This PR takes two steps towards better constraint handling:
## New type: `IConstraintWithId`
Introduces a new type, `IConstraintWithId`. This is the same as an
`IConstraint`, except the constraint id property is required. The idea
is that the list of editable constraints should move towards using this
instead of just `IConstraint`. That should prevent us (on a type-level)
from seeing more of the same kind of errors we saw with the segment
constraints yesterday.
I don't want to go ahead and update all the upstream uses of this to
IConstraintWithId in this PR, so I'll look at that separately.
## API payload constraint replacer
Introduces an api payload constraint "replacer", which we can use for
[JSON.stringify's `replacer`
parameter](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/JSON/stringify#the_replacer_parameter).
The current implementation works both for strategies and for segments
and has been added to edit + create forms for both of these resources.
This has a couple benefits:
1. We can clearly state exactly how we want them to be rendered,
including property order. I've decided to go with context -> operator ->
value(s) as the main one (check the screenie), as I believe this is the
most logical reading order.
2. We can exclude value/values (whichever one doesn't work with the
operator)
3. It doesn't matter how we treat constraints internally, we can still
present the payload how we want
4. Importantly: this only affects the stringification for the
user-facing API payload, so it's very low risk. It does not affect
anything that we actually send to the api.
Here's what it can look like with ordered properties:
<img width="392" alt="image"
src="https://github.com/user-attachments/assets/f46f77c8-0b5a-4ded-b13a-bb567df60bd3"
/>
Prevents the property order from changing when constraints are set from
the editable constraint component. When we render out the API command,
we don't specify the order of properties in the objects, which means
that it can change dramatically, which can be a little jarring.
As it is right now, it first renders in one order when you open the
strategy form:
<img width="299" alt="image"
src="https://github.com/user-attachments/assets/52cf2445-d9eb-402c-b5bc-0fece5fbe822"
/>
And when you navigate to the targeting section of the strategy form, it
changes to:
<img width="299" alt="image"
src="https://github.com/user-attachments/assets/e4cb7006-dcf4-4e88-befb-ccba5b647ddd"
/>
This also applies to constraints in segments.
With this change, the order will remain the same before and after.
Additionally, there's some extra tests around constraint ids being kept
the same and being set if it doesn't exist.
## Further work
This came about as part of the issue we had with constraint editing in
segments being broken as of now. As part of that, I would like to make
some further improvements (such as making the ID required when you use a
list of editable constraints), but that will be in a follow-up. There
are some complications that might not make it a viable option, sadly.
We could also try to stabilize the property order in the API rendering
methods (which I think might be a good idea), but because there's
multiple different ones, this seems to be a faster solution.
This adds constraint ids to segment constraints used in editing
segments. Without them, there was a bug where when you went to edit the
segment, all constraints would be invisibly set to the same constraint.
Addresses and removes all leftover comments related to the flag overview
redesign flag.
There's four changes here:
1. Remove release plan milestone strategy and environment footer.
Dangling files, no references.
2. Delete old code without references in theme.ts
3. Delete legacy playground result chip. Replace all references to it
with references to the new chip. The API is the exact same and the
legacy chip was just passing through everything to the new chip anyway,
so nothing should change.
4. Remove a now-redundant way to supply a default
Deletes the legacy strategy separator and removes all references to it.
Luckily, all references to the separator were in dangling files that
could themselves be deleted directly.
This PR addresses and removes the last comment related to the
addEditStrategy flag. In doing so, I have also removed the remaining
dangling files from the new constraint accordion directory. I believe
that everything that's left in there now is currently in use.
This PR removes more constraint inputs and validators that are not in
use anymore. Additionally, the old constraint components that are still
being used by the project action filter item, have been moved to that
directory. This also goes for ResolveInput which has been simplified to
only the inputs and operators used by actions filter item.
I've done a manual side-by-side comparison of the old and newly
refactored filter item, and it appears to be working the exact same.
This PR offers a little QoL upgrade in cases where you have only a
single available filter: Instead of having to first click "filters" and
then select the only option, we'll now always show the available filter,
whether it's marked as persistent or not.
The filter still gets a 'delete filter' button that clears the filter
(which is more convenient than having to deselect every one of the
options one by one), but the filter won't disappear when you clear it.
