Other `Builder` Flutter wrappers

[FMorschel]

We have this issue opened on the Flutter-IntelliJ tracker:

• Offer more Builders when wrapping widgets flutter/flutter-intellij#4540

Now that #60075 (comment) was closed and we have a single API for creating the Builder-wrappers, what does the Dev-Exp team say about also adding AnimatedBuilder and LayoutBuilder?

If you dislike the idea, both this and that issue can be closed.

CC @bwilkerson

[bwilkerson]

I don't use Flutter on a daily basis so I don't remember, but how many of the 13 existing wrap-with assists show up for an "average" widget? I ask because I have concerns about having the list get too long.

@kenzieschmoll In case you have thoughts about this as well.

[FMorschel]

how many of the 13 existing wrap-with assists show up for an "average" widget?

All of them. They each may have a specific case to not show up. E.g:

• Expanded will not show when the immediate parent (not inside a closure) is not a Flex (Column/Row).
• After the above-mentioned issue was closed, all Builder-like wrappers will show up when the wrapped widget is not a plain Builder.
• Some other rules for each.

---

Edit: Take a look at this comment #56265 (comment) from @DanTup too. On VS Code at least, we can write down the name of the widget to find it faster.

[bwilkerson]

The ability to jump to an item is interesting, but not very discoverable. I have to wonder how many users know about it.

Makes me wonder whether we should alphabetize these actions to make it easier to scan the list.

I also realized that the number of assists we have impacts performance, because we ask each one whether it applies to the cursor position every time the cursor moves. It should be fast to return without generating an assist, but it still potentially has some impact. (We're working now to try to quantify the impact so we know whether it's a real issue in practice.)

[kenzieschmoll]

I have concerns about having the list get too long

I also share this concern. @bwilkerson do we have analytics on how often the current wrap options are being used? I think we should decide on some maximum number of assists and prioritize support for assists that are used frequently.

That being said, I do see the value in LayoutBuilder and AnimatedBuilder. I really wish we could have drop down menu ("Wrap with Builder >") that exposed all the builders we support, but I think we are limited by the IDE.

[bwilkerson]

do we have analytics on how often the current wrap options are being used?

We do. We can discuss off-line.

I think we should decide on some maximum number of assists and prioritize support for assists that are used frequently.

I agree in general, but we have to remember that there's a lag. There are several new wrap-with assists that haven't yet made it to stable, and after they do it will still take time for users to migrate forward. Until that happens they will appear to be less frequently used because not enough users are able to use them.

... but I think we are limited by the IDE.

Yes, we are. We should discuss this off-line as well.