threads / patch / 19148

v3Replace inefficient new String(String) constructor to silence FindBugs

Subject: [JGIT PATCH v3] Replace inefficient new String(String) constructor to silence FindBugs

## tl;dr

2 messages between May 1, 2009 and May 4, 2009. Diffs are folded; open one to read it.

replies: 1people: 2as markdown or json

Shawn O. Pearce· May 1, 2009, 15:54 UTC · lore

FindBugs keeps reporting that our usage of new String(String) is not the most efficient way to construct a string.

http://thread.gmane.org/gmane.comp.version-control.git/113739/focus=113787
Show 11 quoted lines
> I had a specific reason for forcing a new String object here.
>
> The line in question, p, is from the packed-refs file and
> contains the entire SHA-1 in hex form at the beginning of it.
> We've converted that into binary as an ObjectId, it uses 1/4 the
> space of the string portion.
>
> The Ref object, its ObjectId, and its name string, are going to be
> cached in a Map, probably long-term.  We're better off shedding the
> 80 bytes of memory used to hold the hex SHA-1 then risk substring()
> deciding its "faster" to reuse the char[] then to make a copy of it.

Another way to force this new unique String instance with its own private char[] is to use a StringBuilder and append onto it the ref name. This shouldn't be a warning for FindBugs, but it would accomplish the same goal of producing 1 clean copy, with no extra transient temporary array.

Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
CC: Yann Simon <yann.simon.fr@gmail.com>
CC: Matthias Sohn <matthias.sohn@sap.com>
---
 A less ugly version ?
 .../src/org/spearce/jgit/lib/RefDatabase.java      |    9 ++++++++-
 1 files changed, 8 insertions(+), 1 deletions(-)
Show changes to org.spearce.jgit/src/org/spearce/jgit/lib/RefDatabase.java +8 −1
diff --git a/org.spearce.jgit/src/org/spearce/jgit/lib/RefDatabase.java b/org.spearce.jgit/src/org/spearce/jgit/lib/RefDatabase.java
index 87f26bf..a865fba 100644
--- a/org.spearce.jgit/src/org/spearce/jgit/lib/RefDatabase.java
+++ b/org.spearce.jgit/src/org/spearce/jgit/lib/RefDatabase.java
@@ -447,7 +447,7 @@ private synchronized void refreshPackedRefs() {
 
 					final int sp = p.indexOf(' ');
 					final ObjectId id = ObjectId.fromString(p.substring(0, sp));
-					final String name = new String(p.substring(sp + 1));
+					final String name = copy(p.substring(sp + 1));
 					last = new Ref(Ref.Storage.PACKED, name, name, id);
 					newPackedRefs.put(last.getName(), last);
 				}
@@ -469,6 +469,13 @@ private synchronized void refreshPackedRefs() {
 		}
 	}
 
+	private static String copy(final String src) {
+		// Force a deep copy of the underlying char[] so that we can
+		// discard any garbage from any shared char[] within src.
+		//
+		return new StringBuilder(src.length()).append(src).toString();
+	}
+
 	private void lockAndWriteFile(File file, byte[] content) throws IOException {
 		String name = file.getName();
 		final LockFile lck = new LockFile(file);
-- 
1.6.3.rc3.212.g8c698
Sohn, Matthias· May 4, 2009, 21:23 UTC · re: Shawn O. Pearce · lore

RE: [JGIT PATCH v3] Replace inefficient new String(String) constructor to silence FindBugs

Shawn O. Pearce [mailto:spearce@spearce.org] wrote :
 
Show 29 quoted lines
> FindBugs keeps reporting that our usage of new String(String)
> is not the most efficient way to construct a string.
> 
> http://thread.gmane.org/gmane.comp.version-
> control.git/113739/focus=113787
> > I had a specific reason for forcing a new String object here.
> >
> > The line in question, p, is from the packed-refs file and
> > contains the entire SHA-1 in hex form at the beginning of it.
> > We've converted that into binary as an ObjectId, it uses 1/4 the
> > space of the string portion.
> >
> > The Ref object, its ObjectId, and its name string, are going to be
> > cached in a Map, probably long-term.  We're better off shedding the
> > 80 bytes of memory used to hold the hex SHA-1 then risk substring()
> > deciding its "faster" to reuse the char[] then to make a copy of it.
> 
> Another way to force this new unique String instance with its own
> private char[] is to use a StringBuilder and append onto it the
> ref name.  This shouldn't be a warning for FindBugs, but it would
> accomplish the same goal of producing 1 clean copy, with no extra
> transient temporary array.
> 
> Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
> CC: Yann Simon <yann.simon.fr@gmail.com>
> CC: Matthias Sohn <matthias.sohn@sap.com>
> ---
> 
>  A less ugly version ?

I agree, this looks better than the previous proposal but still a simple String copy constructor looks even simpler.

I tried the alternative approach Robin proposed using FindBugs filter mechanisms to suppress the undesired warning. I will post that in my next mail.

-- Matthias

← back to recent threads