Add multi-repo registry onboarding

This commit is contained in:
2026-05-17 23:22:55 +02:00
parent f372d5f0e4
commit 3ddf76252b
7 changed files with 430 additions and 13 deletions

4
.gitignore vendored
View File

@@ -52,6 +52,9 @@ coverage.xml
.pytest_cache/
cover/
# Railiance Fabric local runtime state
.railiance-fabric/
# Translations
*.mo
*.pot
@@ -173,4 +176,3 @@ cython_debug/
# PyPI configuration file
.pypirc

View File

@@ -67,6 +67,13 @@ Feed the running service from this checkout:
railiance-fabric registry sync --repo-slug railiance-fabric .
```
Or register and sync the known local Railiance ecosystem repos from the
onboarding manifest:
```bash
railiance-fabric registry sync-manifest registry/railiance-repos.yaml
```
Ingest a CycloneDX SBOM as queryable library inventory:
```bash
@@ -81,3 +88,6 @@ GET /repositories/{repo_slug}/snapshots
GET /repositories/{repo_slug}/snapshots/diff
GET /search?q=jsonschema
```
See `docs/registry-onboarding.md` for the multi-repo manifest and operating
loop.

View File

@@ -0,0 +1,66 @@
# Registry Onboarding
Multi-repo onboarding uses a repo-owned manifest to register ecosystem
repositories with a running Railiance Fabric registry and to push graph and
library inventory when the local checkout has the required inputs.
## Run The Railiance Manifest
Start the registry:
```bash
railiance-fabric-registry --db .railiance-fabric/registry.sqlite3 --port 8765
```
Sync the known local Railiance ecosystem repos:
```bash
railiance-fabric registry sync-manifest registry/railiance-repos.yaml
```
Use `--json` for automation and `--strict` when unavailable repos or invalid
inputs should fail the command.
## Manifest Shape
```yaml
apiVersion: railiance.fabric/v1alpha1
kind: RegistryOnboardingManifest
registry_url: http://127.0.0.1:8765
repositories:
- slug: railiance-fabric
name: railiance-fabric
path: ..
default_branch: main
state_hub_repo_id: 2c0de614-e468-4eb6-8157-470649ac8c05
declaration_paths:
- ..
sbom: bom.json
```
`path` is the local checkout used for git metadata and default graph discovery.
`declaration_paths` is optional; when omitted, the repo path is scanned for a
`fabric/` directory. Relative paths are resolved from the manifest file.
`sbom` or `sboms` may point to CycloneDX JSON/YAML files. When present, the
command ingests them as queryable library inventory after repository
registration.
## Onboarding Behavior
Each repository entry is registered first. If the checkout is unavailable or has
no Fabric declarations yet, the command leaves the repo registered and reports a
warning. This lets the registry represent known repos before every repo has
adopted local declarations.
When declarations exist, the command validates them, builds a graph snapshot,
and posts it to:
```text
POST /repositories/{repo_slug}/snapshots
```
The first Railiance manifest keeps the seed ecosystem graph in
`railiance-fabric` and registers adjacent repos as known sources. As those repos
adopt repo-local `fabric/` declarations, they can be synced without changing the
registry API.

View File

@@ -10,7 +10,7 @@ import urllib.request
from datetime import datetime, timezone
from pathlib import Path
from .loader import load_yaml
from .loader import declaration_files, load_yaml
from .graph import FabricGraph, build_graph
from .validation import validate_roots
@@ -75,6 +75,12 @@ def build_parser() -> argparse.ArgumentParser:
sync.add_argument("--commit", default=None)
sync.add_argument("--json", action="store_true", help="Print the raw snapshot response.")
sync_manifest = registry_sub.add_parser("sync-manifest", help="Register and sync repos from an onboarding manifest.")
sync_manifest.add_argument("manifest", type=Path)
sync_manifest.add_argument("--registry-url", default=None, help="Override the manifest registry_url.")
sync_manifest.add_argument("--strict", action="store_true", help="Exit non-zero when any repo cannot be synced.")
sync_manifest.add_argument("--json", action="store_true", help="Print the raw manifest sync summary.")
cyclonedx = registry_sub.add_parser("ingest-cyclonedx", help="Ingest a CycloneDX SBOM as library inventory.")
