From 965972a719bc89dabbde803280911dbccd7bcf90 Mon Sep 17 00:00:00 2001
From: Victor Shyba <victor1984@riseup.net>
Date: Thu, 19 Sep 2019 17:07:20 -0300
Subject: [PATCH] match regex from app into ours

---
 lbry/lbry/schema/url.py            |  7 +++++--
 lbry/tests/unit/schema/test_url.py | 31 ++++++++++++++++++++++++++++++
 2 files changed, 36 insertions(+), 2 deletions(-)

diff --git a/lbry/lbry/schema/url.py b/lbry/lbry/schema/url.py
index ef3ccf5e5..e6c83ac71 100644
--- a/lbry/lbry/schema/url.py
+++ b/lbry/lbry/schema/url.py
@@ -4,7 +4,10 @@ from typing import NamedTuple, Tuple
 
 
 def _create_url_regex():
-    # see https://spec.lbry.com/
+    # see https://spec.lbry.com/ and test_url.py
+    invalid_names_regex = \
+        r"[^=&#:$@%?;\"/\\<>%{}|^~`\[\]" \
+        r"\u0000-\u0008\u000b-\u000c\u000e-\u001F\uD800-\uDFFF\uFFFE-\uFFFF]+"
 
     def _named(name, regex):
         return "(?P<" + name + ">" + regex + ")"
@@ -17,7 +20,7 @@ def _create_url_regex():
 
     def _claim(name, prefix=""):
         return _group(
-            _named(name+"_name", prefix + "[^=&#:$@%?/]+") +
+            _named(name+"_name", prefix + invalid_names_regex) +
             _oneof(
                 _group('#' + _named(name+"_claim_id", "[0-9a-f]{1,40}")),
                 _group(':' + _named(name+"_sequence", '[1-9][0-9]*')),
diff --git a/lbry/tests/unit/schema/test_url.py b/lbry/tests/unit/schema/test_url.py
index 83cc203fd..c496f6e5a 100644
--- a/lbry/tests/unit/schema/test_url.py
+++ b/lbry/tests/unit/schema/test_url.py
@@ -55,10 +55,41 @@ class TestURLParsing(unittest.TestCase):
         url('lbry://@test:1/stuff', channel_name='@test', channel_sequence='1', stream_name='stuff')
         url('lbry://@test$1/stuff', channel_name='@test', channel_amount_order='1', stream_name='stuff')
         url(f'lbry://@test#{claim_id}/stuff', channel_name='@test', channel_claim_id=claim_id, stream_name='stuff')
+        # unicode regex edges
+        _url = lambda name: url(name, stream_name=name)
+        _url('\u0009')
+        _url('\u000a')
+        _url('\u000d')
+        _url('\u0020')
+        _url('\uD799')
+        _url('\uE000')
+        _url('\uFFFD')
 
     def test_parser_invalid_urls(self):
         fail = self._fail_url
         fail("lbry://")
+        fail("lbry://\u0000")
+        fail("lbry://\u0008")
+        fail("lbry://\u000b")
+        fail("lbry://\u000c")
+        fail("lbry://\u000e")
+        fail("lbry://\u001f")
+        fail("lbry://\uD800")
+        fail("lbry://\uDFFF")
+        fail("lbry://\uDFFE")
+        fail("lbry://\uFFFF")
+        fail("lbry://;")
+        fail("lbry://\"")
+        fail("lbry://\\")
+        fail("lbry:///")
+        fail("lbry://<") and fail("lbry://>")
+        fail("lbry://{") and fail("lbry://}")
+        fail("lbry://[") and fail("lbry://]")
+        fail("lbry://%")
+        fail("lbry://|")
+        fail("lbry://^")
+        fail("lbry://~")
+        fail("lbry://`")
         fail("lbry://test:3$1")
         fail("lbry://test$1:1")
         fail("lbry://test#x")