Everything in diagrams.css should be scoped to some wrapping css class, because as is it cannot be bundled with the rest of an app's css because of styles like this:

div {
    -webkit-touch-callout: none;
    -webkit-user-select: none;
    -khtml-user-select: none;
    -moz-user-select: none;
    -ms-user-select: none;
    user-select: none;
}

svg {
    width: 100%;
}

Curre

It would be nice if subpatterns could be used within subpatterns.

#[derive(Logos, Copy, Clone, Debug)]
#[logos(subpattern a = r"[a-zA-Z!$%&*/:<=>?^_~]")]
#[logos(subpattern b = r"(?&a)|[0-9+\-.@]")]
pub enum Token {
    #[error]
    Error,
    // ...
}

Currently, this is not supported and causes a regex parse error.

Input test case

Any sv file.

Describe what is wrong or missing

As far as I can tell, the verible-verilog-kythe-extractor tool cannot directly write its output into a file.
The issue is that it force having a potentially huge buffer in memory.

It would be nice to have the capability to output the result of the kythe extraction directly to an output file without having to rely on

Prepare the parser to PHP 8 :

• https://wiki.php.net/rfc/union_types_v2

Work In Progress as RFC are not yet closed

Would it be possible to have the regex parser support character classes like \w within other character classes? I had a regex pattern earlier that used the character class [0-9a-zA-Z_\.-], and I attempted to simplify it with [\w\.\-]. I didn't notice this library doesn't support doing that, and was wondering just how difficult that would be to implement. For the time being i'm just expanding

Bison has %expect declarations which tell the parsing system "I know there are n conflicts, please don't warn me if you find n too". Although undocumented in http://dinosaur.compilertools.net/yacc/index.html, Berkeley Yacc seems to support %expect too, so I think grmtools should add that. See section 3.7.9 of https://www.gnu.org/software/bison/manual/bison.html for more details.

Bison a