Merge pull request 'v26.05.24.2: IR Post/Provenance restore + modal artist fallback' (#8) from dev into main

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

backend/app/services/gallery_service.py

@@ -237,6 +237,15 @@ class GalleryService:
         )
         tags = (await self.session.execute(tag_stmt)).scalars().all()
         neighbors = await self._neighbors(record)
+        # Direct artist FK — used by the modal's ProvenancePanel as a
+        # fallback when ImageProvenance is empty (i.e., filesystem-
+        # imported images without a post-track provenance row). The
+        # source of truth for richer post-level data is still
+        # ImageProvenance/Post; this is just the "we at least know who
+        # made it" line.
+        artist = None
+        if record.artist_id is not None:
+            artist = await self.session.get(Artist, record.artist_id)
         return {
             "id": record.id,
             "path": record.path,
@@ -249,6 +258,10 @@ class GalleryService:
             "created_at": record.created_at.isoformat(),
             "thumbnail_url": thumbnail_url(record.sha256, record.mime),
             "image_url": f"/images/{record.path.split('/images/', 1)[-1]}",
+            "artist": (
+                {"id": artist.id, "name": artist.name, "slug": artist.slug}
+                if artist is not None else None
+            ),
             "tags": [
                 {
                     "id": t.id,

backend/app/services/migrators/ir_ingest.py

@@ -70,7 +70,7 @@ async def migrate_async(
     """
     if data.get("source_app") != "imagerepo":
         raise ValueError("export source_app must be 'imagerepo'")
-    if data.get("schema_version") != 1:
+    if data.get("schema_version") not in (1, 2):
         raise ValueError(f"unsupported schema_version: {data.get('schema_version')}")
 
     counts = _zero_counts()
@@ -126,15 +126,20 @@ async def migrate_async(
         await db.commit()
 
     # Phase 2: write the per-image manifest for tag_apply.py to consume later.
+    # schema_version 2 (added 2026-05-24) carries `image_posts` for
+    # Post + Source + ImageProvenance restore; schema 1 manifests
+    # without it stay valid (tag_apply treats the missing field as []).
     manifest = {
-        "schema_version": 1,
+        "schema_version": data.get("schema_version", 1),
         "image_artist_assignments": data.get("image_artist_assignments", []),
         "image_tag_associations": data.get("image_tag_associations", []),
         "series_pages": data.get("series_pages", []),
+        "image_posts": data.get("image_posts", []),
     }
     counts["rows_processed"] += len(manifest["image_artist_assignments"])
     counts["rows_processed"] += len(manifest["image_tag_associations"])
     counts["rows_processed"] += len(manifest["series_pages"])
+    counts["rows_processed"] += len(manifest["image_posts"])
     if not dry_run:
         manifest_path(images_root).write_text(json.dumps(manifest, indent=2))
 

backend/app/services/migrators/tag_apply.py

@@ -7,6 +7,7 @@ FC's filesystem scan over the mounted IR images dir).
 - image_artist_assignments → ImageRecord.artist_id (find_or_create Artist by slug).
 - image_tag_associations  → image_tag insert (idempotent).
 - series_pages            → series_page insert (idempotent on image_id unique).
+- image_posts (schema v2) → Source + Post + ImageProvenance restore.
 
 Unmatched sha256s are logged into the result's `unmatched` list so the
 Celery task can drop them into MigrationRun.metadata for the operator
@@ -15,15 +16,118 @@ to inspect.
 from __future__ import annotations
 
 import json
+from datetime import datetime
 from pathlib import Path
 
 from sqlalchemy import select
 from sqlalchemy.ext.asyncio import AsyncSession
 
-from ...models import Artist, ImageRecord, SeriesPage, Tag, TagKind, image_tag
+from ...models import (
+    Artist,
+    ImageProvenance,
+    ImageRecord,
+    Post,
+    SeriesPage,
+    Source,
+    Tag,
+    TagKind,
+    image_tag,
+)
 from ...utils.slug import slugify
 from .ir_ingest import manifest_path
 
+# Per-platform artist-profile URL — used as Source.url when restoring
+# IR PostMetadata into FC. Keep this table in sync with
+# backend/app/services/extension_service.py:_PLATFORM_PATTERNS and
+# extension/lib/platforms.js.
+_PLATFORM_PROFILE_URL = {
+    "patreon":       "https://www.patreon.com/{slug}",
+    "subscribestar": "https://www.subscribestar.com/{slug}",
+    "hentaifoundry": "https://www.hentai-foundry.com/user/{slug}",
+}
+
+
+def _profile_url(platform: str, artist_slug: str) -> str | None:
+    fmt = _PLATFORM_PROFILE_URL.get(platform)
+    return fmt.format(slug=artist_slug) if fmt else None
+
+
+async def _find_or_create_source(
+    db: AsyncSession, *, artist_id: int, platform: str, url: str, dry_run: bool,
+) -> int | None:
+    existing = (await db.execute(
+        select(Source.id).where(
+            Source.artist_id == artist_id,
+            Source.platform == platform,
+            Source.url == url,
+        )
+    )).scalar_one_or_none()
+    if existing is not None:
+        return existing
+    if dry_run:
+        return None
+    s = Source(artist_id=artist_id, platform=platform, url=url, enabled=False)
+    db.add(s)
+    await db.flush()
+    return s.id
+
+
+async def _find_or_create_post(
+    db: AsyncSession, *,
+    source_id: int, external_post_id: str,
+    title: str | None, description: str | None, post_url: str | None,
+    post_date_iso: str | None, attachment_count: int, dry_run: bool,
+) -> int | None:
+    existing = (await db.execute(
+        select(Post.id).where(
+            Post.source_id == source_id,
+            Post.external_post_id == external_post_id,
+        )
+    )).scalar_one_or_none()
+    if existing is not None:
+        return existing
+    if dry_run:
+        return None
+    post_date = None
+    if post_date_iso:
+        post_date = datetime.fromisoformat(post_date_iso)
+    p = Post(
+        source_id=source_id,
+        external_post_id=external_post_id,
+        post_title=title,
+        description=description,
+        post_url=post_url,
+        post_date=post_date,
+        attachment_count=attachment_count,
+        raw_metadata={"migrated_from": "imagerepo"},
+    )
+    db.add(p)
+    await db.flush()
+    return p.id
+
+
+async def _ensure_provenance(
+    db: AsyncSession, *,
+    image_id: int, post_id: int, source_id: int, dry_run: bool,
+) -> bool:
+    """Returns True if a new ImageProvenance row was inserted."""
+    existing = (await db.execute(
+        select(ImageProvenance.id).where(
+            ImageProvenance.image_record_id == image_id,
+            ImageProvenance.post_id == post_id,
+            ImageProvenance.source_id == source_id,
+        )
+    )).scalar_one_or_none()
+    if existing is not None:
+        return False
+    if dry_run:
+        return True
+    db.add(ImageProvenance(
+        image_record_id=image_id, post_id=post_id, source_id=source_id,
+    ))
+    await db.flush()
+    return True
+
 
 def _zero_counts() -> dict:
     return {
@@ -167,6 +271,67 @@ async def apply_async(
         ))
         counts["rows_inserted"] += 1
 
+    # 4. Image posts (schema v2) → Source + Post + ImageProvenance.
+    # Restores IR PostMetadata as FC's downloader-track provenance,
+    # so the modal's ProvenancePanel surfaces title/description/
+    # source URL/publish date the same way it does for live
+    # gallery-dl downloads.
+    for entry in manifest.get("image_posts", []):
+        counts["rows_processed"] += 1
+        platform = entry.get("platform")
+        artist_name = entry.get("artist")
+        if not platform or not artist_name:
+            counts["rows_skipped"] += 1
+            continue
+
+        aid = await _ensure_artist_id(db, artist_name, dry_run)
+        if aid is None:
+            counts["rows_skipped"] += 1
+            continue
+
+        url = _profile_url(platform, slugify(artist_name))
+        if url is None:
+            counts["rows_skipped"] += 1
+            continue
+
+        source_id = await _find_or_create_source(
+            db, artist_id=aid, platform=platform, url=url, dry_run=dry_run,
+        )
+        if source_id is None:
+            counts["rows_skipped"] += 1
+            continue
+
+        post_id = await _find_or_create_post(
+            db, source_id=source_id,
+            external_post_id=entry.get("post_id") or "",
+            title=entry.get("title"),
+            description=entry.get("description"),
+            post_url=entry.get("source_url"),
+            post_date_iso=entry.get("published_at"),
+            attachment_count=entry.get("attachment_count") or 0,
+            dry_run=dry_run,
+        )
+        if post_id is None:
+            counts["rows_skipped"] += 1
+            continue
+
+        for sha in entry.get("image_sha256s", []):
+            img_id = await _sha_to_image_id(db, sha)
+            if img_id is None:
+                unmatched.append({
+                    "kind": "post", "sha256": sha,
+                    "post_id": entry.get("post_id"),
+                })
+                continue
+            inserted = await _ensure_provenance(
+                db, image_id=img_id, post_id=post_id,
+                source_id=source_id, dry_run=dry_run,
+            )
+            if inserted:
+                counts["rows_inserted"] += 1
+            else:
+                counts["rows_skipped"] += 1
+
     if not dry_run:
         await db.commit()
     return {"counts": counts, "unmatched": unmatched}

