Feature request: filter urls that match a pattern.

[artfulrobot]

Describe the feature or enhancement

A common URL pattern might be /blog/{title}

A common need might be to know how many visitors to blog pages you've had.

Currently you can't do this because you can only filter by discrete URLs. e.g. you could filter separately on /blog/why-are-red-shoes-so-good and /blog/happiness-comes-to-those-with-nice-shoes but you'd need to add them up and that would include overlaps and anyway if you had dozens of blog entries then that would take ages.

Being able to search for urls that start with blog would pick up all these. Being able to search for blog*shoes might bring up all the blog posts about shoes.

Not sure if it's relevant, but you can do this in Plausible and of course you can do it in GA (at least pre GA4, not sure about 4). Plausible's filter says:

Hint: You can use double asterisks to match any character e.g. /blog** to group all of your blog posts. Or use double asterisks in front and back (e.g. **keyword**) to group all URLs containing a specific keyword.

[BlaineMoore]

This would be a great feature to add.

[github-actions[bot]]

This issue is stale because it has been open for 60 days with no activity.

[iamyashkhandelwal]

Hey @mikecao @artfulrobot @BlaineMoore , I think for this we should introduce content group tags using umami.track();

For example: If we have different url's like:

1. /blog/first-blog
2. /blog/second-blog
3. /blog/third-blog
4. /blog

Then the front-end team of the project should add a custom event on these pages that should look like this: umami.track('set', 'content_group', '/blog');

a. set - we want to set a custom event (content-group in this case). b. content_group - this is our event name c. /blog - this is the group name

Using this the users will have more authority over what they want to include in which content-group and content-grouping can be done at any levels.

I'm open to discussion for this feature.

[artfulrobot]

@iamyashkhandelwal I think that's a different request. URLs are the mainstay of stats reporting and are common to every page request; My request is about making those more searchable when they already represent structured data, as they often do. I feel your request is an additional/secondary way to group certain data, which is fine but is not the same as filtering the core URL property.

(~~Incidentally, my PR works to do this, although it needs updating now~~).

[artfulrobot]

Having recently upgraded, I note that we have some architecture for filters:

but it's lacking a LIKE, or "starts with" etc.

[iamyashkhandelwal]

Yes, It might be considered as a new feature, but I think its useful too. Waiting for your inputs @mikecao @BlaineMoore

[iamyashkhandelwal]

Also @artfulrobot , is the snapshot you have shared released/merged in master?

[artfulrobot]

@iamyashkhandelwal

Yes, It might be considered as a new feature, but I think its useful too.

I don't doubt the usefulness, it's just it's not about filtering existing URL data, it's about suggesting a new events based feature and deserves its own issue so it can be discussed properly.

snapshot..

the screenshot is from v 2.5 I think, from the docker image.

[ScottEAdams]

We are also very interested in this feature. For devs regex match would be perfect. Then for non devs: equals does not equal contains does not contain starts with does not start with ends with does not end with

[MariusVB]

Also very interested in this feature. Can also contribute if needed.

Could we start by adding the commit form @artfulrobot that adds * to the dev syntax to turn it into a LIKE query? https://github.com/artfulrobot/umami/commit/2257ad7b709d39d2dba1478f5c1c0f1ccd0a44d3

[artfulrobot]

Newer implementation ideas drafted at https://github.com/umami-software/umami/pull/2367

[anton-g]

Also very interested in this, and would love to contribute if needed. Is there any progress being made on this somewhere other than https://github.com/umami-software/umami/pull/2367?

[markwoon]

Would love to see this implemented as well. Being able to filter by path-based regular expressions is critical for us.

[franciscao633]

Added additional filters in v2.11.0. Its currently a full wildcard search Contains / Does not contain, but we have something to iterate on if we want to add more patterns.