[PATCH] style patches

[Shatzer, Larry]

Fixes stuff like if( to if (, spaces between = and other corrections.

-- Larry

Index: 
src/main/org/apache/tools/ant/taskdefs/optional/j2ee/AbstractHotDeploymentTool.java
===
RCS file: 
/home/cvspublic/ant/src/main/org/apache/tools/ant/taskdefs/optional/j2ee/AbstractHotDeploymentTool.java,v
retrieving revision 1.8
diff -u -r1.8 AbstractHotDeploymentTool.java
--- 
src/main/org/apache/tools/ant/taskdefs/optional/j2ee/AbstractHotDeploymentTool.java
 19 Jul 2003 08:11:04 -  1.8
+++ 
src/main/org/apache/tools/ant/taskdefs/optional/j2ee/AbstractHotDeploymentTool.java
 19 Sep 2003 21:15:52 -
@@ -1,7 +1,7 @@
 /*
  * The Apache Software License, Version 1.1
  *
- * Copyright (c) 2002 The Apache Software Foundation.  All rights
+ * Copyright (c) 2002-2003 The Apache Software Foundation.  All rights
  * reserved.
  *
  * Redistribution and use in source and binary forms, with or without
@@ -101,9 +101,9 @@
  *  @return A Path object representing the classpath to be used.
  */
 public Path createClasspath() {
-if (classpath == null)
+if (classpath == null) {
 classpath = new Path(task.getProject());
-
+}
 return classpath.createPath();
 }
 
@@ -126,14 +126,17 @@
  *  @exception org.apache.tools.ant.BuildException if the attributes are 
invalid or incomplete.
  */
 public void validateAttributes() throws BuildException {
-if (task.getAction() == null)
+if (task.getAction() == null) {
 throw new BuildException("The \"action\" attribute must be set");
+}
 
-if (!isActionValid())
+if (!isActionValid()) {
 throw new BuildException("Invalid action \"" + task.getAction() + 
"\" passed");
+}
 
-if (classpath == null)
+if (classpath == null) {
 throw new BuildException("The classpath attribute must be set");
+}
 }
 
 /**
Index: 
src/main/org/apache/tools/ant/taskdefs/optional/extension/resolvers/AntResolver.java
===
RCS file: 
/home/cvspublic/ant/src/main/org/apache/tools/ant/taskdefs/optional/extension/resolvers/AntResolver.java,v
retrieving revision 1.5
diff -u -r1.5 AntResolver.java
--- 
src/main/org/apache/tools/ant/taskdefs/optional/extension/resolvers/AntResolver.java
19 Jul 2003 08:11:04 -  1.5
+++ 
src/main/org/apache/tools/ant/taskdefs/optional/extension/resolvers/AntResolver.java
19 Sep 2003 21:22:05 -
@@ -1,7 +1,7 @@
 /*
  * The Apache Software License, Version 1.1
  *
- * Copyright (c) 2002 The Apache Software Foundation.  All rights
+ * Copyright (c) 2002-2003 The Apache Software Foundation.  All rights
  * reserved.
  *
  * Redistribution and use in source and binary forms, with or without
@@ -99,8 +99,7 @@
 final File dir =
 m_antfile.getParentFile().getCanonicalFile();
 ant.setDir(dir);
-}
-catch (final IOException ioe) {
+} catch (final IOException ioe) {
 throw new BuildException(ioe.getMessage(), ioe);
 }
 
Index: src/main/org/apache/tools/ant/types/resolver/ApacheCatalogResolver.java
===
RCS file: 
/home/cvspublic/ant/src/main/org/apache/tools/ant/types/resolver/ApacheCatalogResolver.java,v
retrieving revision 1.11
diff -u -r1.11 ApacheCatalogResolver.java
--- src/main/org/apache/tools/ant/types/resolver/ApacheCatalogResolver.java 
19 Jul 2003 11:20:22 -  1.11
+++ src/main/org/apache/tools/ant/types/resolver/ApacheCatalogResolver.java 
19 Sep 2003 21:23:52 -
@@ -145,11 +145,9 @@
 
 try {
 catalog.parseCatalog(file);
-}
-catch (MalformedURLException ex) {
+} catch (MalformedURLException ex) {
 throw new BuildException(ex);
-}
-catch (IOException ex) {
+} catch (IOException ex) {
 throw new BuildException(ex);
 }
 }
Index: src/main/org/apache/tools/ant/types/Assertions.java
===
RCS file: 
/home/cvspublic/ant/src/main/org/apache/tools/ant/types/Assertions.java,v
retrieving revision 1.5
diff -u -r1.5 Assertions.java
--- src/main/org/apache/tools/ant/types/Assertions.java 1 Aug 2003 06:44:35 
-   1.5
+++ src/main/org/apache/tools/ant/types/Assertions.java 19 Sep 2003 21:33:31 
-
@@ -187,7 +187,7 @@
  * @return
  */
 private int getFinalSize() {
-return assertionList.size()+ (enableSystemAssertions!=null?1:0);
+return assertionList.size() + (enableSystemAssertions != null ? 1 : 0);
 }
 
 /**
@@ -254,16 +254,16 @@
  * @throws CloneNotSupportedException
  */
 protected Object clone() throws CloneNotSupportedException {
-Assertions that=(Assertions) super.clone();
-   

Re: [patch] Having execute a nested if the target is out of date

[Martijn Kruithof]

Hi,
If you make something like this (I don't see a direct benefit), why not 
make uptodate an taskcontainer so that you can wrap any task in there 
without the need of indirection via the sequential task.

Kind regards, Martijn.
Paul Mclachlan wrote:
Basically, I've added the capability for  to have a nested 
 task that only gets executed if the targets are out of 
date.  At present I have to create a whole other  to do the 
dependency check so that the primary target and have an 
unless="blah.uptodate".

This would make my ant script have fewer 'unnecessary' targets & seems 
like a good idea.  (To me, anyway).

- Paul

Index: docs/manual/CoreTasks/uptodate.html
===
RCS file: /home/cvspublic/ant/docs/manual/CoreTasks/uptodate.html,v
retrieving revision 1.11
diff -u -r1.11 uptodate.html
--- docs/manual/CoreTasks/uptodate.html	1 Jun 2002 12:26:33 -	1.11
+++ docs/manual/CoreTasks/uptodate.html	19 Sep 2003 20:11:48 -
@@ -37,7 +37,8 @@
   
 property
 The name of the property to set.
-Yes
+Yes, unless a nested 
+   is present.
   
   
 value
@@ -70,6 +71,10 @@
 The nested  element allows you to specify
 a set of target files to check for being up-to-date with respect to a
 set of source files.
+
+sequential
+The nested  element allows you to specify a
+set of tasks to run when the source files are out of date.
 
 Examples
   
Index: src/main/org/apache/tools/ant/taskdefs/UpToDate.java
===
RCS file: /home/cvspublic/ant/src/main/org/apache/tools/ant/taskdefs/UpToDate.java,v
retrieving revision 1.32
diff -u -r1.32 UpToDate.java
--- src/main/org/apache/tools/ant/taskdefs/UpToDate.java	13 Aug 2003 14:46:15 -	1.32
+++ src/main/org/apache/tools/ant/taskdefs/UpToDate.java	19 Sep 2003 20:12:07 -
@@ -91,6 +91,7 @@
 private File _sourceFile;
 private File _targetFile;
 private Vector sourceFileSets = new Vector();
+private Sequential _sequential = null;
 
 protected Mapper mapperElement = null;
 
@@ -218,14 +219,21 @@
 return upToDate;
 }
 
+public void addSequential( Sequential s ) {
+if( _sequential != null ) {
+throw new BuildException( "Only one nested  permitted", getLocation() );
+}
+
+_sequential = s;
+}
 
 /**
  * Sets property to true if target file(s) have a more recent timestamp
  * than (each of) the corresponding source file(s).
  */
 public void execute() throws BuildException {
-if (_property == null) {
-throw new BuildException("property attribute is required.",
+if (_property == null && _sequential == null) {
+throw new BuildException("property attribute or nested  required.",
  getLocation());
 }
 boolean upToDate = eval();
@@ -237,6 +245,11 @@
 } else {
 log("All target files are up-to-date.",
 Project.MSG_VERBOSE);
+}
+}
+else {
+if (_sequential != null) {
+_sequential.perform();
 }
 }
 }



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

[patch] Having execute a nested if the target is out of date

[Paul Mclachlan]

Basically, I've added the capability for  to have a nested 
 task that only gets executed if the targets are out of 
date.  At present I have to create a whole other  to do the 
dependency check so that the primary target and have an 
unless="blah.uptodate".

This would make my ant script have fewer 'unnecessary' targets & seems 
like a good idea.  (To me, anyway).

- Paul
Index: docs/manual/CoreTasks/uptodate.html
===
RCS file: /home/cvspublic/ant/docs/manual/CoreTasks/uptodate.html,v
retrieving revision 1.11
diff -u -r1.11 uptodate.html
--- docs/manual/CoreTasks/uptodate.html 1 Jun 2002 12:26:33 -   1.11
+++ docs/manual/CoreTasks/uptodate.html 19 Sep 2003 20:11:48 -
@@ -37,7 +37,8 @@
   
 property
 The name of the property to set.
-Yes
+Yes, unless a nested 
+   is present.
   
   
 value
@@ -70,6 +71,10 @@
 The nested  element allows you to specify
 a set of target files to check for being up-to-date with respect to a
 set of source files.
+
+sequential
+The nested  element allows you to specify a
+set of tasks to run when the source files are out of date.
 
