
+            for use in used:
+                print(f'   {use[0].relative_to(module.parent)} {use[1]}')
+            print('')
+
+    print('')
+    print('Unused Errors:')
+    for error, used in errors.items():
+        if not used:
+            print(f' - {error}')
+
+
+def main():
+    parser = argparse.ArgumentParser()
+    parser.add_argument("action", choices=['generate', 'analyze'])
+    args = parser.parse_args()
+    if args.action == "analyze":
+        analyze()
+    elif args.action == "generate":
+        generate(sys.stdout)
+
+
+if __name__ == "__main__":
+    main()

lbry/extras/cli.py

@@ -1,5 +1,6 @@
 import os
 import sys
+import shutil
 import signal
 import pathlib
 import json
@@ -13,9 +14,8 @@ from aiohttp.web import GracefulExit
 from docopt import docopt
 
 from lbry import __version__ as lbrynet_version
-from lbry.extras.daemon.loggly_handler import get_loggly_handler
+from lbry.extras.daemon.daemon import Daemon
 from lbry.conf import Config, CLIConfig
-from lbry.extras.daemon.Daemon import Daemon
 
 log = logging.getLogger('lbry')
 
@@ -101,7 +101,7 @@ class ArgumentParser(argparse.ArgumentParser):
         self._optionals.title = 'Options'
         if group_name is None:
             self.epilog = (
-                f"Run 'lbrynet COMMAND --help' for more information on a command or group."
+                "Run 'lbrynet COMMAND --help' for more information on a command or group."
             )
         else:
             self.epilog = (
@@ -167,16 +167,16 @@ def add_command_parser(parent, command):
 
 
 def get_argument_parser():
-    main = ArgumentParser(
+    root = ArgumentParser(
         'lbrynet', description='An interface to the LBRY Network.', allow_abbrev=False,
     )
-    main.add_argument(
+    root.add_argument(
         '-v', '--version', dest='cli_version', action="store_true",
         help='Show lbrynet CLI version and exit.'
     )
-    main.set_defaults(group=None, command=None)
-    CLIConfig.contribute_to_argparse(main)
-    sub = main.add_subparsers(metavar='COMMAND')
+    root.set_defaults(group=None, command=None)
+    CLIConfig.contribute_to_argparse(root)
+    sub = root.add_subparsers(metavar='COMMAND')
     start = sub.add_parser(
         'start',
         usage='lbrynet start [--config FILE] [--data-dir DIR] [--wallet-dir DIR] [--download-dir DIR] ...',
@@ -186,11 +186,19 @@ def get_argument_parser():
         '--quiet', dest='quiet', action="store_true",
         help='Disable all console output.'
     )
+    start.add_argument(
+        '--no-logging', dest='no_logging', action="store_true",
+        help='Disable all logging of any kind.'
+    )
     start.add_argument(
         '--verbose', nargs="*",
         help=('Enable debug output for lbry logger and event loop. Optionally specify loggers for which debug output '
               'should selectively be applied.')
     )
+    start.add_argument(
+        '--initial-headers', dest='initial_headers',
+        help='Specify path to initial blockchain headers, faster than downloading them on first run.'
+    )
     Config.contribute_to_argparse(start)
     start.set_defaults(command='start', start_parser=start, doc=start.format_help())
 
@@ -212,15 +220,18 @@ def get_argument_parser():
         else:
             add_command_parser(groups[command['group']], command)
 
-    return main
+    return root
 
 
 def ensure_directory_exists(path: str):
     if not os.path.isdir(path):
         pathlib.Path(path).mkdir(parents=True, exist_ok=True)
+    use_effective_ids = os.access in os.supports_effective_ids
+    if not os.access(path, os.W_OK, effective_ids=use_effective_ids):
+        raise PermissionError(f"The following directory is not writable: {path}")
 
 
-LOG_MODULES = ('lbry', 'torba', 'aioupnp')
+LOG_MODULES = 'lbry', 'aioupnp'
 
 
 def setup_logging(logger: logging.Logger, args: argparse.Namespace, conf: Config):
@@ -236,7 +247,6 @@ def setup_logging(logger: logging.Logger, args: argparse.Namespace, conf: Config
             logger.getChild(module_name).addHandler(handler)
 
     logger.getChild('lbry').setLevel(logging.INFO)
-    logger.getChild('torba').setLevel(logging.INFO)
     logger.getChild('aioupnp').setLevel(logging.WARNING)
     logger.getChild('aiohttp').setLevel(logging.CRITICAL)
 
@@ -247,17 +257,13 @@ def setup_logging(logger: logging.Logger, args: argparse.Namespace, conf: Config
         else:
             logger.getChild('lbry').setLevel(logging.DEBUG)
 
-    if conf.share_usage_data:
-        loggly_handler = get_loggly_handler()
-        loggly_handler.setLevel(logging.ERROR)
-        logger.getChild('lbry').addHandler(loggly_handler)
-
 
 def run_daemon(args: argparse.Namespace, conf: Config):
     loop = asyncio.get_event_loop()
     if args.verbose is not None:
         loop.set_debug(True)
-    setup_logging(logging.getLogger(), args, conf)
+    if not args.no_logging:
+        setup_logging(logging.getLogger(), args, conf)
     daemon = Daemon(conf)
 
     def __exit():
@@ -297,6 +303,16 @@ def main(argv=None):
         if args.help:
             args.start_parser.print_help()
         else:
+            if args.initial_headers:
+                ledger_path = os.path.join(conf.wallet_dir, 'lbc_mainnet')
+                ensure_directory_exists(ledger_path)
+                current_size = 0
+                headers_path = os.path.join(ledger_path, 'headers')
+                if os.path.exists(headers_path):
+                    current_size = os.stat(headers_path).st_size
+                if os.stat(args.initial_headers).st_size > current_size:
+                    log.info('Copying header from %s to %s', args.initial_headers, headers_path)
+                    shutil.copy(args.initial_headers, headers_path)
             run_daemon(args, conf)
     elif args.command is not None:
         doc = args.doc

lbry/extras/daemon/analytics.py

@@ -1,8 +1,8 @@
 import asyncio
 import collections
 import logging
-import aiohttp
 import typing
+import aiohttp
 from lbry import utils
 from lbry.conf import Config
 from lbry.extras import system_info
@@ -18,6 +18,7 @@ DOWNLOAD_STARTED = 'Download Started'
 DOWNLOAD_ERRORED = 'Download Errored'
 DOWNLOAD_FINISHED = 'Download Finished'
 HEARTBEAT = 'Heartbeat'
+DISK_SPACE = 'Disk Space'
 CLAIM_ACTION = 'Claim Action'  # publish/create/update/abandon
 NEW_CHANNEL = 'New Channel'
 CREDITS_SENT = 'Credits Sent'
@@ -66,7 +67,7 @@ def _download_properties(conf: Config, external_ip: str, resolve_duration: float
         "node_rpc_timeout": conf.node_rpc_timeout,
         "peer_connect_timeout": conf.peer_connect_timeout,
         "blob_download_timeout": conf.blob_download_timeout,
-        "use_fixed_peers": len(conf.reflector_servers) > 0,
+        "use_fixed_peers": len(conf.fixed_peers) > 0,
         "fixed_peer_delay": fixed_peer_delay,
         "added_fixed_peers": added_fixed_peers,
         "active_peer_count": active_peer_count,
@@ -110,7 +111,6 @@ class AnalyticsManager:
         self.cookies = {}
         self.url = ANALYTICS_ENDPOINT
         self._write_key = utils.deobfuscate(ANALYTICS_TOKEN)
-        self._enabled = conf.share_usage_data
         self._tracked_data = collections.defaultdict(list)
         self.context = _make_context(system_info.get_platform())
         self.installation_id = installation_id
@@ -118,20 +118,24 @@ class AnalyticsManager:
         self.task: typing.Optional[asyncio.Task] = None
         self.external_ip: typing.Optional[str] = None
 
+    @property
+    def enabled(self):
+        return self.conf.share_usage_data
+
     @property
     def is_started(self):
         return self.task is not None
 
     async def start(self):
-        if self._enabled and self.task is None:
-            self.external_ip = await utils.get_external_ip()
+        if self.task is None:
             self.task = asyncio.create_task(self.run())
 
     async def run(self):
         while True:
-            await self._send_heartbeat()
+            if self.enabled:
+                self.external_ip, _ = await utils.get_external_ip(self.conf.lbryum_servers)
+                await self._send_heartbeat()
             await asyncio.sleep(1800)
-            self.external_ip = await utils.get_external_ip()
 
     def stop(self):
         if self.task is not None and not self.task.done():
@@ -154,7 +158,7 @@ class AnalyticsManager:
 
     async def track(self, event: typing.Dict):
         """Send a single tracking event"""
-        if self._enabled:
+        if self.enabled:
             log.debug('Sending track event: %s', event)
             await self._post(event)
 
@@ -166,6 +170,15 @@ class AnalyticsManager:
             })
         )
 
+    async def send_disk_space_used(self, storage_used, storage_limit, is_from_network_quota):
+        await self.track(
+            self._event(DISK_SPACE, {
+                'used': storage_used,
+                'limit': storage_limit,
+                'from_network_quota': is_from_network_quota
+            })
+        )
+
     async def send_server_startup(self):
         await self.track(self._event(SERVER_STARTUP))
 

lbry/extras/daemon/client.py

@@ -1,5 +1,5 @@
-from lbry.conf import Config
 from lbry.extras.cli import execute_command
+from lbry.conf import Config
 
 
 def daemon_rpc(conf: Config, method: str, **kwargs):

lbry/extras/daemon/component.py

@@ -1,7 +1,7 @@
 import asyncio
 import logging
 from lbry.conf import Config
-from lbry.extras.daemon.ComponentManager import ComponentManager
+from lbry.extras.daemon.componentmanager import ComponentManager
 
 log = logging.getLogger(__name__)
 
@@ -9,7 +9,7 @@ log = logging.getLogger(__name__)
 class ComponentType(type):
     def __new__(mcs, name, bases, newattrs):
         klass = type.__new__(mcs, name, bases, newattrs)
-        if name != "Component":
+        if name != "Component" and newattrs['__module__'] != 'lbry.testcase':
             ComponentManager.default_component_classes[klass.component_name] = klass
         return klass
 
@@ -37,7 +37,7 @@ class Component(metaclass=ComponentType):
     def running(self):
         return self._running
 
-    async def get_status(self):
+    async def get_status(self): # pylint: disable=no-self-use
         return
 
     async def start(self):

lbry/extras/daemon/componentmanager.py

@@ -1,7 +1,7 @@
 import logging
 import asyncio
 from lbry.conf import Config
-from lbry.error import ComponentStartConditionNotMet
+from lbry.error import ComponentStartConditionNotMetError
 from lbry.dht.peer import PeerManager
 
 log = logging.getLogger(__name__)
@@ -42,7 +42,7 @@ class ComponentManager:
         self.analytics_manager = analytics_manager
         self.component_classes = {}
         self.components = set()
-        self.started = asyncio.Event(loop=self.loop)
+        self.started = asyncio.Event()
         self.peer_manager = peer_manager or PeerManager(asyncio.get_event_loop_policy().get_event_loop())
 
         for component_name, component_class in self.default_component_classes.items():
@@ -106,7 +106,7 @@ class ComponentManager:
                 staged.update(to_stage)
                 steps.append(step)
             elif components:
-                raise ComponentStartConditionNotMet("Unresolved dependencies for: %s" % components)
+                raise ComponentStartConditionNotMetError(components)
         if reverse:
             steps.reverse()
         return steps
@@ -118,7 +118,7 @@ class ComponentManager:
                 component._setup() for component in stage if not component.running
             ]
             if needing_start:
-                await asyncio.wait(needing_start)
+                await asyncio.wait(map(asyncio.create_task, needing_start))
         self.started.set()
 
     async def stop(self):
@@ -131,7 +131,7 @@ class ComponentManager:
                 component._stop() for component in stage if component.running
             ]
             if needing_stop:
-                await asyncio.wait(needing_stop)
+                await asyncio.wait(map(asyncio.create_task, needing_stop))
 
     def all_components_running(self, *component_names):
         """
@@ -158,11 +158,14 @@ class ComponentManager:
             for component in self.components
         }
 
-    def get_component(self, component_name):
+    def get_actual_component(self, component_name):
         for component in self.components:
             if component.component_name == component_name:
-                return component.component
+                return component
         raise NameError(component_name)
 
+    def get_component(self, component_name):
+        return self.get_actual_component(component_name).component
+
     def has_component(self, component_name):
         return any(component for component in self.components if component_name == component.component_name)

lbry/extras/daemon/components.py

@@ -1,10 +1,10 @@
-import hashlib
+import math
 import os
 import asyncio
 import logging
-import math
 import binascii
 import typing
+
 import base58
 
 from aioupnp import __version__ as aioupnp_version
@@ -13,15 +13,22 @@ from aioupnp.fault import UPnPError
 
 from lbry import utils
 from lbry.dht.node import Node
+from lbry.dht.peer import is_valid_public_ipv4
 from lbry.dht.blob_announcer import BlobAnnouncer
 from lbry.blob.blob_manager import BlobManager
+from lbry.blob.disk_space_manager import DiskSpaceManager
 from lbry.blob_exchange.server import BlobServer
+from lbry.stream.background_downloader import BackgroundDownloader
 from lbry.stream.stream_manager import StreamManager
-from lbry.extras.daemon.Component import Component
+from lbry.file.file_manager import FileManager
+from lbry.extras.daemon.component import Component
 from lbry.extras.daemon.exchange_rate_manager import ExchangeRateManager
 from lbry.extras.daemon.storage import SQLiteStorage
-from lbry.wallet import LbryWalletManager
-from lbry.wallet.header import Headers
+from lbry.torrent.torrent_manager import TorrentManager
+from lbry.wallet import WalletManager
+from lbry.wallet.usage_payment import WalletServerPayer
+from lbry.torrent.tracker import TrackerClient
+from lbry.torrent.session import TorrentSession
 
 log = logging.getLogger(__name__)
 
@@ -29,14 +36,18 @@ log = logging.getLogger(__name__)
 
 DATABASE_COMPONENT = "database"
 BLOB_COMPONENT = "blob_manager"
-HEADERS_COMPONENT = "blockchain_headers"
 WALLET_COMPONENT = "wallet"
+WALLET_SERVER_PAYMENTS_COMPONENT = "wallet_server_payments"
 DHT_COMPONENT = "dht"
 HASH_ANNOUNCER_COMPONENT = "hash_announcer"
-STREAM_MANAGER_COMPONENT = "stream_manager"
+FILE_MANAGER_COMPONENT = "file_manager"
+DISK_SPACE_COMPONENT = "disk_space"
+BACKGROUND_DOWNLOADER_COMPONENT = "background_downloader"
 PEER_PROTOCOL_SERVER_COMPONENT = "peer_protocol_server"
 UPNP_COMPONENT = "upnp"
 EXCHANGE_RATE_MANAGER_COMPONENT = "exchange_rate_manager"
+TRACKER_ANNOUNCER_COMPONENT = "tracker_announcer_component"
+LIBTORRENT_COMPONENT = "libtorrent_component"
 
 
 class DatabaseComponent(Component):
@@ -52,7 +63,7 @@ class DatabaseComponent(Component):
 
     @staticmethod
     def get_current_db_revision():
-        return 12
+        return 15
 
     @property
     def revision_filename(self):
@@ -97,135 +108,9 @@ class DatabaseComponent(Component):
         self.storage = None
 
