Skip to content

Latest commit

 

History

History
110 lines (81 loc) · 4.78 KB

README.md

File metadata and controls

110 lines (81 loc) · 4.78 KB
Raw

BatchUploadTools Build Status

An attempt of unifying the various tools for batch uploads under one repo.

Heavily based on the LSH repo

To install

You can install BatchUploadTools via pip using: pip install git+https://github.com/lokal-profil/BatchUploadTools.git

If it is your first time running pywikibot you will also have to set up a user-config.py file.

Running as a different user:

To run as a different user to your standard pywikibot simply place a modified user-config.py-file in the top directory.

To use a different user for a particular batchupload place the user-config.py in the subdirectory and run the script with -dir:<sub-directory>.

When creating a new batch upload

Extend make_info to create own methods for reading and processing the indata. Any method marked as abstract must be implemented locally. You can make use of the various helper functions in the other classes.

If you are making use of mappings lists on Wikimedia Commons then create a MappingList instance for each such list to manage the creation of the mapping tables, the harvest of the tables when mapped and the preservation of old mappings when new lists are needed for later uploads.

Alternatively you can make use of only the prep-uploader/uploader tools by creating your own indata file. This must then be a json file where each media file is represented by a dictionary entry with the original filename (without the file extension) as the key and the following values:

  • info: the wikitext to be used on the description page (e.g. an information template)
  • filename: the filename to be used on Commons (without the file extension)
  • cats: a list of content categories (without "Category" prefix)
  • meta_cats: a list of meta categories (without "Category" prefix)

Protocol for a batch upload

  1. Load indata to a dictionary
  2. Process the indata to generate mapping lists
  3. Load the indata and the mappings to produce a list of original filenames (of media files) and their final filenames as well as json holding the following for each file:
    • Maintenance categories
    • Content categories
    • File description
    • Output filename
    • Input filename or url to file (as key)
  4. Run the prep-uploader to rename the media files and create the text file for the associated file description page. *
  5. Run the uploader to upload it all

* This step is not needed for upload by url.

Using mapping lists

To generate new tables:

  1. collect the data you wish mapped
  2. create a MappingList instance
  3. use mappings_merger() or multi_table_mappings_merger to combine the collected data with pre-existing data. (set update=False if there is no pre-existing data).
  4. pass the result to save_as_wikitext()

To make use of the mapping list data:

  1. create a MappingList instance
  2. load the existing data using load_old_mappings()
  3. pass the result to consume_entries()

Post-upload processing

To make use of the post-upload processing tools use import batchupload.postUpload.

Usage example:

For usage examples see lokal-profil/upload_batches. In particular SMM-images.

Handling upload errors

In most cases it is worth doing a second pass over any files which trigger an error since it is either a temporary hick-up or the file was actually uploaded. Below follows a list of of common errors and what to do about them (when known).

  1. stashedfilenotfound: Could not find the file in the stash. Seems to primarilly be due to larger files. Solution: Manually upload this using Upload Wizard.
  2. stashfailed: This file contains HTML or script code that may be erroneously interpreted by a web browser. Either you really have html tags in your exif data or you have triggered T143610. Smaller files can often be uploaded unchunked (slow).
  3. stashfailed: Cannot upload this file because Internet Explorer would detect it as "$1", which is a disallowed and potentially dangerous file type No clue yet. See T147720

Structured data on Commons

Basic support for Structured data on Commons is offered by passing expect_sdc to the uploader and providing the data as either a <basename>.sdc file (where <basename> is shared with the .info text file holding the associated file description page) or under the sdc-key if the data is provided as a make_info json file.

The expected format of the data is described at pywikibot-sdc.