Prerequisites

A .csv dump of the join between categorylinks and page, e.g.:

SELECT * FROM page JOIN categorylinks ON categorylinks.cl_from = page.page_id INTO OUTFILE 'E:/wikidumps/categorylinkspage-join.csv' FIELDS TERMINATED BY ',' OPTIONALLY ENCLOSED BY '"' LINES TERMINATED BY '\n'

The article dump, e.g. enwiki-20190101-pages-articles-multistream.xml

lxml:

pip install lxml

bz2file:

pip install bz2file

Usage:

Place the .csv dump and the unzipped .xml article dump in the /resources folder

Set the FILENAME_CSV (.csv dump), FILENAME_ARTICLES_XML (article dump), wantedCategory and maxDepth variables in script.py as needed.

Run script.py

python script.py

Run WikiExtractor with the output from script.py as input (choose SIZE bigger than input file to receive a single output file):

python wikiextractor\WikiExtractor.py output[FILENAME] -o [OUTPUTPATH] -b [SIZE] --json

e.g.:

python wikiextractor\WikiExtractor.py output\Computer_hardware-d10\Computer_hardware_articles-d10.xml.bz2 -o output/Computer_hardware-d10 -b 5G --json