Compare commits

...

39 commits

Author SHA1 Message Date
Mark Beamer Jr
c2230cdefb
Send parent id and comment text to internal-apis for notifications 2020-07-24 15:43:34 -04:00
Mark
1c3a25f82d
Merge pull request #47 from lbryio/fix_notifications
Fix notifications
2020-07-21 11:10:28 -04:00
Mark Beamer Jr
7c26b80971
Fix notifications 2020-07-21 02:41:29 -04:00
Oleg Silkin
3cce89cbac serializes error messages into json format 2020-04-09 16:55:07 -04:00
Oleg Silkin
b4377b2f54 Cleans up & adds todos for future 2020-04-07 16:16:34 -04:00
Oleg Silkin
3b91279cc7 Moves schema from SQLite3 to MySQL 2020-04-07 16:16:15 -04:00
Oleg Silkin
0817b70083 Update README.md 2020-04-07 16:08:29 -04:00
Oleg Silkin
7b7e6c66ac VARCHAR -> CHAR for all ID fields, constraints moved out of table definition 2020-04-07 15:04:50 -04:00
Oleg Silkin
c7e8d274f7 update travis 2020-04-06 19:26:22 -04:00
Oleg Silkin
be45a70c36 sets column names to be lowercase, uses utf8mb4 charset, utf8mb4_unicode_ci collation 2020-04-06 19:15:54 -04:00
Oleg Silkin
d25e03d853 database name gets set to social 2020-04-06 19:13:12 -04:00
Oleg Silkin
80f77218f9
Merge pull request #39 from lbryio/orm-rewrite
Replace database methods with peewee ORM
2020-04-03 17:40:42 -04:00
Oleg Silkin
75c8f82072 Add todos 2020-04-03 16:40:05 -04:00
Oleg Silkin
38e9af24ec Update .gitignore 2020-04-03 16:07:23 -04:00
Oleg Silkin
a84e0e0f84 Add docker-compose.yml 2020-04-03 15:38:29 -04:00
Oleg Silkin
fd05ea1145 Update travis config & add generic yaml config 2020-04-03 15:33:23 -04:00
Oleg Silkin
c27baa89fe rate limit comments 2020-04-01 18:52:36 -04:00
Oleg Silkin
84cafd643f replace oracle driver with pymysql driver 2020-04-01 18:50:09 -04:00
Oleg Silkin
3b6b052000 Removes incorrect method for get_channel_from_comment_id 2020-04-01 18:49:37 -04:00
Oleg Silkin
20f9ccc8c5 Moves to using CharField in place of TextField 2020-04-01 18:49:03 -04:00
Oleg Silkin
08060a71d3 Implements hide_comments method using ORM 2020-04-01 18:13:28 -04:00
Oleg Silkin
c852697c94 reimplements abandon method using ORM; no longer require channel_id param 2020-04-01 18:12:46 -04:00
Oleg Silkin
8f12d997ae json -> yaml 2020-03-31 13:59:37 -04:00
Oleg Silkin
a6f056821f Adds a strict validator which fails upon possible FK violations 2020-03-31 13:59:20 -04:00
Oleg Silkin
bd06d1c992 Improves database configuration 2020-03-31 13:57:30 -04:00
Oleg Silkin
1153711636 Reimplements API methods using ORM 2020-03-31 13:55:44 -04:00
Oleg Silkin
e0b6d16c89 Removes certain sqlite-specific functions from application 2020-03-30 22:00:10 -04:00
Oleg Silkin
dba14460cc Unittests using peewee binds instead of direct sqlite connection 2020-03-30 21:59:26 -04:00
Oleg Silkin
45733d2dc4 Remove queries.py 2020-03-30 21:58:37 -04:00
Oleg Silkin
aee12eba54 removes inherited base database in favor of binding api 2020-03-30 21:58:07 -04:00
Oleg Silkin
a22b4a9162 Adds delete, edit, and hide operations 2020-03-30 18:04:20 -04:00
Oleg Silkin
8138e71668 Sets database dynamically 2020-03-30 18:02:20 -04:00
Oleg Silkin
510f2a5d29 Adds comment create logic 2020-03-27 17:44:51 -04:00
Oleg Silkin
63f2c7e9e0 Moves the clean method to misc.py, moves misc.py 2020-03-27 17:44:22 -04:00
Oleg Silkin
644e5e8477 adds get_channel_from_id & simplifies to all call comment_list 2020-03-27 15:59:57 -04:00
Oleg Silkin
cc20088b06 Implements all search queries 2020-03-27 05:07:56 -04:00
Oleg Silkin
a581425a64 Shifts from JSON configuration to yml based 2020-03-27 01:26:13 -04:00
Oleg Silkin
0529fa7d01 Implements DB model and preliminary select queries 2020-03-27 01:24:34 -04:00
Oleg Silkin
464fc88d8e
Merge pull request #35 from lbryio/disables-anon
Disables Anonymous Comments
2020-03-03 20:49:42 -05:00
25 changed files with 861 additions and 788 deletions

4
.gitignore vendored
View file

@ -1 +1,3 @@
config/conf.json
config/conf.yml
docker-compose.yml

View file

@ -1,7 +1,28 @@
sudo: required
language: python
dist: xenial
python: 3.7
python: 3.8
# for docker-compose
services:
- docker
# to avoid "well it works on my computer" moments
env:
- DOCKER_COMPOSE_VERSION=1.25.4
before_install:
# ensure docker-compose version is as specified above
- sudo rm /usr/local/bin/docker-compose
- curl -L https://github.com/docker/compose/releases/download/${DOCKER_COMPOSE_VERSION}/docker-compose-`uname -s`-`uname -m` > docker-compose
- chmod +x docker-compose
- sudo mv docker-compose /usr/local/bin
# refresh docker images
- sudo apt-get update
before_script:
- docker-compose up -d
jobs:
include:
@ -9,6 +30,5 @@ jobs:
name: "Unit Tests"
install:
- pip install -e .
- mkdir database
script:
- python -m unittest

View file

@ -3,34 +3,36 @@
[![Build Status](https://travis-ci.com/lbryio/comment-server.svg?branch=master)](https://travis-ci.com/lbryio/comment-server)
[![Maintainability](https://api.codeclimate.com/v1/badges/22f420b8b5f2373fd885/maintainability)](https://codeclimate.com/github/lbryio/comment-server/maintainability)
This is the code for the LBRY Comment Server.
Fork it, run it, set it on fire. Up to you.
## Before Installing
Comment Deletion requires having the [`lbry-sdk`](https://github.com/lbryio/lbry-sdk)
Install the [`lbry-sdk`](https://github.com/lbryio/lbry-sdk)
in order to validate & properly delete comments.
## Installation
#### Installing the server:
```bash
$ git clone https://github.com/osilkin98/comment-server
$ git clone https://github.com/lbryio/comment-server
$ cd comment-server
# create a virtual environment
$ virtualenv --python=python3 venv
$ virtualenv --python=python3.8 venv
# Enter the virtual environment
$ source venv/bin/activate
# install the Server as a Executable Target
(venv) $ python setup.py develop
# Install required dependencies
(venv) $ pip install -e .
# Run the server
(venv) $ python src/main.py \
--port=5921 \ # use a different port besides the default
--config=conf.yml \ # provide a custom config file
& \ # detach and run the service in the background
```
### Installing the systemd Service Monitor
@ -70,16 +72,11 @@ To Test the database, simply run:
There are basic tests to run against the server, though they require
that there is a server instance running, though the database
chosen may have to be edited in `config/conf.json`.
chosen may have to be edited in `config/conf.yml`.
Additionally there are HTTP requests that can be send with whatever
software you choose to test the integrity of the comment server.
## Schema
![schema](schema.png)
## Contributing
Contributions are welcome, verbosity is encouraged. Please be considerate

31
config/conf.yml Normal file
View file

@ -0,0 +1,31 @@
---
# for running local-tests without using MySQL for now
testing:
database: sqlite
file: comments.db
pragmas:
journal_mode: wal
cache_size: 64000
foreign_keys: 0
ignore_check_constraints: 1
synchronous: 0
# actual database should be running MySQL
production:
charset: utf8mb4
database: mysql
name: social
user: lbry
password: lbry
host: localhost
port: 3306
mode: production
logging:
format: "%(asctime)s | %(levelname)s | %(name)s | %(module)s.%(funcName)s:%(lineno)d
| %(message)s"
aiohttp_format: "%(asctime)s | %(levelname)s | %(name)s | %(message)s"
datefmt: "%Y-%m-%d %H:%M:%S"
host: localhost
port: 5921
lbrynet: http://localhost:5279

25
docker-compose.yml Normal file
View file

@ -0,0 +1,25 @@
version: "3.7"
services:
###########
## MySQL ##
###########
mysql:
image: mysql/mysql-server:5.7.27
restart: "no"
command: --character_set_server=utf8mb4 --max_allowed_packet=1073741824
ports:
- "3306:3306"
environment:
- MYSQL_ALLOW_EMPTY_PASSWORD=true
- MYSQL_DATABASE=social
- MYSQL_USER=lbry
- MYSQL_PASSWORD=lbry
- MYSQL_LOG_CONSOLE=true
#############
## Adminer ##
#############
adminer:
image: adminer
restart: always
ports:
- 8080:8080