cyclonedx.add_argument("sbom", type=Path)
cyclonedx.add_argument("--registry-url", default="http://127.0.0.1:8765")
@@ -131,6 +137,8 @@ def main(argv: list[str] | None = None) -> int:
if args.command == "registry":
if args.registry_command == "sync":
return _registry_sync(args)
if args.registry_command == "sync-manifest":
return _registry_sync_manifest(args)
if args.registry_command == "ingest-cyclonedx":
return _registry_ingest_cyclonedx(args)
@@ -177,6 +185,166 @@ def _registry_sync(args: argparse.Namespace) -> int:
return 0
def _registry_sync_manifest(args: argparse.Namespace) -> int:
manifest_path = args.manifest.resolve()
manifest = load_yaml(manifest_path)
if not isinstance(manifest, dict):
print(f"ERROR {manifest_path}: manifest must be a YAML mapping", file=sys.stderr)
return 1
repositories = manifest.get("repositories")
if not isinstance(repositories, list):
print(f"ERROR {manifest_path}: manifest requires a repositories list", file=sys.stderr)
return 1
registry_url = args.registry_url or str(manifest.get("registry_url") or "http://127.0.0.1:8765")
results = [
_sync_manifest_repo(registry_url, manifest_path.parent, item)
for item in repositories
]
summary = {
"manifest": str(manifest_path),
"registry_url": registry_url,
"repositories": results,
"counts": {
"total": len(results),
"synced": sum(1 for item in results if item["status"] == "synced"),
"registered": sum(1 for item in results if item["status"] == "registered"),
"errors": sum(1 for item in results if item["status"] == "error"),
},
}
if args.json:
print(json.dumps(summary, indent=2, sort_keys=True))
else:
for item in results:
if item["status"] == "synced":
print(f"synced {item['slug']} snapshot {item['snapshot_id']} ({item['commit']})")
elif item["status"] == "registered":
print(f"registered {item['slug']} ({'; '.join(item['warnings'])})")
else:
print(f"error {item['slug']}: {item['error']}")
counts = summary["counts"]
print(
f"summary: {counts['total']} repo(s), {counts['synced']} synced, "
f"{counts['registered']} registered only, {counts['errors']} error(s)"
)
if args.strict and (summary["counts"]["errors"] or summary["counts"]["registered"]):
return 1
return 0
def _sync_manifest_repo(registry_url: str, manifest_dir: Path, item: object) -> dict[str, object]:
if not isinstance(item, dict):
return {"slug": "<invalid>", "status": "error", "error": "repository entry must be a mapping"}
slug = str(item.get("slug") or "").strip()
if not slug:
return {"slug": "<missing>", "status": "error", "error": "repository entry requires slug"}
repo_path = _manifest_optional_path(item.get("path"), manifest_dir)
result: dict[str, object] = {"slug": slug, "status": "registered", "warnings": []}
try:
repository = _registry_post_checked(
registry_url,
"/repositories",
{
"slug": slug,
"name": item.get("name") or (repo_path.name if repo_path else slug),
"remote_url": item.get("remote_url") or _git_value(repo_path, "config", "--get", "remote.origin.url"),
"default_branch": item.get("default_branch") or "main",
"state_hub_repo_id": item.get("state_hub_repo_id"),
},
)
result["repository"] = repository
except RegistryRequestError as exc:
return {"slug": slug, "status": "error", "error": str(exc), "warnings": []}
if repo_path is None:
result["warnings"].append("no repo path configured")
return result
if not repo_path.is_dir():
result["warnings"].append(f"repo path not found: {repo_path}")
return result
graph_paths = _manifest_paths(item.get("declaration_paths"), manifest_dir) or [repo_path]
if not any(declaration_files(path) for path in graph_paths):
result["warnings"].append("no Fabric declarations found")
try:
_ingest_manifest_sboms(registry_url, manifest_dir, slug, item, result)
except RegistryRequestError as exc:
return {"slug": slug, "status": "error", "error": str(exc), "warnings": result["warnings"]}
return result
report = validate_roots(graph_paths)
if report.errors:
return {
"slug": slug,
"status": "error",
"error": report.summary(),
"warnings": [diagnostic.format() for diagnostic in report.diagnostics],
}
graph = build_graph(graph_paths)
if graph.load_errors:
return {
"slug": slug,
"status": "error",
"error": "; ".join(f"{path}: {message}" for path, message in graph.load_errors),
"warnings": [],
}
try:
snapshot = _registry_post_checked(
registry_url,
f"/repositories/{slug}/snapshots",