 Examples
   
Index: src/main/org/apache/tools/ant/taskdefs/UpToDate.java
===
RCS file: 
/home/cvspublic/ant/src/main/org/apache/tools/ant/taskdefs/UpToDate.java,v
retrieving revision 1.32
diff -u -r1.32 UpToDate.java
--- src/main/org/apache/tools/ant/taskdefs/UpToDate.java13 Aug 2003 
14:46:15 -  1.32
+++ src/main/org/apache/tools/ant/taskdefs/UpToDate.java19 Sep 2003 
20:12:07 -
@@ -91,6 +91,7 @@
 private File _sourceFile;
 private File _targetFile;
 private Vector sourceFileSets = new Vector();
+private Sequential _sequential = null;
 
 protected Mapper mapperElement = null;
 
@@ -218,14 +219,21 @@
 return upToDate;
 }
 
+public void addSequential( Sequential s ) {
+if( _sequential != null ) {
+throw new BuildException( "Only one nested  
permitted", getLocation() );
+}
+
+_sequential = s;
+}
 
 /**
  * Sets property to true if target file(s) have a more recent timestamp
  * than (each of) the corresponding source file(s).
  */
 public void execute() throws BuildException {
-if (_property == null) {
-throw new BuildException("property attribute is required.",
+if (_property == null && _sequential == null) {
+throw new BuildException("property attribute or nested 
 required.",
  getLocation());
 }
 boolean upToDate = eval();
@@ -237,6 +245,11 @@
 } else {
 log("All target files are up-to-date.",
 Project.MSG_VERBOSE);
+}
+}
+else {
+if (_sequential != null) {
+_sequential.perform();
 }
 }
 }

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Volunteering: ejclientjar task missing? I want to write it

[Per Olesen]

Hi,

I've often missed a task for generating a ejb clientjar. Ant already includes 
a "ejbjar" task, but - as far as I know - no way of generating a ejb 
clientjar from a generic ejbjar as input/source.

I hereby volunteer to write one myself !

I've read the guidelines on the web for how to contribute etc., but I'm not 
clear precisely how to start this up.

Should I start by sending a description on this list of how I would implement 
it (how it should work) or should I make a bugzilla report as a request for 
enhancement and then grab it myself (may I even grab an issue? I have not 
contributed to ant before).

I can also just write it like I would like it to work, and then submit the 
patch for it. But then I first get your valuable input after I've done the 
implementation.

Also: Should I develop against 1.6? I can see that there is a proposal for 
starting a release-plan for 1.6. Might not make it into 1.6 then. Is there a 
2.0 on the way? (there is a 2.0 name in bugzilla).

Regards, Per Olesen

-- 
Per Olesen @ Nordija A/S - www.nordija.com - main#: +45 70 20 25 10
email: [EMAIL PROTECTED] - cell#: +45 23 38 95 81


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 21424] - stcheckout task fails to obey "deleteuncontrolled attribute, leaves errant subdirectories in basedir

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=21424

stcheckout task fails to obey "deleteuncontrolled attribute, leaves errant 
subdirectories in basedir





--- Additional Comments From [EMAIL PROTECTED]  2003-09-19 18:11 ---
Just a friendly ping to see how this issue with be dealt with in 1.6.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 23278] New: - Project.setName() does not replaceProperties()

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23278

Project.setName() does not replaceProperties()

   Summary: Project.setName() does not replaceProperties()
   Product: Ant
   Version: 1.5
  Platform: All
OS/Version: Other
Status: UNCONFIRMED
  Severity: Minor
  Priority: Other
 Component: Core
AssignedTo: [EMAIL PROTECTED]
ReportedBy: [EMAIL PROTECTED]


Scenario:  pass ant commandline parameter "-propertyfile "

property file contains:
foo = bar

build.xml


At the time that Project.setName() is invoked by ProjectHelperImpl, the 
project's properties hashtable has already been initialized and the specified 
propertyfile has been loaded.

Despite this, the project name property contains "${foo}" instead of "bar", as 
it would, if Project.replaceProperties() was called.

Query: should this also be applied to Project.setDefault() as well?

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 23272] - CVS task over ssh blocks execution

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23272

CVS task over ssh blocks execution

[EMAIL PROTECTED] changed:

   What|Removed |Added

 Status|NEW |RESOLVED
 Resolution||INVALID



--- Additional Comments From [EMAIL PROTECTED]  2003-09-19 15:56 ---

With the following combination of programs:
Cygwin ssh
Cygwin cvs
ant 1.5.4

I got the expected result. It seems from this that there is no problem with Ant.
In fact, the only thing ant does is call Process.waitFor(), so if that had been
a problem it would have been with the JRE itself. Given that with a particular
set of external apps this is not a problem, I think that I can assume this is a
bug in the command line client that comes with WinCVS 1.3.8.1.

I will go hunt over there for more info on the problem from their website, and
report this bug there.

Sorry to have bothered you all.

Dave.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 23273] New: - Java task interprets it's arguments incorrectly if terminated with \\

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23273

Java task interprets it's arguments incorrectly if terminated with \\

   Summary: Java task interprets it's arguments incorrectly if
terminated with \\
   Product: Ant
   Version: 1.5.4
  Platform: PC
OS/Version: Windows NT/2K
Status: NEW
  Severity: Normal
  Priority: Other
 Component: Core tasks
AssignedTo: [EMAIL PROTECTED]
ReportedBy: [EMAIL PROTECTED]


Consider the following ANT task :





 
Which is used to invoke the following Java class :

public final class MyClass
{
public static void main( String[] args )
{
int loop = -1;
while ( ++loop != args.length )
{
System.out.println ( loop + ")\t" + args[ loop ] );
}
}
}
 
 
 
The expected output would be:
 
 [java] 0)  Some text\\
 [java] 1)  Second argument
 
 
What actually happens is
 
 [java] 0)  Some text\" Second
 [java] 1)  argument
 
  
 
The first argument has a trailing double back-slash character. This has been 
replaced by a single back slash and a double quote character.

ANT has then broken the second argument into two parts, as separated by a space 
character. The first part of this argument has been concatenated with the first 
argument! If there had been no space character in the second argument, then it 
would have been completely concatenated with the first argument.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 23268] - Building MANIFEST-only jar

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23268

Building MANIFEST-only jar

[EMAIL PROTECTED] changed:

   What|Removed |Added

 Status|NEW |RESOLVED
 Resolution||INVALID



--- Additional Comments From [EMAIL PROTECTED]  2003-09-19 14:36 ---
I am closing this bug, as I strongly believe this is some error in the build
file rather than a bug in ant.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 23272] - CVS task over ssh blocks execution

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23272

CVS task over ssh blocks execution





--- Additional Comments From [EMAIL PROTECTED]  2003-09-19 14:35 ---

With a local CVSROOT, there are no issues.

Attached is a run of ant -f build_test.xml -v with CVSROOT=c:\java\cvs. CVS was
run with "cvs -t".

C:\java\build_env>ant -v -f build_test.xml
Apache Ant version 1.5.3 compiled on April 16 2003
Buildfile: build_test.xml
Detected Java version: 1.3 in: c:\java\jdk1.3\jre
Detected OS: Windows 2000
parsing buildfile build_test.xml with URI = file:C:/java/build_env/build_test.xm
l
Project base dir set to: C:\java\build_env
 [property] Loading Environment env.
Build sequence for target `all' is [all]
Complete build sequence is [all]

all:
  [cvs] Executing 'cvs' with arguments:
  [cvs] '-dc:\java\cvs'
  [cvs] '-t'
  [cvs] 'co'
  [cvs] 'test1'
  [cvs]
  [cvs] The ' characters around the executable and arguments are
  [cvs] not part of the command.
  [cvs]
  [cvs]
  [cvs] environment:

  [cvs]
  [cvs] ANT_CMD_LINE_ARGS=-v -f build_test.xml
  [cvs] ANT_HOME=c:\java\apache-ant-1.5.3-1
  [cvs] CVSROOT=c:\java\cvs
  [cvs] JAVA_HOME=c:\java\jdk1.3
  [cvs] OS=Windows_NT
  [cvs]  -> main loop with CVSROOT=c:\java\cvs
  [cvs]  -> do_module (test1, Updating, , )
  [cvs]  -> Create_Admin (., test1, c:\java\cvs/test1, , , 0, 0)
  [cvs]  -> unlink(./CVS/Tag)
  [cvs]  <- Create_Admin
  [cvs]  -> fopen(c:\java\cvs/CVSROOT/history,a)
  [cvs]  -> unlink(./CVS/Entries.Static)
  [cvs] cvs checkout: Updating test1
  [cvs]  -> unlink_file_dir(CVS/,,a)
  [cvs]  -> checkout (c:\java\cvs/test1/a,v, 1.1.1.1, , a)
  [cvs]  -> chmod(a,100666)
  [cvs] U test1/a
  [cvs]  -> Register(a, 1.1.1.1, Fri Sep 19 13:24:26 2003, ,  )
  [cvs]  -> rename(CVS/Entries.Backup,CVS/Entries)
  [cvs]  -> unlink(CVS/Entries.Log)
  [cvs] Executing 'cvs' with arguments:
  [cvs] '-dc:\java\cvs'
  [cvs] '-t'
  [cvs] 'co'
  [cvs] 'test2'
  [cvs]
  [cvs] The ' characters around the executable and arguments are
  [cvs] not part of the command.
  [cvs]
  [cvs]
  [cvs] environment:
  [cvs]

  [cvs] ANT_CMD_LINE_ARGS=-v -f build_test.xml
  [cvs] ANT_HOME=c:\java\apache-ant-1.5.3-1
  [cvs] CVSROOT=c:\java\cvs
  [cvs] JAVA_HOME=c:\java\jdk1.3
  [cvs]  -> main loop with CVSROOT=c:\java\cvs
  [cvs]  -> do_module (test2, Updating, , )
  [cvs]  -> Create_Admin (., test2, c:\java\cvs/test2, , , 0, 0)
  [cvs]  -> unlink(./CVS/Tag)
  [cvs]  <- Create_Admin
  [cvs]  -> fopen(c:\java\cvs/CVSROOT/history,a)
  [cvs]  -> unlink(./CVS/Entries.Static)
  [cvs] cvs checkout: Updating test2
  [cvs]  -> unlink_file_dir(CVS/,,b)
  [cvs]  -> checkout (c:\java\cvs/test2/b,v, 1.1.1.1, , b)
  [cvs]  -> chmod(b,100666)
  [cvs]  -> Register(b, 1.1.1.1, Fri Sep 19 13:24:34 2003, ,  )
  [cvs] U test2/b
  [cvs]  -> rename(CVS/Entries.Backup,CVS/Entries)
  [cvs]  -> unlink(CVS/Entries.Log)

BUILD SUCCESSFUL
Total time: 6 seconds

I believe this confirms that there is a problem in the CVS clients ssh handling.
 But again, when run from the command line, all works as expected, so perhaps
cvs is doing what it should, and ant is not correctly interpreting what's
happenning? I really don't know enough about it to say.

Any advice is appreciated. If this is definitely a CVS bug, then I will go fight
with them :)

Dave.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 23272] - CVS task over ssh blocks execution

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23272

CVS task over ssh blocks execution





--- Additional Comments From [EMAIL PROTECTED]  2003-09-19 14:24 ---

I have confirmed that this also happens with CVS_RSH=ssh (cygwin port of
openssh) as opposed to plink. This implies to me that it is not plink which
causes trouble, but cvs or ant.

