This will have to wait for the corresponding engine PR to go in before its tests can pass (flutter/engine#8530).

It might be best to change the naming for the sake of consistency. Stacks have a fit parameter whose value is StackFit (enum). Since this feature is for Text: maybe Text.fit and TextFit, where TextFit is either expand or tight.

I do think that the word Width should probably be in there somewhere as this specifically deals with the width of the paragraph.

@GaryQian How about TextFitWidth and Text.fitWidth?

Maybe TextWidthFit sounds a bit more natural? I'm not totally convinced on the Fit part either, as it looks like the StackFit is dealing with constraints directly whereas this is just a simple switch for which metric to use.

is this related/the same issue? #15300

Yes, #15300 seems to be the same issue. Adding to the desc.

Hixie called it "longestLine" which also seems to make sense...

OK, I propose that we call the enum TextWidthBasis (because it's what the text widget's width is based on), with values parent and longestLine, and we call the parameter textWidthBasis.

This is ready for a final review.

// The current behavior, unchanged.
// A short single line is wrapped tightly.
// Multiple lines cause the text to take the full width of its parent.
Text(
  string,
  textWidthBasis: TextWidthBasis.parent, //default
)

// For a single line, the behavior is the same! Tightly wrapped.
// For multiple lines, the width is only the width of the longest line.
Text(
  string,
  textWidthBasis: TextWidthBasis.longestLine,
)

Green is parent and pink is longestLine:

The tests are failing because they think longestLine doesn't exist in the engine, but the engine was rolled into flutter/flutter with my change here: 3cd15b5

Do I need to wait for something else?

Sometimes the luci engine buildbots haven't caught up yet, i'd check https://ci.chromium.org/p/flutter/builders/luci.flutter.prod/Linux%20Host%20Engine
and here for all the builders: https://ci.chromium.org/p/flutter/builders

Although it does look like everything is caught up.

It is also strange that everything else worked, but those specific ones failed as if the engine build was bad, many of the other tests should have failed too

@dnfield After merging in your addition of flutter doctor, the same tests are still failing. If I look at the output of the failure, I don't think I see flutter doctor being run, but maybe I'm looking for the wrong thing. Shouldn't I find something with ctrl-f "doctor" in the failing test output?

It looks like it ran from the output. Are we sure 6280d42 is the right engine hash?

Yeah that hash should be fine. The engine commit I need is flutter/engine@3493dcf, which looks to be about 5 days older than 6280d42.

I'm going to go ahead and merge this and see what happens. If anything fails then I'll immediately revert.

SGTM.