-
Notifications
You must be signed in to change notification settings - Fork 22
Continous integration and deployment
There is a script that can be used to generate the TTL files and, if needed, publish them to the code-registry. The actual processing logic is contained in a module which is located in the scripts/ folder.
The script can be used to only generate the TTL files, or to generate the TTL files and subsequently publish them to the codes-registry. For this to happen, one needs to have access the the codes-registry and generate a temporary token, which needs to be passed to the publish function in the script. In order to know which tables to process, the script uses the file wmdr-tables.csv and then opens the files contained in tables_en to obtain a list of items. The script is meant to operate on the current version of the repository present in the local file system. Thus, to generate or publish an older release, one simply checks out the release in question and runs the code (provided the code, which is part of the repository, too, was already part of the code at the time).
The wmds repository is linked to the travis-ci continuous integration engine and will check for each commit if all code-tables are valid and can be turned into valid TTLs. The configuration of travis is done in the file ".travis.yml" in the repository root.
There is also another script that generates the readme.md file from the reference file wmrd-tables.csv . If you have added a new codelist you need to update wmdr-tables.csv and then run python scripts/generate-readme.py
which will update the readme.md file.