simon/ocaml-containers
1
0
Fork 0
mirror of https://github.com/c-cube/ocaml-containers.git synced 2025-12-11 13:38:32 -05:00
Code Issues Projects Releases Packages Wiki Activity Actions

optimize KMP search in CCString.Find (hand-specialize code)

Browse source
This commit is contained in:
Simon Cruanes 2016-03-20 23:20:46 +01:00
parent 71794d8d45
commit 5e30104954
1 changed files with 37 additions and 14 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

51
src/core/CCString.cppo.ml
View file

@ -134,17 +134,14 @@ module Find = struct
[i] index in [s] [i] index in [s]
[j] index in [pattern] [j] index in [pattern]
[len] length of [s] *) [len] length of [s] *)
let kmp_find_ let kmp_find ~pattern s idx =
: type a. dir:a direction -> pattern:a kmp_pattern -> string -> int -> int
= fun ~dir ~pattern s idx ->
let len = length s in let len = length s in
let get = get_ ~dir in
let i = ref idx in let i = ref idx in
let j = ref 0 in let j = ref 0 in
let pat_len = kmp_pattern_length pattern in let pat_len = kmp_pattern_length pattern in
while !j < pat_len && !i + !j < len do while !j < pat_len && !i + !j < len do
let c = get s (!i + !j) in let c = String.get s (!i + !j) in
let expected = get pattern.str !j in let expected = String.get pattern.str !j in
if c = expected if c = expected
then ( then (
(* char matches *) (* char matches *)
@ -168,18 +165,44 @@ module Find = struct
then !i then !i
else -1 else -1
let kmp_find ~pattern s i = kmp_find_ ~dir:Direct ~pattern s i (* proper search function, from the right.
[i] index in [s]
let kmp_rfind ~pattern s i = [j] index in [pattern]
let i = String.length s - i - 1 in [len] length of [s] *)
let res = kmp_find_ ~dir:Reverse ~pattern s i in let kmp_rfind ~pattern s idx =
let len = length s in
let i = ref (len - idx - 1) in
let j = ref 0 in
let pat_len = kmp_pattern_length pattern in
while !j < pat_len && !i + !j < len do
let c = String.get s (len - !i - !j - 1) in
let expected = String.get pattern.str (String.length pattern.str - !j - 1) in
if c = expected
then (
(* char matches *)
incr j;
) else (
let fail_offset = pattern.failure.(!j) in
if fail_offset >= 0
then (
assert (fail_offset < !j);
(* follow the failure link *)
i := !i + !j - fail_offset;
j := fail_offset
) else (
(* beginning of pattern *)
j := 0;
incr i
)
)
done;
(* adjust result: first, [res = string.length s - res -1] to convert (* adjust result: first, [res = string.length s - res -1] to convert
back to real indices; then, what we got is actually the position back to real indices; then, what we got is actually the position
of the end of the pattern, so we subtract the [length of the pattern -1] of the end of the pattern, so we subtract the [length of the pattern -1]
to obtain the real result. *) to obtain the real result. *)
if res = ~-1 if !j = pat_len
then res then len - !i - kmp_pattern_length pattern
else (String.length s - res) - kmp_pattern_length pattern else -1
type 'a pattern = type 'a pattern =
| P_char of char | P_char of char