I will try a similar test on a local repository to see what happens.

Dave.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 23272] - CVS task over ssh blocks execution

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23272

CVS task over ssh blocks execution





--- Additional Comments From [EMAIL PROTECTED]  2003-09-19 13:50 ---
Some additional information: 
ant 1.5.3
WinCVS 1.3.8.1
cvs 1.10 (bundled with above)
putty 0.53

Ant output with ant -v and cvs -t:

C:\java\build_env>ant -v -f build_test.xml
Apache Ant version 1.5.3 compiled on April 16 2003
Buildfile: build_test.xml
Detected Java version: 1.3 in: c:\java\jdk1.3\jre
Detected OS: Windows 2000
parsing buildfile build_test.xml with URI = 
file:C:/java/build_env/build_test.xml
Project base dir set to: C:\java\build_env
 [property] Loading Environment env.
Build sequence for target `all' is [all]
Complete build sequence is [all]

all:
  [cvs] Executing 'cvs' with arguments:
  [cvs] '-d:ext:[EMAIL PROTECTED]:/home/david/cvs'
  [cvs] '-t'
  [cvs] 'co'
  [cvs] 'test1'
  [cvs]
  [cvs] The ' characters around the executable and arguments are
  [cvs] not part of the command.
  [cvs]
  [cvs]
  [cvs] environment:

  [cvs]
  [cvs] ANT_CMD_LINE_ARGS=-v -f build_test.xml
  [cvs] ANT_HOME=c:\java\apache-ant-1.5.3-1
  [cvs] CVSROOT=:ext:[EMAIL PROTECTED]:/home/david/cvs
  [cvs] JAVA_HOME=c:\java\jdk1.3
  [cvs] OS=Windows_NT
  [cvs] Os2LibPath=C:\WINNT\system32\os2\dll;
  [cvs] CVS_RSH=plink
  [cvs]  -> main loop with CVSROOT=:ext:[EMAIL PROTECTED]:/home/david/cvs
  [cvs] S-> do_module (test1, Updating, , )
  [cvs]  -> rename(CVS/Entries.Backup,CVS/Entries)
  [cvs]  -> unlink(CVS/Entries.Log)
  [cvs] S-> do_module (test1, Updating, , )
  [cvs]  -> unlink(CVS/Entries.Static)
  [cvs] cvs server: Updating test1
  [cvs] S-> fopen(/home/david/cvs/CVSROOT/history,a)
  [cvs] S-> unlink_file(./CVS/Entries.Static)
  [cvs] S-> Reader_Lock(/home/david/cvs/test1)
  [cvs] S-> Lock_Cleanup()
  [cvs] S-> rename(CVS/Entries.Backup,CVS/Entries)
  [cvs] S-> unlink_file(CVS/Entries.Log)

Stalls here.

Thanks for the help,
Dave.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 23272] - CVS task over ssh blocks execution

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23272

CVS task over ssh blocks execution





--- Additional Comments From [EMAIL PROTECTED]  2003-09-19 13:41 ---
Created an attachment (id=8291)
CVS repository used with the test

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 23272] - CVS task over ssh blocks execution

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23272

CVS task over ssh blocks execution





--- Additional Comments From [EMAIL PROTECTED]  2003-09-19 13:41 ---
Created an attachment (id=8290)
Ant file exhibiting unexpected behaviour

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 23272] New: - CVS task over ssh blocks execution

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23272

CVS task over ssh blocks execution

   Summary: CVS task over ssh blocks execution
   Product: Ant
   Version: 1.5.3
  Platform: PC
OS/Version: Windows NT/2K
Status: NEW
  Severity: Normal
  Priority: Other
 Component: Core tasks
AssignedTo: [EMAIL PROTECTED]
ReportedBy: [EMAIL PROTECTED]


Hi,

Attaching a bunch of stuff to help reproduce this...

When I use ant to run a number of checkouts and build tasks, all is well on
linux (as usual), but on windows, the cvs tasks seem to block the operation of
the rest of the target.

What I mean by that is, when using the ext access method in CVS, with plink as
the ssh bridge, with this very simple build.xml (attached as well for easy
downloading):



  ANT file showing problems with ant/wincvs/plink

  

  



  



I get the following result:
C:\java\build_env>ant -f build_test.xml
Buildfile: build_test.xml

all:
  [cvs] cvs server: Updating test1
  [cvs] U test1/a


And after several minutes, I kill the ant instance.

Relevant environment variables, for information:
CVSROOT=:ext:[EMAIL PROTECTED]:/path/to/repos
CVS_RSH=plink

ANT_HOME, JAVA_HOME et al are all set to reasonable values.

I am also attaching the tarred up repository I was using. You will see it really
is just a dummy CVS repository. This script runs as expected (that is, both
checkouts proceed to completion, and we get the command line back) on Linux.
Also, when run from the command line, the cvs co test1 and tes2 work fine, and
return control to the command line when finished. 

I will attach some more verbose output with cvs -t and ant -v in a minute.

Thanks,
Dave.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: new Launcher and CLASSPATH

[Stefan Bodewig]

On Fri, 19 Sep 2003, Antoine Lévy-Lambert <[EMAIL PROTECTED]>
wrote:
>Von: Stefan Bodewig [mailto:[EMAIL PROTECTED]

>>One of the reasons for the new launcher has been a command line
>>length limitation.  With this approach the command line would even
>>become longer, I'm afraid.
> 
> This is just adding 5 characters

OK, we could live with that, I guess.

>>The new Launcher doesn't solve that, but that we are now splitting
>>up optional.jar makes solution (2) a lot easier.
>>
> Yes, this is what I am talking about. Can you explain how the
> splitting of optional.jar will make the solution easier ?

If we take the  task as an example (as I'm more familiar with
it):  remove ant-junit.jar from ANT_HOME/lib and then use


  


  


before splitting up optional.jar you had to remove all optional tasks
from ANT_HOME/lib or fiddle with the jar.

> If you have this scenario :
> 
> - custom task
> - java runtime class or interface called by the custom task (example
> javax.naming.Context)
> - other library (JNDI driver)
> 
> then the other library has to be in the CLASSPATH.

or the custom task is neither on the CLASSPATH nor in ANT_HOME/lib.  I
think this still holds true.  Just that with the new Launcher, your
custom task won't work if it is on the CLASSPATH anyway.

Stefan

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

AW: new Launcher and CLASSPATH

[Antoine Lévy-Lambert]

>Von: Stefan Bodewig [mailto:[EMAIL PROTECTED]
>Gesendet: Freitag, 19. September 2003 14:49
>An: [EMAIL PROTECTED]
>Betreff: Re: new Launcher and CLASSPATH
>
>
>On Fri, 19 Sep 2003, Antoine Lévy-Lambert <[EMAIL PROTECTED]>
>wrote:
>
>> I am a bit concerned about the fact that CLASSPATH is ignored by the
>> new launcher.
>
>But the old launcher (Main) will still work the same way.
>
>> There are lots of scripts which invoke ant as script, and set
>> CLASSPATH to include different libraries.
>
>You mean ant the wrapper script or Ant the Java program.  In the later
>case, they'll use Main as their entry point and nothing is going to
>break.  The former case has to be strongly documented.
>
I am meaning ant as wrapper script.


>> 1) change the ant startup script
>> - do not include CLASSPATH in LOCALCLASSPATH
>> - pass the CLASSPATH as suggested next to a -lib flag
>
>One of the reasons for the new launcher has been a command line length
>limitation.  With this approach the command line would even become
>longer, I'm afraid.

No the size of the command line would remain the same

the current command line is

java -cp $CLASSPATH:ant-launcher.jar org.apache.tools.launch.Launcher ...

I want to change it to :

java -cp ant-launcher.jar org.apache.tools.launch.Launcher -lib $CLASSPATH
...

This is just adding 5 characters


>
>> 3) testcases : would using the  task with fork set to true
>> and main class set to Launcher allow us to test that
>> build.sysclasspath behaves as documented ?
>
>To be honest, it never has.
>
>build.sysclasspath=only has always included the jars from ANT_HOME/lib
>if you used the wrapper script (as the script added them to the
>CLASSPATH), so it has always added things to the CLASSPATH.  The new
>Launcher will still do so.

Then we need to change our documentation. I will try to write the testcases,
it is an interesting exercise.

>
>> 4) does the launcher solve the problem that ant 1.5 was requiring in
>> CLASSPATH some libraries which are loaded by the java runtime.
>
>I assume you mean something like
>.
>
>The new Launcher doesn't solve that, but that we are now splitting up
>optional.jar makes solution (2) a lot easier.
>
Yes, this is what I am talking about. Can you explain how the splitting of
optional.jar will make the solution easier ?

If you have this scenario :

- custom task
- java runtime class or interface called by the custom task (example
javax.naming.Context)
- other library (JNDI driver)

then the other library has to be in the CLASSPATH.

Cheers,

Antoine



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 22865] - Zip whenempty attribute is broken

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=22865

Zip whenempty attribute is broken





--- Additional Comments From [EMAIL PROTECTED]  2003-09-19 13:16 ---
I haven't been able to test this yet but I will get around to it sometime over 
the next few days.  (I think) I was using version 1.5.4 when I tried this last.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: new Launcher and CLASSPATH

[Stefan Bodewig]

On Fri, 19 Sep 2003, Antoine Lévy-Lambert <[EMAIL PROTECTED]>
wrote:

> I am a bit concerned about the fact that CLASSPATH is ignored by the
> new launcher.

But the old launcher (Main) will still work the same way.

> There are lots of scripts which invoke ant as script, and set
> CLASSPATH to include different libraries.

You mean ant the wrapper script or Ant the Java program.  In the later
case, they'll use Main as their entry point and nothing is going to
break.  The former case has to be strongly documented.

> 1) change the ant startup script
> - do not include CLASSPATH in LOCALCLASSPATH
> - pass the CLASSPATH as suggested next to a -lib flag

One of the reasons for the new launcher has been a command line length
limitation.  With this approach the command line would even become
longer, I'm afraid.

> 3) testcases : would using the  task with fork set to true
> and main class set to Launcher allow us to test that
> build.sysclasspath behaves as documented ?

To be honest, it never has.

build.sysclasspath=only has always included the jars from ANT_HOME/lib
if you used the wrapper script (as the script added them to the
CLASSPATH), so it has always added things to the CLASSPATH.  The new
Launcher will still do so.

