Use fixtures for PR checks

.github/workflows/python-test.yml

@@ -1,4 +1,4 @@
-name: Python Package using Conda
+name: Tests
 
 on:
   push:
@@ -27,7 +27,7 @@ permissions:
   checks: write          # a neutral (grey) "Backend coverage" check when skipped
 
 jobs:
-  notebooks:
+  tests:
     name: "Run Tests"
     runs-on: ubuntu-latest
     steps:

.github/workflows/test-lint.yml

@@ -1,4 +1,4 @@
-name: Test Lint
+name: Lint tests
 
 # Fast, static check that a PR doesn't introduce the silently-passing test
 # anti-patterns documented in TESTING.md (empty-suppressing guards, stale keys,

scripts/capture_term_info_fixtures.py

@@ -0,0 +1,157 @@
+#!/usr/bin/env python3
+"""Capture / refresh committed term_info render-test fixtures from live SOLR.
+
+The term_info serialization, parity and expression-pattern render tests split
+into a **fixture** variant (PR-blocking, deterministic) and a ``_live`` /
+``data_health`` variant (scheduled, live). The fixture variants read a committed
+complete ``vfb_json`` term_info document out of ``src/test/fixtures/term_info/``.
+This script (re)captures those documents so they stay regenerable rather than
+hand-authored. See ``TESTING.md`` -> "Fixture vs live (data_health)".
+
+A fixture is the raw term_info dict as ``vfb_queries._load_term_info_dict``
+returns it — the exact shape every ``_load_fixture`` / ``_fixture_raw`` /
+``_fixture_term_info`` helper consumes.
+
+Usage::
+
+    python scripts/capture_term_info_fixtures.py            # capture all in FIXTURES
+    python scripts/capture_term_info_fixtures.py <name>...  # only these fixtures
+    python scripts/capture_term_info_fixtures.py --check    # verify, don't write
+
+``--check`` fetches each id and reports whether the live document looks complete
+(non-empty label/title etc.), without writing — a quick way to see, before a
+scheduled run does, which production ``vfb_json`` documents are currently thin.
+
+IMPORTANT: this reads live SOLR, so run it only when the backend is healthy, and
+eyeball the diff before committing — a fixture captured during an outage or an
+incomplete indexer run would bake the wrong "correct" answer into a PR gate.
+Some documents (title-less pubs, split classes missing their ``Expression_pattern``
+type) are known-incomplete in production; those fixtures are hand-completed from
+the PDB and are marked ``hand_edited`` below so this script leaves them alone
+unless explicitly named.
+"""
+import argparse
+import json
+import os
+import sys
+
+FIXTURE_DIR = os.path.join(os.path.dirname(__file__), "..", "src", "test",
+                           "fixtures", "term_info")
+
+#: fixture file name (without .json) -> the term whose term_info it holds.
+#: Keep in sync with the fixtures the render tests load. ``hand_edited`` fixtures
+#: are completed from the PDB because production SOLR serves them incomplete;
+#: this script skips them unless named explicitly (so a bulk recapture during an
+#: outage cannot silently hollow them out).
+FIXTURES = {
+    # pub / split — known-incomplete in production, completed from the PDB.
+    "pub_FBrf0243986":                  {"id": "FBrf0243986",                    "hand_edited": True},
+    "pub_FBrf0242477":                  {"id": "FBrf0242477",                    "hand_edited": True},
+    "split_VFBexp_FBtp0124468FBtp0133404": {"id": "VFBexp_FBtp0124468FBtp0133404", "hand_edited": True},
+    "split_VFBexp_FBtp0129935FBtp0129968": {"id": "VFBexp_FBtp0129935FBtp0129968", "hand_edited": True},
+
+    # term_info_queries_test.py — serialization render tests.
+    "class_FBbt_00048514_taste_mechanosensory": {"id": "FBbt_00048514"},
+    "individual_VFB_00010001_fru":      {"id": "VFB_00010001"},
+    "class_FBbt_00048531":              {"id": "FBbt_00048531"},
+    "neuron_class_FBbt_00048999":       {"id": "FBbt_00048999"},
+    "neuron_class_FBbt_00047030_EPG":   {"id": "FBbt_00047030"},
+    "dataset_Ito2013":                  {"id": "Ito2013"},
+    "license_CC_BY_NC_3_0":             {"id": "VFBlicense_CC_BY_NC_3_0"},
+    "template_VFB_00200000":            {"id": "VFB_00200000"},
+
+    # test_term_info_parity.py — parity / coverage render tests.
+    "class_FBbt_00003748_medulla":      {"id": "FBbt_00003748"},
+    "class_FBbt_00003686_kenyon":       {"id": "FBbt_00003686"},
+    "individual_VFB_00101385_MEon":     {"id": "VFB_00101385"},
+    "gene_FBgn0010339_128up":           {"id": "FBgn0010339"},
+    "gene_FBgn0051882":                 {"id": "FBgn0051882"},
+    "class_FBbt_00000058_related_individuals": {"id": "FBbt_00000058"},
+    "dataset_Cachero2010":              {"id": "Cachero2010"},
+    "template_VFB_00101567_JRC2018U":   {"id": "VFB_00101567"},
+    "neuron_class_FBbt_00100243_MBON":  {"id": "FBbt_00100243"},
+
+    # test_expression_pattern_individual_queries.py — menu / stock render tests.
+    "ep_class_VFBexp_FBtp0060056":      {"id": "VFBexp_FBtp0060056"},
+    "ep_individual_VFB_00020530":       {"id": "VFB_00020530"},
+    "split_individual_VFB_00070031":    {"id": "VFB_00070031"},
+    "split_individual_VFB_00069525":    {"id": "VFB_00069525"},
+}
+
+
+def _load(short_form):
+    from vfbquery.vfb_queries import _load_term_info_dict
+    return _load_term_info_dict(short_form)
+
+
+def _looks_complete(doc):
+    """Cheap completeness heuristic for --check: a real label, and for pubs a
+    non-empty title. Not exhaustive — the tests are the real check."""
+    if not doc:
+        return False, "no term_info document"
+    core = (doc.get("term") or {}).get("core") or {}
+    label, sf = core.get("label") or "", core.get("short_form") or ""
+    if not label or label == sf:
+        return False, "label is empty or just the id"
+    if "pub" in (core.get("types") or []):
+        if not ((doc.get("pub_specific_content") or {}).get("title") or "").strip():
+            return False, "pub has no title"
+    return True, "ok"
+
+
+def capture(names, check):
+    os.makedirs(FIXTURE_DIR, exist_ok=True)
+    rc = 0
+    for name in names:
+        entry = FIXTURES[name]
+        sf = entry["id"]
+        doc = _load(sf)
+        ok, why = _looks_complete(doc)
+        path = os.path.join(FIXTURE_DIR, name + ".json")
+        if check:
+            print(f"[{'ok' if ok else 'INCOMPLETE'}] {name} ({sf}): {why}")
+            if not ok:
+                rc = 1
+            continue
+        if not ok:
+            print(f"SKIP {name} ({sf}): live document {why} — not overwriting. "
+                  f"Recapture when the backend is healthy, or hand-complete it.")
+            rc = 1
+            continue
+        with open(path, "w", encoding="utf-8") as f:
+            json.dump(doc, f, indent=1, ensure_ascii=False)
+        print(f"wrote {os.path.relpath(path)} ({sf})")
+    return rc
+
+
+def main():
+    ap = argparse.ArgumentParser(description=__doc__,
+                                 formatter_class=argparse.RawDescriptionHelpFormatter)
+    ap.add_argument("names", nargs="*", help="fixture names to capture (default: all "
+                    "non-hand-edited)")
+    ap.add_argument("--check", action="store_true",
+                    help="report completeness of the live documents; do not write")
+    ap.add_argument("--include-hand-edited", action="store_true",
+                    help="also (re)capture fixtures marked hand_edited — use only "
+                         "when you intend to rebuild them from live data")
+    args = ap.parse_args()
+
+    if args.names:
+        unknown = [n for n in args.names if n not in FIXTURES]
+        if unknown:
+            ap.error("unknown fixture name(s): %s\nknown: %s"
+                     % (", ".join(unknown), ", ".join(sorted(FIXTURES))))
+        names = args.names
+    else:
+        names = [n for n, e in FIXTURES.items()
+                 if args.include_hand_edited or not e.get("hand_edited")]
+        if not names and not args.include_hand_edited:
+            print("All registered fixtures are hand_edited; nothing to capture "
+                  "automatically. Pass names explicitly or --include-hand-edited.")
+            return 0
+
+    return capture(names, args.check)
+
+
+if __name__ == "__main__":
+    sys.exit(main())

src/test/fixtures/term_info/class_FBbt_00000058_related_individuals.json

@@ -0,0 +1,55 @@
+{
+ "term": {
+  "core": {
+   "iri": "http://purl.obolibrary.org/obo/FBbt_00000058",
+   "symbol": "",
+   "types": [
+    "Entity",
+    "Class",
+    "Deprecated"
+   ],
+   "short_form": "FBbt_00000058",
+   "unique_facets": [
+    "Deprecated"
+   ],
+   "label": "obsolete vitellophage"
+  },
+  "description": [],
+  "comment": [
+   "This term was obsoleted because it is the same as 'yolk nucleus' - mc141031."
+  ]
+ },
+ "query": "Get JSON for Class",
+ "version": "0dcd736",
+ "parents": [],
+ "relationships": [],
+ "related_individuals": [
+  {
+   "relation": {
+    "iri": "http://purl.obolibrary.org/obo/IAO_0100001",
+    "database_cross_reference": [],
+    "label": "term replaced by",
+    "type": "term_replaced_by",
+    "confidence_value": ""
+   },
+   "object": {
+    "symbol": "",
+    "iri": "http://purl.obolibrary.org/obo/FBbt_00000057",
+    "types": [
+     "Entity",
+     "Class",
+     "Anatomy"
+    ],
+    "short_form": "FBbt_00000057",
+    "unique_facets": [
+     "Anatomy"
+    ],
+    "label": "yolk nucleus"
+   }
+  }
+ ],
+ "xrefs": [],
+ "anatomy_channel_image": [],
+ "pub_syn": [],
+ "def_pubs": []
+}