Support full issues resync from cutoff date

Author: sterliakov

.github/workflows/download_issues.yaml

@@ -2,6 +2,11 @@ name: Sync issues
 
 on:
   workflow_dispatch:
+    inputs:
+      since:
+        description: Cutoff date
+        required: false
+        default: detect
   schedule:
     - cron: 20 */4 * * *
 
@@ -27,6 +32,7 @@ jobs:
 
       - name: Fetch issues
         run: |
-          uv run fetch
+          uv run fetch --since "$CHECK_SINCE"
         env:
           GH_ACCESS_TOKEN: ${{ secrets.CUSTOM_GITHUB_PAT }}
+          CHECK_SINCE: ${{ inputs.since || 'detect' }}

src/mypy_issues/__init__.py

@@ -5,6 +5,7 @@
 import logging
 import os
 import sys
+from datetime import datetime
 
 from .apply import (
     LOG as APPLY_LOGGER,
@@ -27,7 +28,7 @@ def fetch_issues() -> None:
     token = _get_gh_token()
 
     ISSUES_LOGGER.setLevel(logging.DEBUG if args.verbose else logging.INFO)
-    asyncio.run(download_snippets(token, limit=args.limit))
+    asyncio.run(download_snippets(token, limit=args.limit, since=args.since))
 
 
 def run_mypy() -> None:
@@ -69,6 +70,14 @@ def run_diff() -> None:
 
 
 def _make_fetch_parser() -> argparse.ArgumentParser:
+    def parse_datetime_or_detect(val: str) -> datetime | None:
+        if val == "detect":
+            return None
+        try:
+            return datetime.fromisoformat(val)
+        except ValueError:
+            parser.error(f"Invalid datetime value: {val!r}")
+
     parser = argparse.ArgumentParser()
     parser.add_argument(
         "-v", "--verbose", action="store_true", dest="verbose", help="Print more output"
@@ -79,6 +88,12 @@ def _make_fetch_parser() -> argparse.ArgumentParser:
         default=None,
         help="Max amount of valid snippets to generate",
     )
+    parser.add_argument(
+        "--since",
+        type=parse_datetime_or_detect,
+        default=None,
+        help="Force synchronization after this cutoff date, YYYY-MM-DD, or 'detect' to only update since last fetch.",
+    )
     return parser
 
 

src/mypy_issues/issues.py

@@ -44,12 +44,13 @@
 T_co = TypeVar("T_co", covariant=True)
 
 
-async def download_snippets(
+async def download_snippets(  # noqa: C901
     token: str,
     *,
     limit: int | None = None,
     org: str = "python",
     repo: str = "mypy",
+    since: datetime | None = None,
 ) -> None:
     SNIPPETS_ROOT.mkdir(exist_ok=True)
 
@@ -59,15 +60,15 @@ async def download_snippets(
     inventory: list[InventoryItem] = []
     issues: dict[int, IssueWithComments] = {}
     seen: set[str] = set()
-    since = None
     removed_count = 0
     if INVENTORY_FILE.is_file() and ISSUES_FILE.is_file():
         inventory = load_inventory()
         issues = load_issues()
-        since = max(
-            (iss.issue.created_at for iss in issues.values()),
-            default=datetime.fromtimestamp(0, tz=UTC),
-        )
+        if since is None:
+            since = max(
+                (iss.issue.created_at for iss in issues.values()),
+                default=datetime.fromtimestamp(0, tz=UTC),
+            )
         removed = {iss.number async for iss in _get_closed_issues(gh, org, repo, since)}
         removed_count = sum(n in issues for n in removed)
         comments = await _get_comments(gh, org, repo, since)