Benjamin Mako Hill
015d4f9164
Nate has given up on getting wikiq to output parquet directly because it
makes too many unpredictable large memory allocations. The workflow now
outputs JSONL in a single pass and then uses spark (wikiq_spark) to index
it as parquet in a second pass.
Remove the parquet output path along with the machinery that existed only
to support it: checkpoint files, resume temp-file merging, namespace
partitioning, and file rotation (--partition-namespaces,
--max-revisions-per-file). Resume support remains for JSONL output, where
the resume point is derived from the last complete line of the output
file. Also remove the parquet tests and baseline files.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
(cherry picked from commit 72a8851373)
When you install this from git, you will need to first clone the repository:
git clone git://projects.mako.cc/mediawiki_dump_tools
From within the repository working directory, initiatlize and set up the submodule like:
git submodule init
git submodule update
Wikimedia dumps are usually in a compressed format such as 7z (most common), gz, or bz2. Wikiq uses your computer's compression software to read these files. Therefore wikiq depends on [7za]{.title-ref}, [gzcat]{.title-ref}, and [zcat]{.title-ref}.
Dependencies
These non-Python dependencies must be installed on your system for wikiq and its associated tests to work.
- 7zip
- ffmpeg
A new diff engine based on [_wikidiff2]{.title-ref} can be used for word-persistence. Wikiq can also output the diffs between each page revision. This requires installing Wikidiff 2 on your system. On Debian or Ubuntu Linux this can be done via.
apt-get install php-wikidiff2
You may have to also run. sudo phpenmod wikidiff2.
Tests ----To run tests:
python -m unittest test.Wikiq_Unit_Test
TODO:
- [] Output metadata about the run. What parameters were used? What versions of deltas?
- [] Url encoding by default