[Python-checkins] cpython: Close #13126: "Simplify" FASTSEARCH() code to help the compiler to emit more

[victor.stinner]

http://hg.python.org/cpython/rev/c5e2ea9e3aa7
changeset:   83192:c5e2ea9e3aa7
user:        Victor Stinner <victor.stinner at gmail.com>
date:        Mon Apr 08 00:26:43 2013 +0200
summary:
  Close #13126: "Simplify" FASTSEARCH() code to help the compiler to emit more
efficient machine code. Patch written by Antoine Pitrou.

Without this change, str.find() was 10% slower than str.rfind() in the worst
case.

files:
  Objects/stringlib/fastsearch.h |  8 +++++---
  1 files changed, 5 insertions(+), 3 deletions(-)


diff --git a/Objects/stringlib/fastsearch.h b/Objects/stringlib/fastsearch.h
--- a/Objects/stringlib/fastsearch.h
+++ b/Objects/stringlib/fastsearch.h
@@ -142,6 +142,8 @@
     mask = 0;
 
     if (mode != FAST_RSEARCH) {
+        const STRINGLIB_CHAR *ss = s + m - 1;
+        const STRINGLIB_CHAR *pp = p + m - 1;
 
         /* create compressed boyer-moore delta 1 table */
 
@@ -156,7 +158,7 @@
 
         for (i = 0; i <= w; i++) {
             /* note: using mlast in the skip path slows things down on x86 */
-            if (s[i+m-1] == p[m-1]) {
+            if (ss[i] == pp[0]) {
                 /* candidate match */
                 for (j = 0; j < mlast; j++)
                     if (s[i+j] != p[j])
@@ -172,13 +174,13 @@
                     continue;
                 }
                 /* miss: check if next character is part of pattern */
-                if (!STRINGLIB_BLOOM(mask, s[i+m]))
+                if (!STRINGLIB_BLOOM(mask, ss[i+1]))
                     i = i + m;
                 else
                     i = i + skip;
             } else {
                 /* skip: check if next character is part of pattern */
-                if (!STRINGLIB_BLOOM(mask, s[i+m]))
+                if (!STRINGLIB_BLOOM(mask, ss[i+1]))
                     i = i + m;
             }
         }

-- 
Repository URL: http://hg.python.org/cpython