MultiplexingMode works for the most simple of cases, but for more complicated usages it breaks down. For example, in C++ there's a convention to put string-versions of protocol buffers in special raw strings (see https://reviews.llvm.org/D35943). But if you use the naive approach of implementing this with MultiplexingMode, then even if you comment out such a string, the highlighting will still occur.

It would be great if there was a reasonable solution that would work for grammar modes, even better if the inner mode doesn't have to be a grammar mode but can be any codemirror mode.

Hello, more than an issues this is actually a question.
I need to create a mode for a specific grammar for the current project I'm working on.

Our grammar is a little bit specific. It's like the where clause of a sql mode. Before to start to write our own grammar I tried this tool using the first example on the README file.

Using that input I got a .js file containing this:

var e = [/^[0-9]+/, /^(?:(?:[a-zA-Z])(?:[a-zA-Z0-9])*)/, /^(?:[\+\-\/\*])/, [1, " ", "\t", "\n"]]
exports.nodes = [
  [1, 3, 1],
  [/^[^]/, 0],
  [3, "number", e[0], -1,
   3, "variable", e[1], -1,
   e[2], -1,
   /^[^]/, -1],
  [3, "number", e[0], 4,
   3, "variable", e[1], 4,
   2, 6, 4, {"name":"ParenExpr"}],
  [e[3], 4,
   e[2], 5,
   0, -1],
  [e[3], 5,
   1, 3, -1],
  ["(", 7],
  [e[3], 7,
   1, 3, 8],
  [e[3], 8,
   ")", -1]
]
exports.start = 0
exports.token = 2

But I'm not sure how to proceed once I get this. How can I create the mode using this??
My guess is eventually I need to call CodeMirror.defineMode(...) but I'm a little bit lost. I was wondering if you can throw me a hint.

And by the way thanks for the great work. I'm going to give some good feedback to my bosses (my current project is not a personal project) and mention them the idea about share some love... in term$$ of... yeah, $$

This line makes it hard to require or import mode.js:

I believe this commit in kayak's fork is working around the same issue: kayak@f97b465

Not sure what changed, but after upgrading to v0.1.3 I get:
(sjql.mode.js contains the built grammar):

error  in ./resources/frontend/components/sjql-editor/grammar/sjql.mode.js

Syntax Error: Unterminated regular expression (76:8)
  
  74 |   [3, "identifier", e[4], -1],
  75 |   [1, 31, -1,
> 76 |    [0, /^(?=LOOKAHEAD(  _lookahead_3_0 -> _lookahead_3_RET[label="[iI][sS][tT](?![a-zA-Z¡-�])|[iI][sS](?![a-zA-Z¡-�]) TOKEN(operator)"];
     |         ^
  77 |   _lookahead_3_0 -> _lookahead_3_RET[label="[eE][nN][tT][hH][äÄ][lL][tT](?![a-zA-Z¡-�])|[cC][oO][nN][tT][aA][iI][nN][sS](?![a-zA-Z¡-�]) TOKEN(operator)"];
  78 |   _lookahead_3_0 -> _lookahead_3_RET[label="[bB][eE][gG][iI][nN][nN][tT](?![a-zA-Z¡-�])|[bB][eE][gG][iI][nN][sS](?![a-zA-Z¡-�]) TOKEN(operator)"];
  79 |   _lookahead_3_0 -> _lookahead_3_RET[label="[eE][nN][dD][eE][tT](?![a-zA-Z¡-�])|[eE][nN][dD][sS](?![a-zA-Z¡-�]) TOKEN(operator)"];

Hello,

Would you considere to provide this library as a npm package ?
It will be great to be able to do that, because it makes it difficult to use it when you are going use it through an angular/react application for example.

Thanks :).