EntriesRepository

Create a new entries repository instance.

Save the given entry payload.

Get the payload of the given entry.

All recorded entry metadata, newest first.

Delete the entries recorded more than the given number of hours ago.

Delete all but the given number of newest entries for the tab. Entries recorded without a tab, such as initial page loads and requests made without the extension, are limited as one group.

Prune expired entries when the prune interval has elapsed.

Ensure the storage directory and its .gitignore exist.

Get the file path of the given entry.

Determine if the given entry id is valid.

Get the path of the index file.

Get the path of the file recording the last prune time.

Normalize an entry's __meta for storage in the index. The full meta is kept so the list endpoint may filter and render without reading every entry file; the fields the storage layer relies on are coerced to known types.

Read the entry index, rebuilding it when it is missing or corrupt.

Rebuild the index from the per-entry files.

Scan the per-entry files and build the index map straight from their __meta.

Apply the given change to the index while holding its exclusive lock.

Delete entry files and remove them from the index in a single rewrite.

Read the metadata from the given entry file.

Read the time of the last prune.

Record the time of the last prune.

Read and decode the given JSON file.

Read the given file under a shared lock.

Get the paths of the per-entry JSON files.