This commit fixes a bug where timeseries queries - such as the main
graph - would fail with imported data in realtime mode. Realtime mode
`date` field is not actually a date, but minutes from now. This would
cause the imported tables join to fail with this error:
```
(Ch.Error Code: 53. DB::Exception: Can't infer common type for joined
columns: date: Int64 at left, s1.date: Date at right. There is no
supertype for types Int64, Date because some of them are
Date/Date32/DateTime/DateTime64 and some of them are not.
(TYPE_MISMATCH)
```
This commit removes imported data from realtime queries, as it doesn't
make sense to include it. Imported data does not have time precision,
and would only appear in the first day the data was imported anyways.
* copy relevant files from b2ace16540
* make it work and set site.funnels to empty list
* make Behaviours a functional component
* add UI for a setup hint and always display conversions by default
* cherry-pick migration commit
* update site schema with new fields
* backend: implement disable-feature action
* switch between tabs in the behaviors section
Introduces template components to build props and funnels on. Both
only show a setup notice atm, and both are behind feature flags.
* extend API for disabling props and funnels
* render feature setup note directly from the Behaviours component
* fix UI behavior when features are hidden
* update setup notices
* add conversions feature switch to Site Settings > Goals
* mix format
* remove IO.inspect
* change setup notice - use buttons + popup confirmation
* optimize for light mode
* restrict access to setup notices
* some styling improvements
* allow super-admins to enable/disable features
* only show conversions (last 30min) in realtime mode
* use shorter display names for tabs
* optimize for mobile screens
* note about sending custom events
* changelog update + fix CI
* change HTTP verb for the disable-feature action
* change UI label for show/hide goals
* Escape domain when constructing favicon URL
A domain may include a slash, and in that case the domain must be
escaped, before it is used as an attribute for the image tag.
* match with 'conn.request_path' instead + test
---------
Co-authored-by: Robert Joonas <robertjoonas16@gmail.com>
* default to source_query.include_imported before true
This fixes a bug where a Stats API aggregate query was trying
to query current period *without imported data* but previous
period *with imported data*.
* remove __internal_visits at a better time
As soon as we have made the db query, we don't need it anymore
* disallow events metric with imported data
* update changelog
* keep default in a better place
* remove unused option
* update test description
Co-authored-by: Vini Brasil <vini@hey.com>
---------
Co-authored-by: Vini Brasil <vini@hey.com>
* add entry page filter condition to bounce rate
* Fix select_merge with dynamic
* fix tests and add one test
* generalize page_filter_condition function
* use dynamic_filter_condition for session filters too
* disable views_per_visit with a page filter
* update changelog
* disable credo for complex function
---------
Co-authored-by: Uku Taht <uku.taht@gmail.com>
* Migration (PR: https://github.com/plausible/analytics/pull/2802)
* Implement Site.Domain interface allowing change and expiry
* Fixup seeds so they work with V2_MIGRATION_DONE=1
* Update Sites.Cache so it's capable of multi-keyed lookups
* Implement worker handling domain change expiration
* Implement domain change UI
* Implement transition period for public APIs
* Exclude v2 tests in primary test run
* Update lib/plausible_web/controllers/site_controller.ex
Co-authored-by: Vini Brasil <vini@hey.com>
* Update lib/plausible_web/controllers/site_controller.ex
Co-authored-by: Vini Brasil <vini@hey.com>
* Update moduledoc
* Update changelog
* Remove remnant from previous implementation attempt
* !fixup
* !fixup
* Implement domain change via Sites API
cc @ukutaht
* Update CHANGELOG
* Credo
* !fixup commit missing tests
* Allow continuous domain change within the same site
---------
Co-authored-by: Vini Brasil <vini@hey.com>
* Use user-agent instead of screen_width to get device type
Co-authored-by: eriknakata <erik.nakata5@gmail.com>
* Fix credo
* Log on unhandled UAInspector device type
* Make 'browser' the default tab in devices report
* Remove device tooltip
* Remove screen_width from ingestion completely
* Remove browserstack harness, run playwright directly
* Select meta key based on OS platform
* Run CI tests in parallel
* Improve device match readability
* Add changelog
---------
Co-authored-by: eriknakata <erik.nakata5@gmail.com>
* Reject unknown imported cities from queries
This commit fixes a bug where the city report returned `N/A` entries.
The functions that build imported data queries were using SQL
`COALESCE`, assuming city data is `NULL` when unknown, when actually its
unknown value is `0`.
This commit addresses the problem using SQL `NULLIF` combined with the
previous `COALESCE` call. With this change both `0` and `NULL` are
treated as unknown.
Since 1cb07efe6d cities can be `NULL`, but
previously we saved `0` as unknown.
Closes#1960
* Add entry to CHANGELOG
* Ignore cyclomatic complexity Credo check
* Make Device section components aware of (not set)
So that no extra sub-filters are possible when the unset
top item is selected.
* Support '(not set)' in breakdown/filters
* Update expectations for export tests
* Add extra tests for returning/filtering by '(not set)'
* Add changelog entry
* Remove ListReport conditional render
* Prevent redundant sub-filters
* Fix filter text rendering
---------
Co-authored-by: Uku Taht <uku.taht@gmail.com>
* Remove the ability to collpase the top graph
This commit removes the ability to collapse the top graph. The graph
collapsed whenever `metric` was falsy. I removed all related code to
that. Metric now defaults to visitors.
We want to add new items to top stats, and this commit will make it
easier to change it. Also, there's currently a bug where top stats is
randomly collapsing, which should be fixed by this commit.
* Refactor graph and top stats loading state
The graph loading state shows and hides the graph conditionally
depending on whether the data is loaded, loading or refreshing.
The current code is a bit difficult to read because its big
conditionals. This commit refactors the loading state making it easier
to read.
This commit also fixes a bug where the graph wasn't fading out when
changing metrics.
* fix subquery for sessions in base_event_query/2
As the 'sessions' table is using the CollapsingMergeTree engine, we have
to select session_id's distinctively. Otherwise we will get multiple rows
(with sign -1 and 1) as long as the background merge hasn't happened.
* update changelog
* use GROUP BY instead of SELECT DISTINCT
* remove comma
* extract blinkingDot function
* position pulsating-circle with tailwind instead
* remove unused function
* extract renderStatName function
* display seconds since last realtime update
Adds a 'Last updated X seconds ago' label to the Current Visitors tooltip.
* small refactor: avoid duplication of this.props and this.state
* show the 'last updated ...' tooltip in historical
* changelog update
* use className utility function
* Fix breakdown API pagination when using event metrics
This commit fixes a bug where the subsequent breakdown API pages had
the same items as the first page. The fix sorts the underlying
ClickHouse query by timestamp, keeping the same order between requests,
as we use OFFSET/LIMIT pagination.
* Fix repeated results assertion
* Add different ORDER BY to each breakdown property
* Ignore XX and T1 countries
* Add fallback if country_code=nil
* Lookup city overrides directly in CityOverrides module
* Changelog
* Add empty moduledoc
* Remove redundant comment
* add a new realtime-update-timer module
* hook to the new 'tick' event in ListReport for auto-updates
This commit fixes the bug where all reports using the `ListReport` component did not
auto-update in realtime mode. Those reports are:
- Pages (Top / Entry / Exit)
- Locations (Countries / Regions / Cities)
- Devices (Screen Sizes / Browsers + versions / OS-s + versions)
* fetch data for ListReports only when scrolled into view
* refactor fetching data in ListReport
* refer to one source of truth for utm tags
* make the 'All' tab in Sources auto-update
* make all UTM tabs in Sources auto-update
* fetch UTM data only when scrolled into view
* auto-update Referrers with the new timer
* auto-update google search terms
* auto-update Conversions
* make countries map auto-update
* auto-update visitor-graph and top stats with new timer
* use new tick event for current visitors (in Historical)
* remove the old timer class
* update changelog
* Visual improvements to automatic realtime updates (#2532)
* minor consistency fix for text color in dark mode
* use FlipMove in goal conversions report
* use FlipMove in ListReports
* set main graph and top stats loading state correctly
* refactor isIntervalValid function
* enforce intervals are valid when set and stored
* remove duplicate data fetching on interval change
Fetching new data is handled by the `fetchGraphData` callback in `updateInterval`
* refactor updateMetric function
* make it clearer why 'metric' can be a faulty value
* extract 'query' and 'site' variables from 'this.props'
* reset interval state only when period is changed
The 'maybeRollbackInterval' function was also used to fetch data. This commit replaces
all those function calls with 'fetchGraphData' which better describes the actual behavior.
We should only worry about rolling back the interval if 'query.period' has changed.
This commit also stops the graph from flickering when it is updated in realtime.
* update names of two variables
* remove unnecessary negation
* make collapsed graph state more explicit
* consider stored invalid intervals when graph mounts
* fix not showing loading spinner regression
* remove interval state from VisitorGraph (#2540)
* Realtime prop breakdown (#2535)
* disable load more in realtime mode
* extract doFetch function
* separate fetchPropBreakdown and fetchNextPage functions
* subscribe for auto-updates in realtime
* improve readability with function name changes
* Cascade delete sent_renewal_notifications table when user is deleted
This commit fixes a bug when deleting a user would trigger a constraint
error.
* Update CHANGELOG.md
* Return empty list when breaking down by event:page without events
This commit fixes a bug with pagination where breaking down by event:page
would always return results despite pagination.
Closes#2255
* Update CHANGELOG.md
* Remove show_noref behaviour
Removes query param show_noref which was used from React to control
whether to show Direct / None traffic or not. The show_noref behaviour
was untested previously.
Closes#2523
* Add changelog entry
* Fix tests
* Removed files I did not mean to check in :)
### Changes
Preparations for `v1.5.1` release.
### Tests
- [x] This PR does not require tests
### Changelog
- [x] Entry has been added to changelog
### Documentation
- [x] This change does not need a documentation update
### Dark mode
- [x] This PR does not change the UI
### Changes
Preparing for `v1.5.0` release.
### Tests
- [x] This PR does not require tests
### Changelog
- [x] Entry has been added to changelog
### Documentation
- [ ] [Docs](https://github.com/plausible/docs) have been updated
### Dark mode
- [x] This PR does not change the UI
### Changes
Hello friends, I'm trying to self-host plausible on my server and have
observed that if I add a file named (for example) `secrets/LISTEN_IP` to
configure the bind IP I'll get the following error:
```
Compiling 1 file (.ex)
** (RuntimeError) Invalid LISTEN_IP '172.19.130.83
' error: :einval
/home/serafeim/plausible/config/runtime.exs:20: (file)
```
(notice that I get the same error when configuring the DATABASE_URL or
whatever comes first)
It seems that for whatever reason the File.read! will insert a newline
at the end of the file. Notice that I have double checked the file with
vim and *it does not contain a newline at the end* or at least I don't
know *how* to make the file not have a newline at the end.
Thus I'm adding a `String.trim()` there to fix that thing; even if I am
doing something wrong and my files *do* contain a newline at the end,
adding a `String.trim()` there definite won't hurt :)
Below you'll find a checklist. For each item on the list, check one
option and delete the other.
### Tests
- [ ] Automated tests have been added
- [x] This PR does not require tests
### Changelog
- [ ] Entry has been added to changelog
- [x] This PR does not make a user-facing change
### Documentation
- [ ] [Docs](https://github.com/plausible/docs) have been updated
- [x] This change does not need a documentation update
### Dark mode
- [ ] The UI has been tested both in dark and light mode
- [x] This PR does not change the UI
Adds a new script extension that allows tracking interactions with specific HTML elements on a website. For example - to track link clicks on one specific `<a>` element, you can tag it like this:
```html
<a href=... class="plausible-event-name=<your_event_name>">
```
And you can also tag the link with custom property names and values:
```html
<a href=... class="plausible-event-name=<your_event_name> plausible-event-<your_custom_prop>=<your_value>">
```
Tagging a link as above will send a custom event with the given name and props, if a `click` or `auxclick` browser event happens, and targets the link element.
The tracking behavior is somewhat different based on the HTML element type:
- `<a>`
- triggers on `click` and `auxclick` events
- intercepts navigation based on the same rules as `outbound-links` and `file-downloads`
- `<form>`
- triggers on `submit` event
- always intercepts navigation (calls `form.submit()` after preventing default and sending the Plausible event)
- other (`<img>`, `<button>`, `<span>`, `<div>`, `<h2>`, etc ...)
- triggers on `click` and `auxclick` events
- does not prevent default to intercept possible navigation. Simply calls Plausible with the event name and props read from the element class list.
