From c2fb556919e6398609eeaefcfdf6c68635884dab Mon Sep 17 00:00:00 2001
From: Daksh <dakshlacksdinner@gmail.com>
Date: Sun, 27 Sep 2026 16:35:11 +0530
Subject: [PATCH] fix(ingest): rewrite only the host when normalizing a tweet
 URL for oEmbed

_fetch_tweet did url.replace("x.com", "twitter.com") over the whole URL,
so "x.com" in the path or query was rewritten too (a ?next= link, or a
handle like dropbox.com_fan becoming dropbotwitter.com_fan), and an
uppercase X.COM host was missed. Match and rewrite the host only, the
same way _detect_url_type classifies it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit 3ec4fdcfca5f34a82d7f163807eb7108c4e5e983)
---
 graphify/ingest.py            | 11 +++++++++--
 tests/test_ingest_url_type.py | 37 +++++++++++++++++++++++++++++++++++
 2 files changed, 46 insertions(+), 2 deletions(-)

diff --git a/graphify/ingest.py b/graphify/ingest.py
index c861d51a2..20962250b 100644
--- a/graphify/ingest.py
+++ b/graphify/ingest.py
@@ -109,8 +109,15 @@ def _html_to_markdown(html: str, url: str) -> str:
 
 def _fetch_tweet(url: str, author: str | None, contributor: str | None) -> tuple[str, str]:
     """Fetch a tweet URL. Returns (content, filename)."""
-    # Normalize to twitter.com for oEmbed
-    oembed_url = url.replace("x.com", "twitter.com")
+    # Normalize to twitter.com for oEmbed. Rewrite the host only: a plain
+    # text replace also rewrote "x.com" wherever it appeared in the path or query.
+    parts = urllib.parse.urlsplit(url)
+    host = (parts.hostname or "").rstrip(".")
+    if _host_is(host, "x.com"):
+        host = host[: -len("x.com")] + "twitter.com"
+        netloc = f"{host}:{parts.port}" if parts.port else host
+        parts = parts._replace(netloc=netloc)
+    oembed_url = urllib.parse.urlunsplit(parts)
     oembed_api = f"https://publish.twitter.com/oembed?url={urllib.parse.quote(oembed_url)}&omit_script=true"
     try:
         data = json.loads(safe_fetch_text(oembed_api))
diff --git a/tests/test_ingest_url_type.py b/tests/test_ingest_url_type.py
index c2cb9e2ba..45d5f11c6 100644
--- a/tests/test_ingest_url_type.py
+++ b/tests/test_ingest_url_type.py
@@ -1,6 +1,8 @@
 """Tests for how graphify.ingest classifies and routes a URL."""
 from __future__ import annotations
 
+import urllib.parse
+
 import pytest
 
 import graphify.ingest as ingest_mod
@@ -66,3 +68,38 @@ def test_a_dropbox_pdf_is_downloaded_not_saved_as_a_tweet_stub(tmp_path, monkeyp
     assert fetched == [url]
     assert out.suffix == ".pdf"
     assert out.read_bytes() == b"%PDF-1.7 test"
+
+
+@pytest.mark.parametrize(
+    ("url", "expected"),
+    [
+        ("https://x.com/user/status/1", "https://twitter.com/user/status/1"),
+        ("https://mobile.x.com/user/status/1", "https://mobile.twitter.com/user/status/1"),
+        ("https://X.COM/user/status/1", "https://twitter.com/user/status/1"),
+        # "x.com" outside the host must survive untouched.
+        (
+            "https://x.com/user/status/1?next=https://x.com/home",
+            "https://twitter.com/user/status/1?next=https://x.com/home",
+        ),
+        (
+            "https://twitter.com/dropbox.com_fan/status/1",
+            "https://twitter.com/dropbox.com_fan/status/1",
+        ),
+    ],
+)
+def test_tweet_oembed_rewrites_only_the_host(url, expected, monkeypatch):
+    """The x.com -> twitter.com rewrite for oEmbed was a replace over the whole
+    URL, so "x.com" inside the path or query (e.g. a dropbox.com handle, a
+    ?next= link) was rewritten too, sending oEmbed a different URL."""
+    requested = []
+
+    def fake_fetch_text(u):
+        requested.append(u)
+        return '{"html": "<p>hi</p>", "author_name": "someone"}'
+
+    monkeypatch.setattr(ingest_mod, "safe_fetch_text", fake_fetch_text)
+
+    ingest_mod._fetch_tweet(url, None, None)
+
+    query = urllib.parse.parse_qs(urllib.parse.urlsplit(requested[0]).query)
+    assert query["url"] == [expected]