 
-class HeadersComponent(Component):
-    component_name = HEADERS_COMPONENT
-    HEADERS_URL = "https://headers.lbry.io/blockchain_headers_latest"
-    CHECKPOINT = ('100b33ca3d0b86a48f0d6d6f30458a130ecb89d5affefe4afccb134d5a40f4c2', 600_000)
-
-    def __init__(self, component_manager):
-        super().__init__(component_manager)
-        self.headers_dir = os.path.join(self.conf.wallet_dir, 'lbc_mainnet')
-        self.headers_file = os.path.join(self.headers_dir, 'headers')
-        self.old_file = os.path.join(self.conf.wallet_dir, 'blockchain_headers')
-        self.headers = Headers(self.headers_file)
-        self.is_downloading_headers = False
-        self._headers_progress_percent = 0
-
-    @property
-    def component(self):
-        return self
-
-    def _round_progress(self, local_height, remote_height):
-        return min(max(math.ceil(float(local_height) / float(remote_height) * 100), 0), 100)
-
-    async def get_status(self) -> dict:
-        progress = None
-        if self.is_downloading_headers:
-            progress = self._headers_progress_percent
-        elif self.component_manager.has_component(WALLET_COMPONENT):
-            wallet_manager = self.component_manager.get_component(WALLET_COMPONENT)
-            if wallet_manager and wallet_manager.ledger.network.remote_height > 0:
-                local_height = wallet_manager.ledger.headers.height
-                remote_height = wallet_manager.ledger.network.remote_height
-                progress = self._round_progress(local_height, remote_height)
-        return {
-            'downloading_headers': True,
-            'download_progress': progress
-        } if progress is not None and progress < 100 else {}
-
-    async def fetch_headers_from_s3(self):
-        local_header_size = self.local_header_file_size()
-        resume_header = {"Range": f"bytes={local_header_size}-"}
-        async with utils.aiohttp_request('get', self.HEADERS_URL, headers=resume_header) as response:
-            if response.status == 406 or response.content_length < self.headers.header_size:  # our file is bigger
-                log.warning("s3 is more out of date than we are")
-                return
-            final_size_after_download = response.content_length + local_header_size
-            if final_size_after_download % self.headers.header_size != 0:
-                log.warning("s3 appears to have corrupted header")
-                return
-            write_mode = "wb"
-            if local_header_size > 0:
-                log.info("Resuming download of %i bytes from s3", response.content_length)
-                write_mode = "a+b"
-            with open(self.headers_file, write_mode) as fd:
-                while not response.content.at_eof():
-                    local_header_size += fd.write(await response.content.readany())
-                    self._headers_progress_percent = self._round_progress(
-                        local_header_size, final_size_after_download
-                    )
-
-    def local_header_file_size(self) -> int:
-        if os.path.isfile(self.headers_file):
-            return os.stat(self.headers_file).st_size
-        return 0
-
-    async def get_downloadable_header_height(self) -> typing.Optional[int]:
-        async with utils.aiohttp_request('HEAD', self.HEADERS_URL) as response:
-            if response.status != 200:
-                log.warning("Header download error, unexpected response code: %s", response.status)
-                return -1
-            return response.content_length // self.headers.header_size
-
-    async def should_download_headers_from_s3(self) -> bool:
-        if self.conf.blockchain_name != "lbrycrd_main":
-            return False
-        s3_headers_depth = self.conf.s3_headers_depth
-        if not s3_headers_depth:
-            return False
-
-        local_height = self.local_header_file_size() // self.headers.header_size
-        remote_height = await self.get_downloadable_header_height()
-        if remote_height is not None:
-            log.info("remote height: %i, local height: %i", remote_height, local_height)
-            if remote_height > (local_height + s3_headers_depth):
-                return True
-        return False
-
-    def verify_checkpoint(self):
-        expected_hash, at_height = self.CHECKPOINT
-        if self.local_header_file_size() // self.headers.header_size < at_height:
-            return False
-        hash = hashlib.sha256()
-        chunk_size = self.headers.header_size * 1000
-        with open(self.headers_file, 'rb') as header_file:
-            data = header_file.read(chunk_size)
-            while data and header_file.tell() <= at_height * self.headers.header_size:
-                hash.update(data)
-                data = header_file.read(chunk_size)
-        return hash.hexdigest() == expected_hash
-
-    async def start(self):
-        if not os.path.exists(self.headers_dir):
-            os.mkdir(self.headers_dir)
-        if os.path.exists(self.old_file):
-            log.warning("Moving old headers from %s to %s.", self.old_file, self.headers_file)
-            os.rename(self.old_file, self.headers_file)
-
-        try:
-            if await self.should_download_headers_from_s3():
-                self.is_downloading_headers = True
-                await self.fetch_headers_from_s3()
-        except Exception as err:
-            log.error("failed to fetch headers from s3: %s", err)
-        finally:
-            self.is_downloading_headers = False
-            # fixme: workaround, this should happen before download but happens after because headers.connect fail
-            if not self.verify_checkpoint():
-                log.info("Checkpoint failed, verifying headers using slower method.")
-                await self.headers.open()
-                await self.headers.repair()
-                await self.headers.close()
-            else:
-                log.info("Header checkpoint verified.")
-
-    async def stop(self):
-        pass
-
-
 class WalletComponent(Component):
     component_name = WALLET_COMPONENT
-    depends_on = [DATABASE_COMPONENT, HEADERS_COMPONENT]
+    depends_on = [DATABASE_COMPONENT]
 
     def __init__(self, component_manager):
         super().__init__(component_manager)
@@ -236,27 +121,53 @@ class WalletComponent(Component):
         return self.wallet_manager
 
     async def get_status(self):
-        if self.wallet_manager and self.wallet_manager.ledger.network.remote_height:
-            local_height = self.wallet_manager.ledger.headers.height
+        if self.wallet_manager is None:
+            return
+        is_connected = self.wallet_manager.ledger.network.is_connected
+        sessions = []
+        connected = None
+        if is_connected:
+            addr, port = self.wallet_manager.ledger.network.client.server
+            connected = f"{addr}:{port}"
+            sessions.append(self.wallet_manager.ledger.network.client)
+
+        result = {
+            'connected': connected,
+            'connected_features': self.wallet_manager.ledger.network.server_features,
+            'servers': [
+                {
+                    'host': session.server[0],
+                    'port': session.server[1],
+                    'latency': session.connection_latency,
+                    'availability': session.available,
+                } for session in sessions
+            ],
+            'known_servers': len(self.wallet_manager.ledger.network.known_hubs),
+            'available_servers': 1 if is_connected else 0
+        }
+
+        if self.wallet_manager.ledger.network.remote_height:
+            local_height = self.wallet_manager.ledger.local_height_including_downloaded_height
+            disk_height = len(self.wallet_manager.ledger.headers)
             remote_height = self.wallet_manager.ledger.network.remote_height
-            best_hash = self.wallet_manager.get_best_blockhash()
-            return {
-                'connected_servers': [
-                    {
-                        'host': session.server[0],
-                        'port': session.server[1],
-                        'latency': round(session.connection_latency, 2),
-                    } for session in self.wallet_manager.ledger.network.session_pool.sessions
-                    if session and session.available
-                ],
+            download_height, target_height = local_height - disk_height, remote_height - disk_height
+            if target_height > 0:
+                progress = min(max(math.ceil(float(download_height) / float(target_height) * 100), 0), 100)
+            else:
+                progress = 100
+            best_hash = await self.wallet_manager.get_best_blockhash()
+            result.update({
+                'headers_synchronization_progress': progress,
                 'blocks': max(local_height, 0),
                 'blocks_behind': max(remote_height - local_height, 0),
                 'best_blockhash': best_hash,
-            }
+            })
+
+        return result
 
     async def start(self):
-        log.info("Starting torba wallet")
-        self.wallet_manager = await LbryWalletManager.from_lbrynet_config(self.conf)
+        log.info("Starting wallet")
+        self.wallet_manager = await WalletManager.from_lbrynet_config(self.conf)
         await self.wallet_manager.start()
 
     async def stop(self):
@@ -264,6 +175,34 @@ class WalletComponent(Component):
         self.wallet_manager = None
 
 
+class WalletServerPaymentsComponent(Component):
+    component_name = WALLET_SERVER_PAYMENTS_COMPONENT
+    depends_on = [WALLET_COMPONENT]
+
+    def __init__(self, component_manager):
+        super().__init__(component_manager)
+        self.usage_payment_service = WalletServerPayer(
+            max_fee=self.conf.max_wallet_server_fee, analytics_manager=self.component_manager.analytics_manager,
+        )
+
+    @property
+    def component(self) -> typing.Optional[WalletServerPayer]:
+        return self.usage_payment_service
+
+    async def start(self):
+        wallet_manager = self.component_manager.get_component(WALLET_COMPONENT)
+        await self.usage_payment_service.start(wallet_manager.ledger, wallet_manager.default_wallet)
+
+    async def stop(self):
+        await self.usage_payment_service.stop()
+
+    async def get_status(self):
+        return {
+            'max_fee': self.usage_payment_service.max_fee,
+            'running': self.usage_payment_service.running
+        }
+
+
 class BlobComponent(Component):
     component_name = BLOB_COMPONENT
     depends_on = [DATABASE_COMPONENT]
@@ -304,7 +243,7 @@ class BlobComponent(Component):
 
 class DHTComponent(Component):
     component_name = DHT_COMPONENT
-    depends_on = [UPNP_COMPONENT]
+    depends_on = [UPNP_COMPONENT, DATABASE_COMPONENT]
 
     def __init__(self, component_manager):
         super().__init__(component_manager)
@@ -338,8 +277,9 @@ class DHTComponent(Component):
         self.external_peer_port = upnp_component.upnp_redirects.get("TCP", self.conf.tcp_port)
         self.external_udp_port = upnp_component.upnp_redirects.get("UDP", self.conf.udp_port)
         external_ip = upnp_component.external_ip
+        storage = self.component_manager.get_component(DATABASE_COMPONENT)
         if not external_ip:
-            external_ip = await utils.get_external_ip()
+            external_ip, _ = await utils.get_external_ip(self.conf.lbryum_servers)
             if not external_ip:
                 log.warning("failed to get external ip")
 
