Perl-based tools for Wikipedia, Wiktionary etc.

Prerequisites

Software:

• Unix-like operating system
• Perl
• Perl modules: MediaWiki::Bot version 3.3.1 or newer, Array::Compare

You may need to force install MediaWiki::Bot, as its tests fail.

Other:

• account in MediaWiki with bot flag

Copy settings.ini.example to settings.ini and set your bot username and password.

Supported wikis

audiosetter.pl supports the following Wiktionaries:

• de
• en
• pl
• simple

Usage scenarios

Adding audio files to Wiktionary

./audio_fetcher.pl
./audiosetter.pl -w en -a
./count_audio.pl -w en

Short description:

• ./audio_fetcher.pl

  Saves information on available pronunciation files to audio/ directory.

• ./audiosetter.pl -w en -a

  Adds pronunciation files on English Wiktionary (-w en) in all available languages (-a). Will take a lot of time. You can kill audiosetter.pl at any time using Ctrl+C. It will save progress in done/ directory and resume without repeating anything when started for the next time.

• ./count_audio.pl -w en

  Prints a MediaWiki table with a summary of work done.

• ./count_audio.pl -w en > /tmp/en.txt && ./audio_errors.pl -w en >> /tmp/en.txt

  Saves a summary of added files and skipped files to /tmp/en.txt.

Running for chosen languages

./audio_fetcher -r de,fi,ru

Refresh audio files for given languages.

./audiosetter.pl -w en -l de,fi,ru

Only run for given languages instead of all.

Running again after all work is done

1. ./audio_fetcher.pl --cleanstart --cleancache

   audio_fetcher.pl caches web pages, so running it again normally won't detect any new files. Use --cleanstart and --cleancache options to fetch new audio files.

2. ./audiosetter.pl --cleanstart -w en -a

   After you run audio_fetcher.pl, run audiosetter.pl for the first time with --cleanstart option. This will reset done/ directory and the count of added files. Otherwise audiosetter.pl will consider all work done and finish without doing anything. Alternatively:

   sed -i -e '/=no_entry/ d' -e '/=no_section/ d' done/done_audio_en.txt && ./audiosetter.pl -w en -a

3. sed -i -e '/=no_pronunciation/ d' -e '/=no_section/ d' -e '/=error/ d' done/done_dewikt_de.txt && ./dewikt_audiosetter_de.pl --recache && ./audio_errors.pl -w de --send

Files

• audio_errors.pl

  prints out a list pronunciation files to be added manually, because audiosetter.pl was unable to add them automatically

• audio_fetcher.pl

  scans for pronunciation files in Wikimedia Commons and writes results in audio/ directory for later use in scripts setting pronunciation on Wiktionary

• audiosetter.pl

  reads pronunciation files from audio/ directory and sets them in Wiktionary

• commons_sort_fixer.pl

  sets category sorting of media files on Wikimedia Commons

• count_audio.pl

  counts how many audio files have been added by audiosetter.pl

• dewikt_audiosetter_de.pl

  adds German pronunciation on German Wiktionary using an improved algorithm

Development

• test/

  Test harness.

Source code, bug reports

GitHub

Copyright

All code created by Derbeth under MIT license.