Files
endoflife-date-release-data/src/unity.py
Marc Wrobel 312ce078bb Improve HTTP handling in scripts (#449)
Provide some helper methods to hide the complexity of parsing HTML, JSON, YAML, XML or Markdown.
2025-06-28 12:26:10 +02:00

27 lines
1.4 KiB
Python

from common import dates, endoflife, http, releasedata
"""Fetches the Unity LTS releases from the Unity website. Non-LTS releases are not listed there, so this automation
is only partial.
This script is cumulative, only the first page is fetched (e.g. the first ten versions). This is because:
- it is too long to fetch all (at least 30s, usually more than a minute),
- this generates too many requests to the unity.com servers,
- fetching multiple pages in parallel is raising a lot of errors and makes the overall process slower (this was tested
during https://github.com/endoflife-date/release-data/pull/194),
- and anyway oldest versions are never updated.
Note that it was assumed that:
- the script is ran regularly enough to keep the versions up to date (once a day or week looks enough),
- there is never more than 10 new LTS versions at a time.
The script will need to be updated if someday those conditions are not met."""
for config in endoflife.list_configs_from_argv():
with releasedata.ProductData(config.product) as product_data:
html = http.fetch_html(config.url)
for release in html.find_all('div', class_='component-releases-item__show__inner-header'):
version = release.find('h4').find('span').text
date = dates.parse_datetime(release.find('time').attrs['datetime'])
product_data.declare_version(version, date)