import2wikidata

Script to import Open Data Zurich fountains into wikidata without overwriting existing data. The script generates Quickstatement commands to be run in the online interface https://tools.wmflabs.org/wikidata-todo/quick_statements.php.

The project is open source under the GNU Affero General Public License, with a profit contribution agreement applying under restricted conditions. See COPYING for information.

Process diagram

Requirements

• Python 3.3 or higher (https://www.python.org/downloads/release/python-365/) make sure, it is added to your path
  • v3.65 works ; most other were very troublesome as per https://github.com/water-fountains/import2wikidata/wiki/Trouble-Shooting (v3.53 on debian9 problematic with gastrodon; on win v3.8 with asyncio and win32api DLL, v3.7 with 'GenericMeta' from 'typing')
• Jupyter (http://jupyter.org/install)
• all packages mentioned in the import lines of the Jupyter notebook zurich_fountains_preprocess.ipynb (use pip install [package name] to install - in debian rather 'pip3')
  • gastrodon , pandas

How to run

• Clone this repository to your local machine git clone https://github.com/mmmatthew/import2wikidata.git
• Start jupyter from a command window in the directory just created jupyter notebook
• In your browser, open http://localhost:8888/notebooks/zurich_fountains_preprocess.ipynb (the port may be different)
• Run the script. It should create a file quickstatement_commands*.txt

How to use the generated commands

• Go to https://tools.wmflabs.org/wikidata-todo/quick_statements.php.
• Authenticate yourself with your Wikidata account.
• Copy and paste the contents of quickstatement_commands*.txt into the blank field, and run the commands
  • before 'run', check the Umlauts (opening with notepad was fine, just doubleclick distorted them)
  • if the wmflabs is slow, do not re-do it a second time, otherwise you generate duplicates
  • you might even not see that pressing the "Do it" button had an effect. Watch twice before you hit it again (duplicates!)
  • also, it takes up to hours until items you see newly inserted in the GUI also arrive in sparql queries
    • ==> wait hours between multiple attempts to run quickstatement_commands*.txt (and redo the notebook first to detect duplicates)

Field mappings

Open Data Zurich fountain properties are mapped to the following wikidata properties

• bezeichnung -> Lde
• brunnenart_txt -> P31
• wasserart_txt -> P5623
• nummer -> P528
• historisches_baujahr -> P571

Type of water can take several values

• Verteilnetz -> Q53633635
• Quellwasser -> Q1881858
• Eigene Versorgung -> Q53634173
• Grundwasser -> Q161598

Contributions welcome

Check the project issues to find out where the current dev focus is.

See also https://github.com/water-fountains/import2wikidata/wiki

See also

https://github.com/opendatabs/csvtowikidata