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

[JGIT PATCH 07/14] Micro-optimize AbstractTreeIterator.pathCompare

From
Shawn O. Pearce <spearce@spearce.org>
Date
Aug 18, 2008, 23:53 UTC
Message-ID
<1219103602-32222-8-git-send-email-spearce@spearce.org>
In-Reply-To
<1219103602-32222-7-git-send-email-spearce@spearce.org>

We were doing far too much work in pathCompare to handle cases that just cannot ever happen, such as if the paths were the same length but had different "last path char" and then somehow had different lengths.

We also had the JVM doing a lot of comparsion ops just to return -1/0/1 when really we can get away with the non-zero result returned to the caller. Issuing just the subtraction and one comparsion to 0 is much quicker, JIT or not.

Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
---
 .../jgit/treewalk/AbstractTreeIterator.java        |   44 ++-----------------
 1 files changed, 5 insertions(+), 39 deletions(-)
diff --git a/org.spearce.jgit/src/org/spearce/jgit/treewalk/AbstractTreeIterator.java b/org.spearce.jgit/src/org/spearce/jgit/treewalk/AbstractTreeIterator.java
index bd75d2d..31257b5 100644
--- a/org.spearce.jgit/src/org/spearce/jgit/treewalk/AbstractTreeIterator.java
+++ b/org.spearce.jgit/src/org/spearce/jgit/treewalk/AbstractTreeIterator.java
@@ -243,45 +243,11 @@ int pathCompare(final AbstractTreeIterator p, final int pMode) {
 				return cmp;
 		}
 
-		if (cPos < aLen) {
-			final int aj = a[cPos] & 0xff;
-			final int lastb = lastPathChar(pMode);
-			if (aj < lastb)
-				return -1;
-			else if (aj > lastb)
-				return 1;
-			else if (cPos == aLen - 1)
-				return 0;
-			else
-				return -1;
-		}
-
-		if (cPos < bLen) {
-			final int bk = b[cPos] & 0xff;
-			final int lasta = lastPathChar(mode);
-			if (lasta < bk)
-				return -1;
-			else if (lasta > bk)
-				return 1;
-			else if (cPos == bLen - 1)
-				return 0;
-			else
-				return 1;
-		}
-
-		final int lasta = lastPathChar(mode);
-		final int lastb = lastPathChar(pMode);
-		if (lasta < lastb)
-			return -1;
-		else if (lasta > lastb)
-			return 1;
-
-		if (aLen == bLen)
-			return 0;
-		else if (aLen < bLen)
-			return -1;
-		else
-			return 1;
+		if (cPos < aLen)
+			return (a[cPos] & 0xff) - lastPathChar(pMode);
+		if (cPos < bLen)
+			return lastPathChar(mode) - (b[cPos] & 0xff);
+		return lastPathChar(mode) - lastPathChar(pMode);
 	}
 
 	private static int lastPathChar(final int mode) {
-- 
1.6.0.87.g2858d
Previous: Shawn O. PearceNext: Shawn O. Pearce
Message 8 of 20 in “TreeWalk D/F conflict detection”
  1. 00/14 TreeWalk D/F conflict detectionShawn O. Pearce, Aug 18, 2008
  2. 01/14 Detect path names which overflow the name length field in the indexShawn O. Pearce, Aug 18, 2008
  3. 02/14 Fix NB.decodeUInt16 to correctly handle the high byteShawn O. Pearce, Aug 18, 2008
  4. 03/14 Add test cases for NB.encode and NB.decode family of routinesShawn O. Pearce, Aug 18, 2008
  5. 04/14 Fix DirCache's skip over null byte padding when reading a DIRC fileShawn O. Pearce, Aug 18, 2008
  6. 05/14 Fix usage of assertEquals in DirCacheIteratorTestShawn O. Pearce, Aug 18, 2008
  7. 06/14 Refactor AbstractTreeIterator.pathCompare to force another modeShawn O. Pearce, Aug 18, 2008
  8. 07/14 Micro-optimize AbstractTreeIterator.pathCompareShawn O. Pearce, Aug 18, 2008
  9. 08/14 Optimize path comparsion within subtrees during TreeWalkShawn O. Pearce, Aug 18, 2008
  10. 09/14 Refactor AbstractTreeIterator semantics to start on first entryShawn O. Pearce, Aug 18, 2008
  11. 10/14 Make all AbstractTreeIterator implementations bi-directionalShawn O. Pearce, Aug 18, 2008
  12. 11/14 Expose beginning of iterator indication from AbstractTreeIteratorShawn O. Pearce, Aug 18, 2008
  13. 12/14 Allow application code to set ObjectIds in DirCacheEntryShawn O. Pearce, Aug 18, 2008
  14. 13/14 Create NameConflictTreeWalk to transparently detect D/F conflictsShawn O. Pearce, Aug 18, 2008
  15. 14/14 Add test case for NameConflictTreeWalkShawn O. Pearce, Aug 18, 2008
  16. Junio C HamanoAug 19, 2008
  17. Robin RosenbergAug 19, 2008
  18. Shawn O. PearceAug 19, 2008
  19. David WoodhouseAug 19, 2008
  20. Shawn O. PearceAug 19, 2008

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.