HTML/JS/CSS syntax highlighter that supports LLVM

Question

I'm currently using google-code-prettify for syntax highlighting. It doesn't seem to support LLVM. Here's what it looks like when formatting LLVM:

It's pretty ugly. Are there any front-end syntax highlighters that support LLVM?

Answer 1

I grabbed the patterns for LLVM assembly syntax highlighting from the LLVM TextMate Bundle project and wrote up a plugin for google-code-prettify .

PR['registerLangHandler'](
    PR['createSimpleLexer'](
        [
         // Whitespace
         [PR['PR_PLAIN'],       /^[\t\n\r \xA0]+/, null, '\t\n\r \xA0'],

         // A double quoted, possibly multi-line, string.
         [PR['PR_STRING'],      /^\"(?:[^\"\\]|\\[\s\S])*(?:\"|$)/, null, '"'],

         // comment.llvm
         [PR['PR_COMMENT'],       /^;[^\r\n]*/, null, ';'],
        ],
        [
         // llvm instructions
         [PR['PR_KEYWORD'],     /^\b(?:add|alloca|and|ashr|bitcast|br|call|eq|exact|extractelement|extractvalue|fadd|fcmp|fdiv|fmul|fpext|fptosi|fptoui|fptrunc|free|frem|fsub|getelementptr|icmp|inbounds|indirectbr|insertelement|insertvalue|inttoptr|invoke|load|lshr|malloc|mul|ne|nsw|nuw|oeq|oge|ogt|ole|olt|one|or|ord|phi|ptrtoint|ret|sdiv|select|sext|sge|sgt|shl|shufflevector|sitofp|sle|slt|srem|store|sub|switch|trunc|udiv|ueq|uge|uge|ugt|ugt|uitofp|ule|ule|ult|ult|une|uno|unreachable|unwind|urem|va_arg|xor|zext)\b/, null],

         // llvm keywords
         [PR['PR_KEYWORD'],     /^\b(?:addrspace|alias|align|alignstack|alwaysinline|appending|asm|blockaddress|byval|c|cc|ccc|coldcc|common|constant|datalayout|declare|default|define|deplibs|dllexport|dllimport|except|extern_weak|external|fastcc|gc|global|hidden|inlinehint|inreg|internal|linkonce|linkonce_odr|metadata|module|naked|nest|noalias|nocapture|noimplicitfloat|noinline|noredzone|noreturn|nounwind|optsize|private|protected|ptx_device|ptx_kernel|readnone|readonly|section|sideeffect|signext|sret|ssp|sspreq|tail|target|thread_local|to|triple|uwtable|volatile|weak|weak_odr|x86_fastcallcc|x86_stdcallcc|zeroext)\b/, null],

         // variable.llvm
         [PR['PR_TYPE'],       /^\s(?:[%@][-a-zA-Z$._][-a-zA-Z$._0-9]*)/],

         // variable.language.llvm
         [PR['PR_TYPE'],       /^\s(?:[%]\d+)/],

         // storage.type.language.llvm
         [PR['PR_PLAIN'],       /^\b(?:i\d+\**)/],

         // variable.metadata.llvm
         [PR['PR_PLAIN'],       /^(!\d+)/],

         // constant.numeric.float.llvm
         [PR['PR_LITERAL'],       /^\b\d+\.\d+\b/],

         // constant.numeric.integer.llvm
         [PR['PR_LITERAL'],       /^\b(?:\d+|0(?:x|X)[a-fA-F0-9]+)\b/],
        ]),
    ['llvm', 'll']);

Here are the links to a demo jsFiddle and a Gist with a demo page and README.

Answer 2

Considering LLVM is a rather obscure programming language, it's doubtful you'll find a syntax highlighter specifically designed for it. However, that doesn't mean you can't cheat a little, and make something work. I'd recommend using SyntaxHighlighter. http://alexgorbatchev.com/SyntaxHighlighter

Though it does not natively support LLVM, it provides many built-in brushes for various other programming languages. I would try it out, and try applying different brushes (syntax highlighters) to your LLVM code. Some will match quite well, while others will appear pretty bland, like in the example you demonstrated.

If you look here, you'll also see that many unofficial brushes exist as well. That is a benefit to using SyntaxHighlighter--if a brush for LLVM is ever designed (or if you decide to design it yourself) it'll be simple to implement.

Good luck!

Answer 3

don't know if you'r a linux guy, since you use Python i suppose you do. Emacs and Vim seem to support LLVM syntax highlighting. See http://llvm.org/docs/GettingStarted.html. If you don't want to use these perhaps you can use the syntax files (with some fiddling) that are provided in other IDE's or Editors. Googled it up so can't help you any further. Success on your quest if this is't enough.

Answer 4

Pyments has support for LLVM. It is based on a python backend and the user can choose between styles.

Output modes:

• HTML
• ANSI sequences (console output)
• LaTeX combined with the package minted
• RTF

http://pygments.org/demo/45402/