> 4) does the launcher solve the problem that ant 1.5 was requiring in
> CLASSPATH some libraries which are loaded by the java runtime.

I assume you mean something like
.

The new Launcher doesn't solve that, but that we are now splitting up
optional.jar makes solution (2) a lot easier.

Stefan

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

cvs commit: ant/docs external.html

[jhm]

jhm 2003/09/19 05:44:15

  Modified:docs external.html
  Log:
  Add Simian Task (used in check.xml). Information by Simon Harris, Redhill 
Consulting.
  
  Revision  ChangesPath
  1.136 +53 -2 ant/docs/external.html
  
  Index: external.html
  ===
  RCS file: /home/cvs/ant/docs/external.html,v
  retrieving revision 1.135
  retrieving revision 1.136
  diff -u -r1.135 -r1.136
  --- external.html 17 Sep 2003 11:29:20 -  1.135
  +++ external.html 19 Sep 2003 12:44:15 -  1.136
  @@ -2543,7 +2543,7 @@
 
 
  -  http://proguard.sourceforge.net/";>http://proguard.sourceforge.net/ 
  +  http://proguard.sourceforge.net/";>http://proguard.sourceforge.net/
   
 
 
  @@ -2555,7 +2555,7 @@
 
 
  -  http://proguard.sourceforge.net/feedback.html";>Feedback Page 
  +  http://proguard.sourceforge.net/feedback.html";>Feedback Page
   
 
 
  @@ -2717,6 +2717,57 @@
 
 Apache Software License
  +  
  +  
  +
  +
  +
  +Simian
  +  
  +Simian (Similarity Analyser) identifies 
duplication in Java,
  +C#, C, CPP, COBOL, JSP, HTML source code and even plain text 
files.
  +  
  +  
  +  
  +  Compatibility:
  +  
  +  
  +  Ant 1.5.1 and later
  +  
  +  
  +  
  +  
  +  URL:
  +  
  +  
  +  http://www.redhillconsulting.com.au/products/simian/";>
  +http://www.redhillconsulting.com.au/products/simian/
  +  
  +  
  +  
  +  
  +  Contact:
  +  
  +  
  +  mailto:[EMAIL PROTECTED]">
  +[EMAIL PROTECTED] (User Mailinglist)
  +
  +  
  +  
  +  
  +  
  +  License:
  +  
  +  
  +  Commercial, Free Licenses available for Non-Commercial Projects
 
 
   
  
  
  

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

cvs commit: ant/xdocs external.xml

[jhm]

jhm 2003/09/19 05:43:49

  Modified:xdocsexternal.xml
  Log:
  Add Simian Task (used in check.xml). Information by Simon Harris, Redhill 
Consulting.
  
  Revision  ChangesPath
  1.100 +36 -8 ant/xdocs/external.xml
  
  Index: external.xml
  ===
  RCS file: /home/cvs/ant/xdocs/external.xml,v
  retrieving revision 1.99
  retrieving revision 1.100
  diff -u -r1.99 -r1.100
  --- external.xml  17 Sep 2003 11:29:20 -  1.99
  +++ external.xml  19 Sep 2003 12:43:49 -  1.100
  @@ -1333,24 +1333,24 @@
   
   
 
  -Compatibility: 
  -Tested with 1.5. Should work with all versions. 
  +Compatibility:
  +Tested with 1.5. Should work with all versions.
 
 
  -URL: 
  +URL:
   
  -  http://proguard.sourceforge.net/";>http://proguard.sourceforge.net/ 
  +  http://proguard.sourceforge.net/";>http://proguard.sourceforge.net/
   
 
 
  -Contact: 
  +Contact:
   
  -  http://proguard.sourceforge.net/feedback.html";>Feedback Page 
  +  http://proguard.sourceforge.net/feedback.html";>Feedback Page
   
 
 
  -License: 
  -GNU General Public License 
  +License:
  +GNU General Public License
 
   
 
  @@ -1430,6 +1430,34 @@
 
   License:
   Apache Software License
  +  
  +
  +  
  +
  +  
  +
  +Simian (Similarity Analyser) identifies duplication in Java,
  +C#, C, CPP, COBOL, JSP, HTML source code and even plain text 
files.
  +
  +
  +  
  +Compatibility:
  +Ant 1.5.1 and later
  +  
  +  
  +URL:
  +http://www.redhillconsulting.com.au/products/simian/";>
  +http://www.redhillconsulting.com.au/products/simian/
  +  
  +  
  +Contact:
  +mailto:[EMAIL PROTECTED]">
  +[EMAIL PROTECTED] (User Mailinglist)
  +
  +  
  +  
  +License:
  +Commercial, Free Licenses available for Non-Commercial 
Projects
 
   
 
  
  
  

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

AW: new Launcher and CLASSPATH

[Antoine Lévy-Lambert]

Hi,

I am a bit concerned about the fact that CLASSPATH is ignored by the new
launcher.
This is a breach of behavioral compatibility.
There are lots of scripts which invoke ant as script, and set CLASSPATH to
include different libraries.

If it is possible to set the extra libraries needed by the users in Main, I
have an idea of a workaround :

1) change the ant startup script
- do not include CLASSPATH in LOCALCLASSPATH
- pass the CLASSPATH as suggested next to a -lib flag

2) process the path passed in the -lib flag in Main

Would this solution be OK ?


3) testcases : would using the  task with fork set to true and main
class set to Launcher allow us to test that build.sysclasspath behaves as
documented ?

4) does the launcher solve the problem that ant 1.5 was requiring in
CLASSPATH some libraries which are loaded by the java runtime. For instance,
with ant 1.5, a custom task using JNDI (javax.naming.Context) requires the
JNDI drivers in the CLASSPATH before ant gets started. Is this problem
supposed to be fixed in ant 1.6 (with the custom task working if the JNDI
drivers are in the nested classpath element of taskdef, rather than in
CLASSPATH ?)

Antoine

-Ursprüngliche Nachricht-
Von: Conor MacNeill [mailto:[EMAIL PROTECTED]
Gesendet: Donnerstag, 18. September 2003 15:37
An: Ant Developers List
Betreff: Re: new Launcher and CLASSPATH


On Wed, 17 Sep 2003 08:29 pm, Stefan Bodewig wrote:
> So far all our documentation more or less states that putting your jar
> containing custom tasks into ANT_HOME/lib or pointing to it in your
> CLASSPATH was equivalent.  This is no longer true if the new launcher
> is used.

True.

>
> As ant.jar cannot be found on the system classloader anymore, any
> class loaded from the system classloader (i.e. classes coming from
> CLASSPATH) can't load the Task class for example.
>
> I see two options:
>
> (1) live with it and document it properly (which means people may have
> to link/copy a lot into ANT_HOME/lib).

Yes. I think we should recommend that Ant be run with an empty CLASSPATH. I
have added an option ot read libs out of a home directory and would like to
add an option to specify arbitrary lib directories with a -lib option or
similar. The only drag is that the command processing would need to be done
in the launcher to extract this info and then pass the rest of the command
line to the Main class.

>
> (2) Fix all tasks that load classes - like Definer - to delegate to
> the classloader created in Launcher instead of the system classloader.

I don't think this will work. No matter which classloader you use to find
the
class, if it is found on the system class loader it will try to load
dependent classes with the same loader.


>  Or the classloader
> created in Launcher must be the parent of the created AntClassloader
> instances (probably the better solution).

I expect that to be the case anyway.

>
> Also, I'm not sure whether build.sysclasspath is going to work as
> expected when used with the new launcher.  It seems as if the code in
> Launcher that resets the java.class.path system property should take
> care of it, but I'm not entirely sure.  If we wanted to test it via
> Gump (which uses build.sysclasspath=only), we first need to find a
> solution for my first issue above.
>

True - may not be compatible with my empty CLASSPATH recommendation above.

Conor




-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: cvs commit: ant/src/main/org/apache/tools/ant/taskdefs/optional/junit JUnitTask.java

[Stefan Bodewig]

On Fri, 19 Sep 2003, Antoine Lévy-Lambert <[EMAIL PROTECTED]>
wrote:

> I am a bit surprised that you solve this issue
> (classpath/includeantruntime) this way.

Please read the bugzilla report.

 will not pass any
additional classpath at al.  This means the user's CLASSPATH is the
only thing passed to the forked VM.  This is so that classes from
ANT_HOME/lib like the specific Xerces version Ant uses do not taint
the classpath - users may want to test with a different XML parser,
DOM version, whatever.

 is there to "help" the
user.  Ant uses a testrunner of its own, therefore some classes of Ant
itself must be available to the forked VM, namely JUnit itself, the
optional.junit package more or less completely and classes from Ant's
core the junit package depends on.

The code up until now located the necessary resources and passed them
in via java -classpath .  But if you do that, CLASSPATH will be
ignored by the forked VM.  I simply added the CLASSPATH to the command
line explicitly.

> The ant runtime is not in the CLASSPATH environment variable in
> ant1.6,

but in one of the jars added by the task.

> but in the system property java.class.path.

Like Xerces and everything else from ANT_HOME/lib that we explicitly
do not want to pass to the forked VM - that's why I've used the
environment variable directly.

Stefan

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

AW: cvs commit: ant/src/main/org/apache/tools/ant/taskdefs/optional/junit JUnitTask.java

[Antoine Lévy-Lambert]

Stefan,

I am a bit surprised that you solve this issue (classpath/includeantruntime)
this way.
The ant runtime is not in the CLASSPATH environment variable in ant1.6, but
in the system property java.class.path.
Or did I understand something wrongly ?
Antoine

-Ursprüngliche Nachricht-
Von: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
Gesendet: Freitag, 19. September 2003 11:18
An: [EMAIL PROTECTED]
Betreff: cvs commit:
ant/src/main/org/apache/tools/ant/taskdefs/optional/junit JUnitTask.java


bodewig 2003/09/19 02:18:16

  Modified:.WHATSNEW
   src/main/org/apache/tools/ant/taskdefs/optional/junit
JUnitTask.java
  Log:
   replaced the CLASSPATH
  instead of adding to it.

  PR: 14971

  Revision  ChangesPath
  1.503 +3 -0  ant/WHATSNEW

  Index: WHATSNEW
  ===
  RCS file: /home/cvs/ant/WHATSNEW,v
  retrieving revision 1.502
  retrieving revision 1.503
  diff -u -r1.502 -r1.503
  --- WHATSNEW  18 Sep 2003 15:47:47 -  1.502
  +++ WHATSNEW  19 Sep 2003 09:18:16 -  1.503
  @@ -242,6 +242,9 @@
   * The socket condition will now close the socket created to test.
 Bugzilla Report 23040.