@@ -352,11 +292,11 @@ class DHTComponent(Component):
             external_ip=external_ip,
             peer_port=self.external_peer_port,
             rpc_timeout=self.conf.node_rpc_timeout,
-            split_buckets_under_index=self.conf.split_buckets_under_index
-        )
-        self.dht_node.start(
-            interface=self.conf.network_interface, known_node_urls=self.conf.known_dht_nodes
+            split_buckets_under_index=self.conf.split_buckets_under_index,
+            is_bootstrap_node=self.conf.is_bootstrap_node,
+            storage=storage
         )
+        self.dht_node.start(self.conf.network_interface, self.conf.known_dht_nodes)
         log.info("Started the dht")
 
     async def stop(self):
@@ -392,41 +332,175 @@ class HashAnnouncerComponent(Component):
         }
 
 
-class StreamManagerComponent(Component):
-    component_name = STREAM_MANAGER_COMPONENT
+class FileManagerComponent(Component):
+    component_name = FILE_MANAGER_COMPONENT
     depends_on = [BLOB_COMPONENT, DATABASE_COMPONENT, WALLET_COMPONENT]
 
     def __init__(self, component_manager):
         super().__init__(component_manager)
-        self.stream_manager: typing.Optional[StreamManager] = None
+        self.file_manager: typing.Optional[FileManager] = None
 
     @property
-    def component(self) -> typing.Optional[StreamManager]:
-        return self.stream_manager
+    def component(self) -> typing.Optional[FileManager]:
+        return self.file_manager
 
     async def get_status(self):
-        if not self.stream_manager:
+        if not self.file_manager:
             return
         return {
-            'managed_files': len(self.stream_manager.streams),
+            'managed_files': len(self.file_manager.get_filtered()),
         }
 
     async def start(self):
         blob_manager = self.component_manager.get_component(BLOB_COMPONENT)
         storage = self.component_manager.get_component(DATABASE_COMPONENT)
         wallet = self.component_manager.get_component(WALLET_COMPONENT)
-        node = self.component_manager.get_component(DHT_COMPONENT)\
+        node = self.component_manager.get_component(DHT_COMPONENT) \
             if self.component_manager.has_component(DHT_COMPONENT) else None
         log.info('Starting the file manager')
         loop = asyncio.get_event_loop()
-        self.stream_manager = StreamManager(
-            loop, self.conf, blob_manager, wallet, storage, node, self.component_manager.analytics_manager
+        self.file_manager = FileManager(
+            loop, self.conf, wallet, storage, self.component_manager.analytics_manager
         )
-        await self.stream_manager.start()
+        self.file_manager.source_managers['stream'] = StreamManager(
+            loop, self.conf, blob_manager, wallet, storage, node,
+        )
+        if self.component_manager.has_component(LIBTORRENT_COMPONENT):
+            torrent = self.component_manager.get_component(LIBTORRENT_COMPONENT)
+            self.file_manager.source_managers['torrent'] = TorrentManager(
+                loop, self.conf, torrent, storage, self.component_manager.analytics_manager
+            )
+        await self.file_manager.start()
         log.info('Done setting up file manager')
 
     async def stop(self):
-        self.stream_manager.stop()
+        await self.file_manager.stop()
+
+
+class BackgroundDownloaderComponent(Component):
+    MIN_PREFIX_COLLIDING_BITS = 8
+    component_name = BACKGROUND_DOWNLOADER_COMPONENT
+    depends_on = [DATABASE_COMPONENT, BLOB_COMPONENT, DISK_SPACE_COMPONENT]
+
+    def __init__(self, component_manager):
+        super().__init__(component_manager)
+        self.background_task: typing.Optional[asyncio.Task] = None
+        self.download_loop_delay_seconds = 60
+        self.ongoing_download: typing.Optional[asyncio.Task] = None
+        self.space_manager: typing.Optional[DiskSpaceManager] = None
+        self.blob_manager: typing.Optional[BlobManager] = None
+        self.background_downloader: typing.Optional[BackgroundDownloader] = None
+        self.dht_node: typing.Optional[Node] = None
+        self.space_available: typing.Optional[int] = None
+
+    @property
+    def is_busy(self):
+        return bool(self.ongoing_download and not self.ongoing_download.done())
+
+    @property
+    def component(self) -> 'BackgroundDownloaderComponent':
+        return self
+
+    async def get_status(self):
+        return {'running': self.background_task is not None and not self.background_task.done(),
+                'available_free_space_mb': self.space_available,
+                'ongoing_download': self.is_busy}
+
+    async def download_blobs_in_background(self):
+        while True:
+            self.space_available = await self.space_manager.get_free_space_mb(True)
+            if not self.is_busy and self.space_available > 10:
+                self._download_next_close_blob_hash()
+            await asyncio.sleep(self.download_loop_delay_seconds)
+
+    def _download_next_close_blob_hash(self):
+        node_id = self.dht_node.protocol.node_id
+        for blob_hash in self.dht_node.stored_blob_hashes:
+            if blob_hash.hex() in self.blob_manager.completed_blob_hashes:
+                continue
+            if utils.get_colliding_prefix_bits(node_id, blob_hash) >= self.MIN_PREFIX_COLLIDING_BITS:
+                self.ongoing_download = asyncio.create_task(self.background_downloader.download_blobs(blob_hash.hex()))
+                return
+
+    async def start(self):
+        self.space_manager: DiskSpaceManager = self.component_manager.get_component(DISK_SPACE_COMPONENT)
+        if not self.component_manager.has_component(DHT_COMPONENT):
+            return
+        self.dht_node = self.component_manager.get_component(DHT_COMPONENT)
+        self.blob_manager = self.component_manager.get_component(BLOB_COMPONENT)
+        storage = self.component_manager.get_component(DATABASE_COMPONENT)
+        self.background_downloader = BackgroundDownloader(self.conf, storage, self.blob_manager, self.dht_node)
+        self.background_task = asyncio.create_task(self.download_blobs_in_background())
+
+    async def stop(self):
+        if self.ongoing_download and not self.ongoing_download.done():
+            self.ongoing_download.cancel()
+        if self.background_task:
+            self.background_task.cancel()
+
+
+class DiskSpaceComponent(Component):
+    component_name = DISK_SPACE_COMPONENT
+    depends_on = [DATABASE_COMPONENT, BLOB_COMPONENT]
+
+    def __init__(self, component_manager):
+        super().__init__(component_manager)
+        self.disk_space_manager: typing.Optional[DiskSpaceManager] = None
+
+    @property
+    def component(self) -> typing.Optional[DiskSpaceManager]:
+        return self.disk_space_manager
+
+    async def get_status(self):
+        if self.disk_space_manager:
+            space_used = await self.disk_space_manager.get_space_used_mb(cached=True)
+            return {
+                'total_used_mb': space_used['total'],
+                'published_blobs_storage_used_mb': space_used['private_storage'],
+                'content_blobs_storage_used_mb': space_used['content_storage'],
+                'seed_blobs_storage_used_mb': space_used['network_storage'],
+                'running': self.disk_space_manager.running,
+            }
+        return {'space_used': '0', 'network_seeding_space_used': '0', 'running': False}
+
+    async def start(self):
+        db = self.component_manager.get_component(DATABASE_COMPONENT)
+        blob_manager = self.component_manager.get_component(BLOB_COMPONENT)
+        self.disk_space_manager = DiskSpaceManager(
+            self.conf, db, blob_manager,
+            analytics=self.component_manager.analytics_manager
+        )
+        await self.disk_space_manager.start()
+
+    async def stop(self):
+        await self.disk_space_manager.stop()
+
+
+class TorrentComponent(Component):
+    component_name = LIBTORRENT_COMPONENT
+
+    def __init__(self, component_manager):
+        super().__init__(component_manager)
+        self.torrent_session = None
+
+    @property
+    def component(self) -> typing.Optional[TorrentSession]:
+        return self.torrent_session
+
+    async def get_status(self):
+        if not self.torrent_session:
+            return
+        return {
+            'running': True,  # TODO: what to return here?
+        }
+
+    async def start(self):
+        self.torrent_session = TorrentSession(asyncio.get_event_loop(), None)
+        await self.torrent_session.bind()  # TODO: specify host/port
+
+    async def stop(self):
+        if self.torrent_session:
+            await self.torrent_session.pause()
 
 
 class PeerProtocolServerComponent(Component):
@@ -443,9 +517,8 @@ class PeerProtocolServerComponent(Component):
 
     async def start(self):
         log.info("start blob server")
-        upnp = self.component_manager.get_component(UPNP_COMPONENT)
         blob_manager: BlobManager = self.component_manager.get_component(BLOB_COMPONENT)
-        wallet: LbryWalletManager = self.component_manager.get_component(WALLET_COMPONENT)
+        wallet: WalletManager = self.component_manager.get_component(WALLET_COMPONENT)
         peer_port = self.conf.tcp_port
         address = await wallet.get_unused_address()
         self.blob_server = BlobServer(asyncio.get_event_loop(), blob_manager, address)
@@ -478,7 +551,7 @@ class UPnPComponent(Component):
         while True:
             if now:
                 await self._maintain_redirects()
-            await asyncio.sleep(360, loop=self.component_manager.loop)
+            await asyncio.sleep(360)
 
     async def _maintain_redirects(self):
         # setup the gateway if necessary
@@ -487,8 +560,6 @@ class UPnPComponent(Component):
                 self.upnp = await UPnP.discover(loop=self.component_manager.loop)
                 log.info("found upnp gateway: %s", self.upnp.gateway.manufacturer_string)
             except Exception as err:
-                if isinstance(err, asyncio.CancelledError):
-                    raise
                 log.warning("upnp discovery failed: %s", err)
                 self.upnp = None
 
@@ -501,14 +572,17 @@ class UPnPComponent(Component):
                     log.info("got external ip from UPnP: %s", external_ip)
             except (asyncio.TimeoutError, UPnPError, NotImplementedError):
                 pass
-
-        if external_ip == "0.0.0.0" or (external_ip and external_ip.startswith("192.")):
-            log.warning("unable to get external ip from UPnP, checking lbry.com fallback")
-            external_ip = await utils.get_external_ip()
+        if external_ip and not is_valid_public_ipv4(external_ip):
+            log.warning("UPnP returned a private/reserved ip - %s, checking lbry.com fallback", external_ip)
+            external_ip, _ = await utils.get_external_ip(self.conf.lbryum_servers)
         if self.external_ip and self.external_ip != external_ip:
             log.info("external ip changed from %s to %s", self.external_ip, external_ip)
         if external_ip:
             self.external_ip = external_ip
+            dht_component = self.component_manager.get_component(DHT_COMPONENT)
+            if dht_component:
+                dht_node = dht_component.component
+                dht_node.protocol.external_ip = external_ip
         # assert self.external_ip is not None   # TODO: handle going/starting offline
 
         if not self.upnp_redirects and self.upnp:  # setup missing redirects
@@ -553,22 +627,24 @@ class UPnPComponent(Component):
                     log.info("refreshed upnp redirect for peer port: %i", tcp_port)
                 except (asyncio.TimeoutError, UPnPError, NotImplementedError):
                     del self.upnp_redirects['TCP']
-            if ('TCP' in self.upnp_redirects
-                and PEER_PROTOCOL_SERVER_COMPONENT not in self.component_manager.skip_components) and (
-                    'UDP' in self.upnp_redirects and DHT_COMPONENT not in self.component_manager.skip_components):
+            if ('TCP' in self.upnp_redirects and
+                    PEER_PROTOCOL_SERVER_COMPONENT not in self.component_manager.skip_components) and \
+                    ('UDP' in self.upnp_redirects and DHT_COMPONENT not in self.component_manager.skip_components):
                 if self.upnp_redirects:
                     log.debug("upnp redirects are still active")
 
     async def start(self):
         log.info("detecting external ip")
         if not self.use_upnp:
-            self.external_ip = await utils.get_external_ip()
+            self.external_ip, _ = await utils.get_external_ip(self.conf.lbryum_servers)
             return
         success = False
         await self._maintain_redirects()
         if self.upnp:
-            if not self.upnp_redirects and not all([x in self.component_manager.skip_components for x in
-                                                    (DHT_COMPONENT, PEER_PROTOCOL_SERVER_COMPONENT)]):
+            if not self.upnp_redirects and not all(
+                    x in self.component_manager.skip_components
+                    for x in (DHT_COMPONENT, PEER_PROTOCOL_SERVER_COMPONENT)
+            ):
                 log.error("failed to setup upnp")
             else:
                 success = True
@@ -577,12 +653,14 @@ class UPnPComponent(Component):
         else:
             log.error("failed to setup upnp")
         if not self.external_ip:
-            self.external_ip = await utils.get_external_ip()
+            self.external_ip, probed_url = await utils.get_external_ip(self.conf.lbryum_servers)
             if self.external_ip:
-                log.info("detected external ip using lbry.com fallback")
+                log.info("detected external ip using %s fallback", probed_url)
         if self.component_manager.analytics_manager:
-            await self.component_manager.analytics_manager.send_upnp_setup_success_fail(
-                success, await self.get_status()
+            self.component_manager.loop.create_task(
+                self.component_manager.analytics_manager.send_upnp_setup_success_fail(
+                    success, await self.get_status()
+                )
             )
         self._maintain_redirects_task = self.component_manager.loop.create_task(
             self._repeatedly_maintain_redirects(now=False)
@@ -593,7 +671,7 @@ class UPnPComponent(Component):
             log.info("Removing upnp redirects: %s", self.upnp_redirects)
             await asyncio.wait([
                 self.upnp.delete_port_mapping(port, protocol) for protocol, port in self.upnp_redirects.items()
-            ], loop=self.component_manager.loop)
+            ])
         if self._maintain_redirects_task and not self._maintain_redirects_task.done():
             self._maintain_redirects_task.cancel()
 
@@ -624,3 +702,49 @@ class ExchangeRateManagerComponent(Component):
 
     async def stop(self):
         self.exchange_rate_manager.stop()
+
+
+class TrackerAnnouncerComponent(Component):
+    component_name = TRACKER_ANNOUNCER_COMPONENT
+    depends_on = [FILE_MANAGER_COMPONENT]
+
+    def __init__(self, component_manager):
+        super().__init__(component_manager)
+        self.file_manager = None
+        self.announce_task = None
+        self.tracker_client: typing.Optional[TrackerClient] = None
+
+    @property
+    def component(self):
+        return self.tracker_client
+
+    @property
+    def running(self):
+        return self._running and self.announce_task and not self.announce_task.done()
+
+    async def announce_forever(self):
+        while True:
+            sleep_seconds = 60.0
+            announce_sd_hashes = []
+            for file in self.file_manager.get_filtered():
+                if not file.downloader:
+                    continue
+                announce_sd_hashes.append(bytes.fromhex(file.sd_hash))
+            await self.tracker_client.announce_many(*announce_sd_hashes)
+            await asyncio.sleep(sleep_seconds)
+
+    async def start(self):
+        node = self.component_manager.get_component(DHT_COMPONENT) \
+            if self.component_manager.has_component(DHT_COMPONENT) else None
+        node_id = node.protocol.node_id if node else None
+        self.tracker_client = TrackerClient(node_id, self.conf.tcp_port, lambda: self.conf.tracker_servers)
+        await self.tracker_client.start()
+        self.file_manager = self.component_manager.get_component(FILE_MANAGER_COMPONENT)
+        self.announce_task = asyncio.create_task(self.announce_forever())
+
+    async def stop(self):
+        self.file_manager = None
+        if self.announce_task and not self.announce_task.done():
+            self.announce_task.cancel()
+        self.announce_task = None
+        self.tracker_client.stop()

