Add the Commons reference-board research tool

tools/refboard.py

@@ -0,0 +1,293 @@
+#!/usr/bin/env python3
+"""Wikimedia Commons reference-board helper for venue documentation.
+
+Lists and fetches licence-cleared photos of a real location and writes the
+`SIDECAR.md` rows and the folder `README.md` gallery rows used by
+`docs/screens/refs/<venue>/`. Research downloads stay under `tmp/` unless
+`--stage refs` copies them into the board folder (still uncommitted; the git
+commit is a deliberate human step).
+
+    python3 tools/refboard.py search "El Capitan Entertainment Centre" \
+        [--months 12] [--radius 500] [--at coord] [--json]
+    python3 tools/refboard.py fetch --from search.json --dir tmp/research/elcapitan \
+        [--max N]
+    python3 tools/refboard.py rows --from meta.json --licence-note ... [--append-sidecar PATH]
+
+Licence gate: CC0 / public domain / CC BY are listed; CC BY-SA rows are marked
+`(review)` and must record the share-alike decision; anything else is skipped.
+
+Rate limits: the Commons API needs a descriptive User-Agent, metadata is fetched
+in one batched `titles=` query, and file downloads are spaced (default 3 s) to
+avoid HTTP 429.
+"""
+
+import argparse
+import json
+import re
+import sys
+import time
+import urllib.parse
+import urllib.request
+from pathlib import Path
+
+ROOT = Path(__file__).resolve().parent.parent
+USER_AGENT = "gig-line-dev-refboard/1.0 (development reference collection; contact: repo)"
+API = "https://commons.wikimedia.org/w/api.php"
+DOWNLOAD_SPACING_S = 3.0
+#Licences that can be committed with the stated attribution; anything else is
+#research-only.
+COMMITTABLE = {
+    "Public domain": "Public domain",
+    "CC0": "CC0",
+    "CC BY 2.0": "CC BY 2.0",
+    "CC BY 3.0": "CC BY 3.0",
+    "CC BY 4.0": "CC BY 4.0",
+}
+# CC BY-SA requires a share-alike review before a file enters the repository, so
+# it stays research-only until `--allow-ccbysa` records the decision.
+REVIEW_ONLY = {
+    "CC BY-SA 2.0": "CC BY-SA 2.0 (review)",
+    "CC BY-SA 3.0": "CC BY-SA 3.0 (review)",
+    "CC BY-SA 4.0": "CC BY-SA 4.0 (review)",
+}
+def _strip_html(text: str) -> str:
+    return re.sub("<[^>]+>", "", text or "").strip()
+
+
+def normalise_licence(raw: str, allow_ccbysa: bool = False):
+    """Return (committed label | None, research-only label | None).
+
+    `clean` is set when the file may enter the repository; `review` is set for
+    share-alike licences that stay under `tmp/` until a review records otherwise.
+    """
+    if raw in COMMITTABLE:
+        return raw, None
+    if raw in REVIEW_ONLY:
+        return (REVIEW_ONLY[raw], None) if allow_ccbysa else (None, REVIEW_ONLY[raw])
+    return None, raw or None
+class Commons:
+    def __init__(self, user_agent: str = USER_AGENT):
+        self.user_agent = user_agent
+        self.last_request = 0.0
+    def _get(self, params: dict, binary: bool = False, spacing: float = 0.0):
+        if spacing:
+            wait = spacing - (time.monotonic() - self.last_request)
+            if wait > 0:
+                time.sleep(wait)
+        self.last_request = time.monotonic()
+        req = urllib.request.Request(
+            API if not binary else params["url"],
+            data=None if binary else urllib.parse.urlencode(params).encode(),
+            headers={"User-Agent": self.user_agent}, method="GET" if binary else "POST")
+        with urllib.request.urlopen(req, timeout=90) as resp:
+            return resp.read() if binary else json.loads(resp.read().decode("utf-8"))
+    def meta_for_titles(self, titles: list[str], allow_ccbysa: bool = False) -> dict:
+        rows = {}
+        for chunk in [titles[i:i + 20] for i in range(0, len(titles), 20)]:
+            data = self._get({
+                "action": "query", "format": "json", "titles": "|".join(chunk),
+                "prop": "imageinfo", "iiprop": "url|size|extmetadata",
+                "iiurlwidth": "1600", "formatversion": "2",
+            })
+            for page in data.get("query", {}).get("pages", []):
+                if "missing" in page or not page.get("imageinfo"):
+                    continue
+                info = page["imageinfo"][0]
+                meta = info.get("extmetadata", {})
+                def val(key):
+                    return _strip_html(meta.get(key, {}).get("value", ""))
+
+                def first_date():
+                    for key in ("DateTimeOriginal", "DateTime", "DateTimeDigitized"):
+                        text = _strip_html(meta.get(key, {}).get("value", ""))
+                        match = re.search(r"\d{4}-\d{2}-\d{2}", text)
+                        if match:
+                            return match.group(0)
+                    return ""
+
+                raw_licence = val("LicenseShortName")