View file

@ -1,84 +0,0 @@
import sqlite3
import time
import faker
from faker.providers import misc
fake = faker.Faker()
fake.add_provider(misc)
if __name__ == '__main__':
song_time = """One, two, three!
My baby don't mess around
'Cause she loves me so
This I know fo sho!
But does she really wanna
But can't stand to see me walk out tha door
Don't try to fight the feeling
Because the thought alone is killin' me right now
Thank God for Mom and Dad
For sticking to together
Like we don't know how
Hey ya! Hey ya!
Hey ya! Hey ya!
Hey ya! Hey ya!
Hey ya! Hey ya!
You think you've got it
Oh, you think you've got it
But got it just don't get it when there's nothin' at all
We get together
Oh, we get together
But separate's always better when there's feelings involved
Know what they say -its
Nothing lasts forever!
Then what makes it, then what makes it
Then what makes it, then what makes it
Then what makes love the exception?
So why, oh, why, oh
Why, oh, why, oh, why, oh
Are we still in denial when we know we're not happy here
Hey ya! (y'all don't want to here me, ya just want to dance) Hey ya!
Don't want to meet your daddy (oh ohh), just want you in my caddy (oh ohh)
Hey ya! (oh, oh!) Hey ya! (oh, oh!)
Don't want to meet your momma, just want to make you cum-a (oh, oh!)
I'm (oh, oh) I'm (oh, oh) I'm just being honest! (oh, oh)
I'm just being honest!
Hey! alright now! alright now, fellas!
Yea?
Now, what cooler than being cool?
Ice cold!
I can't hear ya! I say what's, what's cooler than being cool?
Ice cold!
Alright alright alright alright alright alright alright alright alright alright alright alright alright alright alright alright!
Okay, now ladies!
Yea?
Now we gonna break this thang down for just a few seconds
Now don't have me break this thang down for nothin'
I want to see you on your badest behavior!
Lend me some sugar, I am your neighbor!
Ah! Here we go now,
Shake it, shake it, shake it, shake it, shake it
Shake it, shake it, shake it, shake it
Shake it like a Polaroid picture! Hey ya!
Shake it, shake it, shake it, shake it, shake it
Shake it, shake it, shake it, suga!
Shake it like a Polaroid picture!
Now all the Beyonce's, and Lucy Lu's, and baby dolls
Get on tha floor get on tha floor!
Shake it like a Polaroid picture!
Oh, you! oh, you!
Hey ya!(oh, oh) Hey ya!(oh, oh)
Hey ya!(oh, oh) Hey ya!(oh, oh)
Hey ya!(oh, oh) Hey ya!(oh, oh)"""
song = song_time.split('\n')
claim_id = '2aa106927b733e2602ffb565efaccc78c2ed89df'
run_len = [(fake.sha256(), song_time, claim_id, str(int(time.time()))) for k in range(5000)]
conn = sqlite3.connect('database/default_test.db')
with conn:
curs = conn.executemany("""
INSERT INTO COMMENT(CommentId, Body, LbryClaimId, Timestamp) VALUES (?, ?, ?, ?)
""", run_len)
print(f'rows changed: {curs.rowcount}')

View file

@ -2,11 +2,12 @@ import binascii
import logging
import hashlib
import json
# todo: remove sqlite3 as a dependency
import sqlite3
import asyncio
import aiohttp
from server.validation import is_signature_valid, get_encoded_signature
from src.server.validation import is_signature_valid, get_encoded_signature
logger = logging.getLogger(__name__)

View file

@ -14,15 +14,17 @@ setup(
data_files=[('config', ['config/conf.json',])],
include_package_data=True,
install_requires=[
'pymysql',
'pyyaml',
'Faker>=1.0.7',
'asyncio>=3.4.3',
'aiohttp==3.5.4',
'aiojobs==0.2.2',
'asyncio',
'aiohttp',
'aiojobs',
'ecdsa>=0.13.3',
'cryptography==2.5',
'aiosqlite==0.10.0',
'PyNaCl>=1.3.0',
'requests',
'cython',
'peewee'
]
)

View file

@ -1,76 +1,50 @@
PRAGMA FOREIGN_KEYS = ON;
USE `social`;
ALTER DATABASE `social`
DEFAULT CHARACTER SET utf8mb4
DEFAULT COLLATE utf8mb4_unicode_ci;
-- Although I know this file is unnecessary, I like keeping it around.
DROP TABLE IF EXISTS `CHANNEL`;
CREATE TABLE `CHANNEL` (
`claimid` VARCHAR(40) NOT NULL,
`name` CHAR(255) CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci NOT NULL,
CONSTRAINT `channel_pk` PRIMARY KEY (`claimid`)
)
CHARACTER SET utf8mb4
COLLATE utf8mb4_unicode_ci;
-- I'm not gonna remove it.
DROP TABLE IF EXISTS `COMMENT`;
CREATE TABLE `COMMENT` (
-- should be changed to CHAR(64)
`commentid` CHAR(64) NOT NULL,
-- should be changed to CHAR(40)
`lbryclaimid` CHAR(40) NOT NULL,
-- can be null, so idk if this should be char(40)
`channelid` CHAR(40) DEFAULT NULL,
`body` TEXT
CHARACTER SET utf8mb4
COLLATE utf8mb4_unicode_ci
NOT NULL,
`parentid` CHAR(64) DEFAULT NULL,
`signature` CHAR(128) DEFAULT NULL,
-- 22 chars long is prolly enough
`signingts` VARCHAR(22) DEFAULT NULL,
-- tables
CREATE TABLE IF NOT EXISTS COMMENT
(
CommentId TEXT NOT NULL,
LbryClaimId TEXT NOT NULL,
ChannelId TEXT DEFAULT NULL,
Body TEXT NOT NULL,
ParentId TEXT DEFAULT NULL,
Signature TEXT DEFAULT NULL,
Timestamp INTEGER NOT NULL,
SigningTs TEXT DEFAULT NULL,
IsHidden BOOLEAN NOT NULL DEFAULT FALSE,
CONSTRAINT COMMENT_PRIMARY_KEY PRIMARY KEY (CommentId) ON CONFLICT IGNORE,
CONSTRAINT COMMENT_SIGNATURE_SK UNIQUE (Signature) ON CONFLICT ABORT,
CONSTRAINT COMMENT_CHANNEL_FK FOREIGN KEY (ChannelId) REFERENCES CHANNEL (ClaimId)
ON DELETE NO ACTION ON UPDATE NO ACTION,
CONSTRAINT COMMENT_PARENT_FK FOREIGN KEY (ParentId) REFERENCES COMMENT (CommentId)
ON UPDATE CASCADE ON DELETE NO ACTION -- setting null implies comment is top level
);
-- ALTER TABLE COMMENT ADD COLUMN IsHidden BOOLEAN DEFAULT (FALSE);
-- ALTER TABLE COMMENT ADD COLUMN SigningTs TEXT DEFAULT NULL;
-- DROP TABLE IF EXISTS CHANNEL;
CREATE TABLE IF NOT EXISTS CHANNEL
(
ClaimId TEXT NOT NULL,
Name TEXT NOT NULL,
CONSTRAINT CHANNEL_PK PRIMARY KEY (ClaimId)
ON CONFLICT IGNORE
);
`timestamp` INTEGER NOT NULL,
-- there's no way that the timestamp will ever reach 22 characters
`ishidden` BOOLEAN DEFAULT FALSE,
CONSTRAINT `COMMENT_PRIMARY_KEY` PRIMARY KEY (`commentid`)
-- setting null implies comment is top level
)
CHARACTER SET utf8mb4
COLLATE utf8mb4_unicode_ci;
-- indexes
-- DROP INDEX IF EXISTS COMMENT_CLAIM_INDEX;
-- CREATE INDEX IF NOT EXISTS CLAIM_COMMENT_INDEX ON COMMENT (LbryClaimId, CommentId);
ALTER TABLE COMMENT
ADD CONSTRAINT `comment_channel_fk` FOREIGN KEY (`channelid`) REFERENCES `CHANNEL` (`claimid`)
ON DELETE CASCADE ON UPDATE CASCADE,
ADD CONSTRAINT `comment_parent_fk` FOREIGN KEY (`parentid`) REFERENCES `COMMENT` (`commentid`)
ON UPDATE CASCADE ON DELETE CASCADE
;
-- CREATE INDEX IF NOT EXISTS CHANNEL_COMMENT_INDEX ON COMMENT (ChannelId, CommentId);
-- VIEWS
CREATE VIEW IF NOT EXISTS COMMENTS_ON_CLAIMS AS
SELECT C.CommentId AS comment_id,
C.Body AS comment,
C.LbryClaimId AS claim_id,
C.Timestamp AS timestamp,
CHAN.Name AS channel_name,
CHAN.ClaimId AS channel_id,
('lbry://' || CHAN.Name || '#' || CHAN.ClaimId) AS channel_url,
C.Signature AS signature,
C.SigningTs AS signing_ts,
C.ParentId AS parent_id,
C.IsHidden AS is_hidden
FROM COMMENT AS C
LEFT OUTER JOIN CHANNEL CHAN ON C.ChannelId = CHAN.ClaimId
ORDER BY C.Timestamp DESC;
DROP VIEW IF EXISTS COMMENT_REPLIES;
CREATE VIEW IF NOT EXISTS COMMENT_REPLIES (Author, CommentBody, ParentAuthor, ParentCommentBody) AS
SELECT AUTHOR.Name, OG.Body, PCHAN.Name, PARENT.Body
FROM COMMENT AS OG
JOIN COMMENT AS PARENT
ON OG.ParentId = PARENT.CommentId
JOIN CHANNEL AS PCHAN ON PARENT.ChannelId = PCHAN.ClaimId
JOIN CHANNEL AS AUTHOR ON OG.ChannelId = AUTHOR.ClaimId
ORDER BY OG.Timestamp;
-- this is the default channel for anyone who wants to publish anonymously
-- INSERT INTO CHANNEL
-- VALUES ('9cb713f01bf247a0e03170b5ed00d5161340c486', '@Anonymous');
CREATE INDEX `claim_comment_index` ON `COMMENT` (`lbryclaimid`, `commentid`);
CREATE INDEX `channel_comment_index` ON `COMMENT` (`channelid`, `commentid`);

