Cache GitHub repository metadata

README.md

@@ -46,6 +46,11 @@ do that.  (There's also a more fundamental reason why this
 won't work: since each commit is a separate PR, you have to
 resolve conflicts in *each* PR, not just for the entire stack.)
 
+**What if the repository default branch changed?**  ghstack caches
+repository metadata in `.git/ghstack-repo-info.json` for the local
+checkout.  If ghstack is still using an old default branch name,
+delete that file and rerun ghstack; it will query GitHub again.
+
 **How do I start a new feature?**  Just checkout main on a new
 branch, and start working on a fresh branch.
 

src/ghstack/github_utils.py

@@ -1,5 +1,8 @@
 #!/usr/bin/env python3
 
+import json
+import logging
+import os
 import re
 from typing import Optional
 
@@ -60,6 +63,11 @@ def get_github_repo_name_with_owner(
 )
 
 
+def _repo_info_cache_path(sh: ghstack.shell.Shell) -> str:
+    git_dir = sh.abspath(sh.git("rev-parse", "--git-dir"))
+    return os.path.join(git_dir, "ghstack-repo-info.json")
+
+
 def get_github_repo_info(
     *,
     github: ghstack.github.GitHubEndpoint,
@@ -78,7 +86,20 @@ def get_github_repo_info(
     else:
         name_with_owner = {"owner": repo_owner, "name": repo_name}
 
-    # TODO: Cache this guy
+    cache_path = _repo_info_cache_path(sh)
+    try:
+        with open(cache_path) as f:
+            cached = json.load(f)
+        if (
+            cached.get("name_with_owner") == name_with_owner
+            and cached.get("id")
+            and cached.get("default_branch")
+        ):
+            logging.debug("Using cached repo info from %s", cache_path)
+            return cached
+    except (OSError, json.JSONDecodeError, KeyError):
+        pass
+
     try:
         repo = github.graphql(
             """
@@ -117,13 +138,21 @@ def get_github_repo_info(
             # Re-raise the original error if it's not the repository access issue
             raise
 
-    return {
+    result: GitHubRepoInfo = {
         "name_with_owner": name_with_owner,
         "id": repo["id"],
         "is_fork": repo["isFork"],
         "default_branch": repo["defaultBranchRef"]["name"],
     }
 
+    try:
+        with open(cache_path, "w") as f:
+            json.dump(result, f)
+    except OSError:
+        pass
+
+    return result
+
 
 RE_PR_URL = re.compile(
     r"^https://(?P<github_url>[^/]+)/(?P<owner>[^/]+)/(?P<name>[^/]+)/pull/(?P<number>[0-9]+)/?$"

src/ghstack/test_prelude.py

@@ -119,6 +119,7 @@ def __init__(self, direct: bool) -> None:
         local_dir = tempfile.mkdtemp()
         self.sh = ghstack.shell.Shell(cwd=local_dir, testing=True)
         self.sh.git("clone", upstream_dir, ".")
+        self.sh.git("fetch", "origin", "+refs/heads/*:refs/remotes/origin/*")
         self.direct = direct
 
     def cleanup(self) -> None:

test/land/default_branch_change.py.test

@@ -1,3 +1,4 @@
+import os
 from ghstack.test_prelude import *
 
 init_test()
@@ -15,6 +16,13 @@ get_github().patch(
     name="pytorch",
     default_branch="release",
 )
+# invalidate repo info cache since default branch changed
+cache_path = os.path.join(
+    get_sh().abspath(get_sh().git("rev-parse", "--git-dir")),
+    "ghstack-repo-info.json",
+)
+if os.path.exists(cache_path):
+    os.remove(cache_path)
 
 assert_github_state(
     """\
@@ -63,6 +71,8 @@ get_github().patch(
     name="pytorch",
     default_branch="main",
 )
+if os.path.exists(cache_path):
+    os.remove(cache_path)
 
 assert_github_state(
     """\