Auto-translated map_item updates from 4CAT: twitter-import

[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: 48.65s
• Trigger: push of 7cba15f to 4CAT master (module: twitter-import).

Summary

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

Datasource	Module	Time	Warnings
datasources/twitter-import/search_twitter.py	modules/twitter.js	48.63s	⚠️ 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/twitter-import/search_twitter.py -> modules/twitter.js

• [helpers_to_add[5]] Bare truthy check on a name that looks like a collection. An empty [] or {} is false in Python and true in JavaScript, so this takes the opposite branch on an empty list, and the code inside then indexes element [0] of it. Use Array.isArray(x) && x.length (or Object.keys(x).length). NOTE: the check goes by the name, so a plural holding a string or a boolean is a false positive — ignore the warning if the value is not a list or an object.
• [helpers_to_add[7]] Regex detected. The current LLM translates regex unreliably (escapes, character classes, flags) — please verify the regex behavior against the Python original by hand.
• [helpers_to_add[9]] .toFixed( found. It pads with zeros where Python's str(round(x, n)) does not, turning a coordinate like 4.85 into 4.850000. NOTE: the corrected form String(Number(x.toFixed(n))) contains .toFixed( as well — ignore the warning where the result is already wrapped in Number(.

datasources/twitter-import/search_twitter.py -> modules/twitter.js (48.63s) ⚠️

LLM commentary:

Implemented map_item to dispatch between modern and legacy tweet shapes using py_get for safe dict access. All helper static methods from the Python class are provided as top‑level functions: map_user, get_note_result, get_full_text, get_entities, get_media, _screen_name_from_url, _screen_name_from_media, and get_centroid. Date parsing uses the built‑in Date constructor; formatUtcTimestamp formats the Unix timestamp. All Python truthy fallbacks (or) are translated to JavaScript ||, and dict.get calls become py_get. Missing fields are left as empty strings to match the original Python output. No external imports are required beyond the global helpers supplied by Zeeschuimer.

Python diff

diff --git a/datasources/twitter-import/search_twitter.py b/datasources/twitter-import/search_twitter.py
index 213770e1..79c24e51 100644
--- a/datasources/twitter-import/search_twitter.py
+++ b/datasources/twitter-import/search_twitter.py
@@ -153,7 +153,7 @@ class SearchTwitterViaZeeschuimer(Search):
             "author_bio": author_bio,
             "author_location": author_location,
             "verified": author_verified,
-            "source": strip_tags(tweet["source"]),
+            "source": strip_tags(tweet.get("source", "")),
             "language_guess": tweet["legacy"].get("lang"),
             "possibly_sensitive": "yes" if tweet.get("possibly_sensitive", False) or tweet["legacy"].get("possibly_sensitive", False) else "no",
             "retweet_count": tweet["legacy"]["retweet_count"],