feat: capture coauthors and per-item width/height/tagged_users

Two additions to the metadata sidecars, following up on the 2026-09-01
collab/dedup investigation already documented here:

- coauthors added to the post-level JSON's include list -- a direct,
  native signal for collab detection instead of inferring it from the
  filename/directory identity mismatch also documented that day.
- width, height, width_original, height_original and tagged_users get
  their own per-file sidecar (event: "file", one JSON per downloaded
  media item, named "<filename>.json"), since a carousel's items can
  each have different dimensions and tags -- data the existing
  post-level JSON has no way to represent. owner is deliberately left
  out, same reasoning as audio_user's existing exclusion.

Verified against a live re-fetch of an already-archived carousel:
correct per-item data came back, zero media re-downloaded (skip-archive
still applies; only the new sidecars are new files). 3 new tests.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_011qAds5qr7nZRq5R4yAuxUk

docs/gallery-dl.md

@@ -301,16 +301,31 @@ participant's page shows the post, not just the one whose directory JD2/
 gallery-dl happened to land the JSON in. See that repo's `NOTES.md`,
 2026-09-01 entries, for the full fix.
 
-**`coauthors` is a native, richer signal for this that we don't currently
-capture.** The raw API response carries a `coauthors` array
-(`{"full_name", "id", "username"}` per collaborator) that **excludes the
-post's own owner** — confirmed on both saved examples, including one where
-the owner (`dailyfashion_news`) is an external account with no ARTMS members
-in her own name, present only via `coauthors`. It is not in the metadata
-sidecar's `include` list today. Adding it would give collab detection a
-direct field instead of inferring it from filename/directory identity
-mismatches, and it never needs the "is this the owner" branch `coauthors`
-already excludes for us.
+**`coauthors` is a native, richer signal for this, and is now captured.**
+The raw API response carries a `coauthors` array (`{"full_name", "id",
+"username"}` per collaborator) that **excludes the post's own owner** —
+confirmed on both saved examples, including one where the owner
+(`dailyfashion_news`) is an external account with no ARTMS members in her
+own name, present only via `coauthors`. Added to the post-level metadata
+JSON's `include` list on 2026-09-01: a direct field instead of inferring a
+collab from filename/directory identity mismatches, and it never needs the
+"is this the owner" branch `coauthors` already excludes for us.
+
+**Per-carousel-item fields got their own sidecar, added the same day.**
+`width`, `height`, `width_original`, `height_original` and `tagged_users`
+live on the per-FILE kwdict, not the per-post one the metadata JSON above
+reads — a carousel's items can each have different dimensions and tags,
+which one post-level JSON can't represent. `gdl-sync.py`'s `media_pp` is a
+second `metadata` postprocessor, `event: "file"` (gallery-dl's default when
+omitted), so it runs once per downloaded file and writes `<filename>.json`
+alongside it — e.g. `... - 01.jpg.json` next to `... - 01.jpg`, never
+colliding with the post-level `....json`, which has no per-item number.
+`owner` — a full user object (profile pic URLs, privacy flags) for whoever
+posted that specific item — is deliberately left out, the same reasoning as
+`audio_user` above. Verified against the same two saved examples: correct
+per-item `width`/`height` and `tagged_users` came back on a live re-fetch of
+an already-archived carousel, with zero media re-downloaded (the existing
+skip-archive still applies; only the new sidecars are new files).
 
 ### The shared archive-db dedups media across profiles too, not just within one
 

scripts/gdl-sync.py

@@ -325,6 +325,32 @@ def build_config(rate: str, sleep_request: list[float],
             "post_shortcode", "post_id", "type", "date", "post_date",
             "username", "fullname", "owner_id", "description", "count",
             "likes", "post_url", "sidecar_shortcode",
+            # A Collab's `username`/`fullname`/`owner_id` above are always the
+            # original poster's, never the scraped account's -- see
+            # docs/gallery-dl.md. `coauthors` is the API's own list of every
+            # OTHER collaborator (it excludes the post's owner), a direct
+            # signal instead of inferring a collab from identity mismatches.
+            "coauthors",
+        ],
+    }
+    # Per-carousel-item fields (`width`/`height`/`tagged_users`/`owner`) live
+    # on the per-FILE kwdict, not the per-post one `meta_pp` reads -- a
+    # carousel's items can each have different dimensions and tags, which one
+    # post-level JSON can't represent. `owner` is deliberately left out: it is
+    # a full user object (profile pic URLs, privacy flags) for whoever posted
+    # that specific item, the same reason `audio_user` is excluded above.
+    media_pp = {
+        "name": "metadata",
+        "mode": "json",
+        # event defaults to "file" when omitted -- runs once per downloaded
+        # file, appending ".json" to that file's own full name, e.g.
+        # "... - 01.jpg" gets "... - 01.jpg.json" alongside it. Never
+        # collides with the post-level "....json" above, which has no
+        # per-item number.
+        "include": [
+            "media_id", "shortcode", "num",
+            "width", "height", "width_original", "height_original",
+            "tagged_users",
         ],
     }
 
@@ -363,6 +389,7 @@ def build_config(rate: str, sleep_request: list[float],
             "postprocessors": [
                 {**caption_pp, "filename": stem + ".txt"},
                 {**meta_pp, "filename": stem + ".json"},
+                media_pp,
             ],
         }
 

scripts/test_gdl_sync.py

@@ -232,6 +232,37 @@ class PostUrl(unittest.TestCase):
             self.assertNotIn(str(staging / subcategory), cmd)
 
 
+class MetadataFields(unittest.TestCase):
+    """Extra fields captured from the raw API response, verified against two
+    saved real examples on 2026-09-01 (see docs/gallery-dl.md)."""
+
+    def test_post_level_json_captures_coauthors(self):
+        config = gdl.build_config("1M", [6.0, 10.0], [3.0, 6.0])
+        pps = config["extractor"]["instagram"]["posts"]["postprocessors"]
+        post_json = next(pp for pp in pps if pp.get("filename", "").endswith(".json"))
+        self.assertIn("coauthors", post_json["include"])
+
+    def test_per_item_dimensions_and_tags_get_their_own_sidecar(self):
+        config = gdl.build_config("1M", [6.0, 10.0], [3.0, 6.0])
+        pps = config["extractor"]["instagram"]["posts"]["postprocessors"]
+        # No "filename" of its own -- unlike the other two, which are keyed
+        # by POST_STEM -- because it must vary per carousel item, not per post.
+        media_pp = next(pp for pp in pps if "filename" not in pp)
+        for field in ("width", "height", "width_original", "height_original",
+                      "tagged_users", "shortcode", "num"):
+            self.assertIn(field, media_pp["include"])
+        # `owner` is a full user object (profile pic URLs, privacy flags) for
+        # whoever posted that item -- deliberately excluded, same reasoning
+        # as `audio_user` on the post-level json.
+        self.assertNotIn("owner", media_pp["include"])
+
+    def test_reels_get_the_same_capture_as_posts(self):
+        config = gdl.build_config("1M", [6.0, 10.0], [3.0, 6.0])
+        posts_pps = config["extractor"]["instagram"]["posts"]["postprocessors"]
+        reels_pps = config["extractor"]["instagram"]["reels"]["postprocessors"]
+        self.assertEqual(posts_pps, reels_pps)
+
+
 class UrlsFile(unittest.TestCase):
     def test_reads_every_form_a_person_might_paste(self):
         with tempfile.TemporaryDirectory() as d: