[4cat-to-zeeschuimer-automation-pr]

🤖 This PR was auto-generated by the 4CAT map_item sync workflow. The JavaScript was produced by an LLM and requires human review before merging — including manual fixes for any lint warnings flagged below.

Generation parameters

• Model: gpt-oss-120b (provider: litellm)
• Total LLM time: 35.2s
• Trigger: push of 451220d to 4CAT master (module: xiaohongshu).

Summary

• ✅ 1 translated
• ⚠️ 1 translated with lint warnings (require manual fix)
• ❌ 0 failed
• ❔ 0 skipped

Datasource	Module	Time	Warnings
datasources/xiaohongshu/search_rednote.py	modules/rednote.js	35.19s	⚠️ 3

⚠️ Lint warnings — fix before merging

The following datasources translated successfully but the static lint flagged issues that need human fixes. The auto-generated code was spliced into the JS module as-is; please patch the file directly in this PR.

datasources/xiaohongshu/search_rednote.py -> modules/rednote.js

• [map_item_function] Even without an f prefix, "text {var}" / 'text {var}' are literal text in JavaScript — no interpolation happens. Whenever the original Python used an f-string, the JS must use a template literal (backticks).
• [map_item_function] Literal newline inside a string literal — JS strings can't span lines without escape ("\n") or template literals (`\n`).
• [map_item_function] Regex detected. The current LLM translates regex unreliably (escapes, character classes, flags) — please verify the regex behavior against the Python original by hand.

datasources/xiaohongshu/search_rednote.py -> modules/rednote.js (35.19s) ⚠️

LLM commentary:

Implemented all static helper methods from the Python class as plain functions: map_item_from_json_api_explore, getVideoUrl, map_item_from_json_embedded, and map_item_from_html. Used the global helpers (MappedItem, MissingMappedField, MapItemException, normalize_url_encoding, py_get, formatUtcTimestamp) as required. Regex for hashtags mirrors the Python pattern using matchAll and captures the group. Image handling follows the same fallback logic, returning a MissingMappedField when no images are found. Timestamp conversion uses formatUtcTimestamp for human‑readable strings and Math.floor for unix seconds. All .get calls are replaced with py_get, and f‑strings become template literals. No additional imports are needed.

Python diff

diff --git a/datasources/xiaohongshu/search_rednote.py b/datasources/xiaohongshu/search_rednote.py
index 27c5c930..13350a0d 100644
--- a/datasources/xiaohongshu/search_rednote.py
+++ b/datasources/xiaohongshu/search_rednote.py
@@ -107,7 +107,7 @@ class SearchRedNote(Search):
         # permalinks need this token to work, else you get a 404 not found
         xsec_bit = f"?xsec_token={post['xsec_token']}" if post.get("xsec_token") else ""
         if item.get("video", {}).get("media"):
-            video_url = item["video"]["media"]["stream"]["h264"][0]["master_url"]
+            video_url = SearchRedNote.get_video_url(item["video"]["media"].get("stream") or {})
         else:
             video_url = MissingMappedField("")
 
@@ -136,6 +136,29 @@ class SearchRedNote(Search):
             "unix_timestamp": int(timestamp / 1000) if timestamp else MissingMappedField(""),
         })
 
+    @staticmethod
+    def get_video_url(streams):
+        """
+        Get the URL of a post's video
+
+        RedNote offers each video in several versions, grouped by video
+        format. Older data names these groups after the format ("h264",
+        "h265"); newer data uses codes instead ("EF4", "EF5"). H.264 plays in
+        the most browsers and video players, so use it when it is there, and
+        otherwise the first version of any format that has a URL.
+
+        :param dict streams:  The post's `video.media.stream` object
+        :return str|MissingMappedField:
+        """
+        h264 = ("h264", "EF4")
+        formats = [*h264, *[name for name in streams if name not in h264]]
+        for name in formats:
+            for stream in streams.get(name) or []:
+                if stream.get("master_url"):
+                    return stream["master_url"]
+
+        return MissingMappedField("")
+
     @staticmethod
     def map_item_from_json_embedded(item):
         """