Vim：错误格式匹配所有内容

发布于 2024-09-19 06:09:37 字数 748 浏览 4 评论 0原文

我试图将以下错误与 efm 匹配：

AssertionError: 1 == 2
    at /home/strager/projects/blah/blah.js:13:37

错误消息可以是任何内容（即它并不总是与 AssertionError: .* 或 .*Error: .* 格式匹配）。一般格式为：

errormessage
    at filename:line:column

我的问题是错误消息与任意行匹配；我需要将错误消息限制为仅一行，并且仅在其后跟匹配的“at”行时才匹配。

我尝试过以下efm：

set efm=%Z\ \ \ \ at\ %f:%l:%c,%E%m
" %Z    at %f:%l:%c,%E%m

这几乎可以工作，但除了错误之外，它还匹配状态行（例如，错误之前和之后的非错误）。如何强制 %E%m ... %Z 总共只有两行（一行用于错误消息，一行用于 at 行）？如果需要，我可以使用 makeprg 的标准 UNIX 工具。

原文

I am trying to match the following error with efm:

AssertionError: 1 == 2
    at /home/strager/projects/blah/blah.js:13:37

The error message can be anything (i.e. it doesn't always match the AssertionError: .* or .*Error: .* formats). The general format is:

errormessage
    at filename:line:column

My problem is that the error message matches any line; I need to restrict the error message to one line only, and only match if it's followed by a matching "at" line.

I have tried the following efm:

set efm=%Z\ \ \ \ at\ %f:%l:%c,%E%m
" %Z    at %f:%l:%c,%E%m

This almost works, but it matches status lines (e.g. non-errors before and after the error) in addition to the errors. How can I force %E%m ... %Z to be only two lines total (one for the error message, and one for the at line)? I have access to the standard UNIX tools for makeprg if needed.

分享到QQ

分享到微博

如果你对这篇内容有疑问，欢迎到本站社区发帖提问参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。

发布评论

需要登录才能够评论，你可以免费注册一个本站的账号。

神仙妹妹 2024-09-26 06:09:37

这有效吗？

set efm=%Z\ \ \ \ at\ %f:%l:%c,%E%m,%-G%.%#

%-G%.%# 告诉 vim 忽略与其他模式不匹配的整行。

Does this work?

set efm=%Z\ \ \ \ at\ %f:%l:%c,%E%m,%-G%.%#

The %-G%.%# tells vim to ignore entire lines that do not match the other patterns.

回复收藏 0 原文

债姬 2024-09-26 06:09:37

您真的想花时间学习一种在其他地方没有应用的晦涩模式语言吗？除非有人付钱让你编写 vim 编译器插件，否则我不会（而且我喜欢 vim！）。既然您愿意打开工具箱，只需在检查器周围编写一个包装器并输出易于解析的格式即可。例如：

#!/usr/bin/perl
use warnings;
use strict;

open my $fh, '-|', 'compiler', @_ or die $!;

my $last_line = <$fh> // exit;
while (defined(my $line = <$fh>)) {
    my($file, $l, $c) = $line =~ /^    at (.+?):(\d+):(\d+)$/;
    print "$file:$l:$c: $last_line" if defined($file);
    $last_line = $line;
}

Do you really want to spend your time learning an obscure pattern language that has no application anywhere else? Unless someone is paying you to write vim compiler plug-ins, I wouldn't (and I love vim!). Since you're willing to open the toolbox, just write a wrapper around your checker and spit out an easy-to-parse format. Eg:

#!/usr/bin/perl
use warnings;
use strict;

open my $fh, '-|', 'compiler', @_ or die $!;

my $last_line = <$fh> // exit;
while (defined(my $line = <$fh>)) {
    my($file, $l, $c) = $line =~ /^    at (.+?):(\d+):(\d+)$/;
    print "$file:$l:$c: $last_line" if defined($file);
    $last_line = $line;
}

回复收藏 0 原文