Question

我使用Vim / gVim在javascript（节点）中编程。我在我的filetype插件中将jslint连接为makeprg。这是错误格式：

efm=%-P%f,
        \%A%>%\\s%\\?#%*\\d\ %m,%Z%.%#Line\ %l\\,\ Pos\ %c,
        \%-G%f\ is\ OK.,%-Q

这是jslint的输出：

routes/pcr.js
#1 'db' was used before it was defined.
db.collection('pcrs', function (err, collection) { // Line 11, Pos 5
#2 'db' was used before it was defined.
db.collection('pcrs', function (err, collection) { // Line 23, Pos 5
#3 'BSON' was used before it was defined.
collection.findOne({'_id': new BSON.ObjectID(id)}, function (err, item) { // Line 24, Pos 40

以下是quickfix窗口的输出：

routes/pcr.js|11 col 5| 'db' was used before it was defined.
routes/pcr.js|23 col 5| 'db' was used before it was defined.
routes/pcr.js|24 col 40| 'BSON' was used before it was defined.

在列号之后，我想留下来说出2位数字（我希望文件不超过99个错误！）所以它看起来像：

routes/pcr.js|11 col  5| 'db' was used before it was defined.
routes/pcr.js|23 col  5| 'db' was used before it was defined.
routes/pcr.js|24 col 40| 'BSON' was used before it was defined.

我想这也会影响0-9行。是否可以有条件地填充输出？

Answer 1

合理的数字肯定会很好，但我认为这需要Vim的源代码补丁。

quickfix窗口中的信息来自Vim的内部数据结构（格式见:help getqflist()），Vim决定如何将其可视化。

Answer 2

:help quickfix-window提及重新格式化错误列表。

以下设置适用于我（更新）：

au BufRead quickfix setl modifiable
            \| silent exe "%!perl -ple '
                \my ($file, $pos, $msg) = split qr{[|]}, $_, 3;
                \my $aligned_pos = sub {
                \  my @p = split qr{[ ]}, shift;
                \  return                                        if @p == 0;
                \  return sprintf q{\\%3s}, @p                   if @p == 1;
                \  return sprintf q{\\%3s \\%s}, @p              if @p == 2;
                \  return sprintf q{\\%3s \\%s \\%2s}, @p        if @p == 3;
                \  return sprintf q{\\%3s \\%s \\%2s \\%-8s}, @p if @p == 4;
                \  return join q{ }, @p;
                \}->($pos);
                \$_ = join q{|}, $file, $aligned_pos, $msg;
            \'"
            \| setl nomodifiable

在Vim中对齐/对齐errorformat输出

2 个答案: