The_Great_Archive/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2026-09-12 19:50:57 +05:00
Code Issues Actions 4 Packages Projects Releases Wiki Activity

Keep static manifest schema consistent

Browse Source
This commit is contained in:
Nick Sweeting 2026-08-30 16:11:48 -07:00
parent e874ee4201
commit c498fc174c
No known key found for this signature in database
2 changed files with 15 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
archivebox/core/models.py
View File

@ -459,6 +459,7 @@ class SnapshotQuerySet(models.QuerySet):
template = "static_index.html" if with_headers else "minimal_index.html"
snapshot_list = list(self.iterator(chunk_size=500))
manifest_records = []
for snapshot in snapshot_list:
outputs = snapshot.discover_outputs(include_filesystem_fallback=True)
output_paths = [str(output.get("path") or "") for output in outputs]
@ -467,9 +468,14 @@ class SnapshotQuerySet(models.QuerySet):
]
snapshot._public_favicon_paths = [path for path in output_paths if path in ("favicon/favicon.ico", "favicon.ico")]
snapshot.write_html_details()
if with_headers:
# Use the same portable schema as the JSON export. Rendering
# above has already populated result-count caches, archive_size
# reuses the sealed output_size field, and tags are prefetched.
manifest_records.append(snapshot.to_dict(extended=True, static_export=True))
if with_headers:
manifest = "".join(f"{json.dumps(snapshot.to_json(), ensure_ascii=False, sort_keys=True)}\n" for snapshot in snapshot_list)
manifest = "".join(f"{to_json(record, indent=None, sort_keys=True)}\n" for record in manifest_records)
atomic_write(str(CONSTANTS.DATA_DIR / CONSTANTS.JSONL_INDEX_FILENAME), manifest)
return render_to_string(

9
archivebox/tests/test_cli_list.py
View File

@ -40,8 +40,15 @@ def test_static_export_creates_detail_page_for_unarchived_snapshot(snapshot):
assert 'href="./index.json"' not in html
root_manifest = CONSTANTS.DATA_DIR / "index.jsonl"
assert root_manifest.exists()
manifest_records = parse_jsonl_output(root_manifest.read_text())
manifest_records = [json.loads(line) for line in root_manifest.read_text().splitlines() if line.strip()]
assert [record["id"] for record in manifest_records] == [str(snapshot.id)]
# JSON and JSONL are alternate containers for one static-export schema;
# consumers must not see TYPE/tags/archive paths change by file format.
assert manifest_records[0]["TYPE"] == "core.models.Snapshot"
assert "type" not in manifest_records[0]
assert isinstance(manifest_records[0]["tags"], list)
assert manifest_records[0]["archive_path"] == static_path
assert manifest_records[0]["archive_url"] == f"./{static_path}/index.html"
assert detail_path.exists()
detail_html = detail_path.read_text()
assert f"/snapshot/{snapshot.id.hex}" not in detail_html