+                clean, review = normalise_licence(raw_licence, allow_ccbysa)
+                rows[page["title"]] = {
+                    "title": page["title"],
+                    "licence": raw_licence,
+                    "licence_url": val("LicenseUrl"),
+                    "artist": val("Artist") or "unknown",
+                    "credit": val("Credit")[:120],
+                    "date": first_date(),
+                    "description": val("ImageDescription")[:300],
+                    "page": info.get("descriptionurl", ""),
+                    "width": info.get("thumbwidth") or info.get("width", 0),
+                    "height": info.get("thumbheight") or info.get("height", 0),
+                    "url": info.get("thumburl") or info.get("url", ""),
+                    "clean": clean,
+                    "review": review,
+                }
+        return rows
+    def search(self, query: str, limit: int = 12) -> list[str]:
+        data = self._get({"action": "query", "format": "json", "generator": "search",
+                          "gsrsearch": query, "gsrnamespace": "6", "gsrlimit": str(limit),
+                          "formatversion": "2"})
+        return [p["title"] for p in data.get("query", {}).get("pages", [])
+                if not p.get("missing")]
+    def geosearch(self, coord: str, radius_m: int = 500, limit: int = 60) -> list[str]:
+        data = self._get({"action": "query", "format": "json", "generator": "geosearch",
+                          "ggscoord": coord, "ggsradius": str(radius_m),
+                          "ggslimit": str(limit), "ggsnamespace": "6",
+                          "formatversion": "2"})
+        return [p["title"] for p in data.get("query", {}).get("pages", [])
+                if not p.get("missing")]
+    def category(self, category: str, limit: int = 50) -> list[str]:
+        data = self._get({"action": "query", "format": "json", "generator": "categorymembers",
+                          "gcmtitle": category, "gcmtype": "file", "gcmlimit": str(limit),
+                          "formatversion": "2"})
+        return [p["title"] for p in data.get("query", {}).get("pages", [])
+                if not p.get("missing")]
+def collect(source: str, argument: str, limit: int) -> list[str]:
+    api = Commons()
+    if source == "search":
+        return api.search(argument, limit)
+    if source == "geo":
+        coord, radius = (argument.split(",") + [str(500)])[:2] if "," in argument \
+            else (argument, "500")
+        return api.geosearch(argument if "," in argument else coord, int(radius), limit)
+    if source == "category":
+        return api.category(argument if argument.startswith("Category:")
+                            else f"Category:{argument}", limit)
+    raise SystemExit(f"unknown source {source}")
+def _age_months(date: str, today: str) -> int | None:
+    if not re.match(r"^\d{4}-\d{2}", date):
+        return None
+    years = int(today[:4]) - int(date[:4])
+    months = int(today[5:7]) - int(date[5:7])
+    return years * 12 + months
+def cmd_search(args) -> int:
+    titles = collect(args.source, args.query, args.limit)
+    api = Commons()
+    rows = api.meta_for_titles(titles, allow_ccbysa=args.allow_ccbysa)
+    today = time.strftime("%Y-%m-%d")
+    result = []
+    for title in titles:
+        row = rows.get(title)
+        if not row:
+            continue
+        row["age_months"] = _age_months(row["date"], today)
+        row["recent"] = (args.months <= 0 or
+                         (row["age_months"] is not None and row["age_months"] <= args.months))
+        row["usable"] = bool(row["clean"] and row["url"]
+                             and int(row["width"] or 0) >= args.min_width)
+        row["research_only"] = not row["usable"] and bool(row["review"] and row["url"]
+                                                          and int(row["width"] or 0) >= args.min_width)
+        result.append(row)
+    result.sort(key=lambda r: (not r["recent"], r["date"]), reverse=False)
+    result.sort(key=lambda r: r["date"], reverse=True)
+    if args.json:
+        print(json.dumps(result, indent=2))
+        return 0
+    for row in result:
+        age = "?" if row["age_months"] is None else row["age_months"]
+        flags = "recent" if row["recent"] else "      "
+        flags += " commit" if row["usable"] else " review" if row["research_only"] else "  skip"
+        print(f"{row['date'] or '?':10s} {age:>4}mo {flags:12s} "
+              f"{row['licence'][:20]:20s} {row['title'][:56]}")
+    return 0
+def cmd_fetch(args) -> int:
+    rows = json.loads(Path(args.manifest).read_text(encoding="utf-8"))
+    usable = [r for r in rows if r.get("usable")
+              or (args.include_review and r.get("research_only"))]
+    if args.only_recent:
+        usable = [r for r in usable if r.get("recent")]
