Fix initial and auto value for flex

[iegik]

This PR to handle the initial and auto value for flex CSS property

flex: initial → flex: 0 1 auto
flex: auto → flex: 1 1 auto

#48

[iegik]

tests: #55

[luisrudge]

Hi, thanks for the PR. Can you add a bit of context on this change? 😬

This says the default is 0

[iegik]

Here https://github.com/luisrudge/postcss-flexbugs-fixes/pull/54/files#diff-4346a0d83ae11ca744ad6aed47d5e8fcL4 is all about flex css property, not 'flex-basis`.

flex: initial - This is the default value. The item is sized according to its width and height properties. It shrinks to its minimum size to fit the container, but does not grow to absorb any extra free space in the flex container. This is equivalent to setting flex: 0 1 auto.

flex-basis: <length> | auto; /* default auto */

• https://css-tricks.com/snippets/css/a-guide-to-flexbox/#article-header-id-12

[iegik]

Ok, I leave default value to 'auto', but also leave 0 if it appears.

[iegik]

https://codepen.io/iegik/pen/OBKEML

[luisrudge]

I guess this conflicts with #22

@silvenon what do you think about this PR? 🤔

[silvenon]

Initial value is not the same as default value. The default value for flex-basis is the one being set when you're using the flex shorthand without covering flex-basis, initial value is what was there initially.

• flex: initial → flex: 0 1 auto
• flex: 1 → flex: 1 1 0%

Initial values are outside of the control of the plugin because it can work only on present rules, not absent ones, unless someone explicitly writes something like flex: initial. Maybe you could add a transformation for that and modify the readme to recommend globally setting flex: initial instead of flex-basis: 1. I don't know if this plugin already has that transformation, it might be handy.

[iegik]

@silvenon what about "Safari seems to hate '0%' and the others seems to make do with a nice value when basis is missing" ?
IE11 waits 3 props, not 2. ('flex: 1 → flex: 1 1')

Should we normalize flexBasis to 0?

[luisrudge]

@iegik are you talking about this issue? #48

If so, let's fix both scenarios but in different PRs. Please update this PR to handle only the 'initial' value (and add a test for it), and let's create another PR that fixes #48

Thanks 🎉

[luisrudge]

@iegik thanks! can you add a test for this, please? 🎉

[iegik]

breaks bug 4, 8.1.a

[iegik]

Bad practice to declarate rule in index.
These checks chould be repeated in each ./bugs/bunny file!

[axelboc]

@iegik I see you've renamed the PR to focus on flexbug #8, but the code still seems to do a lot more. May I ask where you're at? It's now a fairly simple problem, which shouldn't affect the other bugs. Let me know if you need a hand ;)

[iegik]

@axelboc No, this PR is focused to FlexBug #6

• added default values for initial and auto
• removed tests, that not affects to this bug

The problem now is, that changing default flex values in bug 6 also affects to bug 4 and 8.1.a tests.

[iegik]

I think, that this PR and #56 should be merged as one PR

[adrienharnay]

Hey, wouldn't this PR revert the fix mentioned in #45 and #46 ?