Maybe a longshot but is it anyway possible in ktfmt to not keep chained methods on single line (when under 100 chars)? As in always keep newlines/put method chains with line breaks. This doesn't work too well for modifiers in compose for us

I put up a proposal for 1.0.0 that includes this request there @Javier. Thanks for all your contributions 🙂 https://github.com/facebook/ktfmt/discussions/451#discussioncomment-9305741

Is there specific setup I need to do in order to get the ktfmt tests to run in intellij in order to contribute? Running any tests results in really weird errors.

I'm trying to configure my project to just check the staged files in a git pre-commit hook, but I'm struggling on the syntax for --include-only. Can somebody provide an example of how to pass the results of

git diff --name-only --cached

to the pre-commit script?

I’m having trouble understanding how BlankLineWanted, genSym(), and breakOp all work together. How do I tell if an expression like

blah(
  name = a.b.c.d.e.f.g.h
)

is going to break at any point? It seems that this is partially handled by breaking the expression into parts and then grouping, but I don’t understand the grouping. Also the doc comments are quite confusing in places. or here where I’m confused as to what the groupings actually accomplish. Why is a paired with b, but c is paired with d, with no group between b and c?

Hey, I spend some time yesterday to add an MaxWidth override to the IDEA plugin: https://github.com/facebook/ktfmt/pull/492 Waiting for the maintainers to pick it up 😊 It seems that this particular setting is a dealbreaker for many. I can imagine that the plugin settings could get more detailed. I see this one option as a first step. Any feedback is welcome!

i just upgraded to latest ktfmt and it reformats code from

throw OrmException(
                    "error binding parameter"
                )

to

throw OrmException(
                    "error binding parameter")

putting the closing bracket on the same line. is that documented somewhere?

is there going to be a new ktfmt release anytime soon, since in the latest release the kotlin style is broken?

The IntelliJ Gradle plugin is also broken on IJ 242, which became stable today

@Nivaldo H Bondança would you be interested if I contributed a "custom" style setting for the IJ plugin, to set the few parameters available? This way one could align the style to @gammax's Gradle plugin.

@christophsturm, to answer your question regarding why we allow configuration on gradle/maven plugin, that is there because those configurations are shareable as they are usually committed (and I think it would be a mistake not to) to the repository. This should prevent bikeshedding discussions about the code formatting, unless people are affecting that configuration file (

build.gradle

/

pom.xml

)

New release 0.52 is rolling out now! https://github.com/facebook/ktfmt/releases/tag/v0.52

Or not... Release failed and I don't have time to look into this now and for the next couple of weeks. I'll see if someone else from the team can take a look, but feel free to chime in in case you know what's up https://github.com/facebook/ktfmt/actions/runs/10388724255/job/28765014936

when i set manageTrailingCommas to false, they are not added, but also not removed. how can i set it to always remove them?

Hi all 🙂

Just as a followup and for visibility, have you ever experienced different outputs on Windows and Max? https://github.com/facebook/ktfmt/issues/516

Would be great to figure out what's going on. In this case Ivan got different import ordering on Windows compared to macOS

Oh yeah, thank you for mentioning it. The diff was indeed a couple of imports.

it looks like the import sorting is done as below, though not obvious why that would be different on Win/Mac. Link.

fun canonicalText(importDirective: KtImportDirective) =
        importDirective.importedFqName?.asString() +
            " " +
            importDirective.alias?.text?.replace("`", "") +
            " " +
            if (importDirective.isAllUnder) "*" else ""

    val sortedImports = importList.imports.sortedBy(::canonicalText).distinctBy(::canonicalText)

Well, it would seem odd to me that this would affect import ordering and this hasn't been noticed yet, but there is this section a few lines down:

return code.replaceRange(
        importList.startOffset,
        importList.endOffset,                   importsWithComments.joinToString(separator = "\n") { imprt -> imprt.text } + "\n")

(Sorry if the formatting is messed up there, currently on mobile) Adding

\n

instead of something like

System.lineSeparator()

seems a bit suspicious to me.

But as mentioned, it's more like something that caught my attention and might be working fine anyways, but since Windows is using CR+LF instead of just LF...

that’s even trickier - while Windows uses CRLF the given file may not - or may perhaps (sadly) have mixed line endings. In any event, it looks like the sorting is done prior build the code string that include the eol character.

Hi, any update on this https://github.com/facebook/ktfmt/issues/516 please? @gammax PLEEZ 🙏

Is ktfmt compatible with the latest 2024.3? I think it is always crashing by default but I do not see any open issue about it 👀

ktfmt 0.54 is not working with Multi-dollar string interpolation. It is flagging it as

error: Expecting an expression

. Is this a known issue? Sample project to replicate where it's configured through spotless: DynamicVariableGenerator.kt

Hi folks, not sure if anyone noticed, but ktfmt is (indirectly?) using some deprecated Unsafe APIs that will be removed from the JDK https://github.com/facebook/ktfmt/issues/533

I'm a ktfmt noob. Is this really the way fluent chains are supposed to look like? Even if the chain is longer than the max width?

We want to add a special case formatting to multiline strings with

trimIndent()

/

trimMargin()

and wanted to hear your thoughts about it. We'll use the poll results to guide our approach. Feel free to suggest alternatives as well, but do keep in mind this has to work with all the different use cases, which the represented code here for sure does not cover. Option 1

val a =
    """
        a multiline
          string
        """
        .trimIndent()
        .nextMethod()

val b =
    """
        |a multiline
        |  string
        |"""
        .trimMargin()
        .nextMethod()

Option 2

val a =
    """
    a multiline
      string
    """
        .trimIndent()
        .nextMethod()

val b =
    """
    |a multiline
    |  string
    |"""
        .trimMargin()
        .nextMethod()

Option 3

val a =
    """
    a multiline
      string
    """
    .trimIndent()
        .nextMethod()

val b =
    """
    |a multiline
    |  string
    |"""
    .trimMargin()
        .nextMethod()

Option 4

val a =
    """
    a multiline
      string
    """.trimIndent()
        .nextMethod()

val b =
    """
    |a multiline
    |  string
    |""".trimMargin()
        .nextMethod()

Updating from .46 to .57 noticed the biggest change was trailing commas... any way to opt out of them? 😅