README.md

@@ -15,18 +15,20 @@ fetch it — which was measured to be the difference between one step and three.
 |---|---|
 | `with-nixpkgs.sh` | Runs a command with one nixpkgs package on PATH, pinned to the *consuming* repo's `flake.lock`. Avoids `nix shell nixpkgs#x`, which re-resolves the registry and refetches a channel tarball whenever the branch moves. |
 | `report_job_log.py` | Posts the tail of a build log as a PR comment. Exists because `actions/jobs/{id}/logs` returns 500 for every id on Gitea 1.25.2, so a red job otherwise says only that it failed. Takes `--status`; see below. |
-| `sync_blocked_label.py` | Keeps `Status/Blocked` in step with Gitea's dependency graph. Resolves the label from the repo *or the organisation*, and never touches an issue marked `Status/On Hold` or `Status/Abandoned`. |
+| `sync_blocked_label.py` | Keeps `Status/Blocked` in step with Gitea's dependency graph. Resolves the label from the repo *or the organisation*, never touches an issue marked `Status/On Hold` or `Status/Abandoned`, and leaves the label alone (rather than guessing) on a blocker whose state a repo-scoped token cannot resolve — see weblib-ci#11. |
 
 All three are standard library / plain bash only. They are *run*, not built, so
 this repo has no flake.
 
-`test_report_job_log.py` covers the reporter. It is standard library and
-offline — the forge it posts to is an `http.server` on localhost that keeps
-what it is sent, so a test reads the comment back rather than trusting an exit
-status of 0, which this script returns even when the POST failed.
+`test_report_job_log.py` covers the reporter and `test_sync_blocked_label.py`
+covers the reconciler. Both are standard library and offline — the forge each
+talks to is an `http.server` on localhost that keeps what it is sent, so a
+test reads state back rather than trusting an exit status of 0 (which the
+reporter in particular returns even when its POST failed).
 
 ```bash
-python3 test_report_job_log.py      # 21 tests, ~1.2s, no network
+python3 test_report_job_log.py       # 22 tests, ~1.2s, no network
+python3 test_sync_blocked_label.py   # 12 tests, no network
 ```
 
 There is no workflow running it: this repo has no `.gitea/workflows` at all,

sync_blocked_label.py

@@ -13,12 +13,23 @@ hand.
 **An issue with no dependencies at all is never touched.** Only issues that
 have at least one dependency are managed:
 
-* any blocker not positively closed -> ensure the label is present
-* every blocker closed              -> ensure the label is absent
+* any blocker positively open   -> ensure the label is present
+* every blocker positively closed -> ensure the label is absent
+* any blocker's state cannot be confirmed either way -> leave the label
+  exactly as it is, and say so in the log
 
 The asymmetry is deliberate. A dependency in *another* repo is not fully
 visible to the Actions token, which is scoped to one repo, so "not open" and
-"closed" are not the same statement.
+"closed" are not the same statement. Gitea does not error or omit such an
+entry -- it substitutes a zero-value stub (`id: 0`, every other field blank or
+null, including `state: ""`), so `d.get("state") != "closed"` used to treat it
+as "open forever". That kept a cross-repo `Status/Blocked` from ever being
+removed (weblib-ci#11) and, worse, made the reconciler *reassert* the label
+after a human removed it by hand (weblib-archive#30, 2026-09-09): every 15
+minutes it re-ran the same "not closed" test against the same unreadable stub
+and put the label back. Treating "cannot tell" as its own outcome -- change
+nothing, log the issue number -- fixes both without guessing an answer the
+token cannot support. See `is_resolved()`.
 