+    out = Path(args.dir)
+    out.mkdir(parents=True, exist_ok=True)
+    api = Commons()
+    manifest = []
+    for row in usable[:args.max]:
+        name = re.sub(r"^File:", "", row["title"])
+        name = re.sub(r"[^A-Za-z0-9._-]+", "_", name)
+        path = out / name
+        if path.exists() and not args.force:
+            print(f"skip existing {path.name}")
+        else:
+            data = api._get({"url": row["url"]}, binary=True,
+                            spacing=DOWNLOAD_SPACING_S)
+            path.write_bytes(data)
+            print(f"fetched {path.name} {len(data) // 1024} KB")
+        manifest.append({**row, "file": path.name})
+    (out / "meta.json").write_text(json.dumps(manifest, indent=2) + "\n")
+    print(f"meta -> {out / 'meta.json'} ({len(manifest)} entries)")
+    return 0
+def _markdown_filename(title: str, date: str = "") -> str:
+    stem = re.sub(r"^File:", "", title)
+    stem = re.sub(r"\.[A-Za-z0-9]+$", "", stem)
+    slug = re.sub(r"[^A-Za-z0-9]+", "-", stem).strip("-").lower()
+    return f"{slug}-{date or 'unknown'}.jpg"
+def cmd_rows(args) -> int:
+    rows = json.loads(Path(args.manifest).read_text(encoding="utf-8"))
+    sidecar = []
+    readme = []
+    for row in rows:
+        name = _markdown_filename(row["title"], row["date"] or "unknown")
+        licence = row["clean"] or row["licence"]
+        sidecar.append(
+            f"| `{name}` | {_strip_html(row['description'])[:80] or row['title']} | "
+            f"view | <{row['page']}> | {row['artist'][:40]} | {licence} | "
+            f"{row['date'] or 'n.d.'} | purpose | no |")
+        readme.append(
+            f"| ![{row['title']}]({name}) | **{row['title'][:40]}** |")
+    print("== SIDECAR rows (paste into the Photographs table)\n")
+    print("\n".join(sidecar))
+    print("\n== README rows (paste into the board gallery)\n")
+    print("\n".join(readme))
+    if args.append_sidecar:
+        target = Path(args.append_sidecar)
+        if not target.is_file():
+            return 1
+        text = target.read_text(encoding="utf-8")
+        marker = "| `"
+        additions = []
+        existing = {m.group(1) for m in re.finditer(r"^\| `([^`]+)`", text, re.M)}
+        for row, line in zip(rows, sidecar):
+            name = line.split("|")[1].strip().strip("`")
+            if name not in existing:
+                additions.append(line)
+        if additions:
+            last = [l for l in text.splitlines() if l.startswith(marker)][-1]
+            text = text.replace(last, last + "\n" + "\n".join(additions))
+            target.write_text(text)
+            print(f"appended {len(additions)} row(s) to {target}")
+    return 0
+def build_parser():
+    parser = argparse.ArgumentParser(description=__doc__,
+                                     formatter_class=argparse.RawDescriptionHelpFormatter)
+    sub = parser.add_subparsers(dest="cmd", required=True)
+    search = sub.add_parser("search", help="list Commons candidates with licence/date")
+    search.add_argument("query", help="search text, coord 'lat,lon' with --source geo, "
+                                      "or category with --source category")
+    search.add_argument("--source", choices=("search", "geo", "category"), default="search")
+    search.add_argument("--limit", type=int, default=20)
+    search.add_argument("--months", type=int, default=0,
+                        help="flag files captured within this many months (0 = all)")
+    search.add_argument("--min-width", type=int, default=1200)
+    search.add_argument("--allow-ccbysa", action="store_true",
+                        help="treat CC BY-SA as committable after a recorded review "
+                             "(default: research-only)")
+    search.add_argument("--json", action="store_true")
+    search.set_defaults(handler=cmd_search)
+    fetch = sub.add_parser("fetch", help="download the usable rows of a manifest")
+    fetch.add_argument("--from", dest="manifest", required=True)
+    fetch.add_argument("--dir", required=True)
+    fetch.add_argument("--max", type=int, default=12)
+    fetch.add_argument("--only-recent", action="store_true")
+    fetch.add_argument("--force", action="store_true")
+    fetch.add_argument("--include-review", action="store_true",
+                       help="also download review-only (CC BY-SA) rows for study under tmp/")
+    fetch.set_defaults(handler=cmd_fetch)
+    rows = sub.add_parser("rows", help="emit SIDECAR/README table rows from a manifest")
+    rows.add_argument("--from", dest="manifest", required=True)
+    rows.add_argument("--append-sidecar", default="")
+    rows.set_defaults(handler=cmd_rows)
+    return parser
+def main(argv=None) -> int:
+    args = build_parser().parse_args(argv)
+    return args.handler(args)
+if __name__ == "__main__":
+    sys.exit(main())