From dfab09b17d0799d37d7b504d5c01018cde756993 Mon Sep 17 00:00:00 2001
From: Anders Kaseorg <anders@zulip.com>
Date: Sat, 12 Sep 2020 19:46:59 -0700
Subject: [PATCH] markdown: Replace hyperlink requirement with urllib.parse.

The previous code only worked by accident and hyperlink 20.0.0 breaks
it.

>>> hyperlink.parse("example.com").replace(scheme="https")
DecodedURL(url=URL.from_text('https:example.com'))

Signed-off-by: Anders Kaseorg <anders@zulip.com>
---
 requirements/common.in          |  1 -
 requirements/dev.txt            |  2 +-
 requirements/prod.txt           |  6 +-----
 version.py                      |  2 +-
 zerver/lib/markdown/__init__.py | 11 +++++++----
 5 files changed, 10 insertions(+), 12 deletions(-)

diff --git a/requirements/common.in b/requirements/common.in
index 9a5053c619..f523022a30 100644
--- a/requirements/common.in
+++ b/requirements/common.in
@@ -18,7 +18,6 @@ Jinja2
 Markdown
 importlib-metadata;python_version<"3.8"  # for Markdown
 Pygments
-hyperlink
 jsx-lexer
 
 # Needed for manage.py
diff --git a/requirements/dev.txt b/requirements/dev.txt
index aa98b126a3..c4adf38d6e 100644
--- a/requirements/dev.txt
+++ b/requirements/dev.txt
@@ -368,7 +368,7 @@ hyperframe==3.2.0 \
 hyperlink==19.0.0 \
     --hash=sha256:4288e34705da077fada1111a24a0aa08bb1e76699c9ce49876af722441845654 \
     --hash=sha256:ab4a308feb039b04f855a020a6eda3b18ca5a68e6d8f8c899cbe9e653721d04f \
-    # via -r requirements/common.in, twisted
+    # via twisted
 idna==2.8 \
     --hash=sha256:c357b3f628cf53ae2c4c05627ecc484553142ca23264e593d327bcde5e9c3407 \
     --hash=sha256:ea8b7f6188e6fa117537c3df7da9fc686d485087abf6ac197f9c46432f7e4a3c \
diff --git a/requirements/prod.txt b/requirements/prod.txt
index 10f9daf82b..6bebc8a352 100644
--- a/requirements/prod.txt
+++ b/requirements/prod.txt
@@ -257,14 +257,10 @@ hyperframe==3.2.0 \
     --hash=sha256:05f0e063e117c16fcdd13c12c93a4424a2c40668abfac3bb419a10f57698204e \
     --hash=sha256:4dcab11967482d400853b396d042038e4c492a15a5d2f57259e2b5f89a32f755 \
     # via h2, hyper
-hyperlink==19.0.0 \
-    --hash=sha256:4288e34705da077fada1111a24a0aa08bb1e76699c9ce49876af722441845654 \
-    --hash=sha256:ab4a308feb039b04f855a020a6eda3b18ca5a68e6d8f8c899cbe9e653721d04f \
-    # via -r requirements/common.in
 idna==2.8 \
     --hash=sha256:c357b3f628cf53ae2c4c05627ecc484553142ca23264e593d327bcde5e9c3407 \
     --hash=sha256:ea8b7f6188e6fa117537c3df7da9fc686d485087abf6ac197f9c46432f7e4a3c \
-    # via hyperlink, requests
+    # via requests
 ijson==3.1.post0 \
     --hash=sha256:07cd88c9224392726128673fea7a90276746d3bf56d11635e3457769eda6295a \
     --hash=sha256:0ae656f3ca3bd833dc548583f13ba205eaee7c15d27496091549de4523931e00 \
diff --git a/version.py b/version.py
index bafce853fe..cb7ee65964 100644
--- a/version.py
+++ b/version.py
@@ -44,4 +44,4 @@ API_FEATURE_LEVEL = 32
 #   historical commits sharing the same major version, in which case a
 #   minor version bump suffices.
 
-PROVISION_VERSION = '102.0'
+PROVISION_VERSION = '103.0'
diff --git a/zerver/lib/markdown/__init__.py b/zerver/lib/markdown/__init__.py
index af20f7e5e5..132d81c62c 100644
--- a/zerver/lib/markdown/__init__.py
+++ b/zerver/lib/markdown/__init__.py
@@ -26,6 +26,7 @@ from typing import (
     Union,
 )
 from typing.re import Match, Pattern
+from urllib.parse import urlsplit
 from xml.etree import ElementTree as etree
 from xml.etree.ElementTree import Element, SubElement
 
@@ -36,7 +37,6 @@ import markdown
 import requests
 from django.conf import settings
 from django.db.models import Q
-from hyperlink import parse
 from markdown.extensions import codehilite, nl2br, sane_lists, tables
 from typing_extensions import TypedDict
 
@@ -2077,9 +2077,12 @@ def topic_links(realm_filters_key: int, topic_name: str) -> List[str]:
         link_match = re.match(get_web_link_regex(), sub_string)
         if link_match:
             url = link_match.group('url')
-            url_object = parse(url)
-            if not url_object.scheme:
-                url = url_object.replace(scheme='https').to_text()
+            result = urlsplit(url)
+            if not result.scheme:
+                if not result.netloc:
+                    i = (result.path + "/").index("/")
+                    result = result._replace(netloc=result.path[:i], path=result.path[i:])
+                url = result._replace(scheme="https").geturl()
             matches.append(url)
 
     return matches