Additionally, because the `state` of the filter item will be undefined
if it has no value, I've added a `preventAutoOpen` prop to the
underlying Filter component.
~~I don't understand why we want to auto-open the filter, but it was
added by @kwasniew in https://github.com/Unleash/unleash/pull/5611, so
it appears to be deliberate.~~ The fact that we auto-open the filter
when state is undefined makes this a little tricky. I realized during
this that the reason we do it is that we want the filter to auto-open
when you select it from the dropdown. Maybe there's a better way to do
that than useEffect, but maybe not 🤷🏼
Further, the filter handling logic is quite complex (what filters to
show, ordering, etc), so I've moved as much of that into the multifilter
component, leaving the singlefilter as simple as possible.
I'm ... not particularly proud of the code here, so I'm happy to take
any suggestions for improvements. Happy to throw it all away if you have
a better way to achieve this goal.
## Rendered
The lifecycle insights use the persistent, single filter, the
performance insights do not:
<img width="1629" alt="image"
src="https://github.com/user-attachments/assets/b8599883-97dc-428e-a98f-ad59ad1c74ab"
/>
<img width="1556" alt="image"
src="https://github.com/user-attachments/assets/eacdc4bf-bc60-4e26-a88c-8be7dc5e31be"
/>
This PR continues the cleanup after removing the addEditStrategy flag
(part 2 of ???). The primary purpose of this PR is to delete and remove
all references to the LegacyConstraintAccordion.
I've gone and updated all references to the legacy files in external
components and verified manually that they still work.
Most of the files in this PR are changing references. I've extracted two
bits into more general constants/utils:
1. Constraint IDs are a symbol. it was exported as a const from the
previous createEmptyConstraint file. I've moved it into constants.
2. formatOperatorDescription was similarly used all over the place, so
I've placed it in the shared utils directory.
In reviewing this, you can ignore any changes in the legacy constraint
accordion folder, because that's all been deleted. Instead, focus on the
changes in the other files. It's primarily just import updates, but
would be good to get a second set of eyes, anyway.
Fixes a console log about invalid dom nesting:
<img width="1256" alt="image"
src="https://github.com/user-attachments/assets/0849103c-6901-4b64-a124-00eaf8cc7dde"
/>
I've changed the offending div to a span. We set `display: flex` on it,
anyway, so it shouldn't make a difference.
I've also removed some redundant functions and braces that we don't
need.
Removes all usages of flag addEditStrategy and refactors code where
necessary.
This is only the first step of the cleanup. After this, there's still
lots of code to be removed. I've got a different PR that removes ~5k
lines of code (https://github.com/Unleash/unleash/pull/10105) that I
want to reach in pieces to make sure that everythnig works on the way
there.
Adds skeleton loading indicators for the lifecycle trend tile numbers:
- total flag count
- median stats
In doing so, I have added the `data-loading` attribute to the
PrettifyLargeNumber component (to avoid having to wrap it in a separate
element for that alone), and have added refs to the InsightsSection
component.
The loading indicators look better in dark mode than in light mode,
because they use the same background color as the text box in light
mode, so only the big number is visible. There is a task in Linear to
look into fixing this.
<img width="1548" alt="image"
src="https://github.com/user-attachments/assets/9e58d681-724e-45cb-baa1-b824dda48008"
/>
<img width="1554" alt="image"
src="https://github.com/user-attachments/assets/7738fac0-5660-464f-8d10-1bf2eacc9ca0"
/>
---------
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Switches the "flags per user" box with a "total number of flags" label
for the number in the box and an additional description available via
the help icon.
To accurately label the help text and link it to the figure, I've added
a tooltipId prop to the HelpIcon component.
<img width="551" alt="image"
src="https://github.com/user-attachments/assets/f3227e74-5976-454e-9b7d-db0d05927261"
/>
---------
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Update the title of the insights / analytics page when the menu item
changes
While the side menu item has already changed, this change also updates
the page header and title.
Also fixes an error with a prop that shouldn't have been forwarded.
Adds caching via localstorage to the flag creation form, so that if you
(accidentally) close the form before submitting it, you'll retain (most)
of the same data when you reopen it.
Specifically, we'll store:
- name
- description
- type
- tags
- impression data
We can't store the project as it is now, because it gets overridden by
whatever is in the URL. However, this is probably a good thing. It means
that if you navigate to a different project and open the feature
creation form there, it'll retain everything from the last one, but
it'll use the current project.
