refactor(manifest-extends): thread resolved repos through recursion
lint / lint (push) Successful in 1m32s
test / unit (pull_request) Successful in 28s
test / integration (pull_request) Successful in 16s

Replace the lossy _entry_to_raw round-trip with a repos_cache threaded
alongside the ManifestBottle cache in _resolve_one_bottle. Each bottle's
effective git-gate.repos is stored as raw dicts keyed by name, so a child
field-merges directly against its parent's raw repos instead of
reconstructing them from parsed ManifestGitEntry objects.

_resolve_repos_raw now owns the union/clear/inherit semantics on plain
dicts; _merge_bottles just injects the precomputed merged set before
parsing. Drops _entry_to_raw entirely, removing the maintenance hazard
where a new ManifestGitEntry field would silently vanish from inherited
repos.

Addresses review feedback on #238.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NgEFTXcWZjA8n7ntq2zHQQ
This commit is contained in:
2026-06-19 22:53:27 -04:00
parent ff7a52c1d2
commit 8f21f4df19
+54 -43
View File
@@ -5,16 +5,20 @@ from __future__ import annotations
from typing import TYPE_CHECKING from typing import TYPE_CHECKING
if TYPE_CHECKING: if TYPE_CHECKING:
from .manifest import ManifestBottle, ManifestGitEntry from .manifest import ManifestBottle
from .manifest_egress import ManifestEgressConfig from .manifest_egress import ManifestEgressConfig
def resolve_bottles(raws: dict[str, dict[str, object]]) -> dict[str, ManifestBottle]: def resolve_bottles(raws: dict[str, dict[str, object]]) -> dict[str, ManifestBottle]:
"""Apply `extends:` chains and return resolved ManifestBottle objects.""" """Apply `extends:` chains and return resolved ManifestBottle objects."""
cache: dict[str, ManifestBottle] = {} cache: dict[str, ManifestBottle] = {}
# Per-bottle effective git-gate.repos, as raw dicts keyed by repo name.
# Threaded alongside `cache` so a child can field-merge against its
# parent's repos without reconstructing them from parsed entries.
repos_cache: dict[str, dict[str, object]] = {}
for name in raws: for name in raws:
if name not in cache: if name not in cache:
_resolve_one_bottle(name, raws, cache, ()) _resolve_one_bottle(name, raws, cache, repos_cache, ())
return cache return cache
@@ -22,6 +26,7 @@ def _resolve_one_bottle(
name: str, name: str,
raws: dict[str, dict[str, object]], raws: dict[str, dict[str, object]],
cache: dict[str, ManifestBottle], cache: dict[str, ManifestBottle],
repos_cache: dict[str, dict[str, object]],
seen: tuple[str, ...], seen: tuple[str, ...],
) -> ManifestBottle: ) -> ManifestBottle:
from .manifest import ManifestBottle, ManifestError from .manifest import ManifestBottle, ManifestError
@@ -41,6 +46,7 @@ def _resolve_one_bottle(
if parent_name_raw is None: if parent_name_raw is None:
bottle = ManifestBottle.from_dict(name, child_raw) bottle = ManifestBottle.from_dict(name, child_raw)
cache[name] = bottle cache[name] = bottle
repos_cache[name] = _resolve_repos_raw({}, child_raw)
return bottle return bottle
if not isinstance(parent_name_raw, str): if not isinstance(parent_name_raw, str):
@@ -60,26 +66,33 @@ def _resolve_one_bottle(
f"bottle '{name}' extends '{parent_name}' which is not " f"bottle '{name}' extends '{parent_name}' which is not "
f"defined. Available bottles: {avail}" f"defined. Available bottles: {avail}"
) )
parent = _resolve_one_bottle(parent_name, raws, cache, seen + (name,)) parent = _resolve_one_bottle(
bottle = _merge_bottles(parent, child_raw, name) parent_name, raws, cache, repos_cache, seen + (name,)
)
merged_repos_raw = _resolve_repos_raw(repos_cache[parent_name], child_raw)
bottle = _merge_bottles(parent, child_raw, merged_repos_raw, name)
cache[name] = bottle cache[name] = bottle
repos_cache[name] = merged_repos_raw
return bottle return bottle
def _merge_bottles( def _merge_bottles(
parent: ManifestBottle, parent: ManifestBottle,
child_raw: dict[str, object], child_raw: dict[str, object],
merged_repos_raw: dict[str, object],
name: str, name: str,
) -> ManifestBottle: ) -> ManifestBottle:
"""Apply PRD 0025 merge rules.""" """Apply PRD 0025 merge rules."""
from .manifest import ManifestBottle, ManifestGitUser from .manifest import ManifestBottle, ManifestGitUser
from .manifest_egress import validate_egress_routes from .manifest_egress import validate_egress_routes
from .manifest_util import as_json_object