217
src/database/models.py Normal file
View file

@ -0,0 +1,217 @@
import json
import time
import logging
import math
import typing
from peewee import *
import nacl.hash
from src.server.validation import is_valid_base_comment
from src.misc import clean
class Channel(Model):
claim_id = FixedCharField(column_name='claimid', primary_key=True, max_length=40)
name = CharField(column_name='name', max_length=255)
class Meta:
table_name = 'CHANNEL'
class Comment(Model):
comment = TextField(column_name='body')
channel = ForeignKeyField(
backref='comments',
column_name='channelid',
field='claim_id',
model=Channel,
null=True
)
comment_id = FixedCharField(column_name='commentid', primary_key=True, max_length=64)
is_hidden = BooleanField(column_name='ishidden', constraints=[SQL("DEFAULT 0")])
claim_id = FixedCharField(max_length=40, column_name='lbryclaimid')
parent = ForeignKeyField(
column_name='ParentId',
field='comment_id',
model='self',
null=True,
backref='replies'
)
signature = FixedCharField(max_length=128, column_name='signature', null=True, unique=True)
signing_ts = TextField(column_name='signingts', null=True)
timestamp = IntegerField(column_name='timestamp')
class Meta:
table_name = 'COMMENT'
indexes = (
(('channel', 'comment_id'), False),
(('claim_id', 'comment_id'), False),
)
FIELDS = {
'comment': Comment.comment,
'comment_id': Comment.comment_id,
'claim_id': Comment.claim_id,
'timestamp': Comment.timestamp,
'signature': Comment.signature,
'signing_ts': Comment.signing_ts,
'is_hidden': Comment.is_hidden,
'parent_id': Comment.parent.alias('parent_id'),
'channel_id': Channel.claim_id.alias('channel_id'),
'channel_name': Channel.name.alias('channel_name'),
'channel_url': ('lbry://' + Channel.name + '#' + Channel.claim_id).alias('channel_url')
}
def comment_list(claim_id: str = None, parent_id: str = None,
top_level: bool = False, exclude_mode: str = None,
page: int = 1, page_size: int = 50, expressions=None,
select_fields: list = None, exclude_fields: list = None) -> dict:
fields = FIELDS.keys()
if exclude_fields:
fields -= set(exclude_fields)
if select_fields:
fields &= set(select_fields)
attributes = [FIELDS[field] for field in fields]
query = Comment.select(*attributes)
# todo: allow this process to be more automated, so it can just be an expression
if claim_id:
query = query.where(Comment.claim_id == claim_id)
if top_level:
query = query.where(Comment.parent.is_null())
if parent_id:
query = query.where(Comment.ParentId == parent_id)
if exclude_mode:
show_hidden = exclude_mode.lower() == 'hidden'
query = query.where((Comment.is_hidden == show_hidden))
if expressions:
query = query.where(expressions)
total = query.count()
query = (query
.join(Channel, JOIN.LEFT_OUTER)
.order_by(Comment.timestamp.desc())
.paginate(page, page_size))
items = [clean(item) for item in query.dicts()]
# has_hidden_comments is deprecated
data = {
'page': page,
'page_size': page_size,
'total_pages': math.ceil(total / page_size),
'total_items': total,
'items': items,
'has_hidden_comments': exclude_mode is not None and exclude_mode == 'hidden',
}
return data
def get_comment(comment_id: str) -> dict:
try:
comment = comment_list(expressions=(Comment.comment_id == comment_id), page_size=1).get('items').pop()
except IndexError:
raise ValueError(f'Comment does not exist with id {comment_id}')
else:
return comment
def create_comment_id(comment: str, channel_id: str, timestamp: int):
# We convert the timestamp from seconds into minutes
# to prevent spammers from commenting the same BS everywhere.
nearest_minute = str(math.floor(timestamp / 60))
# don't use claim_id for the comment_id anymore so comments
# are not unique to just one claim
prehash = b':'.join([
comment.encode(),
channel_id.encode(),
nearest_minute.encode()
])
return nacl.hash.sha256(prehash).decode()
def create_comment(comment: str = None, claim_id: str = None,
parent_id: str = None, channel_id: str = None,
channel_name: str = None, signature: str = None,
signing_ts: str = None) -> dict:
if not is_valid_base_comment(
comment=comment,
claim_id=claim_id,
parent_id=parent_id,
channel_id=channel_id,
channel_name=channel_name,
signature=signature,
signing_ts=signing_ts
):
raise ValueError('Invalid Parameters given for comment')
channel, _ = Channel.get_or_create(name=channel_name, claim_id=channel_id)
if parent_id and not claim_id:
parent: Comment = Comment.get_by_id(parent_id)
claim_id = parent.claim_id
timestamp = int(time.time())
comment_id = create_comment_id(comment, channel_id, timestamp)
new_comment = Comment.create(
claim_id=claim_id,
comment_id=comment_id,
comment=comment,
parent=parent_id,
channel=channel,
signature=signature,
signing_ts=signing_ts,
timestamp=timestamp
)
return get_comment(new_comment.comment_id)
def delete_comment(comment_id: str) -> bool:
try:
comment: Comment = Comment.get_by_id(comment_id)
except DoesNotExist as e:
raise ValueError from e
else:
return 0 < comment.delete_instance(True, delete_nullable=True)
def edit_comment(comment_id: str, new_comment: str, new_sig: str, new_ts: str) -> bool:
try:
comment: Comment = Comment.get_by_id(comment_id)
except DoesNotExist as e:
raise ValueError from e
else:
comment.comment = new_comment
comment.signature = new_sig
comment.signing_ts = new_ts
# todo: add a 'last-modified' timestamp
comment.timestamp = int(time.time())
return comment.save() > 0
def set_hidden_flag(comment_ids: typing.List[str], hidden=True) -> bool:
# sets `is_hidden` flag for all `comment_ids` to the `hidden` param
update = (Comment
.update(is_hidden=hidden)
.where(Comment.comment_id.in_(comment_ids)))
return update.execute() > 0
if __name__ == '__main__':
logger = logging.getLogger('peewee')
logger.addHandler(logging.StreamHandler())
logger.setLevel(logging.DEBUG)
comments = comment_list(
page_size=20,
expressions=((Comment.timestamp < 1583272089) &
(Comment.claim_id ** '420%'))
)
print(json.dumps(comments, indent=4))