lbry/extras/daemon/exchange_rate_manager.py

@@ -0,0 +1,248 @@
+import json
+import time
+import asyncio
+import logging
+from statistics import median
+from decimal import Decimal
+from typing import Optional, Iterable, Type
+from aiohttp.client_exceptions import ContentTypeError, ClientConnectionError
+from lbry.error import InvalidExchangeRateResponseError, CurrencyConversionError
+from lbry.utils import aiohttp_request
+from lbry.wallet.dewies import lbc_to_dewies
+
+log = logging.getLogger(__name__)
+
+
+class ExchangeRate:
+    def __init__(self, market, spot, ts):
+        if not int(time.time()) - ts < 600:
+            raise ValueError('The timestamp is too dated.')
+        if not spot > 0:
+            raise ValueError('Spot must be greater than 0.')
+        self.currency_pair = (market[0:3], market[3:6])
+        self.spot = spot
+        self.ts = ts
+
+    def __repr__(self):
+        return f"Currency pair:{self.currency_pair}, spot:{self.spot}, ts:{self.ts}"
+
+    def as_dict(self):
+        return {'spot': self.spot, 'ts': self.ts}
+
+
+class MarketFeed:
+    name: str = ""
+    market: str = ""
+    url: str = ""
+    params = {}
+    fee = 0
+
+    update_interval = 300
+    request_timeout = 50
+
+    def __init__(self):
+        self.rate: Optional[float] = None
+        self.last_check = 0
+        self._last_response = None
+        self._task: Optional[asyncio.Task] = None
+        self.event = asyncio.Event()
+
+    @property
+    def has_rate(self):
+        return self.rate is not None
+
+    @property
+    def is_online(self):
+        return self.last_check+self.update_interval+self.request_timeout > time.time()
+
+    def get_rate_from_response(self, json_response):
+        raise NotImplementedError()
+
+    async def get_response(self):
+        async with aiohttp_request(
+                'get', self.url, params=self.params,
+                timeout=self.request_timeout, headers={"User-Agent": "lbrynet"}
+        ) as response:
+            try:
+                self._last_response = await response.json(content_type=None)
+            except ContentTypeError as e:
+                self._last_response = {}
+                log.warning("Could not parse exchange rate response from %s: %s", self.name, e.message)
+                log.debug(await response.text())
+            return self._last_response
+
+    async def get_rate(self):
+        try:
+            data = await self.get_response()
+            rate = self.get_rate_from_response(data)
+            rate = rate / (1.0 - self.fee)
+            log.debug("Saving rate update %f for %s from %s", rate, self.market, self.name)
+            self.rate = ExchangeRate(self.market, rate, int(time.time()))
+            self.last_check = time.time()
+            return self.rate
+        except asyncio.TimeoutError:
+            log.warning("Timed out fetching exchange rate from %s.", self.name)
+        except json.JSONDecodeError as e:
+            msg = e.doc if '<html>' not in e.doc else 'unexpected content type.'
+            log.warning("Could not parse exchange rate response from %s: %s", self.name, msg)
+            log.debug(e.doc)
+        except InvalidExchangeRateResponseError as e:
+            log.warning(str(e))
+        except ClientConnectionError as e:
+            log.warning("Error trying to connect to exchange rate %s: %s", self.name, str(e))
+        except Exception as e:
+            log.exception("Exchange rate error (%s from %s):", self.market, self.name)
+        finally:
+            self.event.set()
+
+    async def keep_updated(self):
+        while True:
+            await self.get_rate()
+            await asyncio.sleep(self.update_interval)
+
+    def start(self):
+        if not self._task:
+            self._task = asyncio.create_task(self.keep_updated())
+
+    def stop(self):
+        if self._task and not self._task.done():
+            self._task.cancel()
+        self._task = None
+        self.event.clear()
+
+
+class BaseBittrexFeed(MarketFeed):
+    name = "Bittrex"
+    market = None
+    url = None
+    fee = 0.0025
+
+    def get_rate_from_response(self, json_response):
+        if 'lastTradeRate' not in json_response:
+            raise InvalidExchangeRateResponseError(self.name, 'result not found')
+        return 1.0 / float(json_response['lastTradeRate'])
+
+
+class BittrexBTCFeed(BaseBittrexFeed):
+    market = "BTCLBC"
+    url = "https://api.bittrex.com/v3/markets/LBC-BTC/ticker"
+
+
+class BittrexUSDFeed(BaseBittrexFeed):
+    market = "USDLBC"
+    url = "https://api.bittrex.com/v3/markets/LBC-USD/ticker"
+
+
+class BaseCoinExFeed(MarketFeed):
+    name = "CoinEx"
+    market = None
+    url = None
+
+    def get_rate_from_response(self, json_response):
+        if 'data' not in json_response or \
+           'ticker' not in json_response['data'] or \
+           'last' not in json_response['data']['ticker']:
+            raise InvalidExchangeRateResponseError(self.name, 'result not found')
+        return 1.0 / float(json_response['data']['ticker']['last'])
+
+
+class CoinExBTCFeed(BaseCoinExFeed):
+    market = "BTCLBC"
+    url = "https://api.coinex.com/v1/market/ticker?market=LBCBTC"
+
+
+class CoinExUSDFeed(BaseCoinExFeed):
+    market = "USDLBC"
+    url = "https://api.coinex.com/v1/market/ticker?market=LBCUSDT"
+
+
+class BaseHotbitFeed(MarketFeed):
+    name = "hotbit"
+    market = None
+    url = "https://api.hotbit.io/api/v1/market.last"
+
+    def get_rate_from_response(self, json_response):
+        if 'result' not in json_response:
+            raise InvalidExchangeRateResponseError(self.name, 'result not found')
+        return 1.0 / float(json_response['result'])
+
+
+class HotbitBTCFeed(BaseHotbitFeed):
+    market = "BTCLBC"
+    params = {"market": "LBC/BTC"}
+
+
+class HotbitUSDFeed(BaseHotbitFeed):
+    market = "USDLBC"
+    params = {"market": "LBC/USDT"}
+
+
+class UPbitBTCFeed(MarketFeed):
+    name = "UPbit"
+    market = "BTCLBC"
+    url = "https://api.upbit.com/v1/ticker"
+    params = {"markets": "BTC-LBC"}
+
+    def get_rate_from_response(self, json_response):
+        if "error" in json_response or len(json_response) != 1 or 'trade_price' not in json_response[0]:
+            raise InvalidExchangeRateResponseError(self.name, 'result not found')
+        return 1.0 / float(json_response[0]['trade_price'])
+
+
+FEEDS: Iterable[Type[MarketFeed]] = (
+    BittrexBTCFeed,
+    BittrexUSDFeed,
+    CoinExBTCFeed,
+    CoinExUSDFeed,
+#    HotbitBTCFeed,
+#    HotbitUSDFeed,
+#    UPbitBTCFeed,
+)
+
+
+class ExchangeRateManager:
+    def __init__(self, feeds=FEEDS):
+        self.market_feeds = [Feed() for Feed in feeds]
+
+    def wait(self):
+        return asyncio.wait(
+            [feed.event.wait() for feed in self.market_feeds],
+        )
+
+    def start(self):
+        log.info("Starting exchange rate manager")
+        for feed in self.market_feeds:
+            feed.start()
+
+    def stop(self):
+        log.info("Stopping exchange rate manager")
+        for source in self.market_feeds:
+            source.stop()
+
+    def convert_currency(self, from_currency, to_currency, amount):
+        log.debug(
+            "Converting %f %s to %s, rates: %s",
+            amount, from_currency, to_currency,
+            [market.rate for market in self.market_feeds]
+        )
+        if from_currency == to_currency:
+            return round(amount, 8)
+
+        rates = []
+        for market in self.market_feeds:
+            if (market.has_rate and market.is_online and
+                    market.rate.currency_pair == (from_currency, to_currency)):
+                rates.append(market.rate.spot)
+
+        if rates:
+            return round(amount * Decimal(median(rates)), 8)
+
+        raise CurrencyConversionError(
+            f'Unable to convert {amount} from {from_currency} to {to_currency}')
+
+    def to_dewies(self, currency, amount) -> int:
+        converted = self.convert_currency(currency, "LBC", amount)
+        return lbc_to_dewies(str(converted))
+
+    def fee_dict(self):
+        return {market: market.rate.as_dict() for market in self.market_feeds}

lbry/extras/daemon/json_response_encoder.py

@@ -6,11 +6,11 @@ from json import JSONEncoder
 
 from google.protobuf.message import DecodeError
 
-from torba.client.wallet import Wallet
-from torba.client.bip32 import PubKey
 from lbry.schema.claim import Claim
-from lbry.wallet.ledger import MainNetLedger, Account
-from lbry.wallet.transaction import Transaction, Output
+from lbry.schema.support import Support
+from lbry.torrent.torrent_manager import TorrentSource
+from lbry.wallet import Wallet, Ledger, Account, Transaction, Output
+from lbry.wallet.bip32 import PublicKey
 from lbry.wallet.dewies import dewies_to_lbc
 from lbry.stream.managed_stream import ManagedStream
 