 That matters because `Status/Blocked` is also applied by hand for reasons the
 graph knows nothing about -- weblib-archive#29 and #30 are blocked on a
@@ -113,6 +124,21 @@ def label_id(forge, repo, name):
     return None
 
 
+def is_resolved(dep):
+    """False for a dependency entry the token could not actually read.
+
+    Gitea's `/dependencies` response omits nothing -- an entry the token
+    cannot see still gets an element, but one where every field is Go's
+    zero value: `id: 0`, `state: ""`, `repository: null`, etc. `id` is the
+    field that can never legitimately be 0 for a real issue, so it is the
+    one thing safe to key on -- `state` happens to be empty today, but a
+    future Gitea populating it while still stubbing the rest would silently
+    revive the bug this guards against if the check were on `state` instead.
+    Measured against a live cross-repo case, 2026-09-09: see weblib-ci#11.
+    """
+    return dep.get("id", 0) != 0
+
+
 def open_items(forge, repo):
     """Open issues *and* pull requests. They share the /issues endpoint."""
     out = []
@@ -149,19 +175,36 @@ def reconcile(forge, repo, label_name, dry_run):
         if not deps:
             # No edges, no opinion. See the docstring.
             continue
-        # Fail safe: only an *explicitly closed* dependency counts as done.
-        # A cross-repo blocker is not fully visible to the Actions token, which
-        # is scoped to one repo -- weblib-archive#30 is blocked by
-        # weblib/cfbypass#1, and the runner read that entry as not-open and
-        # took the label off while cfbypass#1 was still open. Anything whose
-        # state we cannot positively confirm keeps the issue blocked.
-        blockers = [d for d in deps if d.get("state") != "closed"]
+        # Three-way split, not two. A dependency the token could not read
+        # comes back as a zero-value stub (see is_resolved) rather than being
+        # omitted, and its blank `state` must not be read as "open" or
+        # "closed" -- either guess has burned us. An *open* blocker anywhere
+        # in the list is decisive on its own (positive evidence, add/keep the
+        # label) regardless of what else is unresolvable; only "should this
+        # come off" needs every entry to be positively closed.
+        open_blockers = [d for d in deps if is_resolved(d) and d.get("state") != "closed"]
+        unresolved = [d for d in deps if not is_resolved(d)]
         labelled = any(l["name"] == label_name for l in item.get("labels") or [])
-        if blockers and not labelled:
-            action, detail = "add", f"blocked by {', '.join('#%d' % d['number'] for d in blockers)}"
+        if unresolved and not open_blockers:
+            # Cannot confirm every blocker is closed, and no *known* open one
+            # to justify keeping/adding the label on its own merits either.
+            # The reconciler's job is to keep the label in step with what it
+            # can see; when it cannot see part of the graph it has no
+            # business asserting either state. This is what replaces the old
+            # "not closed -> still blocked" fail-safe, which kept a
+            # cross-repo label stuck forever and then started *reasserting*
+            # it over a human's hand-removal (weblib-archive#30, weblib-ci#11)
+            # -- silence was the actual defect, not the direction of the
+            # guess.
+            print(f"  unknown {repo}#{number}: {len(unresolved)} blocker(s) unresolvable "
+                  f"with this token -- leaving {label_name!r} as-is ({'present' if labelled else 'absent'})",
+                  flush=True)
+            continue
+        elif open_blockers and not labelled:
+            action, detail = "add", f"blocked by {', '.join('#%d' % d['number'] for d in open_blockers)}"
             if not dry_run:
                 forge.post(f"repos/{repo}/issues/{number}/labels", {"labels": [lid]})
-        elif not blockers and labelled:
+        elif not open_blockers and not unresolved and labelled:
             action, detail = "remove", "every blocker is closed"
             if not dry_run:
                 # PUT the whole set minus this label, rather than

test_sync_blocked_label.py

@@ -0,0 +1,300 @@
+#!/usr/bin/env python3
+"""Tests for `sync_blocked_label.py`.
+
+Standard library only, like everything else here, and offline: the Gitea it
+talks to is a `http.server` on localhost that keeps whatever state a test
+hands it, so assertions read that state back rather than trusting an exit
+code.
+
+Run them with any python3:
+
+    python3 test_sync_blocked_label.py
+
+The point of most of them is the three-way split in `reconcile()` --
+positively open, positively closed, or unresolvable -- because collapsing it
+back to two is exactly how weblib-ci#11 happened. A cross-repo dependency the
+Actions token cannot read comes back from Gitea as a zero-value stub
+(`id: 0`, `state: ""`, `repository: null`, ...) rather than being omitted or
+erroring, and the old code's `d.get("state") != "closed"` read that as "still
+open" forever -- which first meant the label could never come off, and then
+(weblib-archive#30, 2026-09-09) meant the reconciler *reasserted* the label
+every 15 minutes over a human's hand-removal.
+"""
+
+import contextlib
+import io
+import json
+import os
+import threading
+import unittest
+from http.server import BaseHTTPRequestHandler, HTTPServer
+
+import sync_blocked_label as sbl
+
+HERE = os.path.dirname(os.path.abspath(__file__))
+
+
+def resolved(number, state, repo="weblib/cfbypass"):
+    """A dependency entry the token *could* read -- the shape a same-repo
+    blocker, or a cross-repo one under a token that can see it, comes back
+    as."""
+    owner, name = repo.split("/")
+    return {
+        "id": 1000 + number, "number": number, "state": state,
+        "repository": {"owner": owner, "name": name, "full_name": repo},
+    }
+
+
+def unresolvable():
+    """The zero-value stub Gitea substitutes for an entry the token cannot
+    read. Measured live against weblib-archive#68 and #30, 2026-09-09 --
+    every field blank or zero, `state` included."""
+    return {
+        "id": 0, "url": "", "html_url": "", "number": 0, "user": None,
+        "state": "", "repository": None, "labels": None, "milestone": None,
+    }
+
+
+class FakeGitea(BaseHTTPRequestHandler):
+    """Just enough of the API to drive `reconcile()`.
+
+    Class-level state, reset per test in `setUp`:
+      - `labels`: {"repo"/"org": [label dicts]}
+      - `issues`: [issue dicts] -- the open items in the one repo under test
+      - `deps`: {issue number: [dependency entries]}
+      - `label_calls`: every POST/PUT to a labels endpoint, in order
+    """
+
+    labels_repo = []
+    labels_org = []
+    issues = []
+    deps = {}
+    label_calls = []
+
+    def _send(self, code, payload):
+        raw = json.dumps(payload).encode()
+        self.send_response(code)
+        self.send_header("Content-Type", "application/json")
+        self.send_header("Content-Length", str(len(raw)))
+        self.end_headers()
+        self.wfile.write(raw)
+
+    def do_GET(self):
+        cls = type(self)
+        if self.path.startswith("/api/v1/repos/") and self.path.endswith(
+                "/labels?limit=100"):
+            return self._send(200, cls.labels_repo)
+        if self.path.startswith("/api/v1/orgs/") and "/labels" in self.path:
+            return self._send(200, cls.labels_org)
+        if "/issues?state=open" in self.path:
+            page = int(self.path.rsplit("page=", 1)[1])
+            return self._send(200, cls.issues if page == 1 else [])
+        if "/dependencies" in self.path:
+            number = int(self.path.rsplit("/issues/", 1)[1].split("/")[0])
+            return self._send(200, cls.deps.get(number, []))
+        self._send(404, {"message": "no route: " + self.path})
+
+    def _record_and_ack(self, method):
+        length = int(self.headers.get("Content-Length") or 0)
+        payload = json.loads(self.rfile.read(length) or b"{}")
+        type(self).label_calls.append({
+            "method": method, "path": self.path, "labels": payload["labels"],
+        })
+        self._send(200, [])
+
+    def do_POST(self):
+        self._record_and_ack("POST")
+
+    def do_PUT(self):
+        self._record_and_ack("PUT")
+
+    def log_message(self, *a):          # keep the test output readable
+        pass
+
+
+BLOCKED_LABEL = {"id": 145, "name": "Status/Blocked"}
+ON_HOLD_LABEL = {"id": 152, "name": "Status/On Hold"}
+
+
+def issue(number, *label_dicts):
+    return {"number": number, "labels": list(label_dicts)}
+
+
+class ReconcileTestCase(unittest.TestCase):
+
+    def setUp(self):
+        FakeGitea.labels_repo = []
+        FakeGitea.labels_org = [BLOCKED_LABEL]
+        FakeGitea.issues = []
+        FakeGitea.deps = {}
+        FakeGitea.label_calls = []
+        self.server = HTTPServer(("127.0.0.1", 0), FakeGitea)
+        self.host = "http://127.0.0.1:%d" % self.server.server_address[1]
+        self.thread = threading.Thread(
+            target=self.server.serve_forever, kwargs={"poll_interval": 0.01},
+            daemon=True)
+        self.thread.start()
+        self.addCleanup(self.server.server_close)
+        self.addCleanup(self.server.shutdown)
+        self.forge = sbl.Forge(self.host, "t0ken")
+
+    def reconcile(self, dry_run=False):
+        out = io.StringIO()
+        with contextlib.redirect_stdout(out):
+            changed = sbl.reconcile(self.forge, "weblib/weblib-archive",
+                                    "Status/Blocked", dry_run)
+        return changed, out.getvalue()
+
+    def calls_for(self, number):
+        needle = f"/issues/{number}/labels"
+        return [c for c in FakeGitea.label_calls if c["path"].endswith(needle)]
+
+
+class TestNoDependencyEdges(ReconcileTestCase):
+
+    def test_untouched(self):
+        """An issue with no dependency edges is never touched -- it may be
+        blocked by hand for a reason the graph does not know about
+        (weblib-archive#29/#30's original 'blocked on a decision' case)."""
+        FakeGitea.issues = [issue(1, BLOCKED_LABEL)]
+        FakeGitea.deps = {1: []}
+        changed, _ = self.reconcile()
+        self.assertEqual(changed, [])
+        self.assertEqual(self.calls_for(1), [])
+
+
+class TestSameRepoUnchanged(ReconcileTestCase):
+    """Pins the pre-existing, already-correct behaviour so a fix aimed at the
+    cross-repo case cannot quietly change the path that already worked --
+    same-repo dependencies always resolve fully (no id:0 stub ever shows up
+    for them), so `unresolved` is always empty here regardless of the fix."""
+
+    def test_open_blocker_adds_the_label(self):
+        FakeGitea.issues = [issue(5)]
+        FakeGitea.deps = {5: [resolved(1, "open", "weblib/weblib-archive")]}
+        changed, log = self.reconcile()
+        self.assertEqual(changed, [(5, "add", "blocked by #1")])
+        self.assertEqual(self.calls_for(5),
+                         [{"method": "POST", "path": "/api/v1/repos/weblib/weblib-archive/issues/5/labels",
+                           "labels": [145]}])
+        self.assertIn("add    Status/Blocked on weblib/weblib-archive#5", log)
+
+    def test_closed_blocker_removes_the_label(self):
+        FakeGitea.issues = [issue(6, BLOCKED_LABEL)]
+        FakeGitea.deps = {6: [resolved(2, "closed", "weblib/weblib-archive")]}
+        changed, _ = self.reconcile()
+        self.assertEqual(changed, [(6, "remove", "every blocker is closed")])
+        self.assertEqual(self.calls_for(6),
+                         [{"method": "PUT", "path": "/api/v1/repos/weblib/weblib-archive/issues/6/labels",
+                           "labels": []}])
+
+    def test_already_correct_is_a_no_op_either_direction(self):
+        FakeGitea.issues = [
+            issue(7, BLOCKED_LABEL),  # open blocker, already labelled
+            issue(8),                 # closed blocker, already unlabelled
+        ]
+        FakeGitea.deps = {
+            7: [resolved(1, "open", "weblib/weblib-archive")],
+            8: [resolved(1, "closed", "weblib/weblib-archive")],
+        }
+        changed, _ = self.reconcile()
+        self.assertEqual(changed, [])
+        self.assertEqual(FakeGitea.label_calls, [])
+
+
+class TestUnresolvableCrossRepoBlocker(ReconcileTestCase):
+    """The fix. Reproduces weblib-archive#68/#30 exactly: a dependency entry
+    that comes back as Gitea's zero-value stub because the Actions token
+    cannot read the other repo."""
+
+    def test_never_removes_a_label_it_cannot_confirm_is_clear(self):
+        """The original bug: every blocker closed, cross-repo, label stuck
+        forever."""
+        FakeGitea.issues = [issue(30, BLOCKED_LABEL)]
+        FakeGitea.deps = {30: [unresolvable(), unresolvable()]}
+        changed, log = self.reconcile()
+        self.assertEqual(changed, [])
+        self.assertEqual(self.calls_for(30), [])
+        self.assertIn("unknown weblib/weblib-archive#30", log)
+        self.assertIn("2 blocker(s) unresolvable", log)
+        self.assertIn("present", log)
+
+    def test_never_reasserts_a_label_a_human_removed(self):
+        """The sharper bug found live on 2026-09-09: same dependency graph
+        as above, but a human had already taken the label off. The old code
+        put it straight back on the next run; this must leave it off."""
+        FakeGitea.issues = [issue(30)]                    # label removed by hand
+        FakeGitea.deps = {30: [unresolvable(), unresolvable()]}
+        changed, log = self.reconcile()
+        self.assertEqual(changed, [])
+        self.assertEqual(self.calls_for(30), [])
+        self.assertIn("absent", log)
+
+    def test_a_known_open_blocker_still_wins_over_an_unresolvable_one(self):
+        """Positive evidence of an open blocker is decisive on its own --
+        this is not a reason to *loosen* the guard, only to stop guessing
+        where there is no evidence either way."""
+        FakeGitea.issues = [issue(9)]
+        FakeGitea.deps = {9: [resolved(1, "open", "weblib/weblib-archive"),
+                              unresolvable()]}
+        changed, _ = self.reconcile()
+        self.assertEqual(changed, [(9, "add", "blocked by #1")])
+
+    def test_an_unresolvable_blocker_blocks_removal_even_with_other_closed(self):
+        """One confirmed-closed same-repo blocker plus one unresolvable
+        cross-repo one must not remove the label -- the unresolvable one
+        might still be open."""
+        FakeGitea.issues = [issue(10, BLOCKED_LABEL)]
+        FakeGitea.deps = {10: [resolved(1, "closed", "weblib/weblib-archive"),
+                               unresolvable()]}
+        changed, log = self.reconcile()
+        self.assertEqual(changed, [])
+        self.assertEqual(self.calls_for(10), [])
+        self.assertIn("unknown weblib/weblib-archive#10", log)
+
+    def test_is_resolved_keys_on_id_not_on_state(self):
+        """bit, 2026-09-09: key on `id == 0`, not on the empty `state` --
+        `state` is a symptom that happens to be visible today; a future
+        Gitea could populate it while still stubbing the rest of the
+        entry."""
+        self.assertFalse(sbl.is_resolved(unresolvable()))
+        self.assertFalse(sbl.is_resolved({"id": 0, "state": "closed"}))
+        self.assertTrue(sbl.is_resolved({"id": 7, "state": ""}))
+
+
+class TestHandsOff(ReconcileTestCase):
+
+    def test_on_hold_is_never_touched_even_with_open_deps(self):
+        FakeGitea.issues = [issue(11, ON_HOLD_LABEL)]
+        FakeGitea.deps = {11: [resolved(1, "open", "weblib/weblib-archive")]}
+        changed, log = self.reconcile()
+        self.assertEqual(changed, [])
+        self.assertEqual(self.calls_for(11), [])
+        self.assertIn("hands off weblib/weblib-archive#11", log)
+
+
+class TestNoLabelInRepo(ReconcileTestCase):
+
+    def test_skips_everything_and_calls_nothing(self):
+        FakeGitea.labels_repo = []
+        FakeGitea.labels_org = []
+        FakeGitea.issues = [issue(1)]
+        changed, log = self.reconcile()
+        self.assertEqual(changed, [])
+        self.assertEqual(FakeGitea.label_calls, [])
+        self.assertIn("skipped: no 'Status/Blocked' label in this repo", log)
+
+
+class TestDryRun(ReconcileTestCase):
+
+    def test_dry_run_reports_but_does_not_call(self):
+        FakeGitea.issues = [issue(5)]
+        FakeGitea.deps = {5: [resolved(1, "open", "weblib/weblib-archive")]}
+        changed, log = self.reconcile(dry_run=True)
+        self.assertEqual(changed, [(5, "add", "blocked by #1")])
+        self.assertEqual(FakeGitea.label_calls, [])
+        self.assertIn("would add", log)
+
+
+if __name__ == "__main__":
+    unittest.main(verbosity=2)