View file

@ -1,290 +0,0 @@
import atexit
import logging
import math
import sqlite3
import time
import typing
import nacl.hash
from src.database.schema import CREATE_TABLES_QUERY
logger = logging.getLogger(__name__)
SELECT_COMMENTS_ON_CLAIMS = """
SELECT comment, comment_id, claim_id, timestamp, is_hidden, parent_id,
channel_name, channel_id, channel_url, signature, signing_ts
FROM COMMENTS_ON_CLAIMS
"""
def clean(thing: dict) -> dict:
if 'is_hidden' in thing:
thing.update({'is_hidden': bool(thing['is_hidden'])})
return {k: v for k, v in thing.items() if v is not None}
def obtain_connection(filepath: str = None, row_factory: bool = True):
connection = sqlite3.connect(filepath)
if row_factory:
connection.row_factory = sqlite3.Row
return connection
def get_claim_comments(conn: sqlite3.Connection, claim_id: str, parent_id: str = None,
page: int = 1, page_size: int = 50, top_level=False):
with conn:
if top_level:
results = [clean(dict(row)) for row in conn.execute(
SELECT_COMMENTS_ON_CLAIMS + " WHERE claim_id = ? AND parent_id IS NULL LIMIT ? OFFSET ?",
(claim_id, page_size, page_size * (page - 1))
)]
count = conn.execute(
"SELECT COUNT(*) FROM COMMENTS_ON_CLAIMS WHERE claim_id = ? AND parent_id IS NULL",
(claim_id,)
)
elif parent_id is None:
results = [clean(dict(row)) for row in conn.execute(
SELECT_COMMENTS_ON_CLAIMS + "WHERE claim_id = ? LIMIT ? OFFSET ? ",
(claim_id, page_size, page_size * (page - 1))
)]
count = conn.execute(
"SELECT COUNT(*) FROM COMMENTS_ON_CLAIMS WHERE claim_id = ?",
(claim_id,)
)
else:
results = [clean(dict(row)) for row in conn.execute(
SELECT_COMMENTS_ON_CLAIMS + "WHERE claim_id = ? AND parent_id = ? LIMIT ? OFFSET ? ",
(claim_id, parent_id, page_size, page_size * (page - 1))
)]
count = conn.execute(
"SELECT COUNT(*) FROM COMMENTS_ON_CLAIMS WHERE claim_id = ? AND parent_id = ?",
(claim_id, parent_id)
)
count = tuple(count.fetchone())[0]
return {
'items': results,
'page': page,
'page_size': page_size,
'total_pages': math.ceil(count / page_size),
'total_items': count,
'has_hidden_comments': claim_has_hidden_comments(conn, claim_id)
}
def get_claim_hidden_comments(conn: sqlite3.Connection, claim_id: str, hidden=True, page=1, page_size=50):
with conn:
results = conn.execute(
SELECT_COMMENTS_ON_CLAIMS + "WHERE claim_id = ? AND is_hidden IS ? LIMIT ? OFFSET ?",
(claim_id, hidden, page_size, page_size * (page - 1))
)
count = conn.execute(
"SELECT COUNT(*) FROM COMMENTS_ON_CLAIMS WHERE claim_id = ? AND is_hidden IS ?", (claim_id, hidden)
)
results = [clean(dict(row)) for row in results.fetchall()]
count = tuple(count.fetchone())[0]
return {
'items': results,
'page': page,
'page_size': page_size,
'total_pages': math.ceil(count / page_size),
'total_items': count,
'has_hidden_comments': claim_has_hidden_comments(conn, claim_id)
}
def claim_has_hidden_comments(conn, claim_id):
with conn:
result = conn.execute(
"SELECT COUNT(DISTINCT is_hidden) FROM COMMENTS_ON_CLAIMS WHERE claim_id = ? AND is_hidden IS 1",
(claim_id,)
)
return bool(tuple(result.fetchone())[0])
def insert_comment(conn: sqlite3.Connection, claim_id: str, comment: str,
channel_id: str = None, signature: str = None, signing_ts: str = None, **extra) -> str:
timestamp = int(time.time())
prehash = b':'.join((claim_id.encode(), comment.encode(), str(timestamp).encode(),))
comment_id = nacl.hash.sha256(prehash).decode()
with conn:
curs = conn.execute(
"""
INSERT INTO COMMENT(CommentId, LbryClaimId, ChannelId, Body, ParentId,
Timestamp, Signature, SigningTs, IsHidden)
VALUES (:comment_id, :claim_id, :channel_id, :comment, NULL,
:timestamp, :signature, :signing_ts, 0) """,
{
'comment_id': comment_id,
'claim_id': claim_id,
'channel_id': channel_id,
'comment': comment,
'timestamp': timestamp,
'signature': signature,
'signing_ts': signing_ts
}
)
logging.info('attempted to insert comment with comment_id [%s] | %d rows affected', comment_id, curs.rowcount)
return comment_id
def insert_reply(conn: sqlite3.Connection, comment: str, parent_id: str,
channel_id: str = None, signature: str = None,
signing_ts: str = None, **extra) -> str:
timestamp = int(time.time())
prehash = b':'.join((parent_id.encode(), comment.encode(), str(timestamp).encode(),))
comment_id = nacl.hash.sha256(prehash).decode()
with conn:
curs = conn.execute(
"""
INSERT INTO COMMENT
(CommentId, LbryClaimId, ChannelId, Body, ParentId, Signature, Timestamp, SigningTs, IsHidden)
SELECT :comment_id, LbryClaimId, :channel_id, :comment, :parent_id, :signature, :timestamp, :signing_ts, 0
FROM COMMENT WHERE CommentId = :parent_id
""", {
'comment_id': comment_id,
'parent_id': parent_id,
'timestamp': timestamp,
'comment': comment,
'channel_id': channel_id,
'signature': signature,
'signing_ts': signing_ts
}
)
logging.info('attempted to insert reply with comment_id [%s] | %d rows affected', comment_id, curs.rowcount)
return comment_id
def get_comment_or_none(conn: sqlite3.Connection, comment_id: str) -> dict:
with conn:
curry = conn.execute(SELECT_COMMENTS_ON_CLAIMS + "WHERE comment_id = ?", (comment_id,))
thing = curry.fetchone()
return clean(dict(thing)) if thing else None
def get_comment_ids(conn: sqlite3.Connection, claim_id: str, parent_id: str = None, page=1, page_size=50):
""" Just return a list of the comment IDs that are associated with the given claim_id.
If get_all is specified then it returns all the IDs, otherwise only the IDs at that level.
if parent_id is left null then it only returns the top level comments.
For pagination the parameters are:
get_all XOR (page_size + page)
"""
with conn:
if parent_id is None:
curs = conn.execute("""
SELECT comment_id FROM COMMENTS_ON_CLAIMS
WHERE claim_id = ? AND parent_id IS NULL LIMIT ? OFFSET ?
""", (claim_id, page_size, page_size * abs(page - 1),)
)
else:
curs = conn.execute("""
SELECT comment_id FROM COMMENTS_ON_CLAIMS
WHERE claim_id = ? AND parent_id = ? LIMIT ? OFFSET ?
""", (claim_id, parent_id, page_size, page_size * abs(page - 1),)
)
return [tuple(row)[0] for row in curs.fetchall()]
def get_comments_by_id(conn, comment_ids: typing.Union[list, tuple]) -> typing.Union[list, None]:
""" Returns a list containing the comment data associated with each ID within the list"""
# format the input, under the assumption that the
placeholders = ', '.join('?' for _ in comment_ids)
with conn:
return [clean(dict(row)) for row in conn.execute(
SELECT_COMMENTS_ON_CLAIMS + f'WHERE comment_id IN ({placeholders})',
tuple(comment_ids)
)]
def delete_comment_by_id(conn: sqlite3.Connection, comment_id: str) -> bool:
with conn:
curs = conn.execute("DELETE FROM COMMENT WHERE CommentId = ?", (comment_id,))
return bool(curs.rowcount)
def insert_channel(conn: sqlite3.Connection, channel_name: str, channel_id: str):
with conn:
curs = conn.execute('INSERT INTO CHANNEL(ClaimId, Name) VALUES (?, ?)', (channel_id, channel_name))
return bool(curs.rowcount)
def get_channel_id_from_comment_id(conn: sqlite3.Connection, comment_id: str):
with conn:
channel = conn.execute(
"SELECT channel_id, channel_name FROM COMMENTS_ON_CLAIMS WHERE comment_id = ?", (comment_id,)
).fetchone()
return dict(channel) if channel else {}
def get_claim_ids_from_comment_ids(conn: sqlite3.Connection, comment_ids: list):
with conn:
cids = conn.execute(
f""" SELECT CommentId as comment_id, LbryClaimId AS claim_id FROM COMMENT
WHERE CommentId IN ({', '.join('?' for _ in comment_ids)}) """,
tuple(comment_ids)
)
return {row['comment_id']: row['claim_id'] for row in cids.fetchall()}
def hide_comments_by_id(conn: sqlite3.Connection, comment_ids: list) -> bool:
with conn:
curs = conn.cursor()
curs.executemany(
"UPDATE COMMENT SET IsHidden = 1 WHERE CommentId = ?",
[[c] for c in comment_ids]
)
return bool(curs.rowcount)
def edit_comment_by_id(conn: sqlite3.Connection, comment_id: str, comment: str,
signature: str, signing_ts: str) -> bool:
with conn:
curs = conn.execute(
"""
UPDATE COMMENT
SET Body = :comment, Signature = :signature, SigningTs = :signing_ts
WHERE CommentId = :comment_id
""",
{
'comment': comment,
'signature': signature,
'signing_ts': signing_ts,
'comment_id': comment_id
})
logger.info("updated comment with `comment_id`: %s", comment_id)
return bool(curs.rowcount)
class DatabaseWriter(object):
_writer = None
def __init__(self, db_file):
if not DatabaseWriter._writer:
self.conn = obtain_connection(db_file)
DatabaseWriter._writer = self
atexit.register(self.cleanup)
logging.info('Database writer has been created at %s', repr(self))
else:
logging.warning('Someone attempted to insantiate DatabaseWriter')
raise TypeError('Database Writer already exists!')
def cleanup(self):
logging.info('Cleaning up database writer')
self.conn.close()
DatabaseWriter._writer = None
@property
def connection(self):
return self.conn
def setup_database(db_path):
with sqlite3.connect(db_path) as conn:
conn.executescript(CREATE_TABLES_QUERY)
def backup_database(conn: sqlite3.Connection, back_fp):
with sqlite3.connect(back_fp) as back:
conn.backup(back)

View file

@ -1,76 +0,0 @@
PRAGMAS = """
PRAGMA FOREIGN_KEYS = ON;
"""
CREATE_COMMENT_TABLE = """
CREATE TABLE IF NOT EXISTS COMMENT (
CommentId TEXT NOT NULL,
LbryClaimId TEXT NOT NULL,
ChannelId TEXT DEFAULT NULL,
Body TEXT NOT NULL,
ParentId TEXT DEFAULT NULL,
Signature TEXT DEFAULT NULL,
Timestamp INTEGER NOT NULL,
SigningTs TEXT DEFAULT NULL,
IsHidden BOOLEAN NOT NULL DEFAULT 0,
CONSTRAINT COMMENT_PRIMARY_KEY PRIMARY KEY (CommentId) ON CONFLICT IGNORE,
CONSTRAINT COMMENT_SIGNATURE_SK UNIQUE (Signature) ON CONFLICT ABORT,
CONSTRAINT COMMENT_CHANNEL_FK FOREIGN KEY (ChannelId) REFERENCES CHANNEL (ClaimId)
ON DELETE NO ACTION ON UPDATE NO ACTION,
CONSTRAINT COMMENT_PARENT_FK FOREIGN KEY (ParentId) REFERENCES COMMENT (CommentId)
ON UPDATE CASCADE ON DELETE NO ACTION -- setting null implies comment is top level
);
"""
CREATE_COMMENT_INDEXES = """
CREATE INDEX IF NOT EXISTS CLAIM_COMMENT_INDEX ON COMMENT (LbryClaimId, CommentId);
CREATE INDEX IF NOT EXISTS CHANNEL_COMMENT_INDEX ON COMMENT (ChannelId, CommentId);
"""
CREATE_CHANNEL_TABLE = """
CREATE TABLE IF NOT EXISTS CHANNEL (
ClaimId TEXT NOT NULL,
Name TEXT NOT NULL,
CONSTRAINT CHANNEL_PK PRIMARY KEY (ClaimId)
ON CONFLICT IGNORE
);
"""
CREATE_COMMENTS_ON_CLAIMS_VIEW = """
CREATE VIEW IF NOT EXISTS COMMENTS_ON_CLAIMS AS SELECT
C.CommentId AS comment_id,
C.Body AS comment,
C.LbryClaimId AS claim_id,
C.Timestamp AS timestamp,
CHAN.Name AS channel_name,
CHAN.ClaimId AS channel_id,
('lbry://' || CHAN.Name || '#' || CHAN.ClaimId) AS channel_url,
C.Signature AS signature,
C.SigningTs AS signing_ts,
C.ParentId AS parent_id,
C.IsHidden AS is_hidden
FROM COMMENT AS C
LEFT OUTER JOIN CHANNEL CHAN ON C.ChannelId = CHAN.ClaimId
ORDER BY C.Timestamp DESC;
"""
# not being used right now but should be kept around when Tom finally asks for replies
CREATE_COMMENT_REPLIES_VIEW = """
CREATE VIEW IF NOT EXISTS COMMENT_REPLIES (Author, CommentBody, ParentAuthor, ParentCommentBody) AS
SELECT AUTHOR.Name, OG.Body, PCHAN.Name, PARENT.Body
FROM COMMENT AS OG
JOIN COMMENT AS PARENT
ON OG.ParentId = PARENT.CommentId
JOIN CHANNEL AS PCHAN ON PARENT.ChannelId = PCHAN.ClaimId
JOIN CHANNEL AS AUTHOR ON OG.ChannelId = AUTHOR.ClaimId
ORDER BY OG.Timestamp;
"""
CREATE_TABLES_QUERY = (
PRAGMAS +
CREATE_COMMENT_TABLE +
CREATE_COMMENT_INDEXES +
CREATE_CHANNEL_TABLE +
CREATE_COMMENTS_ON_CLAIMS_VIEW +
CREATE_COMMENT_REPLIES_VIEW
)