@@ -27,17 +27,22 @@ def encode_txo_doc():
         'address': "address of who can spend the txo",
         'confirmations': "number of confirmed blocks",
         'is_change': "payment to change address, only available when it can be determined",
+        'is_received': "true if txo was sent from external account to this account",
+        'is_spent': "true if txo is spent",
         'is_mine': "payment to one of your accounts, only available when it can be determined",
-        'type': "one of 'claim', 'support' or 'payment'",
+        'type': "one of 'claim', 'support' or 'purchase'",
         'name': "when type is 'claim' or 'support', this is the claim name",
-        'claim_id': "when type is 'claim' or 'support', this is the claim id",
+        'claim_id': "when type is 'claim', 'support' or 'purchase', this is the claim id",
         'claim_op': "when type is 'claim', this determines if it is 'create' or 'update'",
         'value': "when type is 'claim' or 'support' with payload, this is the decoded protobuf payload",
         'value_type': "determines the type of the 'value' field: 'channel', 'stream', etc",
         'protobuf': "hex encoded raw protobuf version of 'value' field",
         'permanent_url': "when type is 'claim' or 'support', this is the long permanent claim URL",
+        'claim': "for purchase outputs only, metadata of purchased claim",
+        'reposted_claim': "for repost claims only, metadata of claim being reposted",
         'signing_channel': "for signed claims only, metadata of signing channel",
         'is_channel_signature_valid': "for signed claims only, whether signature is valid",
+        'purchase_receipt': "metadata for the purchase transaction associated with this claim"
     }
 
 
@@ -105,23 +110,25 @@ def encode_file_doc():
         'metadata': '(dict) None if claim is not found else the claim metadata',
         'channel_claim_id': '(str) None if claim is not found or not signed',
         'channel_name': '(str) None if claim is not found or not signed',
-        'claim_name': '(str) None if claim is not found else the claim name'
+        'claim_name': '(str) None if claim is not found else the claim name',
+        'reflector_progress': '(int) reflector upload progress, 0 to 100',
+        'uploading_to_reflector': '(bool) set to True when currently uploading to reflector'
     }
 
 
 class JSONResponseEncoder(JSONEncoder):
 
-    def __init__(self, *args, ledger: MainNetLedger, include_protobuf=False, **kwargs):
+    def __init__(self, *args, ledger: Ledger, include_protobuf=False, **kwargs):
         super().__init__(*args, **kwargs)
         self.ledger = ledger
         self.include_protobuf = include_protobuf
 
-    def default(self, obj):  # pylint: disable=method-hidden
+    def default(self, obj):  # pylint: disable=method-hidden,arguments-renamed,too-many-return-statements
         if isinstance(obj, Account):
             return self.encode_account(obj)
         if isinstance(obj, Wallet):
             return self.encode_wallet(obj)
-        if isinstance(obj, ManagedStream):
+        if isinstance(obj, (ManagedStream, TorrentSource)):
             return self.encode_file(obj)
         if isinstance(obj, Transaction):
             return self.encode_transaction(obj)
@@ -129,7 +136,9 @@ class JSONResponseEncoder(JSONEncoder):
             return self.encode_output(obj)
         if isinstance(obj, Claim):
             return self.encode_claim(obj)
-        if isinstance(obj, PubKey):
+        if isinstance(obj, Support):
+            return obj.to_dict()
+        if isinstance(obj, PublicKey):
             return obj.extended_key_string()
         if isinstance(obj, datetime):
             return obj.strftime("%Y%m%dT%H:%M:%S")
@@ -152,6 +161,8 @@ class JSONResponseEncoder(JSONEncoder):
         }
 
     def encode_output(self, txo, check_signature=True):
+        if not txo:
+            return
         tx_height = txo.tx_ref.height
         best_height = self.ledger.headers.height
         output = {
@@ -159,14 +170,24 @@ class JSONResponseEncoder(JSONEncoder):
             'nout': txo.position,
             'height': tx_height,
             'amount': dewies_to_lbc(txo.amount),
-            'address': txo.get_address(self.ledger),
+            'address': txo.get_address(self.ledger) if txo.has_address else None,
             'confirmations': (best_height+1) - tx_height if tx_height > 0 else tx_height,
-            'timestamp': self.ledger.headers[tx_height]['timestamp'] if 0 < tx_height <= best_height else None
+            'timestamp': self.ledger.headers.estimated_timestamp(tx_height)
         }
-        if txo.is_change is not None:
-            output['is_change'] = txo.is_change
-        if txo.is_my_account is not None:
-            output['is_mine'] = txo.is_my_account
+        if txo.is_spent is not None:
+            output['is_spent'] = txo.is_spent
+        if txo.is_my_output is not None:
+            output['is_my_output'] = txo.is_my_output
+        if txo.is_my_input is not None:
+            output['is_my_input'] = txo.is_my_input
+        if txo.sent_supports is not None:
+            output['sent_supports'] = dewies_to_lbc(txo.sent_supports)
+        if txo.sent_tips is not None:
+            output['sent_tips'] = dewies_to_lbc(txo.sent_tips)
+        if txo.received_tips is not None:
+            output['received_tips'] = dewies_to_lbc(txo.received_tips)
+        if txo.is_internal_transfer is not None:
+            output['is_internal_transfer'] = txo.is_internal_transfer
 
         if txo.script.is_claim_name:
             output['type'] = 'claim'
@@ -176,6 +197,13 @@ class JSONResponseEncoder(JSONEncoder):
             output['claim_op'] = 'update'
         elif txo.script.is_support_claim:
             output['type'] = 'support'
+        elif txo.script.is_return_data:
+            output['type'] = 'data'
+        elif txo.purchase is not None:
+            output['type'] = 'purchase'
+            output['claim_id'] = txo.purchased_claim_id
+            if txo.purchased_claim is not None:
+                output['claim'] = self.encode_output(txo.purchased_claim)
         else:
             output['type'] = 'payment'
 
@@ -191,23 +219,30 @@ class JSONResponseEncoder(JSONEncoder):
                 output['short_url'] = output['meta'].pop('short_url')
             if 'canonical_url' in output['meta']:
                 output['canonical_url'] = output['meta'].pop('canonical_url')
-            if txo.script.is_claim_name or txo.script.is_update_claim:
-                try:
-                    output['value'] = txo.claim
+            if txo.claims is not None:
+                output['claims'] = [self.encode_output(o) for o in txo.claims]
+            if txo.reposted_claim is not None:
+                output['reposted_claim'] = self.encode_output(txo.reposted_claim)
+        if txo.script.is_claim_name or txo.script.is_update_claim or txo.script.is_support_claim_data:
+            try:
+                output['value'] = txo.signable
+                if self.include_protobuf:
+                    output['protobuf'] = hexlify(txo.signable.to_bytes())
+                if txo.purchase_receipt is not None:
+                    output['purchase_receipt'] = self.encode_output(txo.purchase_receipt)
+                if txo.script.is_claim_name or txo.script.is_update_claim:
                     output['value_type'] = txo.claim.claim_type
-                    if self.include_protobuf:
-                        output['protobuf'] = hexlify(txo.claim.to_bytes())
                     if txo.claim.is_channel:
                         output['has_signing_key'] = txo.has_private_key
-                    if check_signature and txo.claim.is_signed:
-                        if txo.channel is not None:
-                            output['signing_channel'] = self.encode_output(txo.channel)
-                            output['is_channel_signature_valid'] = txo.is_signed_by(txo.channel, self.ledger)
-                        else:
-                            output['signing_channel'] = {'channel_id': txo.claim.signing_channel_id}
-                            output['is_channel_signature_valid'] = False
-                except DecodeError:
-                    pass
+                if check_signature and txo.signable.is_signed:
+                    if txo.channel is not None:
+                        output['signing_channel'] = self.encode_output(txo.channel)
+                        output['is_channel_signature_valid'] = txo.is_signed_by(txo.channel, self.ledger)
+                    else:
+                        output['signing_channel'] = {'channel_id': txo.signable.signing_channel_id}
+                        output['is_channel_signature_valid'] = False
+            except DecodeError:
+                pass
         return output
 
     def encode_claim_meta(self, meta):
@@ -216,7 +251,7 @@ class JSONResponseEncoder(JSONEncoder):
                 if isinstance(value, int):
                     meta[key] = dewies_to_lbc(value)
         if 0 < meta.get('creation_height', 0) <= self.ledger.headers.height:
-            meta['creation_timestamp'] = self.ledger.headers[meta['creation_height']]['timestamp']
+            meta['creation_timestamp'] = self.ledger.headers.estimated_timestamp(meta['creation_height'])
         return meta
 
     def encode_input(self, txi):
@@ -232,7 +267,8 @@ class JSONResponseEncoder(JSONEncoder):
         result['is_default'] = self.ledger.accounts[0] == account
         return result
 
