git/list[1] front-page[2] threads[3] people[4] search[5] about
 

[PATCH v2 8/9] wildmatch: advance faster in <asterisk> + <literal> patterns

From
Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Date
Dec 28, 2012, 04:10 UTC
Message-ID
<1356667854-8686-9-git-send-email-pclouds@gmail.com>
In-Reply-To
<1356667854-8686-1-git-send-email-pclouds@gmail.com>

compat, '*/*/*' on linux-2.6.git file list 2000 times, before: wildmatch 7s 985049us fnmatch 2s 735541us or 34.26% faster

and after: wildmatch 4s 492549us fnmatch 0s 888263us or 19.77% slower

Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
---
 t/t3070-wildmatch.sh |  6 ++++++
 wildmatch.c          | 21 +++++++++++++++++++++
 2 files changed, 27 insertions(+)
diff --git a/t/t3070-wildmatch.sh b/t/t3070-wildmatch.sh
index 4cdb13b..dcbc8bc 100755
--- a/t/t3070-wildmatch.sh
+++ b/t/t3070-wildmatch.sh
@@ -207,6 +207,9 @@ match 0 x foo '*/*/*'
 match 0 x foo/bar '*/*/*'
 match 1 x foo/bba/arr '*/*/*'
 match 0 x foo/bb/aa/rr '*/*/*'
+match 1 x abcXdefXghi '*X*i'
+match 0 x ab/cXd/efXg/hi '*X*i'
+match 1 x ab/cXd/efXg/hi '*/*X*/*/*i'
 
 pathmatch 1 foo foo
 pathmatch 0 foo fo
@@ -226,5 +229,8 @@ pathmatch 0 foo '*/*/*'
 pathmatch 0 foo/bar '*/*/*'
 pathmatch 1 foo/bba/arr '*/*/*'
 pathmatch 1 foo/bb/aa/rr '*/*/*'
+pathmatch 1 abcXdefXghi '*X*i'
+pathmatch 1 ab/cXd/efXg/hi '*/*X*/*/*i'
+pathmatch 1 ab/cXd/efXg/hi '*Xg*i'
 
 test_done
diff --git a/wildmatch.c b/wildmatch.c
index f6d45d5..40eda08 100644
--- a/wildmatch.c
+++ b/wildmatch.c
@@ -132,6 +132,27 @@ static int dowild(const uchar *p, const uchar *text, unsigned int flags)
 			while (1) {
 				if (t_ch == '\0')
 					break;
+				/*
+				 * Try to advance faster when an asterisk is
+				 * followed by a literal. We know in this case
+				 * that the the string before the literal
+				 * must belong to "*".
+				 */
+				if (!is_glob_special(*p)) {
+					p_ch = *p;
+					if ((flags & WM_CASEFOLD) && ISUPPER(p_ch))
+						p_ch = tolower(p_ch);
+					while ((t_ch = *text) != '\0' &&
+					       (!(flags & WM_PATHNAME) || t_ch != '/')) {
+						if ((flags & WM_CASEFOLD) && ISUPPER(t_ch))
+							t_ch = tolower(t_ch);
+						if (t_ch == p_ch)
+							break;
+						text++;
+					}
+					if (t_ch != p_ch)
+						return WM_NOMATCH;
+				}
 				if ((matched = dowild(p, text, flags)) != WM_NOMATCH) {
 					if (!match_slash || matched != WM_ABORT_TO_STARSTAR)
 						return matched;
-- 
1.8.0.rc2.23.g1fb49df
Previous: Nguyễn Thái Ngọc DuyNext: Nguyễn Thái Ngọc Duy
Message 9 of 10 in “fnmatch replacement step 1”
  1. 0/9 fnmatch replacement step 1Nguyễn Thái Ngọc Duy, Dec 28, 2012
  2. 1/9 compat/fnmatch: respect NO_FNMATCH* even on glibcNguyễn Thái Ngọc Duy, Dec 28, 2012
  3. 2/9 wildmatch: replace variable 'special' with better named onesNguyễn Thái Ngọc Duy, Dec 28, 2012
  4. 3/9 wildmatch: rename constants and update prototypeNguyễn Thái Ngọc Duy, Dec 28, 2012
  5. 4/9 wildmatch: make dowild() take arbitrary flagsNguyễn Thái Ngọc Duy, Dec 28, 2012
  6. 5/9 wildmatch: support "no FNM_PATHNAME" modeNguyễn Thái Ngọc Duy, Dec 28, 2012
  7. 6/9 test-wildmatch: add "perf" command to compare wildmatch and fnmatchNguyễn Thái Ngọc Duy, Dec 28, 2012
  8. 7/9 wildmatch: make a special case for "*/" with FNM_PATHNAMENguyễn Thái Ngọc Duy, Dec 28, 2012
  9. 8/9 wildmatch: advance faster in <asterisk> + <literal> patternsNguyễn Thái Ngọc Duy, Dec 28, 2012
  10. 9/9 Makefile: add USE_WILDMATCH to use wildmatch as fnmatchNguyễn Thái Ngọc Duy, Dec 28, 2012

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.