View file

@ -1,3 +1,4 @@
# TODO: scrap notification routines from these files & supply them in handles
import logging
import sqlite3
from asyncio import coroutine
@ -7,7 +8,7 @@ from src.server.validation import is_valid_base_comment
from src.server.validation import is_valid_credential_input
from src.server.validation import validate_signature_from_claim
from src.server.validation import body_is_valid
from src.server.misc import get_claim_from_id
from src.misc import get_claim_from_id
from src.server.external import send_notifications
from src.server.external import send_notification
import src.database.queries as db
@ -84,13 +85,17 @@ async def hide_comments(app, pieces: list) -> list:
# TODO: Amortize this process
claims = {}
comments_to_hide = []
# go through a list of dict objects
for p in pieces:
# maps the comment_id from the piece to a claim_id
claim_id = comment_cids[p['comment_id']]
# resolve the claim from its id
if claim_id not in claims:
claim = await get_claim_from_id(app, claim_id)
if claim:
claims[claim_id] = claim
# get the claim's signing channel, then use it to validate the hidden comment
channel = claims[claim_id].get('signing_channel')
if validate_signature_from_claim(channel, p['signature'], p['signing_ts'], p['comment_id']):
comments_to_hide.append(p)

7
src/definitions.py Normal file
View file

@ -0,0 +1,7 @@
import os
SRC_DIR = os.path.dirname(os.path.abspath(__file__))
ROOT_DIR = os.path.dirname(SRC_DIR)
CONFIG_FILE = os.path.join(ROOT_DIR, 'config', 'conf.yml')
LOGGING_DIR = os.path.join(ROOT_DIR, 'logs')
DATABASE_DIR = os.path.join(ROOT_DIR, 'database')

View file

@ -1,13 +1,19 @@
import argparse
import yaml
import logging
import logging.config
import os
import sys
from src.server.app import run_app
from src.settings import config
from src.definitions import LOGGING_DIR, CONFIG_FILE, DATABASE_DIR
def config_logging_from_settings(conf):
def setup_logging_from_config(conf: dict):
# set the logging directory here from the settings file
if not os.path.exists(LOGGING_DIR):
os.mkdir(LOGGING_DIR)
_config = {
"version": 1,
"disable_existing_loggers": False,
@ -32,7 +38,7 @@ def config_logging_from_settings(conf):
"level": "DEBUG",
"formatter": "standard",
"class": "logging.handlers.RotatingFileHandler",
"filename": conf['path']['debug_log'],
"filename": os.path.join(LOGGING_DIR, 'debug.log'),
"maxBytes": 10485760,
"backupCount": 5
},
@ -40,7 +46,7 @@ def config_logging_from_settings(conf):
"level": "ERROR",
"formatter": "standard",
"class": "logging.handlers.RotatingFileHandler",
"filename": conf['path']['error_log'],
"filename": os.path.join(LOGGING_DIR, 'error.log'),
"maxBytes": 10485760,
"backupCount": 5
},
@ -48,7 +54,7 @@ def config_logging_from_settings(conf):
"level": "NOTSET",
"formatter": "aiohttp",
"class": "logging.handlers.RotatingFileHandler",
"filename": conf['path']['server_log'],
"filename": os.path.join(LOGGING_DIR, 'server.log'),
"maxBytes": 10485760,
"backupCount": 5
}
@ -70,15 +76,42 @@ def config_logging_from_settings(conf):
logging.config.dictConfig(_config)
def get_config(filepath):
with open(filepath, 'r') as cfile:
config = yaml.load(cfile, Loader=yaml.Loader)
return config
def setup_db_from_config(config: dict):
mode = config['mode']
if config[mode]['database'] == 'sqlite':
if not os.path.exists(DATABASE_DIR):
os.mkdir(DATABASE_DIR)
config[mode]['db_file'] = os.path.join(
DATABASE_DIR, config[mode]['name']
)
def main(argv=None):
argv = argv or sys.argv[1:]
parser = argparse.ArgumentParser(description='LBRY Comment Server')
parser.add_argument('--port', type=int)
parser.add_argument('--config', type=str)
parser.add_argument('--mode', type=str)
args = parser.parse_args(argv)
config_logging_from_settings(config)
config = get_config(CONFIG_FILE) if not args.config else args.config
setup_logging_from_config(config)
if args.mode:
config['mode'] = args.mode
setup_db_from_config(config)
if args.port:
config['port'] = args.port
config_logging_from_settings(config)
run_app(config)

View file

@ -16,7 +16,11 @@ async def get_claim_from_id(app, claim_id, **kwargs):
def clean_input_params(kwargs: dict):
for k, v in kwargs.items():
if type(v) is str and k is not 'comment':
if type(v) is str and k != 'comment':
kwargs[k] = v.strip()
if k in ID_LIST:
kwargs[k] = v.lower()
def clean(thing: dict) -> dict:
return {k: v for k, v in thing.items() if v is not None}

View file

@ -1,7 +1,6 @@
# cython: language_level=3
import asyncio
import logging
import pathlib
import signal
import time
@ -9,81 +8,68 @@ import aiojobs
import aiojobs.aiohttp
from aiohttp import web
from src.database.queries import obtain_connection, DatabaseWriter
from src.database.queries import setup_database, backup_database
from peewee import *
from src.server.handles import api_endpoint, get_api_endpoint
from src.database.models import Comment, Channel
MODELS = [Comment, Channel]
logger = logging.getLogger(__name__)
async def setup_db_schema(app):
if not pathlib.Path(app['db_path']).exists():
logger.info(f'Setting up schema in {app["db_path"]}')
setup_database(app['db_path'])
else:
logger.info(f'Database already exists in {app["db_path"]}, skipping setup')
def setup_database(app):
config = app['config']
mode = config['mode']
# switch between Database objects
if config[mode]['database'] == 'mysql':
app['db'] = MySQLDatabase(
database=config[mode]['name'],
user=config[mode]['user'],
host=config[mode]['host'],
password=config[mode]['password'],
port=config[mode]['port'],
charset=config[mode]['charset'],
)
elif config[mode]['database'] == 'sqlite':
app['db'] = SqliteDatabase(
config[mode]['file'],
pragmas=config[mode]['pragmas']
)
async def database_backup_routine(app):
try:
while True:
await asyncio.sleep(app['config']['backup_int'])
with app['reader'] as conn:
logger.debug('backing up database')
backup_database(conn, app['backup'])
except asyncio.CancelledError:
pass
# bind the Model list to the database
app['db'].bind(MODELS, bind_refs=False, bind_backrefs=False)
async def start_background_tasks(app):
# Reading the DB
app['reader'] = obtain_connection(app['db_path'], True)
app['waitful_backup'] = asyncio.create_task(database_backup_routine(app))
# Scheduler to prevent multiple threads from writing to DB simulataneously
app['comment_scheduler'] = await aiojobs.create_scheduler(limit=1, pending_limit=0)
app['db_writer'] = DatabaseWriter(app['db_path'])
app['writer'] = app['db_writer'].connection
app['db'].connect()
app['db'].create_tables(MODELS)
# for requesting to external and internal APIs
app['webhooks'] = await aiojobs.create_scheduler(pending_limit=0)
async def close_database_connections(app):
logger.info('Ending background backup loop')
app['waitful_backup'].cancel()
await app['waitful_backup']
app['reader'].close()
app['writer'].close()
app['db_writer'].cleanup()
app['db'].close()
async def close_schedulers(app):
logger.info('Closing comment_scheduler')
await app['comment_scheduler'].close()
logger.info('Closing scheduler for webhook requests')
await app['webhooks'].close()
class CommentDaemon:
def __init__(self, config, db_file=None, backup=None, **kwargs):
def __init__(self, config, **kwargs):
app = web.Application()
app['config'] = config
# configure the config
app['config'] = config
self.config = app['config']
self.config = config
self.host = config['host']
self.port = config['port']
# configure the db file
if db_file:
app['db_path'] = db_file
app['backup'] = backup
else:
app['db_path'] = config['path']['database']
app['backup'] = backup or (app['db_path'] + '.backup')
setup_database(app)
# configure the order of tasks to run during app lifetime
app.on_startup.append(setup_db_schema)
app.on_startup.append(start_background_tasks)
app.on_shutdown.append(close_schedulers)
app.on_cleanup.append(close_database_connections)
@ -105,20 +91,19 @@ class CommentDaemon:
await self.app_runner.setup()
self.app_site = web.TCPSite(
runner=self.app_runner,
host=host or self.config['host'],
port=port or self.config['port'],
host=host or self.host,
port=port or self.port,
)
await self.app_site.start()
logger.info(f'Comment Server is running on {self.config["host"]}:{self.config["port"]}')
logger.info(f'Comment Server is running on {self.host}:{self.port}')
async def stop(self):
await self.app_runner.shutdown()
await self.app_runner.cleanup()
def run_app(config, db_file=None):
comment_app = CommentDaemon(config=config, db_file=db_file, close_timeout=5.0)
def run_app(config):
comment_app = CommentDaemon(config=config)
loop = asyncio.get_event_loop()
def __exit():

View file

@ -1,3 +1,5 @@
import json
import logging
import aiohttp
@ -32,8 +34,11 @@ def make_error(error, exc=None) -> dict:
async def report_error(app, exc, body: dict):
try:
if 'slack_webhook' in app['config']:
body_dump = json.dumps(body, indent=4)
exec_name = type(exc).__name__
exec_body = str(exc)
message = {
"text": f"Got `{type(exc).__name__}`: `\n{str(exc)}`\n```{body}```"
"text": f"Got `{exec_name}`: `\n{exec_body}`\n```{body_dump}```"
}
async with aiohttp.ClientSession() as sesh:
async with sesh.post(app['config']['slack_webhook'], json=message) as resp:

View file

@ -37,6 +37,10 @@ def create_notification_batch(action: str, comments: List[dict]) -> List[dict]:
}
if comment.get('channel_id'):
event['channel_id'] = comment['channel_id']
if comment.get('parent_id'):
event['parent_id'] = comment['parent_id']
if comment.get('comment'):
event['comment'] = comment['comment']
events.append(event)
return events

View file

@ -1,16 +1,24 @@
import asyncio
import logging
import time
import typing
from aiohttp import web
from aiojobs.aiohttp import atomic
from peewee import DoesNotExist
import src.database.queries as db
from src.database.writes import abandon_comment, create_comment
from src.database.writes import hide_comments
from src.database.writes import edit_comment
from src.server.misc import clean_input_params
from src.server.external import send_notification
from src.server.validation import validate_signature_from_claim
from src.misc import clean_input_params, get_claim_from_id
from src.server.errors import make_error, report_error
from src.database.models import Comment, Channel
from src.database.models import get_comment
from src.database.models import comment_list
from src.database.models import create_comment
from src.database.models import edit_comment
from src.database.models import delete_comment
from src.database.models import set_hidden_flag
logger = logging.getLogger(__name__)
@ -20,51 +28,198 @@ def ping(*args):
return 'pong'
def handle_get_channel_from_comment_id(app, kwargs: dict):
return db.get_channel_id_from_comment_id(app['reader'], **kwargs)
def handle_get_channel_from_comment_id(app: web.Application, comment_id: str) -> dict:
comment = get_comment(comment_id)
return {
'channel_id': comment['channel_id'],
'channel_name': comment['channel_name']
}
def handle_get_comment_ids(app, kwargs):
return db.get_comment_ids(app['reader'], **kwargs)
def handle_get_comment_ids(
app: web.Application,
claim_id: str,
parent_id: str = None,
page: int = 1,
page_size: int = 50,
flattened=False
) -> dict:
results = comment_list(
claim_id=claim_id,
parent_id=parent_id,
top_level=(parent_id is None),
page=page,
page_size=page_size,
select_fields=['comment_id', 'parent_id']
)
if flattened:
results.update({
'items': [item['comment_id'] for item in results['items']],
'replies': [(item['comment_id'], item.get('parent_id'))
for item in results['items']]
})
return results
def handle_get_claim_comments(app, kwargs):
return db.get_claim_comments(app['reader'], **kwargs)
def handle_get_comments_by_id(
app: web.Application,
comment_ids: typing.Union[list, tuple]
) -> dict:
expression = Comment.comment_id.in_(comment_ids)
return comment_list(expressions=expression, page_size=len(comment_ids))
def handle_get_comments_by_id(app, kwargs):
return db.get_comments_by_id(app['reader'], **kwargs)
def handle_get_claim_comments(
app: web.Application,
claim_id: str,
parent_id: str = None,
page: int = 1,
page_size: int = 50,
top_level: bool = False
) -> dict:
return comment_list(
claim_id=claim_id,
parent_id=parent_id,
page=page,
page_size=page_size,
top_level=top_level
)
def handle_get_claim_hidden_comments(app, kwargs):
return db.get_claim_hidden_comments(app['reader'], **kwargs)
def handle_get_claim_hidden_comments(
app: web.Application,
claim_id: str,
hidden: bool,
page: int = 1,
page_size: int = 50,
) -> dict:
exclude = 'hidden' if hidden else 'visible'
return comment_list(
claim_id=claim_id,
exclude_mode=exclude,
page=page,
page_size=page_size
)
async def handle_abandon_comment(app, params):
return {'abandoned': await abandon_comment(app, **params)}
async def handle_abandon_comment(
app: web.Application,
comment_id: str,
signature: str,
signing_ts: str,
**kwargs,
) -> dict:
comment = get_comment(comment_id)
try:
channel = await get_claim_from_id(app, comment['channel_id'])
except DoesNotExist:
raise ValueError('Could not find a channel associated with the given comment')
else:
if not validate_signature_from_claim(channel, signature, signing_ts, comment_id):
raise ValueError('Abandon signature could not be validated')
await app['webhooks'].spawn(send_notification(app, 'DELETE', comment))
with app['db'].atomic():
return {
'abandoned': delete_comment(comment_id)
}
async def handle_hide_comments(app, params):
return {'hidden': await hide_comments(app, **params)}
async def handle_hide_comments(app: web.Application, pieces: list, hide: bool = True) -> dict:
# let's get all the distinct claim_ids from the list of comment_ids
pieces_by_id = {p['comment_id']: p for p in pieces}
comment_ids = list(pieces_by_id.keys())
comments = (Comment
.select(Comment.comment_id, Comment.claim_id)
.where(Comment.comment_id.in_(comment_ids))
.tuples())
# resolve the claims and map them to their corresponding comment_ids
claims = {}
for comment_id, claim_id in comments:
try:
# try and resolve the claim, if fails then we mark it as null
# and remove the associated comment from the pieces
if claim_id not in claims:
claims[claim_id] = await get_claim_from_id(app, claim_id)
# try to get a public key to validate
if claims[claim_id] is None or 'signing_channel' not in claims[claim_id]:
raise ValueError(f'could not get signing channel from claim_id: {claim_id}')
# try to validate signature
else:
channel = claims[claim_id]['signing_channel']
piece = pieces_by_id[comment_id]
is_valid_signature = validate_signature_from_claim(
claim=channel,
signature=piece['signature'],
signing_ts=piece['signing_ts'],
data=piece['comment_id']
)
if not is_valid_signature:
raise ValueError(f'could not validate signature on comment_id: {comment_id}')
except ValueError:
# remove the piece from being hidden
pieces_by_id.pop(comment_id)
# remaining items in pieces_by_id have been able to successfully validate
with app['db'].atomic():
set_hidden_flag(list(pieces_by_id.keys()), hidden=hide)
query = Comment.select().where(Comment.comment_id.in_(comment_ids)).objects()
result = {
'hidden': [c.comment_id for c in query if c.is_hidden],
'visible': [c.comment_id for c in query if not c.is_hidden],
}
return result
async def handle_edit_comment(app, params):
if await edit_comment(app, **params):
return db.get_comment_or_none(app['reader'], params['comment_id'])
async def handle_edit_comment(app, comment: str = None, comment_id: str = None,
signature: str = None, signing_ts: str = None, **params) -> dict:
current = get_comment(comment_id)
channel_claim = await get_claim_from_id(app, current['channel_id'])
if not validate_signature_from_claim(channel_claim, signature, signing_ts, comment):
raise ValueError('Signature could not be validated')
with app['db'].atomic():
if not edit_comment(comment_id, comment, signature, signing_ts):
raise ValueError('Comment could not be edited')
updated_comment = get_comment(comment_id)
await app['webhooks'].spawn(send_notification(app, 'UPDATE', updated_comment))
return updated_comment
# TODO: retrieve stake amounts for each channel & store in db
async def handle_create_comment(app, comment: str = None, claim_id: str = None,
parent_id: str = None, channel_id: str = None, channel_name: str = None,
signature: str = None, signing_ts: str = None) -> dict:
with app['db'].atomic():
comment = create_comment(
comment=comment,
claim_id=claim_id,
parent_id=parent_id,
channel_id=channel_id,
channel_name=channel_name,
signature=signature,
signing_ts=signing_ts
)
await app['webhooks'].spawn(send_notification(app, 'CREATE', comment))
return comment
METHODS = {
'ping': ping,
'get_claim_comments': handle_get_claim_comments,
'get_claim_hidden_comments': handle_get_claim_hidden_comments,
'get_claim_comments': handle_get_claim_comments, # this gets used
'get_claim_hidden_comments': handle_get_claim_hidden_comments, # this gets used
'get_comment_ids': handle_get_comment_ids,
'get_comments_by_id': handle_get_comments_by_id,
'get_channel_from_comment_id': handle_get_channel_from_comment_id,
'create_comment': create_comment,
'get_comments_by_id': handle_get_comments_by_id, # this gets used
'get_channel_from_comment_id': handle_get_channel_from_comment_id, # this gets used
'create_comment': handle_create_comment, # this gets used
'delete_comment': handle_abandon_comment,
'abandon_comment': handle_abandon_comment,
'hide_comments': handle_hide_comments,
'edit_comment': handle_edit_comment
'abandon_comment': handle_abandon_comment, # this gets used
'hide_comments': handle_hide_comments, # this gets used
'edit_comment': handle_edit_comment # this gets used
}
@ -78,17 +233,19 @@ async def process_json(app, body: dict) -> dict:
start = time.time()
try:
if asyncio.iscoroutinefunction(METHODS[method]):
result = await METHODS[method](app, params)
result = await METHODS[method](app, **params)
else:
result = METHODS[method](app, params)
response['result'] = result
result = METHODS[method](app, **params)
except Exception as err:
logger.exception(f'Got {type(err).__name__}:')
logger.exception(f'Got {type(err).__name__}:\n{err}')
if type(err) in (ValueError, TypeError): # param error, not too important
response['error'] = make_error('INVALID_PARAMS', err)
else:
response['error'] = make_error('INTERNAL', err)
await app['webhooks'].spawn(report_error(app, err, body))
else:
response['result'] = result
finally:
end = time.time()