-    def encode_wallet(self, wallet):
+    @staticmethod
+    def encode_wallet(wallet):
         return {
             'id': wallet.id,
             'name': wallet.name
@@ -242,26 +278,32 @@ class JSONResponseEncoder(JSONEncoder):
         output_exists = managed_stream.output_file_exists
         tx_height = managed_stream.stream_claim_info.height
         best_height = self.ledger.headers.height
-        return {
-            'streaming_url': managed_stream.stream_url,
+        is_stream = hasattr(managed_stream, 'stream_hash')
+        if is_stream:
+            total_bytes_lower_bound = managed_stream.descriptor.lower_bound_decrypted_length()
+            total_bytes = managed_stream.descriptor.upper_bound_decrypted_length()
+        else:
+            total_bytes_lower_bound = total_bytes = managed_stream.torrent_length
+        result = {
+            'streaming_url': None,
             'completed': managed_stream.completed,
-            'file_name': managed_stream.file_name if output_exists else None,
-            'download_directory': managed_stream.download_directory if output_exists else None,
-            'download_path': managed_stream.full_path if output_exists else None,
+            'file_name': None,
+            'download_directory': None,
+            'download_path': None,
             'points_paid': 0.0,
             'stopped': not managed_stream.running,
-            'stream_hash': managed_stream.stream_hash,
-            'stream_name': managed_stream.descriptor.stream_name,
-            'suggested_file_name': managed_stream.descriptor.suggested_file_name,
-            'sd_hash': managed_stream.descriptor.sd_hash,
-            'mime_type': managed_stream.mime_type,
-            'key': managed_stream.descriptor.key,
-            'total_bytes_lower_bound': managed_stream.descriptor.lower_bound_decrypted_length(),
-            'total_bytes': managed_stream.descriptor.upper_bound_decrypted_length(),
+            'stream_hash': None,
+            'stream_name': None,
+            'suggested_file_name': None,
+            'sd_hash': None,
+            'mime_type': None,
+            'key': None,
+            'total_bytes_lower_bound': total_bytes_lower_bound,
+            'total_bytes': total_bytes,
             'written_bytes': managed_stream.written_bytes,
-            'blobs_completed': managed_stream.blobs_completed,
-            'blobs_in_stream': managed_stream.blobs_in_stream,
-            'blobs_remaining': managed_stream.blobs_remaining,
+            'blobs_completed': None,
+            'blobs_in_stream': None,
+            'blobs_remaining': None,
             'status': managed_stream.status,
             'claim_id': managed_stream.claim_id,
             'txid': managed_stream.txid,
@@ -273,11 +315,42 @@ class JSONResponseEncoder(JSONEncoder):
             'channel_name': managed_stream.channel_name,
             'claim_name': managed_stream.claim_name,
             'content_fee': managed_stream.content_fee,
+            'purchase_receipt': self.encode_output(managed_stream.purchase_receipt),
             'added_on': managed_stream.added_on,
             'height': tx_height,
             'confirmations': (best_height + 1) - tx_height if tx_height > 0 else tx_height,
-            'timestamp': self.ledger.headers[tx_height]['timestamp'] if 0 < tx_height <= best_height else None
+            'timestamp': self.ledger.headers.estimated_timestamp(tx_height),
+            'is_fully_reflected': False,
+            'reflector_progress': False,
+            'uploading_to_reflector': False
         }
+        if is_stream:
+            result.update({
+                'streaming_url': managed_stream.stream_url,
+                'stream_hash': managed_stream.stream_hash,
+                'stream_name': managed_stream.stream_name,
+                'suggested_file_name': managed_stream.suggested_file_name,
+                'sd_hash': managed_stream.descriptor.sd_hash,
+                'mime_type': managed_stream.mime_type,
+                'key': managed_stream.descriptor.key,
+                'blobs_completed': managed_stream.blobs_completed,
+                'blobs_in_stream': managed_stream.blobs_in_stream,
+                'blobs_remaining': managed_stream.blobs_remaining,
+                'is_fully_reflected': managed_stream.is_fully_reflected,
+                'reflector_progress': managed_stream.reflector_progress,
+                'uploading_to_reflector': managed_stream.uploading_to_reflector
+            })
+        else:
+            result.update({
+                'streaming_url': f'file://{managed_stream.full_path}',
+            })
+        if output_exists:
+            result.update({
+                'file_name': managed_stream.file_name,
+                'download_directory': managed_stream.download_directory,
+                'download_path': managed_stream.full_path,
+            })
+        return result
 
     def encode_claim(self, claim):
         encoded = getattr(claim, claim.claim_type).to_dict()

lbry/extras/daemon/migrator/dbmigrator.py

@@ -31,6 +31,14 @@ def migrate_db(conf, start, end):
             from .migrate10to11 import do_migration
         elif current == 11:
             from .migrate11to12 import do_migration
+        elif current == 12:
+            from .migrate12to13 import do_migration
+        elif current == 13:
+            from .migrate13to14 import do_migration
+        elif current == 14:
+            from .migrate14to15 import do_migration
+        elif current == 15:
+            from .migrate15to16 import do_migration
         else:
             raise Exception(f"DB migration of version {current} to {current+1} is not available")
         try:

lbry/extras/daemon/migrator/migrate12to13.py

@@ -0,0 +1,80 @@
+import os
+import sqlite3
+
+
+def do_migration(conf):
+    db_path = os.path.join(conf.data_dir, "lbrynet.sqlite")
+    connection = sqlite3.connect(db_path)
+    cursor = connection.cursor()
+
+    current_columns = []
+    for col_info in cursor.execute("pragma table_info('file');").fetchall():
+        current_columns.append(col_info[1])
+    if 'bt_infohash' in current_columns:
+        connection.close()
+        print("already migrated")
+        return
+
+    cursor.executescript("""
+        pragma foreign_keys=off;
+
+        create table if not exists torrent (
+            bt_infohash char(20) not null primary key,
+            tracker text,
+            length integer not null,
+            name text not null
+        );
+
+        create table if not exists torrent_node ( -- BEP-0005
+            bt_infohash char(20) not null references torrent,
+            host text not null,
+            port integer not null
+        );
+
+        create table if not exists torrent_tracker ( -- BEP-0012
+            bt_infohash char(20) not null references torrent,
+            tracker text not null
+        );
+
+        create table if not exists torrent_http_seed ( -- BEP-0017
+            bt_infohash char(20) not null references torrent,
+            http_seed text not null
+        );
+
+        create table if not exists new_file (
+            stream_hash char(96) references stream,
+            bt_infohash char(20) references torrent,
+            file_name text,
+            download_directory text,
+            blob_data_rate real not null,
+            status text not null,
+            saved_file integer not null,
+            content_fee text,
+            added_on integer not null
+        );
+
+        create table if not exists new_content_claim (
+            stream_hash char(96) references stream,
+            bt_infohash char(20) references torrent,
+            claim_outpoint text unique not null references claim
+        );
+
+        insert into new_file (stream_hash, bt_infohash, file_name, download_directory, blob_data_rate, status,
+            saved_file, content_fee, added_on) select
+                stream_hash, NULL, file_name, download_directory, blob_data_rate, status, saved_file, content_fee,
+                added_on
+            from file;
+
+        insert or ignore into new_content_claim (stream_hash, bt_infohash, claim_outpoint)
+            select stream_hash, NULL, claim_outpoint from content_claim;
+
+        drop table file;
+        drop table content_claim;
+        alter table new_file rename to file;
+        alter table new_content_claim rename to content_claim;
+
+        pragma foreign_keys=on;
+    """)
+
+    connection.commit()
+    connection.close()

lbry/extras/daemon/migrator/migrate13to14.py

@@ -0,0 +1,21 @@
+import os
+import sqlite3
+
+
+def do_migration(conf):
+    db_path = os.path.join(conf.data_dir, "lbrynet.sqlite")
+    connection = sqlite3.connect(db_path)
+    cursor = connection.cursor()
+
+    cursor.executescript("""
+        create table if not exists peer (
+            node_id char(96) not null primary key,
+            address text not null,
+            udp_port integer not null,
+            tcp_port integer,
+            unique (address, udp_port)
+        );
+    """)
+
+    connection.commit()
+    connection.close()

lbry/extras/daemon/migrator/migrate14to15.py

@@ -0,0 +1,16 @@
+import os
+import sqlite3
+
+
+def do_migration(conf):
+    db_path = os.path.join(conf.data_dir, "lbrynet.sqlite")
+    connection = sqlite3.connect(db_path)
+    cursor = connection.cursor()
+
+    cursor.executescript("""
+        alter table blob add column added_on integer not null default 0;
+        alter table blob add column is_mine integer not null default 1;
+    """)
+
+    connection.commit()
+    connection.close()