This OCaml library interfaces with the C library PCRE, providing Perl-compatible regular expressions for string matching.

PCRE-OCaml offers:

• Pattern searching
• Subpattern extraction
• String splitting by patterns
• Pattern substitution

Reasons to choose PCRE-OCaml:

• The PCRE library by Philip Hazel is mature and stable, implementing nearly all Perl regular expression features. High-level OCaml functions (split, replace, etc.) are compatible with Perl functions, as much as OCaml allows. Some developers find Perl-style regex syntax more intuitive and powerful than the Emacs-style regex used in OCaml's Str module.

• PCRE-OCaml is reentrant and thread-safe, unlike the Str module. This reentrancy offers convenience, eliminating concerns about library state.

• High-level replacement and substitution functions in OCaml are faster than those in the Str module. When compiled to native code, they can even outperform Perl's C-based functions.

• Returned data is unique, allowing safe destructive updates without side effects.

• The library interface uses labels and default arguments for enhanced programming comfort.

Please run:

odig odoc pcre2

Or:

dune build @doc

Consult the API for details.

Functions support two flag types:

1. Convenience flags: Readable and concise, translated internally on each call. Example:

   let rex = Pcre.regexp ~flags:[`ANCHORED; `CASELESS] "some pattern" in
   (* ... *)

   These are easy to use but may incur overhead in loops. For performance optimization, consider the next approach.

2. Internal flags: Predefined and translated from convenience flags for optimal loop performance. Example:

   let iflags = Pcre.cflags [`ANCHORED; `CASELESS] in
   for i = 1 to 1000 do
     let rex = Pcre.regexp ~iflags "some pattern constructed at runtime" in
     (* ... *)
   done

   Translating flags outside loops saves cycles. Avoid creating regex in loops:

   for i = 1 to 1000 do
     let chunks = Pcre.split ~pat:"[ \t]+" "foo bar" in
     (* ... *)
   done

   Instead, predefine the regex:

   let rex = Pcre.regexp "[ \t]+" in
   for i = 1 to 1000 do
     let chunks = Pcre.split ~rex "foo bar" in
     (* ... *)
   done

Functions use optional arguments with intuitive defaults. For instance, Pcre.split defaults to whitespace as the pattern. The examples directory contains applications demonstrating PCRE-OCaml's functionality.

PCRE includes a DFA match function for restarting partial matches with new input, exposed via pcre_dfa_exec. While not suitable for extracting submatches or splitting strings, it's useful for streaming and search tasks.

Example of a partial match restarted:

utop # open Pcre;;
utop # let rex = regexp "12+3";;
val rex : regexp = <abstr>
utop # let workspace = Array.make 40 0;;
val workspace : int array =
  [| ... |]
utop # pcre_dfa_exec ~rex ~flags:[`PARTIAL] ~workspace "12222";;
Exception: Pcre.Error Partial.
utop # pcre_dfa_exec ~rex ~flags:[`PARTIAL; `DFA_RESTART] ~workspace "2222222";;
Exception: Pcre.Error Partial.
utop # pcre_dfa_exec ~rex ~flags:[`PARTIAL; `DFA_RESTART] ~workspace "2222222";;
Exception: Pcre.Error Partial.
utop # pcre_dfa_exec ~rex ~flags:[`PARTIAL; `DFA_RESTART] ~workspace "223xxxx";;
- : int array = [|0; 3; 0|]

Refer to the pcre_dfa_exec documentation and the dfa_restart example for more information.

Submit bug reports, feature requests, and contributions via the GitHub issue tracker.

For the latest information, visit: https://mmottl.github.io/pcre-ocaml