View file

@ -51,11 +51,31 @@ def claim_id_is_valid(claim_id: str) -> bool:
# default to None so params can be treated as kwargs; param count becomes more manageable
def is_valid_base_comment(comment: str = None, claim_id: str = None, parent_id: str = None, **kwargs) -> bool:
return comment and body_is_valid(comment) and \
((claim_id and claim_id_is_valid(claim_id)) or # parentid is used in place of claimid in replies
(parent_id and comment_id_is_valid(parent_id))) \
and is_valid_credential_input(**kwargs)
def is_valid_base_comment(
comment: str = None,
claim_id: str = None,
parent_id: str = None,
strict: bool = False,
**kwargs,
) -> bool:
try:
assert comment and body_is_valid(comment)
# strict mode assumes that the parent_id might not exist
if strict:
assert claim_id and claim_id_is_valid(claim_id)
assert parent_id is None or comment_id_is_valid(parent_id)
# non-strict removes reference restrictions
else:
assert claim_id or parent_id
if claim_id:
assert claim_id_is_valid(claim_id)
else:
assert comment_id_is_valid(parent_id)
except AssertionError:
return False
else:
return is_valid_credential_input(**kwargs)
def is_valid_credential_input(channel_id: str = None, channel_name: str = None,

View file

@ -1,17 +0,0 @@
# cython: language_level=3
import json
import pathlib
root_dir = pathlib.Path(__file__).parent.parent
config_path = root_dir / 'config' / 'conf.json'
def get_config(filepath):
with open(filepath, 'r') as cfile:
conf = json.load(cfile)
for key, path in conf['path'].items():
conf['path'][key] = str(root_dir / path)
return conf
config = get_config(config_path)

View file

@ -1,18 +1,13 @@
import sqlite3
from random import randint
import faker
from faker.providers import internet
from faker.providers import lorem
from faker.providers import misc
from src.database.queries import get_comments_by_id
from src.database.queries import get_comment_ids
from src.database.queries import get_claim_comments
from src.database.queries import get_claim_hidden_comments
from src.database.writes import create_comment_or_error
from src.database.queries import hide_comments_by_id
from src.database.queries import delete_comment_by_id
from src.database.models import create_comment
from src.database.models import delete_comment
from src.database.models import comment_list, get_comment
from src.database.models import set_hidden_flag
from test.testcase import DatabaseTestCase
fake = faker.Faker()
@ -27,26 +22,25 @@ class TestDatabaseOperations(DatabaseTestCase):
self.claimId = '529357c3422c6046d3fec76be2358004ba22e340'
def test01NamedComments(self):
comment = create_comment_or_error(
conn=self.conn,
comment = create_comment(
claim_id=self.claimId,
comment='This is a named comment',
channel_name='@username',
channel_id='529357c3422c6046d3fec76be2358004ba22abcd',
signature=fake.uuid4(),
signature='22'*64,
signing_ts='aaa'
)
self.assertIsNotNone(comment)
self.assertNotIn('parent_in', comment)
previous_id = comment['comment_id']
reply = create_comment_or_error(
conn=self.conn,
reply = create_comment(
claim_id=self.claimId,
comment='This is a named response',
channel_name='@another_username',
channel_id='529357c3422c6046d3fec76be2358004ba224bcd',
parent_id=previous_id,
signature=fake.uuid4(),
signature='11'*64,
signing_ts='aaa'
)
self.assertIsNotNone(reply)
@ -54,34 +48,32 @@ class TestDatabaseOperations(DatabaseTestCase):
def test02AnonymousComments(self):
self.assertRaises(
sqlite3.IntegrityError,
create_comment_or_error,
conn=self.conn,
ValueError,
create_comment,
claim_id=self.claimId,
comment='This is an ANONYMOUS comment'
)
def test03SignedComments(self):
comment = create_comment_or_error(
conn=self.conn,
comment = create_comment(
claim_id=self.claimId,
comment='I like big butts and i cannot lie',
channel_name='@sirmixalot',
channel_id='529357c3422c6046d3fec76be2358005ba22abcd',
signature=fake.uuid4(),
signature='24'*64,
signing_ts='asdasd'
)
self.assertIsNotNone(comment)
self.assertIn('signing_ts', comment)
previous_id = comment['comment_id']
reply = create_comment_or_error(
conn=self.conn,
reply = create_comment(
claim_id=self.claimId,
comment='This is a LBRY verified response',
channel_name='@LBRY',
channel_id='529357c3422c6046d3fec76be2358001ba224bcd',
parent_id=previous_id,
signature=fake.uuid4(),
signature='12'*64,
signing_ts='sfdfdfds'
)
self.assertIsNotNone(reply)
@ -90,75 +82,109 @@ class TestDatabaseOperations(DatabaseTestCase):
def test04UsernameVariations(self):
self.assertRaises(
AssertionError,
callable=create_comment_or_error,
conn=self.conn,
ValueError,
create_comment,
claim_id=self.claimId,
channel_name='$#(@#$@#$',
channel_id='529357c3422c6046d3fec76be2358001ba224b23',
comment='this is an invalid username'
comment='this is an invalid username',
signature='1' * 128,
signing_ts='123'
)
valid_username = create_comment_or_error(
conn=self.conn,
valid_username = create_comment(
claim_id=self.claimId,
channel_name='@' + 'a' * 255,
channel_id='529357c3422c6046d3fec76be2358001ba224b23',
comment='this is a valid username'
comment='this is a valid username',
signature='1'*128,
signing_ts='123'
)
self.assertIsNotNone(valid_username)
self.assertRaises(AssertionError,
callable=create_comment_or_error,
conn=self.conn,
self.assertRaises(
ValueError,
create_comment,
claim_id=self.claimId,
channel_name='@' + 'a' * 256,
channel_id='529357c3422c6046d3fec76be2358001ba224b23',
comment='this username is too long'
comment='this username is too long',
signature='2' * 128,
signing_ts='123'
)
self.assertRaises(
AssertionError,
callable=create_comment_or_error,
conn=self.conn,
ValueError,
create_comment,
claim_id=self.claimId,
channel_name='',
channel_id='529357c3422c6046d3fec76be2358001ba224b23',
comment='this username should not default to ANONYMOUS'
comment='this username should not default to ANONYMOUS',
signature='3' * 128,
signing_ts='123'
)
self.assertRaises(
AssertionError,
callable=create_comment_or_error,
conn=self.conn,
ValueError,
create_comment,
claim_id=self.claimId,
channel_name='@',
channel_id='529357c3422c6046d3fec76be2358001ba224b23',
comment='this username is too short'
comment='this username is too short',
signature='3' * 128,
signing_ts='123'
)
def test05HideComments(self):
comm = create_comment_or_error(self.conn, 'Comment #1', self.claimId, '1'*40, '@Doge123', 'a'*128, '123')
comment = get_comments_by_id(self.conn, [comm['comment_id']]).pop()
comm = create_comment(
comment='Comment #1',
claim_id=self.claimId,
channel_id='1'*40,
channel_name='@Doge123',
signature='a'*128,
signing_ts='123'
)
comment = get_comment(comm['comment_id'])
self.assertFalse(comment['is_hidden'])
success = hide_comments_by_id(self.conn, [comm['comment_id']])
success = set_hidden_flag([comm['comment_id']])
self.assertTrue(success)
comment = get_comments_by_id(self.conn, [comm['comment_id']]).pop()
comment = get_comment(comm['comment_id'])
self.assertTrue(comment['is_hidden'])
success = hide_comments_by_id(self.conn, [comm['comment_id']])
success = set_hidden_flag([comm['comment_id']])
self.assertTrue(success)
comment = get_comments_by_id(self.conn, [comm['comment_id']]).pop()
comment = get_comment(comm['comment_id'])
self.assertTrue(comment['is_hidden'])
def test06DeleteComments(self):
comm = create_comment_or_error(self.conn, 'Comment #1', self.claimId, '1'*40, '@Doge123', 'a'*128, '123')
comments = get_claim_comments(self.conn, self.claimId)
match = list(filter(lambda x: comm['comment_id'] == x['comment_id'], comments['items']))
self.assertTrue(match)
deleted = delete_comment_by_id(self.conn, comm['comment_id'])
# make sure that the comment was created
comm = create_comment(
comment='Comment #1',
claim_id=self.claimId,
channel_id='1'*40,
channel_name='@Doge123',
signature='a'*128,
signing_ts='123'
)
comments = comment_list(self.claimId)
match = [x for x in comments['items'] if x['comment_id'] == comm['comment_id']]
self.assertTrue(len(match) > 0)
deleted = delete_comment(comm['comment_id'])
self.assertTrue(deleted)
comments = get_claim_comments(self.conn, self.claimId)
match = list(filter(lambda x: comm['comment_id'] == x['comment_id'], comments['items']))
# make sure that we can't find the comment here
comments = comment_list(self.claimId)
match = [x for x in comments['items'] if x['comment_id'] == comm['comment_id']]
self.assertFalse(match)
deleted = delete_comment_by_id(self.conn, comm['comment_id'])
self.assertFalse(deleted)
self.assertRaises(
ValueError,
delete_comment,
comment_id=comm['comment_id'],
)
class ListDatabaseTest(DatabaseTestCase):
@ -169,61 +195,75 @@ class ListDatabaseTest(DatabaseTestCase):
def testLists(self):
for claim_id in self.claim_ids:
with self.subTest(claim_id=claim_id):
comments = get_claim_comments(self.conn, claim_id)
comments = comment_list(claim_id)
self.assertIsNotNone(comments)
self.assertGreater(comments['page_size'], 0)
self.assertIn('has_hidden_comments', comments)
self.assertFalse(comments['has_hidden_comments'])
top_comments = get_claim_comments(self.conn, claim_id, top_level=True, page=1, page_size=50)
top_comments = comment_list(claim_id, top_level=True, page=1, page_size=50)
self.assertIsNotNone(top_comments)
self.assertEqual(top_comments['page_size'], 50)
self.assertEqual(top_comments['page'], 1)
self.assertGreaterEqual(top_comments['total_pages'], 0)
self.assertGreaterEqual(top_comments['total_items'], 0)
comment_ids = get_comment_ids(self.conn, claim_id, page_size=50, page=1)
comment_ids = comment_list(claim_id, page_size=50, page=1)
with self.subTest(comment_ids=comment_ids):
self.assertIsNotNone(comment_ids)
self.assertLessEqual(len(comment_ids), 50)
matching_comments = get_comments_by_id(self.conn, comment_ids)
matching_comments = (comment_ids)
self.assertIsNotNone(matching_comments)
self.assertEqual(len(matching_comments), len(comment_ids))
def testHiddenCommentLists(self):
claim_id = 'a'*40
comm1 = create_comment_or_error(self.conn, 'Comment #1', claim_id, '1'*40, '@Doge123', 'a'*128, '123')
comm2 = create_comment_or_error(self.conn, 'Comment #2', claim_id, '1'*40, '@Doge123', 'b'*128, '123')
comm3 = create_comment_or_error(self.conn, 'Comment #3', claim_id, '1'*40, '@Doge123', 'c'*128, '123')
comm1 = create_comment(
'Comment #1',
claim_id,
channel_id='1'*40,
channel_name='@Doge123',
signature='a'*128,
signing_ts='123'
)
comm2 = create_comment(
'Comment #2', claim_id,
channel_id='1'*40,
channel_name='@Doge123',
signature='b'*128,
signing_ts='123'
)
comm3 = create_comment(
'Comment #3', claim_id,
channel_id='1'*40,
channel_name='@Doge123',
signature='c'*128,
signing_ts='123'
)
comments = [comm1, comm2, comm3]
comment_list = get_claim_comments(self.conn, claim_id)
self.assertIn('items', comment_list)
self.assertIn('has_hidden_comments', comment_list)
self.assertEqual(len(comments), comment_list['total_items'])
self.assertIn('has_hidden_comments', comment_list)
self.assertFalse(comment_list['has_hidden_comments'])
hide_comments_by_id(self.conn, [comm2['comment_id']])
listed_comments = comment_list(claim_id)
self.assertEqual(len(comments), listed_comments['total_items'])
self.assertFalse(listed_comments['has_hidden_comments'])
default_comments = get_claim_hidden_comments(self.conn, claim_id)
self.assertIn('has_hidden_comments', default_comments)
set_hidden_flag([comm2['comment_id']])
hidden = comment_list(claim_id, exclude_mode='hidden')
hidden_comments = get_claim_hidden_comments(self.conn, claim_id, hidden=True)
self.assertIn('has_hidden_comments', hidden_comments)
self.assertEqual(default_comments, hidden_comments)
self.assertTrue(hidden['has_hidden_comments'])
self.assertGreater(len(hidden['items']), 0)
hidden_comment = hidden_comments['items'][0]
visible = comment_list(claim_id, exclude_mode='visible')
self.assertFalse(visible['has_hidden_comments'])
self.assertNotEqual(listed_comments['items'], visible['items'])
# make sure the hidden comment is the one we marked as hidden
hidden_comment = hidden['items'][0]
self.assertEqual(hidden_comment['comment_id'], comm2['comment_id'])
visible_comments = get_claim_hidden_comments(self.conn, claim_id, hidden=False)
self.assertIn('has_hidden_comments', visible_comments)
self.assertNotIn(hidden_comment, visible_comments['items'])
hidden_ids = [c['comment_id'] for c in hidden_comments['items']]
visible_ids = [c['comment_id'] for c in visible_comments['items']]
hidden_ids = [c['comment_id'] for c in hidden['items']]
visible_ids = [c['comment_id'] for c in visible['items']]
composite_ids = hidden_ids + visible_ids
listed_comments = comment_list(claim_id)
all_ids = [c['comment_id'] for c in listed_comments['items']]
composite_ids.sort()
comment_list = get_claim_comments(self.conn, claim_id)
all_ids = [c['comment_id'] for c in comment_list['items']]
all_ids.sort()
self.assertEqual(composite_ids, all_ids)

View file

@ -9,13 +9,18 @@ from faker.providers import internet
from faker.providers import lorem
from faker.providers import misc
from src.settings import config
from src.main import get_config, CONFIG_FILE
from src.server import app
from src.server.validation import is_valid_base_comment
from test.testcase import AsyncioTestCase
config = get_config(CONFIG_FILE)
config['mode'] = 'testing'
config['testing']['file'] = ':memory:'
if 'slack_webhook' in config:
config.pop('slack_webhook')
@ -71,10 +76,10 @@ def create_test_comments(values: iter, **default):
class ServerTest(AsyncioTestCase):
db_file = 'test.db'
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
config['mode'] = 'testing'
config['testing']['file'] = ':memory:'
self.host = 'localhost'
self.port = 5931
@ -85,11 +90,10 @@ class ServerTest(AsyncioTestCase):
@classmethod
def tearDownClass(cls) -> None:
print('exit reached')
os.remove(cls.db_file)
async def asyncSetUp(self):
await super().asyncSetUp()
self.server = app.CommentDaemon(config, db_file=self.db_file)
self.server = app.CommentDaemon(config)
await self.server.start(host=self.host, port=self.port)
self.addCleanup(self.server.stop)
@ -135,14 +139,16 @@ class ServerTest(AsyncioTestCase):
test_all = create_test_comments(replace.keys(), **{
k: None for k in replace.keys()
})
test_all.reverse()
for test in test_all:
with self.subTest(test=test):
nulls = 'null fields: ' + ', '.join(k for k, v in test.items() if not v)
with self.subTest(test=nulls):
message = await self.post_comment(**test)
self.assertTrue('result' in message or 'error' in message)
if 'error' in message:
self.assertFalse(is_valid_base_comment(**test))
self.assertFalse(is_valid_base_comment(**test, strict=True))
else:
self.assertTrue(is_valid_base_comment(**test))
self.assertTrue(is_valid_base_comment(**test, strict=True))
async def test04CreateAllReplies(self):
claim_id = '1d8a5cc39ca02e55782d619e67131c0a20843be8'
@ -220,7 +226,8 @@ class ListCommentsTest(AsyncioTestCase):
super().__init__(*args, **kwargs)
self.host = 'localhost'
self.port = 5931
self.db_file = 'list_test.db'
config['mode'] = 'testing'
config['testing']['file'] = ':memory:'
self.claim_id = '1d8a5cc39ca02e55782d619e67131c0a20843be8'
self.comment_ids = None
@ -231,10 +238,6 @@ class ListCommentsTest(AsyncioTestCase):
async def post_comment(self, **params):
return await jsonrpc_post(self.url, 'create_comment', **params)
def tearDown(self) -> None:
print('exit reached')
os.remove(self.db_file)
async def create_lots_of_comments(self, n=23):
self.comment_list = [{key: self.replace[key]() for key in self.replace.keys()} for _ in range(23)]
for comment in self.comment_list:
@ -244,7 +247,7 @@ class ListCommentsTest(AsyncioTestCase):
async def asyncSetUp(self):
await super().asyncSetUp()
self.server = app.CommentDaemon(config, db_file=self.db_file)
self.server = app.CommentDaemon(config)
await self.server.start(self.host, self.port)
self.addCleanup(self.server.stop)

View file

@ -1,12 +1,38 @@
import os
import pathlib
import unittest
from asyncio.runners import _cancel_all_tasks # type: ignore
from unittest.case import _Outcome
import asyncio
from asyncio.runners import _cancel_all_tasks # type: ignore
from peewee import *
from src.database.queries import obtain_connection, setup_database
from src.database.models import Channel, Comment
test_db = SqliteDatabase(':memory:')
MODELS = [Channel, Comment]
class DatabaseTestCase(unittest.TestCase):
def __init__(self, methodName='DatabaseTest'):
super().__init__(methodName)
def setUp(self) -> None:
super().setUp()
test_db.bind(MODELS, bind_refs=False, bind_backrefs=False)
test_db.connect()
test_db.create_tables(MODELS)
def tearDown(self) -> None:
# drop tables for next test
test_db.drop_tables(MODELS)
# close connection
test_db.close()
class AsyncioTestCase(unittest.TestCase):
@ -117,21 +143,3 @@ class AsyncioTestCase(unittest.TestCase):
self.loop.run_until_complete(maybe_coroutine)
class DatabaseTestCase(unittest.TestCase):
db_file = 'test.db'
def __init__(self, methodName='DatabaseTest'):
super().__init__(methodName)
if pathlib.Path(self.db_file).exists():
os.remove(self.db_file)
def setUp(self) -> None:
super().setUp()
setup_database(self.db_file)
self.conn = obtain_connection(self.db_file)
self.addCleanup(self.conn.close)
self.addCleanup(os.remove, self.db_file)
def tearDown(self) -> None:
self.conn.close()