Home/ Questions/Q 8309915
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-06-08T19:16:15+00:00

I am writing a lexer in OCaml for a small language, I have a

  • 0

I am writing a lexer in OCaml for a small language, I have a part of its grammar as follows:

tab-character = %x0009
eom-character = %x0019
space-character = %x0020
underscore = %x005F
single-quote = %x0027 ; '

I try to write let single_quote = [%x0027] in lexer.mll, but I get an illegal character error while compilation,

Could anyone tell me how to specify such numbers as %x0027?

The compilation works with the statement let single_quote = ['''], I have also seen another way to specify a character, for instance let black = ['\009'],

Does anyone know if there is any difference between these 3 ways?

Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team

    I think you’re asking about character literals in OCaml. They are defined in Section 6.1 of the OCaml manual.

    For single quote you can write '\'' or '\x27' or '\039'. All of these are equivalent. All character literals are written in single quotes–there’s no form beginning with a percent (%).

    Regular expressions in ocamllex can contain character literals (as above), which denote single characters, or string literals (in double quotes), which denote a sequence of characters. String literals follow the same pattern as character literals. A string containing just a single quote would be "'" or "\'" or "\x27" or "\039".

    I hope this helps.

    Edit:

    Yes, ['\x09'] and '\x09' and "\x09" are the same when considered as regular expressions. A set with one thing in it denotes the same thing as that one thing by itself. Similarly, a sequence of length 1 is the same as just the one thing.

    The value 0x3000 in Unicode represents an “Ideographic Space” (used in Chinese, Japanese, and Korean I guess). Handling Unicode in OCaml is a whole other topic. There is a Unicode library for OCaml called Camomile. I’ve never used it, but it is well regarded from what I’ve seen. I don’t think ocamllex works with Unicode. Just googling quickly I see a lexer generator named ulex that handles Unicode. There are probably others, this is just the top Google hit.

    (Whoops, I see Jonathan Protzenko already recommended ulex. Sorry for extra noise.)

    • 0