  +*  replaced the CLASSPATH
instead
  +  of adding to it.  Bugzilla Report 14971.
  +
   Other changes:
   --
   * All tasks can be used outside of s.  Note that some tasks



  1.83  +12 -1
ant/src/main/org/apache/tools/ant/taskdefs/optional/junit/JUnitTask.java

  Index: JUnitTask.java
  ===
  RCS file:
/home/cvs/ant/src/main/org/apache/tools/ant/taskdefs/optional/junit/JUnitTas
k.java,v
  retrieving revision 1.82
  retrieving revision 1.83
  diff -u -r1.82 -r1.83
  --- JUnitTask.java17 Sep 2003 08:30:30 -  1.82
  +++ JUnitTask.java19 Sep 2003 09:18:16 -  1.83
  @@ -717,6 +717,17 @@
   cmd.createArgument().setValue("haltOnFailure="
 + test.getHaltonfailure());
   if (includeAntRuntime) {
  +Vector v = Execute.getProcEnvironment();
  +Enumeration e = v.elements();
  +while (e.hasMoreElements()) {
  +String s = (String) e.nextElement();
  +if (s.startsWith("CLASSPATH=")) {
  +cmd.createClasspath(getProject()).createPath()
  +.append(new Path(getProject(),
  + s.substring(10 //
"CLASSPATH=".length()
  + )));
  +}
  +}
   log("Implicitly adding " + antRuntimeClasses + " to
CLASSPATH",
   Project.MSG_VERBOSE);
   cmd.createClasspath(getProject()).createPath()




-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 23268] - Building MANIFEST-only jar

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23268

Building MANIFEST-only jar





--- Additional Comments From [EMAIL PROTECTED]  2003-09-19 11:03 ---
1) do you use the update="true" attribute of the jar task ?
2) can you attach a sample of your build file, and the output of ant -verbose so
that we understand what is getting wrong.

Maybe you have badly specified the fileset that you want to update in your
jarfile, and I would guess you have not used the update attribute, or it is set
to false.

Cheers,

Antoine

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 23268] New: - Building MANIFEST-only jar

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23268

Building MANIFEST-only jar

   Summary: Building MANIFEST-only jar
   Product: Ant
   Version: 1.5.4
  Platform: PC
OS/Version: Windows NT/2K
Status: NEW
  Severity: Normal
  Priority: Other
 Component: Build Process
AssignedTo: [EMAIL PROTECTED]
ReportedBy: [EMAIL PROTECTED]


when i try to pach my class files into a jar file using ant the jar file will be
created with a manifest file only. no class files are added. when i execute the
same procedure on the command shell everything works fine. ant produces the
following echo:
Building MANIFEST-only jar

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

cvs commit: ant/docs/manual/CoreTasks import.html

[peterreilly]

peterreilly2003/09/19 02:43:43

  Modified:docs/manual/CoreTasks import.html
  Log:
  state imports are only allowed as top-level tasks
  
  Revision  ChangesPath
  1.6   +21 -14ant/docs/manual/CoreTasks/import.html
  
  Index: import.html
  ===
  RCS file: /home/cvs/ant/docs/manual/CoreTasks/import.html,v
  retrieving revision 1.5
  retrieving revision 1.6
  diff -u -r1.5 -r1.6
  --- import.html   22 Aug 2003 17:32:37 -  1.5
  +++ import.html   19 Sep 2003 09:43:43 -  1.6
  @@ -5,24 +5,31 @@
 Import Task
   
   
  -Import
  -Description
  -Imports another build file into the current project.
  -
  -On execution it will read another Ant file into
  -the same Project. This means that it basically works like the http://ant.apache.org/faq.html#xml-entity-include";>Entity
  -Includes as explained in the Ant FAQ, as if the imported file was
  -contained in the importing file, minus the top 
  -tag.
  -
  +  Import
  +  Description
  +  
  +Imports another build file into the current project.
  +  
  +  
  +On execution it will read another Ant file into
  +the same Project. This means that it basically works like the 
  +http://ant.apache.org/faq.html#xml-entity-include";>Entity
  +  Includes as explained in the Ant FAQ, as if the imported file was
  +contained in the importing file, minus the top 

  +tag.
  +  
  +  
  +The import task may only be used as a top-level task. This means that
  +it may not be used in a target.
  +  
  +  
   There are two further functional aspects that pertain to this task and
  -that are not possible with entity includes:
  +that are not possible with entity includes:
   
 target overriding
  -  special properties
  -  
  +  special properties
   
  +  
   Target overriding
   
   If a target in the main file is also present in at least one of the
  
  
  

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

cvs commit: ant/src/testcases/org/apache/tools/ant/taskdefs ImportTest.java

[peterreilly]

peterreilly2003/09/19 02:36:23

  Modified:src/main/org/apache/tools/ant/taskdefs ImportTask.java
   src/testcases/org/apache/tools/ant/taskdefs ImportTest.java
  Added:   src/etc/testcases/taskdefs/import/subdir
importinsequential-inner.xml importinsequential.xml
  Log:
  Dissallow  within targets
  remove some commented out code in ImportTask
  
  Revision  ChangesPath
  1.1  
ant/src/etc/testcases/taskdefs/import/subdir/importinsequential-inner.xml
  
  Index: importinsequential-inner.xml
  ===
  

  
  

  

  
  
  
  1.1  
ant/src/etc/testcases/taskdefs/import/subdir/importinsequential.xml
  
  Index: importinsequential.xml
  ===
  

  

  
  
  
  
  1.16  +7 -27 
ant/src/main/org/apache/tools/ant/taskdefs/ImportTask.java
  
  Index: ImportTask.java
  ===
  RCS file: 
/home/cvs/ant/src/main/org/apache/tools/ant/taskdefs/ImportTask.java,v
  retrieving revision 1.15
  retrieving revision 1.16
  diff -u -r1.15 -r1.16
  --- ImportTask.java   17 Jul 2003 10:39:07 -  1.15
  +++ ImportTask.java   19 Sep 2003 09:36:22 -  1.16
  @@ -116,20 +116,20 @@
   if (file == null) {
   throw new BuildException("import requires file attribute");
   }
  -
  +if (getOwningTarget() == null
  +|| !"".equals(getOwningTarget().getName())) {
  +throw new BuildException("import only allowed as a top-level 
task");
  +}
  +
   ProjectHelper helper =
   (ProjectHelper) 
getProject().getReference("ant.projectHelper");
   Vector importStack = helper.getImportStack();
  +
   if (importStack.size() == 0) {
   // this happens if ant is used with a project
   // helper that doesn't set the import.
   throw new BuildException("import requires support in 
ProjectHelper");
   }
  -//ProjectHelper2.AntXmlContext context;
  -//
context=(ProjectHelper2.AntXmlContext)project.getReference("ant.parsing.context");
  -
  -//File buildFile=context.buildFile;
  -//File buildFileParent=context.buildFileParent;
   
   if (getLocation() == null || getLocation().getFileName() == null) {
   throw new BuildException("Unable to get location of import 
task");
  @@ -137,7 +137,7 @@
   
   File buildFile = new File(getLocation().getFileName());
   buildFile = new File(buildFile.getAbsolutePath());
  -//System.out.println("Importing from " + currentSource);
  +
   File buildFileParent = new File(buildFile.getParent());
   
   getProject().log("Importing file " + file + " from "
  @@ -164,26 +164,6 @@
   + importedFile + "\n", Project.MSG_WARN);
   return;
   }
  -
  -//// Add parent build file to the map to avoid cycles...
  -//String parentFilename = getPath(buildFile);
  -//if (!context.importedFiles.containsKey(parentFilename)) {
  -//context.importedFiles.put(parentFilename, buildFile);
  -//}
  -//
  -//// Make sure we import the file only once
  -//String importedFilename = getPath(importedFile);
  -//if (context.importedFiles.containsKey(importedFilename)) {
  -//project.log("\nSkipped already imported file:\n   "+
  -//importedFilename+"\n",Project.MSG_WARN);
  -//return;
  -//} else {
  -//context.importedFiles.put(importedFilename, importedFile);
  -//}
  -
  -//context.ignoreProjectTag=true;
  -//context.helper.parse(project, importedFile,
  -//new ProjectHelper2.RootHandler(context));
   
   helper.parse(getProject(), importedFile);
   }
  
  
  
  1.6   +18 -1 
ant/src/testcases/org/apache/tools/ant/taskdefs/ImportTest.java
  
  Index: ImportTest.java
  ===
  RCS file: 
/home/cvs/ant/src/testcases/org/apache/tools/ant/taskdefs/ImportTest.java,v
  retrieving revision 1.5
  retrieving revision 1.6
  diff -u -r1.5 -r1.6
  --- ImportTest.java   18 Sep 2003 09:02:23 -  1.5
  +++ ImportTest.java   19 Sep 2003 09:36:22 -  1.6
  @@ -91,15 +91,32 @@
   "Unnamed2.xmlUnnamed1.xmlSkipped already imported file");
   }
   
  +// allow this as imported in targets are only tested when a target is run
   public void testImportInTargetNoEffect() {
   
configureProject("src/etc/testcases/taskdefs/import/subdir/importintarget.xml");
   expectPropertyUnset("no-import", "foo");
   assertTrue(null == getProject().getReference("baz"));
  

cvs commit: ant/src/testcases/org/apache/tools/ant/filters ConcatFilterTest.java

[jhm]

jhm 2003/09/19 02:24:13

  Modified:docs/manual/CoreTypes filterchain.html
   src/etc/testcases/filters concat.xml
   src/main/org/apache/tools/ant/filters ConcatFilter.java
   src/testcases/org/apache/tools/ant/filters
ConcatFilterTest.java
  Log:
  ConcatFilter: rename attributes after->append and before->prepend; added 
documentation.
  