{
"commit": item.get("commit") or _git_value(repo_path, "rev-parse", "HEAD") or "working-tree",
"generated_at": _utc_now(),
"graph": graph.to_export(),
},
)
result.update(
{
"status": "synced",
"snapshot_id": snapshot["id"],
"commit": snapshot["commit"],
}
)
_ingest_manifest_sboms(registry_url, manifest_dir, slug, item, result)
except RegistryRequestError as exc:
return {"slug": slug, "status": "error", "error": str(exc), "warnings": result["warnings"]}
return result
def _ingest_manifest_sboms(
registry_url: str,
manifest_dir: Path,
slug: str,
item: dict[str, object],
result: dict[str, object],
) -> None:
sbom_paths = _manifest_paths(item.get("sboms"), manifest_dir)
single_sbom = _manifest_optional_path(item.get("sbom"), manifest_dir)
if single_sbom:
sbom_paths.insert(0, single_sbom)
ingested: list[dict[str, object]] = []
for sbom_path in sbom_paths:
if not sbom_path.is_file():
result.setdefault("warnings", []).append(f"SBOM not found: {sbom_path}")
continue
payload = load_yaml(sbom_path)
if not isinstance(payload, dict):
result.setdefault("warnings", []).append(f"SBOM must be an object: {sbom_path}")
continue
ingested.append(
_registry_post_checked(
registry_url,
f"/repositories/{slug}/libraries/cyclonedx",
payload,
)
)
if ingested:
result["libraries"] = ingested
def _registry_ingest_cyclonedx(args: argparse.Namespace) -> int:
payload = load_yaml(args.sbom)
if not isinstance(payload, dict):
@@ -194,7 +362,19 @@ def _registry_ingest_cyclonedx(args: argparse.Namespace) -> int:
return 0
class RegistryRequestError(Exception):
pass
def _registry_post(registry_url: str, path: str, payload: dict[str, object]) -> dict[str, object]:
try:
return _registry_post_checked(registry_url, path, payload)
except RegistryRequestError as exc:
print(f"ERROR {exc}", file=sys.stderr)
raise SystemExit(1) from exc
def _registry_post_checked(registry_url: str, path: str, payload: dict[str, object]) -> dict[str, object]:
data = json.dumps({key: value for key, value in payload.items() if value is not None}).encode("utf-8")
request = urllib.request.Request(
registry_url.rstrip("/") + path,
@@ -207,17 +387,38 @@ def _registry_post(registry_url: str, path: str, payload: dict[str, object]) ->
body = json.loads(response.read())
except urllib.error.HTTPError as exc:
detail = exc.read().decode("utf-8", errors="replace")
print(f"ERROR registry request failed ({exc.code}): {detail}", file=sys.stderr)
raise SystemExit(1) from exc
raise RegistryRequestError(f"registry request failed ({exc.code}): {detail}") from exc
except urllib.error.URLError as exc:
print(f"ERROR cannot reach registry at {registry_url}: {exc}", file=sys.stderr)
raise SystemExit(1) from exc
raise RegistryRequestError(f"cannot reach registry at {registry_url}: {exc}") from exc
if not isinstance(body, dict):
print("ERROR registry returned a non-object response", file=sys.stderr)
raise SystemExit(1)
raise RegistryRequestError("registry returned a non-object response")
return body
def _manifest_optional_path(value: object, manifest_dir: Path) -> Path | None:
if not isinstance(value, str) or not value.strip():
return None
path = Path(value).expanduser()
return path if path.is_absolute() else (manifest_dir / path).resolve()
def _manifest_paths(value: object, manifest_dir: Path) -> list[Path]:
if value is None:
return []
if isinstance(value, str):
values = [value]
elif isinstance(value, list):
values = value
else:
return []
paths: list[Path] = []
for item in values:
path = _manifest_optional_path(item, manifest_dir)
if path is not None:
paths.append(path)
return paths
def _primary_repo_path(paths: list[Path]) -> Path:
if not paths:
return Path(".").resolve()
@@ -229,7 +430,9 @@ def _slugify(value: str) -> str:
return re.sub(r"-+", "-", re.sub(r"[^a-z0-9]", "-", value.lower())).strip("-") or "repo"
def _git_value(repo_path: Path, *args: str) -> str | None:
def _git_value(repo_path: Path | None, *args: str) -> str | None:
if repo_path is None:
return None
try:
result = subprocess.run(
["git", *args],

View File

@@ -0,0 +1,79 @@
apiVersion: railiance.fabric/v1alpha1
kind: RegistryOnboardingManifest
registry_url: http://127.0.0.1:8765
repositories:
- slug: railiance-fabric
name: railiance-fabric
path: ..
default_branch: main
state_hub_repo_id: 2c0de614-e468-4eb6-8157-470649ac8c05
declaration_paths:
- ..