frontend/src/components/modal/ProvenancePanel.vue

@@ -43,6 +43,21 @@
           class="fc-prov__desc" v-html="e.post.description_html"
         />
       </article>
+
+      <!-- Fallback: image has artist_id set (folder-derived from
+           filesystem import or set by tag_apply) but no ImageProvenance
+           row exists. Show a minimal "by <artist>" line so the modal
+           isn't blank for IR-imported images that aren't in any post. -->
+      <article
+        v-if="showArtistFallback"
+        class="fc-prov__card"
+      >
+        <div class="fc-prov__meta">
+          <RouterLink :to="`/artist/${fallbackArtist.slug}`">
+            by {{ fallbackArtist.name }}
+          </RouterLink>
+        </div>
+      </article>
     </template>
 
     <div v-if="attachments.length" class="fc-prov__attach">
@@ -79,13 +94,27 @@ const state = computed(() =>
   modal.currentImageId == null ? null : prov.imageProv(modal.currentImageId)
 )
 
-// Show the panel while loading or on error, or when there is >=1 entry.
+const fallbackArtist = computed(() => modal.current?.artist || null)
-// Hide entirely when the image has zero provenance.
+
+const showArtistFallback = computed(() => {
+  const st = state.value
+  // Only show the fallback when provenance is genuinely empty (no
+  // entries, not loading, no error) AND the image has a direct
+  // artist_id we can surface.
+  if (!st || st.loading || st.error) return false
+  if (Array.isArray(st.entries) && st.entries.length > 0) return false
+  return fallbackArtist.value != null
+})
+
+// Show the panel while loading or on error, when there is >=1 entry,
+// or when the artist fallback applies (image_record.artist_id set but
+// no ImageProvenance row, e.g. filesystem-imported IR images).
 const show = computed(() => {
   const st = state.value
   if (!st) return false
   if (st.loading || st.error) return true
-  return Array.isArray(st.entries) && st.entries.length > 0
+  if (Array.isArray(st.entries) && st.entries.length > 0) return true
+  return showArtistFallback.value
 })
 
 const attachments = computed(() => state.value?.attachments || [])

tests/test_tag_apply.py

@@ -2,9 +2,19 @@
 import json
 
 import pytest
-from sqlalchemy import select
+from sqlalchemy import func, select
 
-from backend.app.models import Artist, ImageRecord, SeriesPage, Tag, TagKind, image_tag
+from backend.app.models import (
+    Artist,
+    ImageProvenance,
+    ImageRecord,
+    Post,
+    SeriesPage,
+    Source,
+    Tag,
+    TagKind,
+    image_tag,
+)
 from backend.app.services.migrators import tag_apply
 
 pytestmark = pytest.mark.integration
@@ -21,14 +31,19 @@ async def _seed_image(db, sha: str, *, suffix="x"):
     return img
 
 
-def _write_manifest(tmp_path, *, artist_assignments=None, tag_associations=None, series_pages=None):
+def _write_manifest(
+    tmp_path, *,
+    artist_assignments=None, tag_associations=None,
+    series_pages=None, image_posts=None, schema_version=2,
+):
     d = tmp_path / "_migration_state"
     d.mkdir(parents=True, exist_ok=True)
     (d / "ir_tag_manifest.json").write_text(json.dumps({
-        "schema_version": 1,
+        "schema_version": schema_version,
         "image_artist_assignments": artist_assignments or [],
         "image_tag_associations": tag_associations or [],
         "series_pages": series_pages or [],
+        "image_posts": image_posts or [],
     }))
 