  Revision  ChangesPath
  1.13  +45 -1 ant/docs/manual/CoreTypes/filterchain.html
  
  Index: filterchain.html
  ===
  RCS file: /home/cvs/ant/docs/manual/CoreTypes/filterchain.html,v
  retrieving revision 1.12
  retrieving revision 1.13
  diff -u -r1.12 -r1.13
  --- filterchain.html  23 Jul 2003 11:32:12 -  1.12
  +++ filterchain.html  19 Sep 2003 09:24:13 -  1.13
  @@ -103,6 +103,7 @@
   TabsToSpaces
   TailFilter
   DeleteCharacters
  +ConcatFilter
   TokenFilter
   
   FilterReader
  @@ -818,6 +819,49 @@
   
   
   
  +ConcatFilter
  +  This filter prepends or appends the content file to the filtered 
files.
  +  since Ant 1.6
  +
  +  
  +Parameter Name
  +Parameter Value
  +Required
  +  
  +  
  +prepend
  +
  +  The name of the file which content should be prepended to the file.
  +
  +No
  +  
  +  
  +append
  +
  +  The name of the file which content should be appended to the file.
  +
  +No
  +  
  +
  +
  +
  +Examples:
  +
  +Do nothing:
  +
  +
  +
  +
  +
  +
  +Adds a license text before each java source:
  +
  +
  +
  +
  +
  +
  +
   
   TokenFilter
   This filter tokenizes the inputstream into strings and passes these
  @@ -1293,4 +1337,4 @@
   
   
   Copyright © 2002-2003 Apache Software Foundation. All 
rights
  -Reserved.
  +Reserved.
  \ No newline at end of file
  
  
  
  1.2   +21 -21ant/src/etc/testcases/filters/concat.xml
  
  Index: concat.xml
  ===
  RCS file: /home/cvs/ant/src/etc/testcases/filters/concat.xml,v
  retrieving revision 1.1
  retrieving revision 1.2
  diff -u -r1.1 -r1.2
  --- concat.xml18 Sep 2003 15:39:26 -  1.1
  +++ concat.xml19 Sep 2003 09:24:13 -  1.2
  @@ -3,8 +3,8 @@
   
 
   
  -
  -
  +
  +
   
   
 
  @@ -26,35 +26,35 @@
   
 
   
  -  
  +  
   
  +  tofile="result/concat.filterReaderPrepend.test">
 
   
  -  
  +  
   
 
   
 
   
  -  
  +  
   
  +  tofile="result/concat.filterReaderAppend.test">
 
   
  -  
  +  
   
 
   
 
   
  -  
  +  
   
  +  tofile="result/concat.filterReaderPrependAppend.test">
 
   
  -  
  -  
  +  
  +  
   
 
   
  @@ -70,32 +70,32 @@
   
 
   
  -  
  +  
   
   
  +  tofile="result/concat.concatfilterPrepend.test">
 
  -
  +
 
   
 
   
  -  
  +  
   
   
  +  tofile="result/concat.concatfilterAppend.test">
 
  -
  +
 
   
 
   
  -  
  +  
   
   
  +  tofile="result/concat.concatfilterPrependAppend.test">
 
  -
  +
 
   
 
  
  
  
  1.2   +42 -42
ant/src/main/org/apache/tools/ant/filters/ConcatFilter.java
  
  Index: ConcatFilter.java
  ===
  RCS file: 
/home/cvs/ant/src/main/org/apache/tools/ant/filters/ConcatFilter.java,v
  retrieving revision 1.1
  retrieving revision 1.2
  diff -u -r1.1 -r1.2
  --- ConcatFilter.java 18 Sep 2003 15:39:26 -  1.1
  +++ ConcatFilter.java 19 Sep 2003 09:24:13 -  1.2
  @@ -67,7 +67,7 @@
* 
* 
* 
  - * 
  + * 
* 
* 
* 
  @@ -83,16 +83,16 @@
   implements ChainableReader {
   
   /** File to add before the content. */
  -private File before;
  +private File prepend;
   
   /** File to add after the content. */
  -private File after;
  +private File append;
   
  -/** Reader for before-file. */
  -private Reader beforeReader = new EmptyReader();
  +/** Reader for prepend-file. */
  +private Reader prependReader = new EmptyReader();
   
  -/** Reader for after-file. */
  -private Reader afterReader = new EmptyReader();
  +/** Reader for append-file. */
  +private Reader appendReader = new EmptyReader();
   
   /**
* Constructor for "dummy" instances.
  @@ -134,49 +134,49 @@
   
   int ch = -1;
   
  -// The readers return -1 i

DO NOT REPLY [Bug 14971] - task replaces classpath when includeantruntime && fork="yes" is used

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=14971

 task replaces classpath when includeantruntime && fork="yes" is used

[EMAIL PROTECTED] changed:

   What|Removed |Added

 Status|NEW |RESOLVED
 Resolution||FIXED



--- Additional Comments From [EMAIL PROTECTED]  2003-09-19 09:21 ---
Ant will now grab the content of the CLASSPATH environment variable and prepend
that to the classpath argument if you use includeantruntime="true".

Fixed with nightly build 2003-09-20.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

cvs commit: ant/src/main/org/apache/tools/ant/taskdefs/optional/junit JUnitTask.java

[bodewig]

bodewig 2003/09/19 02:18:16

  Modified:.WHATSNEW
   src/main/org/apache/tools/ant/taskdefs/optional/junit
JUnitTask.java
  Log:
   replaced the CLASSPATH
  instead of adding to it.
  
  PR: 14971
  
  Revision  ChangesPath
  1.503 +3 -0  ant/WHATSNEW
  
  Index: WHATSNEW
  ===
  RCS file: /home/cvs/ant/WHATSNEW,v
  retrieving revision 1.502
  retrieving revision 1.503
  diff -u -r1.502 -r1.503
  --- WHATSNEW  18 Sep 2003 15:47:47 -  1.502
  +++ WHATSNEW  19 Sep 2003 09:18:16 -  1.503
  @@ -242,6 +242,9 @@
   * The socket condition will now close the socket created to test.
 Bugzilla Report 23040.
   
  +*  replaced the CLASSPATH instead
  +  of adding to it.  Bugzilla Report 14971.
  +
   Other changes:
   --
   * All tasks can be used outside of s.  Note that some tasks
  
  
  
  1.83  +12 -1 
ant/src/main/org/apache/tools/ant/taskdefs/optional/junit/JUnitTask.java
  
  Index: JUnitTask.java
  ===
  RCS file: 
/home/cvs/ant/src/main/org/apache/tools/ant/taskdefs/optional/junit/JUnitTask.java,v
  retrieving revision 1.82
  retrieving revision 1.83
  diff -u -r1.82 -r1.83
  --- JUnitTask.java17 Sep 2003 08:30:30 -  1.82
  +++ JUnitTask.java19 Sep 2003 09:18:16 -  1.83
  @@ -717,6 +717,17 @@
   cmd.createArgument().setValue("haltOnFailure="
 + test.getHaltonfailure());
   if (includeAntRuntime) {
  +Vector v = Execute.getProcEnvironment();
  +Enumeration e = v.elements();
  +while (e.hasMoreElements()) {
  +String s = (String) e.nextElement();
  +if (s.startsWith("CLASSPATH=")) {
  +cmd.createClasspath(getProject()).createPath()
  +.append(new Path(getProject(), 
  + s.substring(10 // 
"CLASSPATH=".length()
  + )));
  +}
  +}
   log("Implicitly adding " + antRuntimeClasses + " to CLASSPATH",
   Project.MSG_VERBOSE);
   cmd.createClasspath(getProject()).createPath()
  
  
  

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: Should cause an error when used inside a ?

[peter reilly]

On Friday 19 September 2003 09:29, Stefan Bodewig wrote:
> On Fri, 19 Sep 2003, peter reilly <[EMAIL PROTECTED]> wrote:
> > I have a patch ready to check if the import is
> > in the top-level target and fail the build
> > if so.

A "not" missing above ;-).

> >
> > I think we should do this and wait for 1.7 to
> > allow  within a target.
>
> +1 as the feature seems to be at least controversial.
>
> Please simply change the names of the test methods I've introduced
> yesterday so we can reuse the tests if we change our minds.

Ok.

Peter


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

cvs commit: ant/docs/manual/CoreTasks apply.html exec.html

[bodewig]

bodewig 2003/09/19 01:44:23

  Modified:docs/manual/CoreTasks apply.html exec.html
  Log:
  Update and sync  and  docs
  
  Revision  ChangesPath
  1.23  +64 -4 ant/docs/manual/CoreTasks/apply.html
  
  Index: apply.html
  ===
  RCS file: /home/cvs/ant/docs/manual/CoreTasks/apply.html,v
  retrieving revision 1.22
  retrieving revision 1.23
  diff -u -r1.22 -r1.23
  --- apply.html24 Jun 2003 09:16:10 -  1.22
  +++ apply.html19 Sep 2003 08:44:22 -  1.23
  @@ -47,6 +47,17 @@
   Yes, if you specify a nested mapper
 
 
  +spawn
  +whether or not you want the commands to be spawned
  +Default is false.
  +If you spawn a command, its output will not be logged by ant.
  +The input, output, error, and result property settings are not active 
when spawning a process.
  +since Ant 1.6
  +
  +No
  +  
  +
  +  
   dir
   the directory in which the command should be 
executed.
   No
  @@ -75,8 +86,25 @@
 
 
   output
  -the file to which the output of the command should be
  -  redirected.
  +the file to which the output of the command
  +should be redirected.  If the error stream is not also redirected
  +to a file or property, it will appear in this output.
  +No
  +  
  +  
  +error
  +The file to which the standard error of the
  +command should be redirected.  since Ant 1.6
  +No
  +  
  +  
  +logError
  +This attribute is used when you wish to see error
  +output in Ant's log and you are redirecting output to a
  +file/property. The error output will not be included in the output
  +file/property. If you redirect error with the "error" or
  +"errorProperty" attributes, this will have no effect.
  +since Ant 1.6
   No
 
 
  @@ -88,8 +116,30 @@
 
 
   outputproperty
  -the name of a property in which the output of the 
  -  command should be stored.
  +the name of a property in which the output of the
  +command should be stored.  Unless the error stream is redirected
  +to a separate file or stream, this property will include the error
  +output.
  +No
  +  
  +  
  +errorproperty
  +The name of a property in which the standard error of 
the 
  +  command should be stored.  since Ant 1.6
  +No
  +  
  +  
  +input
  +A file from which the executed command's standard
  +input is taken. This attribute is mutually exclusive with the
  +inputstring attribute.  since Ant 1.6
  +No
  +  
  +  
  +inputstring
  +A string which serves as the input stream for the
  +executed command. This attribute is mutually exclusive with the
  +input attribute.  since Ant 1.6
   No
 
