Re: [PATCH 2/5] doc: fix synopsis analysis logic

[Jean-Noël Avila <jn.avila@xxxxxxx>]

Le 05/04/2025 à 00:39, Martin Ågren a écrit :
> On Sun, 30 Mar 2025 at 19:16, Jean-Noël Avila via GitGitGadget
> <gitgitgadget@xxxxxxxxx> wrote:
>>
>> The synopsis analysis logic was not able to handle backslashes and stars
>> which are used in the synopsis of the git-rm command. This patch fixes the
>> issue by updating the regular expression used to match the keywords.
> 
>> --- a/Documentation/asciidoctor-extensions.rb.in
>> +++ b/Documentation/asciidoctor-extensions.rb.in
>> @@ -50,7 +50,7 @@ module Git
>>        def process parent, reader, attrs
>>          outlines = reader.lines.map do |l|
>>            l.gsub(/(\.\.\.?)([^\]$.])/, '`\1`\2')
>> -           .gsub(%r{([\[\] |()>]|^)([-a-zA-Z0-9:+=~@,/_^\$]+)}, '\1{empty}`\2`{empty}')
>> +           .gsub(%r{([\[\] |()>]|^)([-a-zA-Z0-9:+=~@,/_^\$\\\*]+)}, '\1{empty}`\2`{empty}')
>>             .gsub(/(<[-a-zA-Z0-9.]+>)/, '__\\1__')
>>             .gsub(']', ']{empty}')
>>          end
>> @@ -72,7 +72,7 @@ module Git
>>            %(<inlineequation><alt><![CDATA[#{equation = node.text}]]></alt><mathphrase><![CDATA[#{equation}]]></mathphrase></inlineequation>)
>>          elsif type == :monospaced
>>            node.text.gsub(/(\.\.\.?)([^\]$.])/, '<literal>\1</literal>\2')
>> -              .gsub(%r{([\[\s|()>.]|^|\]|&gt;)(\.?([-a-zA-Z0-9:+=~@,/_^\$]+\.{0,2})+)}, '\1<literal>\2</literal>')
>> +              .gsub(%r{([\[\s|()>.]|^|\]|&gt;)(\.?([-a-zA-Z0-9:+=~@,/_^\$\\\*]+\.{0,2})+)}, '\1<literal>\2</literal>')
>>                .gsub(/(&lt;[-a-zA-Z0-9.]+&gt;)/, '<emphasis>\1</emphasis>')
>>          else
>>            open, close, supports_phrase = QUOTE_TAGS[type]
>> @@ -100,7 +100,7 @@ module Git
>>        def convert_inline_quoted node
>>          if node.type == :monospaced
>>            node.text.gsub(/(\.\.\.?)([^\]$.])/, '<code>\1</code>\2')
>> -              .gsub(%r{([\[\s|()>.]|^|\]|&gt;)(\.?([-a-zA-Z0-9:+=~@,/_^\$]+\.{0,2})+)}, '\1<code>\2</code>')
>> +              .gsub(%r{([\[\s|()>.]|^|\]|&gt;)(\.?([-a-zA-Z0-9:+=~@,/_^\$\\\*]+\.{0,2})+)}, '\1<code>\2</code>')
>>                .gsub(/(&lt;[-a-zA-Z0-9.]+&gt;)/, '<em>\1</em>')
> 
> This seems to introduce some extra spacing in the rendered man pages, e.g.,
> "The bundle.*  keys" or "Fileglobs (e.g.  *.c)". (Asciidoctor 2.0.18.) I
> haven't dug into the regexes so see what might be the cause.
> 

The xml regex seems ok for this. The docbook output is as follows:

(...)
pairs in this list are in the <literal>bundle.*</literal> namespace (see
(...)

The manpage output seems also correct:

(...)
would accept (with the
\fB\-\-file\fR
option)\&. The key\-value pairs in this list are in the
\fBbundle\&.*\fR
namespace (see
(...)

Strangely, the --file above is rendered correctly, but the bundle.*
below not. I do not know TROFF to assess what is going on. Has '*' a
special behavior in a bold span?

JN