# git-gate.repos: union merge at the raw level so child entries can # git-gate.repos: when the child declares repos, inject the already
# omit fields present on the same-named parent entry (issue #237). # name-merged repo set (computed by _resolve_repos_raw) so the child
# After this, child.git already contains the fully merged repo set. # parses with the full inherited+overridden list (issue #237).
if _child_declares_git_gate_repos(child_raw): if _child_declares_git_gate_repos(child_raw):
child_raw = _merge_git_repos_raw(parent.git, child_raw) git_raw = as_json_object(child_raw.get("git-gate", {}), "child git-gate")
child_raw = {**child_raw, "git-gate": {**git_raw, "repos": merged_repos_raw}}
# Parse the child's declared fields into a ManifestBottle (with the # Parse the child's declared fields into a ManifestBottle (with the
# usual defaults for anything missing). Validation runs the same # usual defaults for anything missing). Validation runs the same
@@ -98,10 +111,9 @@ def _merge_bottles(
email=child.git_user.email or parent.git_user.email, email=child.git_user.email or parent.git_user.email,
) )
# git-gate.repos: raw union merge already applied above; child.git # git-gate.repos: when declared, child.git already holds the merged
# has the result. An explicit empty repos dict ({}) still clears # set (an explicit empty dict clears parent, leaving child.git empty).
# parent (the raw merge returns early for that case, so child.git # When omitted, the parent's entries are inherited verbatim.
# stays empty).
if _child_declares_git_gate_repos(child_raw): if _child_declares_git_gate_repos(child_raw):
merged_git = child.git merged_git = child.git
else: else:
@@ -137,44 +149,43 @@ def _merge_bottles(
) )
def _entry_to_raw(entry: "ManifestGitEntry") -> dict[str, object]: def _resolve_repos_raw(
"""Convert a ManifestGitEntry back to its YAML-equivalent raw dict.""" parent_repos: dict[str, object],
raw: dict[str, object] = {"url": entry.Upstream}
if entry.KnownHostKey:
raw["host_key"] = entry.KnownHostKey
key: dict[str, object] = {"provider": entry.Key.provider}
if entry.Key.provider == "static":
key["path"] = entry.Key.path
else:
key["forge_token_env"] = entry.Key.forge_token_env
if entry.Key.api_url:
key["api_url"] = entry.Key.api_url
raw["key"] = key
return raw
def _merge_git_repos_raw(
parent: "tuple[ManifestGitEntry, ...]",
child_raw: dict[str, object], child_raw: dict[str, object],
) -> dict[str, object]: ) -> dict[str, object]:
"""Union-merge parent and child git-gate repos at the raw dict level. """Compute a bottle's effective git-gate.repos as raw dicts.
For repos present in both, parent fields provide defaults and child Repos are keyed by name. When the child omits git-gate.repos it
fields win. Repos present in only one side pass through unchanged. inherits the parent's set verbatim; an explicit empty dict clears it.
An empty child repos dict is returned as-is to preserve the Otherwise parent and child unite by name, with same-name entries
"clear parent repos" semantics.""" field-merged (parent fields are defaults, child fields win)."""
from .manifest_util import as_json_object from .manifest_util import as_json_object
git_raw = as_json_object(child_raw.get("git-gate", {}), "child git-gate") if not _child_declares_git_gate_repos(child_raw):
child_repos = as_json_object(git_raw.get("repos", {}), "child git-gate.repos") return parent_repos
child_repos = _declared_repos_raw(child_raw)
if not child_repos: if not child_repos:
return child_raw return {}
parent_repos = {e.Name: _entry_to_raw(e) for e in parent} # Parent entries keep their order; child-only names are appended.
merged_repos: dict[str, object] = { names = list(parent_repos) + [n for n in child_repos if n not in parent_repos]
n: {**parent_repos.get(n, {}), **child_repos.get(n, {})} return {
for n in set(child_repos) | set(parent_repos) name: {
**as_json_object(parent_repos.get(name, {}), "parent git-gate repo"),
**as_json_object(child_repos.get(name, {}), "child git-gate repo"),
} }
return {**child_raw, "git-gate": {**git_raw, "repos": merged_repos}} for name in names
}
def _declared_repos_raw(child_raw: dict[str, object]) -> dict[str, object]:
"""Return the child's explicitly declared git-gate.repos as raw dicts,
or an empty dict when none are declared."""
from .manifest_util import as_json_object
if not _child_declares_git_gate_repos(child_raw):
return {}
git_raw = as_json_object(child_raw.get("git-gate", {}), "child git-gate")
return as_json_object(git_raw.get("repos", {}), "child git-gate.repos")
def _child_declares_git_gate_repos(child_raw: dict[str, object]) -> bool: def _child_declares_git_gate_repos(child_raw: dict[str, object]) -> bool: