Adds `./mach bhr-aggregate`, a thin wrapper around aggregation/bhr_collection.py so developers can run the daily BHR job with one command instead of setting up the Python path by hand. The command takes a build date and an output directory (both required), plus optional sample size, output tag, thread filter, and billing project. It validates the sample size up front so a typo like 5 cannot trigger a full read by mistake. google-cloud-bigquery is installed automatically through a new mach virtualenv site (python/sites/bhr-aggregate.txt). For GCP credentials the command checks them first and, if they are missing or expired, prints how to log in and exits. Pass --login to have it run gcloud for you instead, so it never blocks waiting on a browser in a headless setting. Note that --date is the build date directly, unlike the old python_mozetl CLI which subtracted 4 days. Differential Revision: https://phabricator.services.mozilla.com/D304160
3 lines
57 B
Plaintext
3 lines
57 B
Plaintext
requires-python:>=3.9
|
|
pypi:google-cloud-bigquery==3.41.0
|