101 lines (79 loc) · 3.35 KB

Changelog

All notable changes to this project should be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[0.4.0] - 2019-08-22

Added

http: Rewrite specific URL patterns to DBpedia resource URIs:
- http://<locale>.dbpedia.org/page/<slug>
- https?://<locale>.wikipedia.org/wiki/<slug>
http: Allow lookup of multiple URI clusters at once with the uris parameter.
http: log in the Uvicorn logger that app startup is ready.

Changed

Install the latest RocksDB release from GitHub archive instead of git (see #6).
Update python to v3.7 and install the latest version of python packages.
Added mypy to development dependencies.

[0.3.3] - 2019-05-07

Fixed

db: lock contention (RocksIOError, see #8), fixed by reusing the admin DB connection in restore.create_backup().

[0.3.2] - 2019-05-06

Added

http: purge old data DBs on Gunicorn/Uvicorn initialization.

Changed

http: let Gunicorn preload the app (instead of per worker).

[0.3.1] - 2019-05-05

Added

loader: ability to restore from backup by registering them in the admin DB.
loader: CLI to interactively select a backup to restore.

Changed

loader: hot-swap DB when the current DB never completed loading.

Fixed

db.get_connection(): RocksDB API change.

[0.3.0] - 2019-05-01

Changed

loader: find and fetch latest snapshot via Databus SPARQL query (see #5).

[0.2.2] - 2019-03-23

Changed

http: Gunicorn performance tuning.
loader: RocksDB performance tuning.

[0.2.1] - 2018-12-06

Changed

http: move meta info to its own top-level key.
Docs: updated readme.

Fixed

loader: don't overlook decompressed chunks that are smaller than one line.
loader: only split on the secondary seperator once.

[0.2.0] - 2018-12-05

Added

Docs: deployment and maintenance notes.
parser: parse RDF literals from ntriple lines.
loader: progress bars.
http: include singleton IDs and meta info in response (see #1).

Changed

Adjust DBp Global namespace to global.dbpedia.org/id/.
loader: only load and backup if a new snapshot is available.
loader: load compressed TSV snapshot instead of ntriples part files.
loader: separate concerns with producer/consumer pattern.
http: replace ApiStar with Starlette.
Restructured Docker configuration.

Fixed

pinned python version at 3.6.
loader: close the event loop after load and backup are done.

[0.1.2] - 2018-06-24

Added

Docs: installation and usage notes.

Changed

loader: only load and backup if new parts are available.

Fixed

Docker image path.
loader: close the event loop after load and backup are done.

[0.1.1] - 2018-06-23

Changed

http: run with Gunicorn server in production.
loader: don't attempt manual compaction (prone to stalling).

Fixed

loader: admin part key encoding.

[0.1.0] - 2018-06-23

Added

URI cluster loader from Spark output.
http interface based on ApiStar.