- slug: railiance-infra
name: railiance-infra
path: /home/worsch/railiance-infra
remote_url: http://92.205.130.254:32166/coulomb/railiance-infra.git
default_branch: main
state_hub_repo_id: 485187c0-4fad-42f7-984c-5e317a66c5de
- slug: railiance-cluster
name: railiance-cluster
path: /home/worsch/railiance-cluster
remote_url: http://92.205.130.254:32166/coulomb/railiance-cluster.git
default_branch: main
state_hub_repo_id: c3fd0dd2-a0de-415c-8e3a-e37406f4b8f8
- slug: railiance-platform
name: railiance-platform
path: /home/worsch/railiance-platform
remote_url: http://92.205.130.254:32166/coulomb/railiance-platform.git
default_branch: main
state_hub_repo_id: 5115e0c5-009f-4168-b155-9943fe2ab9a7
- slug: railiance-apps
name: railiance-apps
path: /home/worsch/railiance-apps
remote_url: http://92.205.130.254:32166/coulomb/railiance-apps.git
default_branch: main
state_hub_repo_id: cb06310e-4381-428b-b8f3-d3ea8ac2f70d
- slug: railiance-enablement
name: railiance-enablement
path: /home/worsch/railiance-enablement
remote_url: http://92.205.130.254:32166/coulomb/railiance-enablement.git
default_branch: main
state_hub_repo_id: bc978be9-ca72-42bb-a451-d4737b779c5b
- slug: flex-auth
name: flex-auth
path: /home/worsch/flex-auth
default_branch: main
state_hub_repo_id: fda8ad85-a7d7-4055-8f21-902a533e59df
- slug: key-cape
name: KeyCape
path: /home/worsch/key-cape
default_branch: main
state_hub_repo_id: 8a99bb74-1ec0-4478-ac70-35a7cddb0e3c
- slug: artifact-store
name: artifact-store
path: /home/worsch/artifact-store
remote_url: gitea-remote:coulomb/artifact-store.git
default_branch: main
state_hub_repo_id: e1575e7c-e1b0-4b46-98be-78bd49e06318
- slug: repo-scoping
name: repo-scoping
path: /home/worsch/repo-scoping
remote_url: gitea-remote:coulomb/repo-scoping.git
default_branch: main
state_hub_repo_id: de749ff1-a4a4-42f5-8ed4-6b8c05b18bd9
- slug: the-custodian
name: The Custodian
path: /home/worsch/the-custodian
remote_url: http://gitea.local/worsch/the-custodian
default_branch: main
state_hub_repo_id: 56ae522c-b47e-4748-8239-f61b332fe69d

View File

@@ -231,6 +231,63 @@ def test_registry_http_service_serves_queries(tmp_path: Path) -> None:
thread.join(timeout=5)
def test_registry_sync_manifest_registers_multiple_repos(tmp_path: Path) -> None:
store = RegistryStore(tmp_path / "registry.sqlite3")
store.init_schema()
class Handler(RegistryHandler):
pass
Handler.store = store
server = ThreadingHTTPServer(("127.0.0.1", 0), Handler)
thread = threading.Thread(target=server.serve_forever, daemon=True)
thread.start()
try:
base_url = f"http://127.0.0.1:{server.server_port}"
sbom_path = tmp_path / "bom.json"
sbom_path.write_text(json.dumps(_cyclonedx_bom()), encoding="utf-8")
manifest_path = tmp_path / "manifest.yaml"
manifest_path.write_text(
"\n".join(
[
"apiVersion: railiance.fabric/v1alpha1",
"kind: RegistryOnboardingManifest",
"repositories:",
" - slug: railiance-fabric",
" name: Railiance Fabric",
f" path: {Path('.').resolve()}",
" commit: manifest-commit",
f" sbom: {sbom_path.name}",
" - slug: missing-repo",
" name: Missing Repo",
f" path: {tmp_path / 'missing-repo'}",
"",
]
),
encoding="utf-8",
)
assert cli_main(
[
"registry",
"sync-manifest",
str(manifest_path),
"--registry-url",
base_url,
]
) == 0
repositories = {repo["slug"]: repo for repo in store.list_repositories()}
assert set(repositories) == {"missing-repo", "railiance-fabric"}
assert store.latest_snapshot("railiance-fabric")["commit"] == "manifest-commit"
assert store.list_snapshots("missing-repo") == []
assert store.list_libraries(repo_slug="railiance-fabric")[0]["name"] == "jsonschema"
finally:
server.shutdown()
server.server_close()
thread.join(timeout=5)
def _post_json(url: str, payload: dict) -> dict:
request = urllib.request.Request(
url,

View File

@@ -4,7 +4,7 @@ type: workplan
title: "Multi-Repo Registry Onboarding"
domain: railiance
repo: railiance-fabric
status: active
status: completed
owner: codex
topic_slug: railiance
planning_priority: high
@@ -29,7 +29,7 @@ model.
```task
id: RAIL-FAB-WP-0006-T01
status: todo
status: done
priority: high
state_hub_task_id: "4a17b588-daa2-4b79-b17a-b5806d96a111"
```
@@ -41,7 +41,7 @@ artifact, and SBOM inputs for registry sync.
```task
id: RAIL-FAB-WP-0006-T02
status: todo
status: done
priority: high
state_hub_task_id: "d993b52a-d253-4869-9de5-a7e1f1b268ce"
```
@@ -53,7 +53,7 @@ repositories into the registry service.
```task
id: RAIL-FAB-WP-0006-T03
status: todo
status: done
priority: medium
state_hub_task_id: "e256b295-eacd-4321-b386-a08c143cdc77"
```