 
  @@ -164,6 +214,16 @@
   false as well.
   No, default is true
 
  +  
  +resolveExecutable
  +When this attribute is true, the name of the
  +executable if resolved firstly against the project basedir and if
  +that does not exist, against the execution directory if
  +specified. On Unix systems, if you only want to allow execution of
  +commands in the user's path, set this to false.  since Ant
  +1.6
  +No, default is false
  +   
 
   maxparallel
   Limit the amount of parallelism by passing at
  
  
  
  1.34  +10 -8 ant/docs/manual/CoreTasks/exec.html
  
  Index: exec.html
  ===
  RCS file: /home/cvs/ant/docs/manual/CoreTasks/exec.html,v
  retrieving revision 1.33
  retrieving revision 1.34
  diff -u -r1.33 -r1.34
  --- exec.html 6 Sep 2003 15:48:01 -   1.33
  +++ exec.html 19 Sep 2003 08:44:22 -  1.34
  @@ -73,7 +73,8 @@
   whether or not you want the command to be spawned
   Default is false.
   If you spawn a command, its output will not be logged by ant.
  -The input, output, error, and result property settings are not active 
when spawning a process.
  +The input, output, error, and result property settings are not active 
when spawning a process.
  +since Ant 1.6
   
   No
 
  @@ -85,8 +86,8 @@
 
 
   error
  -The file to which the standard error of the command 
should be
  -  redirected. 
  +The file to which the standard error of the
  +command should be redirected.  since Ant 1.6
   No
 
 
  @@ -95,7 +96,7 @@
log and you are redirecting output to a file/property. 
The error 
output will not be included in the output 
file/property. If you 
redirect error with the "error" or 
"errorProperty"
  - attributes, this will have no effect.
  + attributes, this will have no effect.  since Ant 
1.6
   No
 
 
  @@ -114,21 +115

DO NOT REPLY [Bug 23265] New: - attribute "includes" required if "dir" only is present

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23265

attribute "includes" required if "dir" only is present

   Summary: attribute "includes" required if "dir" only is present
   Product: Ant
   Version: 1.5.3
  Platform: Other
   URL: http://ant.apache.org/manual/CoreTasks/chmod.html
OS/Version: Other
Status: NEW
  Severity: Enhancement
  Priority: Other
 Component: Documentation
AssignedTo: [EMAIL PROTECTED]
ReportedBy: [EMAIL PROTECTED]


without combining this with 'includes="**/*"' nothing happened to "the files
held by dir".

p.s.: see also http://nagoya.apache.org/bugzilla/show_bug.cgi?id=22445 that
hopefully soon will make its way out of CVS on the site...
==> besides a 700 example, also in the table-row "perm |the new permissions."
the word "octal" could appear beneficially

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 22446] - uses the "inputstring" only first time

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=22446

 uses the "inputstring" only first time

[EMAIL PROTECTED] changed:

   What|Removed |Added

 Status|NEW |RESOLVED
 Resolution||FIXED



--- Additional Comments From [EMAIL PROTECTED]  2003-09-19 08:33 ---
Yes, anything but use the same input for all files doesn't seem to make sense.

Supposed to be fixed in CVS now, works for me but should be tested - in 
particular
the case where the task writes output and runs on an OS that locks files in
FileOutputStream#open like Windows, for example.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

cvs commit: ant/src/main/org/apache/tools/ant/taskdefs Execute.java ExecuteOn.java

[bodewig]

bodewig 2003/09/19 01:31:06

  Modified:src/main/org/apache/tools/ant/taskdefs Execute.java
ExecuteOn.java
  Log:
  Restart stream reading and writing for each execution in .
  
  PR: 22446
  
  Revision  ChangesPath
  1.68  +8 -2  ant/src/main/org/apache/tools/ant/taskdefs/Execute.java
  
  Index: Execute.java
  ===
  RCS file: /home/cvs/ant/src/main/org/apache/tools/ant/taskdefs/Execute.java,v
  retrieving revision 1.67
  retrieving revision 1.68
  diff -u -r1.67 -r1.68
  --- Execute.java  18 Sep 2003 14:21:41 -  1.67
  +++ Execute.java  19 Sep 2003 08:31:06 -  1.68
  @@ -348,10 +348,16 @@
*/
   public Execute(ExecuteStreamHandler streamHandler,
  ExecuteWatchdog watchdog) {
  -this.streamHandler = streamHandler;
  +setStreamHandler(streamHandler);
   this.watchdog = watchdog;
   }
   
  +/**
  + * @since Ant 1.6
  + */
  +public void setStreamHandler(ExecuteStreamHandler streamHandler) {
  +this.streamHandler = streamHandler;
  +}
   
   /**
* Returns the commandline used to create a subprocess.
  
  
  
  1.46  +16 -1 ant/src/main/org/apache/tools/ant/taskdefs/ExecuteOn.java
  
  Index: ExecuteOn.java
  ===
  RCS file: 
/home/cvs/ant/src/main/org/apache/tools/ant/taskdefs/ExecuteOn.java,v
  retrieving revision 1.45
  retrieving revision 1.46
  diff -u -r1.45 -r1.46
  --- ExecuteOn.java3 Sep 2003 16:47:17 -   1.45
  +++ ExecuteOn.java19 Sep 2003 08:31:06 -  1.46
  @@ -333,6 +333,11 @@
   log(Commandline.describeCommand(command),
   Project.MSG_VERBOSE);
   exe.setCommandline(command);
  +if (haveExecuted) {
  +// need to reset the stream handler to restart
  +// reading of pipes
  +exe.setStreamHandler(createHandler());
  +}
   runExecute(exe);
   haveExecuted = true;
   }
  @@ -376,6 +381,11 @@
   log(Commandline.describeCommand(command),
   Project.MSG_VERBOSE);
   exe.setCommandline(command);
  +if (haveExecuted) {
  +// need to reset the stream handler to restart
  +// reading of pipes
  +exe.setStreamHandler(createHandler());
  +}
   runExecute(exe);
   haveExecuted = true;
   }
  @@ -595,6 +605,11 @@
   String[] command = getCommandline(cs, cb);
   log(Commandline.describeCommand(command), 
Project.MSG_VERBOSE);
   exe.setCommandline(command);
  +if (currentOffset > 0) {
  +// need to reset the stream handler to restart
  +// reading of pipes
  +exe.setStreamHandler(createHandler());
  +}
   runExecute(exe);
   
   stillToDo -= currentAmount;
  @@ -615,4 +630,4 @@
   return new String[] {"file", "dir", "both"};
   }
   }
  -}
  \ No newline at end of file
  +}
  
  
  

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: Should cause an error when used inside a ?

[Stefan Bodewig]

On Fri, 19 Sep 2003, peter reilly <[EMAIL PROTECTED]> wrote:
> I have a patch ready to check if the import is
> in the top-level target and fail the build
> if so.
> 
> I think we should do this and wait for 1.7 to
> allow  within a target.

+1 as the feature seems to be at least controversial.

Please simply change the names of the test methods I've introduced
yesterday so we can reuse the tests if we change our minds.

Stefan

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: Should cause an error when used inside a ?

[peter reilly]

I have a patch ready to check if the import is
in the top-level target and fail the build
if so.

I think we should do this and wait for 1.7 to
allow  within a target. (In this case
the import would make sure that the imported
file did not contain targets (recursively)).

Peter


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: [VOTE]: Getting 1.6 out the door

[Stefan Bodewig]

On Fri, 19 Sep 2003, peter reilly <[EMAIL PROTECTED]> wrote:
> On Friday 19 September 2003 07:50, Stefan Bodewig wrote:
>> On Thu, 18 Sep 2003, Steve Loughran <[EMAIL PROTECTED]> wrote:
>> > I'd also like Ant 1.6 to have try/catch; if we cannot easily
>> > reuse the ant contrib one,
>>
>> Have we even tried to?
> 
> The  issue was getting permission from the author.

authors.

The other thing should be asking the ant-contrib project admins
whether it was OK with them IMHO.  You and I are mere mortals unlike
Matt and Curt 8-)

> (One Stefan Bodewig with mods by Dan Ritchey from the javadoc).

It's been more the other way around.  The initial code has been Dan's
and I've modified it (quite a bit - the initial code could just catch
the exception but didn't have catch or finally blocks).

> In any case I think we should hold off this debate until
> 1.7 (if  why not  ?).

fine with me.

Stefan

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 22865] - Zip whenempty attribute is broken

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=22865

Zip whenempty attribute is broken





--- Additional Comments From [EMAIL PROTECTED]  2003-09-19 07:52 ---
ping

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: [VOTE]: Getting 1.6 out the door

[peter reilly]

On Friday 19 September 2003 07:50, Stefan Bodewig wrote:
> On Thu, 18 Sep 2003, Steve Loughran <[EMAIL PROTECTED]> wrote:
> > I'd also like Ant 1.6 to have try/catch; if we cannot easily reuse
> > the ant contrib one,
>
> Have we even tried to?

The  issue was getting permission from the author. (One
Stefan Bodewig with mods by Dan Ritchey from the javadoc).

In any case I think we should hold off this debate until
1.7 (if  why not  ?).

Peter
>
> > is it too hard to write our own?
>
> Given that I've written significant parts of the ant-contrib one, I'd
> say no.  But the result would look very very close to the one in
> ant-contrib 8-)
>
> Stefan
>
> -
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 23165] - javadoc task should allow multiple "link" parameters

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23165

javadoc task should allow multiple "link" parameters

[EMAIL PROTECTED] changed:

   What|Removed |Added

 Status|NEW |RESOLVED
 Resolution||FIXED



--- Additional Comments From [EMAIL PROTECTED]  2003-09-19 07:49 ---
fixed in the docs of nightly build 2003-09-20.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

cvs commit: ant/docs/manual/CoreTasks javadoc.html

[bodewig]

bodewig 2003/09/19 00:46:31

  Modified:docs/manual/CoreTasks javadoc.html
  Log:
  Add pointers from attributes to nested elements, PR 23165
  
  Revision  ChangesPath
  1.26  +17 -9 ant/docs/manual/CoreTasks/javadoc.html
  
  Index: javadoc.html
  ===
  RCS file: /home/cvs/ant/docs/manual/CoreTasks/javadoc.html,v
  retrieving revision 1.25
  retrieving revision 1.26
  diff -u -r1.25 -r1.26
  --- javadoc.html  14 May 2003 15:01:12 -  1.25
  +++ javadoc.html  19 Sep 2003 07:46:31 -  1.26
  @@ -67,7 +67,8 @@
 
 
   sourcefiles
  -Comma separated list of source files
  +Comma separated list of source files -- see also
  +the nested source element.
   all
 
 
  @@ -85,7 +86,7 @@
 
   packagenames
   Comma separated list of package files (with terminating
  -  wildcard)
  +  wildcard) -- see also the nested package element.
   all
   No
 
  @@ -246,21 +247,25 @@
 
 
   link
  -Create links to javadoc output at the given URL
  +Create links to javadoc output at the given URL
  +-- see also the nested link element.
   1.2+
   No
 
 
   linkoffline
   Link to docs at  using package list at
  -   - separate the URLs by using a space character.
  + - separate the URLs by using a space character -- see
  +also the nested link element.
   1.2+
   No
 
 
   group
   Group specified packages together in overview
  -  page.  The format is as described below.
  +page.  The format is as described below -- see also the nested
  +group element.
   1.2+
   No
 
  @@ -333,7 +338,9 @@
 
 
   doclet
  -Specifies the class file that starts the doclet used in 
generating the documentation.
  +Specifies the class file that starts the doclet
  +used in generating the documentation -- see also the nested
  +doclet element.
   1.2+
   No
 
  @@ -354,8 +361,9 @@
 
   additionalparam
   Lets you add additional parameters to the javadoc
  -command line. Useful for doclets. Parameters containing
  -spaces need to be quoted using ".
  +command line. Useful for doclets. Parameters containing spaces
  +need to be quoted using " -- see also the nested
  +arg element.
   all
   No
 
  @@ -369,7 +377,7 @@
 
   excludepackagenames
   comma separated list of packages you don't want
  -  docs for.
  +  docs for -- see also the nested excludepackage 
element.
   all
   No
 
  
  
  

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 23253] - updating link for netcomponents.jar

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23253

updating link for netcomponents.jar

[EMAIL PROTECTED] changed:

   What|Removed |Added

 Status|NEW |RESOLVED
 Resolution||INVALID



--- Additional Comments From [EMAIL PROTECTED]  2003-09-19 07:41 ---
All released versions of Ant require NetComponents and won't work with 
commons-net.
The link is thus perfectly correct.

We are aware of commons-net (of course) and Ant's CVS HEAD uses it instead of
the old NetComponents package.  Likewise the links inside CVS HEAD have already
been updated.

Thanks.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: [VOTE]: Getting 1.6 out the door

[Stefan Bodewig]

On Thu, 18 Sep 2003, Steve Loughran <[EMAIL PROTECTED]> wrote:

> I'd also like Ant 1.6 to have try/catch; if we cannot easily reuse
> the ant contrib one,

Have we even tried to?

> is it too hard to write our own?

Given that I've written significant parts of the ant-contrib one, I'd
say no.  But the result would look very very close to the one in
ant-contrib 8-)

Stefan

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: seems the last gump build has been on September 9th

[Stefan Bodewig]

On Thu, 18 Sep 2003, Antoine Lévy-Lambert <[EMAIL PROTECTED]>
wrote:

> Is there any particular reason ?

Sam posted to the gump list that his machine - the one that does the
"official" nightly Gump runs - for a reason unknown to him refuses to
upload the results right now.  He's investigating the problem.

 to stay informed ;-)

Stefan

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: cvs commit: ant/src/testcases/org/apache/tools/ant/filters Co ncatFilterTest.java

[Stefan Bodewig]

On Thu, 18 Sep 2003, Jan Materne <[EMAIL PROTECTED]> wrote:

> More interesting for me was the update from my notebook via mobile
> phone :-)

How long did it take?

Stefan

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

RE: [patch] more typos!

[Shatzer, Larry]

Actually, I ran a spell check. ;)

I should just submit the big patch, but it's like 8,000 lines long, and this
way, people can over look my changes, and make sure I don't delete a double
at when it should be there (or reworded), or misspell xerces as xerxes.

I have a few other areas I'm working on that I need to submit too. Mostly
all with either the javadoc or regular docs.

People always complain about the lack of, or the sub par documentation open
source products have. I like to try to fix that where I can.

-- Larry

-Original Message-
From: Steve Loughran
To: Ant Developers List
Sent: 9/18/2003 9:15 PM
Subject: Re: [patch] more typos!

Shatzer, Larry wrote:

> Next round of typos...
> 
> -- Larry
> 

Larry, I am most impressed. Are you really reading *all* the javadocs 
and correcting their typos. Wow


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Antidote TODO (WAS:RE: antidote diff)

[Christoph Wilhelms]

Hi Jack!

> > It's an old, but just in some points outdated TODO-list. 
> > Feel free to pic up a task!

I re to my own words first: The TODO-List is old and not updated for a
long time! Additionally it is not by me, but by Simeon. To be honest I
do not understand all points by myself, but I can give advice to some of
them! Let me try to! BTW: Checking and updating that TODO, so that we
can put it on the projects webpage would be a task :-).

> This is the same checked-in TODO list that is in the CVS tree, right?

yup

> Maybe the first step is to update the TODO list :-)
> 
> Let's look at some items from TODO List:
> 
> * Cleanup build.xml file; make antidote specific.
> 
>  This has been done already, right?

I think so!
 
> * Rewrite ACSFactory to use it's own parser rather than 
> the implementation
>   specific com.sun.xml.tree.SimpleElementFactory class 
> (only available
>   in JAXP from Sun).
> 
>  Done? Still TODO?

Has to be rececked, but generally it works fin IMHO

> * Add editors for defining file sets, and other "sets".
> 
>  This has been done already, right?

Particualry! IMHO there could be mor comfort in defining filesets, what
do you think!?!

> * Write wizzard framework.
> 
>  The framework is there, but there are currently no 
> wizards, right?

There is a "not working" example wizard for an entire build-file. IMO
the wizarg could be used to create targets in addition...

> * Implement a build progress reporter.
> 
>  What would that look like, compared to the console that 
> currently exists?

I thinks the console is ok, but Sim meant to write a log to a file in
addition. Obviously the log line formtting has to be definable
interactively and stored in preferences...

> * Implement a "Worker Thread" pattern that allows workers to have
>   their work done in a thread property registered with 
> Antidote, and
>   provide support to that worker to provide progress updates via
>   the GUI. Should also provide support for hour-glass cursor
>   handling, and AWT event blocking until task is completed. This
>   would be used for things such as loading files or other tasks
>   that the user must wait for completion.
> 
>  Has this been done or is it still TODO? I think it is still TODO.

It IS a to do! There is no way to stop Ant once it runs from Antidote!

> * Add menu option to select the compiler to use, which then sets
>   the "build.compiler" property. Better yet, create a generic menu
>   building capability that allows the setting of a property from a
>   list of options.
> 
>  Still TODO, right?

It is a todo! I'd propose not just setting the complier, put the option
to set all Ant -D parameters from the gui and make them storable! 

> * Add ability to put an "all" or "don't care" specifyer 
> on the action
>   "enableOn" and "disableOn" properties.
> 
>  Still TODO, right?

It is not implemented IIRC! Additionally I am not really shure what it
should do...

> * Add ability to view task dependencies more fully.
> 
>  Describe this facility?

It should say "target-dependency". It would be nice to be able to show a
grafic with all the tragets and arrows between them to show the
dependency (depends). Would be really a feature for many guys using Ant,
I know...

> * Add better editors for specific tasks.
> 
>  Examples?

Not only for tasks, but properties in addition. We meant editors just
like the dependency choose!

> * Add a Progress Monitor for file loading (especially for 
> slow boxen like
>   mine) .
> 
>  Still TODO, right?

Yes, it is! Progress for loading files and for the time the
introspection rus would be nice!

> * Implement some for of refid hyperlinking functionality.
> 
>  Describe this facility?

Not shure what Sim means... Yould someone else drop in here...!?!

> * Implement context sensitive menus for the console 
> window, allowing
>   an error to be selected and invoked in IDE.
> 
>  "Jump-to" in Antidote itself?

I think integration into IDEs is no BIG project aim anymore or at leas
ATM...

> * Write preferences framwork, including persistence support.
> 
>  Trivial, still TODO? I can borrow this from my own open 
> source code.

Would be nice, if the code is under an Apache compatible OSS License
(like Apache or FreeBSD) LGPL is NOT ok!!
I'd prefer a small, fast an clean implementation as always ;-).

> * Provide some sort of class path debugging support.
> 
>  Describe this facility?

Not 100% shure, but displaying the Ant-build-classpath and introspect
the jars one by one... Search for Classes and multiple occurrences in
the classpath... And so on!

> * Add "syntax" colorization to the console window {done},
>   with a preferences editor for setting up the styles {not-done}.
> 
>  Describe this facility?

Just make the syntax highlighting in the console window configurable via

Re: [VOTE]: Getting 1.6 out the door

[Steve Loughran]

Antoine Lévy-Lambert wrote:
Hi,
I would like to propose a release plan for voting :
- features included in 1.6 : all the features currently present in head
- freeze date for 1.6 branch : Monday, September 22 13:00 GMT
- availability of ANT_16_B1 binaries : within one week of the freeze of the
branch.

+1.
Before then I'd like to
 -fix  as previously discussed
 -test  support in java/junit to see it looks good.
I'd also like Ant 1.6 to have try/catch; if we cannot easily reuse the 
ant contrib one, is it too hard to write our own?

The exact time will depend on whether I will have trouble with practical
issues from the build down to the signing of the jar files to the update of
the web site, of bugzilla, ...
- release manager : myself
(I hope this is OK, although I am not a PMC member).
So, if this release plan is voted, on Monday, September 22 I will create the
ANT_16_BRANCH tag.
The CVS Head will then become ant 1.7alpha
Cheers,
Antoine


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: [patch] more typos!

[Steve Loughran]

Shatzer, Larry wrote:
Next round of typos...
-- Larry
Larry, I am most impressed. Are you really reading *all* the javadocs 
and correcting their typos. Wow

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 23262] New: - fixCRLF adds extraneous character at the end of file.

[bugzilla]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=23262

fixCRLF adds extraneous character at the end of file.

   Summary: fixCRLF adds extraneous character at the end of file.
   Product: Ant
   Version: 1.5.4
  Platform: Other
OS/Version: Other
Status: NEW
  Severity: Normal
  Priority: Other
 Component: Core tasks
AssignedTo: [EMAIL PROTECTED]
ReportedBy: [EMAIL PROTECTED]


The fixcrlf task adds an extra newline character to the end of each file to be 
converted that does not already have a newline character at the end. This is 
causing a lot of problems when a diff is done on files that are converted and 
files that are not converted. Ideally the task should not be modifying the 
content of the file except for replacing newline characters appropriately.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]