pages: a site can be org/repo#branch/subdir; a hook pulls every site on that repo

pages/server.py

@@ -1,7 +1,8 @@
 """Teahub pages: serve each org's public `site` repo by hostname.
 
-SITES maps hostnames to `org/repo` on GITEA (public repos only; cloned over https, no token). A clone lives under
-/data/<org>__<repo>; `POST /_hook/<org>/<repo>` (a Gitea webhook) or the periodic loop pulls it. Static files only:
+SITES maps hostnames to `org/repo`, or `org/repo#branch/subdir`, on GITEA (public repos only; cloned over https, no
+token). A clone lives under /data/<org>__<repo>__<branch>; `POST /_hook/<org>/<repo>` (a Gitea webhook) or the
+periodic loop pulls it. Static files only:
 a path maps to the file, a directory to its index.html, anything missing to 404.html or a plain 404. No listing, no
 execution, nothing writes except git. stdlib only.
 """
@@ -26,21 +27,35 @@ TYPES = {".html": "text/html; charset=utf-8", ".css": "text/css; charset=utf-8",
 LOCKS: dict[str, threading.Lock] = {}
 
 
-def clone_dir(repo: str) -> Path:
-    return DATA / repo.replace("/", "__")
+def parse(site: str) -> tuple[str, str, str]:
+    """'org/repo#branch/sub/dir' -> (repo, branch, subdir); branch '' means the default branch."""
+    repo, _, rest = site.partition("#")
+    branch, _, sub = rest.partition("/")
+    return repo, branch, sub.strip("/")
 
 
-def sync(repo: str) -> str:
+def clone_dir(site: str) -> Path:
+    repo, branch, _ = parse(site)
+    return DATA / (repo.replace("/", "__") + (f"__{branch}" if branch else ""))
+
+
+def site_root(site: str) -> Path:
+    return clone_dir(site) / parse(site)[2]
+
+
+def sync(site: str) -> str:
     """Clone or fast-forward one site repo. Returns the short HEAD or an error line."""
-    d = clone_dir(repo)
-    lock = LOCKS.setdefault(repo, threading.Lock())
+    repo, branch, _ = parse(site)
+    d = clone_dir(site)
+    lock = LOCKS.setdefault(site, threading.Lock())
     with lock:
         try:
             if (d / ".git").exists():
                 subprocess.run(["git", "-C", str(d), "fetch", "-q", "--depth", "1", "origin"], check=True, capture_output=True, timeout=120)
-                subprocess.run(["git", "-C", str(d), "reset", "-q", "--hard", "origin/HEAD"], check=True, capture_output=True, timeout=60)
+                subprocess.run(["git", "-C", str(d), "reset", "-q", "--hard", f"origin/{branch}" if branch else "origin/HEAD"], check=True, capture_output=True, timeout=60)
             else:
-                subprocess.run(["git", "clone", "-q", "--depth", "1", f"{GITEA}/{repo}.git", str(d)], check=True, capture_output=True, timeout=180)
+                subprocess.run(["git", "clone", "-q", "--depth", "1", *(["--branch", branch] if branch else []), f"{GITEA}/{repo}.git", str(d)],
+                               check=True, capture_output=True, timeout=180)
             return subprocess.run(["git", "-C", str(d), "rev-parse", "--short", "HEAD"], capture_output=True, text=True).stdout.strip()
         except subprocess.CalledProcessError as e:
             return f"error: {e.stderr.decode(errors='replace')[-200:]}"
@@ -67,11 +82,12 @@ class Handler(http.server.BaseHTTPRequestHandler):
 
     def do_POST(self):
         parts = urlsplit(self.path).path.strip("/").split("/")
-        if len(parts) == 3 and parts[0] == "_hook" and f"{parts[1]}/{parts[2]}" in SITES.values():
+        hit = [s for s in SITES.values() if len(parts) == 3 and parts[0] == "_hook" and parse(s)[0] == f"{parts[1]}/{parts[2]}"]
+        if hit:
             self.rfile.read(int(self.headers.get("content-length") or 0))
-            out = sync(f"{parts[1]}/{parts[2]}")
-            body = out.encode()
-            self.send_response(200 if not out.startswith("error") else 502)
+            outs = [sync(s) for s in dict.fromkeys(hit)]  # every site on that repo (any branch)
+            body = "\n".join(outs).encode()
+            self.send_response(200 if not any(o.startswith("error") for o in outs) else 502)
             self.send_header("content-type", "text/plain"); self.send_header("content-length", str(len(body))); self.end_headers()
             self.wfile.write(body)
             return
@@ -83,7 +99,7 @@ class Handler(http.server.BaseHTTPRequestHandler):
     def do_GET(self, head: bool = False):
         path = urlsplit(self.path).path
         if path == "/healthz":
-            body = json.dumps({"ok": True, "sites": {h: (clone_dir(r) / ".git").exists() for h, r in SITES.items()}}).encode()
+            body = json.dumps({"ok": True, "sites": {h: site_root(r).is_dir() for h, r in SITES.items()}}).encode()
             self.send_response(200); self.send_header("content-type", "application/json"); self.send_header("content-length", str(len(body))); self.end_headers()
             if not head:
                 self.wfile.write(body)
@@ -92,7 +108,7 @@ class Handler(http.server.BaseHTTPRequestHandler):
         if not repo:
             self.send_error(404, "no site for this host")
             return
-        root = clone_dir(repo).resolve()
+        root = site_root(repo).resolve()
         rel = unquote(path).lstrip("/")
         target = (root / rel).resolve() if rel else root
         if root not in target.parents and target != root: