RFR: 8296420: javac has long lines in its command-line help
Jonathan Gibbons
jjg at openjdk.org
Tue Apr 4 17:40:08 UTC 2023
On Tue, 4 Apr 2023 16:16:51 GMT, Archie L. Cobbs <duke at openjdk.org> wrote:
> The issue states that the output of `javac --help` and `javac --help-extra` should not overflow 80 columns.
>
> This patch updates `javac.properties` to add newlines, etc. to address this issue, and adds a regression test.
Nice work; nice test!
One optional suggestion to improve the test.
src/jdk.compiler/share/classes/com/sun/tools/javac/resources/javac.properties line 102:
> 100: Options to pass to annotation processors
> 101: javac.opt.implicit=\
> 102: Specify whether or not to generate class files\n\
There's been a lot of discussion (elsewhere) of late regarding the phrase _whether or not_.
Generally, the consensus seems to be that the correct form in cases like this is just _whether_ and that _whether or not_ idiomatically means something else. With that in mind, I'd suggest dropping _or not_ and then see how that affects the needs for any line-wrapping.
src/jdk.compiler/share/classes/com/sun/tools/javac/resources/javac.properties line 335:
> 333: are found for an implicitly compiled class
> 334: javac.opt.preview=\
> 335: Enable preview language features. To be used in conjunction\n\
I'd consider breaking "early", after the initial period.
test/langtools/tools/javac/options/HelpOutputColumnWidthTest.java line 88:
> 86: if (tooLongLine != null)
> 87: throw new Exception("output line too long: \"" + tooLongLine.trim() + "\"");
> 88: }
This is good. It would be even better if you removed `findFirst` and then reported all the lines if the resulting list is non-empty.
-------------
Marked as reviewed by jjg (Reviewer).
PR Review: https://git.openjdk.org/jdk/pull/13329#pullrequestreview-1371473526
PR Review Comment: https://git.openjdk.org/jdk/pull/13329#discussion_r1157564897
PR Review Comment: https://git.openjdk.org/jdk/pull/13329#discussion_r1157567049
PR Review Comment: https://git.openjdk.org/jdk/pull/13329#discussion_r1157569385
More information about the compiler-dev
mailing list