This pull request adds a mutator to the url field to ensure we've normalized the URL before we save it to the database. This is a follow-up from https://github.com/DoSomething/bertly/pull/74#discussion_r435407693 that I'd like to merge in before we run the backfill (since some links will be re-normalized differently).
How should this be reviewed?
👀
Any background context you want to provide?
This uses the normalize-url package. It uses slightly different normalization rules than the url-normalize package we'd been using in Python.
In particular, we now trim trailing /s when possible & sort and URL-encode query parameters. (This should not have any impact on applications or metrics, but wanted to flag!)
What's this PR do?
This pull request adds a mutator to the
urlfield to ensure we've normalized the URL before we save it to the database. This is a follow-up from https://github.com/DoSomething/bertly/pull/74#discussion_r435407693 that I'd like to merge in before we run the backfill (since some links will be re-normalized differently).How should this be reviewed?
👀
Any background context you want to provide?
This uses the
normalize-urlpackage. It uses slightly different normalization rules than theurl-normalizepackage we'd been using in Python.In particular, we now trim trailing
/s when possible & sort and URL-encode query parameters. (This should not have any impact on applications or metrics, but wanted to flag!)Relevant tickets
References Pivotal #165507221.
Checklist