The stored data is cleared when you successfully create a feature, so
that you don't get dangling data.
The data is also stored in a shared cache for all projects, so that you
don't have different caches per project.
The behavior of seeding the form is hidden behind a flag (that doesn't
exist yet). We'll still read and write to the cache if the flag is off,
but we won't use it to populate the feature form, so it has no
discernible impact on the user.
## Bug detected 🐛 ... and squashed
Working on this, I came to realize that there was a bug in how the
config button and use feature form hooks interacted. We (in this case
probably me) have assumed that it's fine to use a set for any option
checking in the config buttons. Also, we're using a set to store tags in
the feature form. But objects aren't compared by value in JS, so the set
will happily accept multiple instances of the same tag. Likewise, these
tags won't show up as selected in the dropdown because when the dropdown
checks if the set `has` the value, it's using reference equality.
To get around this, I have normalized the values of the Tags set to
strings (`<type>:<value>`), which are easily comparable.
We can iterate on this later if we need to.
## `useLocalStorageState`
In doing this, I have also made a change to the useLocalStorageState
hook:
the exposed "setState" function now writes to the localstorage
immediately. This is because the useEffect method might not have time to
save the data if the component unmounts (this was the case with the flag
dialog).
However, I have kept the useEffect because it gets run on component
mount and then only when it changes. This means that we will get double
saves to localstorage, but it'll be with the same data, so it's benign.
I've tried out two other uses of the hook (event timeline props and
environment columns in the project flags table) and see no discernible
difference in behavior.
## `useFeatureForm`
I have also made a change to the useFeatureForm hook and removed a
`useEffect` that would reset the name to whatever you passed in as the
initial name if you cleared it out. This essentially meant that you
couldn't clear the name completely, because it would just refill with
the initial name.
As far as I can tell, there is no need to have this sticking around
anymore. The hook is only used in two places: the flag creation dialog
and the flag edit page. The flag edit page doesn't allow you to change
the name anyway and it was causing issues in the dialog. It's likely a
holdover from the way something worked 3 years ago. Both the dialog and
the edit screen seem to work just fine with this change.
I have also changed the function parameters from ordered parameters to
an object. There's so many of them that even you don't think it's a good
idea to use objects when you have multiple params with the same type,
it's just nigh-impossible to remember the order by now.
## Minor changes
Additionally, I came across three issues that were causing react errors,
and have fixed them.
1. we'd forgotten to interpolate a variable and just used the variable
name in a string instead
2. an html attribute that doesn't exist (`aria-role` instead of `role`)
3. Providing a disabled button inside a tooltip. I've seen this one
around for ages and it prevented tooltips from working on disabled
buttons. The solution was wrapping it in a span.
This is the first pass at the full lifecycle tiles. It adds the tile
header and current and historical median data.
I have also added large number handling to all the number instances in
the tile: in the header, the graph, and the median data. In doing so, I
exposed the algorithm we use in the PrettifyLargeNumber component.
Returning a react component isn't always a valid option (such as in the
chart). This does mean that you don't get a tooltip when you use the
function directly, but in things like the chart and the median
measurement that makes sense to me.
I've decided to return "No data" if the median days value is 0 or lower.
There's no data for historical medians yet, so I'm using the same number
for now.
<img width="1538" alt="image"
src="https://github.com/user-attachments/assets/72e6a90a-6b84-47ce-af02-59596a7ff91f"
/>
Adds aria label and description to the lifecycle trend charts
The label explains that it's a bar chart, which stage it's describing,
and
the number of flags in each category.
The description provides more information about the split between new
flags this week and older flags.
Adds lifecycle trend graphs to the insights page.
The graphs are each placed within their own boxes. The boxes do not have
any more information in them yet.
Also, because the data returned from the API is still all zeroes, I've
used mock data that matches the sketches.
Finally, the chart configuration and how it's split into a
LifecycleChart that lazy loads a LifecycleChartComponent is based on the
LineChart and LineChartComponent that we use elsewhere on the insights
page.
Light mode:
<img width="1562" alt="image"
src="https://github.com/user-attachments/assets/6dd11168-be24-42d4-aa97-a7a55651fa0e"
/>
We might want to tweak some colors in dark mode, but maybe not? 🤷🏼