 
@@ -154,3 +169,130 @@ async def test_dry_run_makes_no_changes(db, tmp_path):
         select(image_tag.c.tag_id).where(image_tag.c.image_record_id == img_id)
     )).all()
     assert len(rows) == 0
+
+
+# --- Phase 4: image_posts (schema v2) -----------------------------------
+
+
+_POST_ENTRY = {
+    "platform": "patreon",
+    "post_id": "10001",
+    "artist": "Maewix",
+    "title": "Test Post Title",
+    "description": "<p>HTML description</p>",
+    "source_url": "https://www.patreon.com/posts/test-10001",
+    "attachment_count": 1,
+    "published_at": "2026-01-15T12:00:00+00:00",
+}
+
+
+@pytest.mark.asyncio
+async def test_image_posts_creates_source_post_provenance(db, tmp_path):
+    sha = "d" * 64
+    await _seed_image(db, sha, suffix="d")
+    await db.commit()
+
+    _write_manifest(tmp_path, image_posts=[
+        {**_POST_ENTRY, "image_sha256s": [sha]},
+    ])
+
+    result = await tag_apply.apply_async(db, images_root=tmp_path, dry_run=False)
+    assert result["counts"]["rows_inserted"] >= 1
+
+    # Async Core-DML assertions — column selects, not ORM attribute access.
+    artist_count = (await db.execute(
+        select(func.count(Artist.id)).where(Artist.slug == "maewix")
+    )).scalar_one()
+    assert artist_count == 1
+
+    source_count = (await db.execute(
+        select(func.count(Source.id)).where(
+            Source.platform == "patreon",
+            Source.url == "https://www.patreon.com/maewix",
+        )
+    )).scalar_one()
+    assert source_count == 1
+
+    post_count = (await db.execute(
+        select(func.count(Post.id)).where(Post.external_post_id == "10001")
+    )).scalar_one()
+    assert post_count == 1
+
+    img_id = (await db.execute(
+        select(ImageRecord.id).where(ImageRecord.sha256 == sha)
+    )).scalar_one()
+    prov_count = (await db.execute(
+        select(func.count(ImageProvenance.id))
+        .where(ImageProvenance.image_record_id == img_id)
+    )).scalar_one()
+    assert prov_count == 1
+
+
+@pytest.mark.asyncio
+async def test_image_posts_idempotent_on_rerun(db, tmp_path):
+    sha = "e" * 64
+    await _seed_image(db, sha, suffix="e")
+    await db.commit()
+
+    _write_manifest(tmp_path, image_posts=[
+        {**_POST_ENTRY, "image_sha256s": [sha]},
+    ])
+
+    await tag_apply.apply_async(db, images_root=tmp_path, dry_run=False)
+    result2 = await tag_apply.apply_async(db, images_root=tmp_path, dry_run=False)
+
+    # Re-run: nothing should be newly inserted; all skipped.
+    assert result2["counts"]["rows_inserted"] == 0
+    assert result2["counts"]["rows_skipped"] >= 1
+
+    prov_count = (await db.execute(
+        select(func.count(ImageProvenance.id))
+    )).scalar_one()
+    assert prov_count == 1
+
+
+@pytest.mark.asyncio
+async def test_image_posts_unknown_sha_unmatched(db, tmp_path):
+    # No image seeded; sha doesn't exist.
+    _write_manifest(tmp_path, image_posts=[
+        {**_POST_ENTRY, "image_sha256s": ["f" * 64]},
+    ])
+    result = await tag_apply.apply_async(db, images_root=tmp_path, dry_run=False)
+    unmatched_kinds = [u["kind"] for u in result["unmatched"]]
+    assert "post" in unmatched_kinds
+
+
+@pytest.mark.asyncio
+async def test_image_posts_unknown_platform_skipped(db, tmp_path):
+    sha = "1" * 64
+    await _seed_image(db, sha, suffix="1")
+    await db.commit()
+
+    _write_manifest(tmp_path, image_posts=[
+        {**_POST_ENTRY, "platform": "tumblr", "image_sha256s": [sha]},
+    ])
+    result = await tag_apply.apply_async(db, images_root=tmp_path, dry_run=False)
+    assert result["counts"]["rows_skipped"] >= 1
+
+    # No Post / Source / ImageProvenance should have been created.
+    src_count = (await db.execute(
+        select(func.count(Source.id)).where(Source.platform == "tumblr")
+    )).scalar_one()
+    assert src_count == 0
+
+
+@pytest.mark.asyncio
+async def test_image_posts_dry_run_makes_no_writes(db, tmp_path):
+    sha = "2" * 64
+    await _seed_image(db, sha, suffix="2")
+    await db.commit()
+
+    _write_manifest(tmp_path, image_posts=[
+        {**_POST_ENTRY, "image_sha256s": [sha]},
+    ])
+
+    await tag_apply.apply_async(db, images_root=tmp_path, dry_run=True)
+    prov_count = (await db.execute(
+        select(func.count(ImageProvenance.id))
+    )).scalar_one()
+    assert prov_count == 0