Hi there.

I would like to understand the expected result of a concat step that
'finds' no source files.

I believe that prior to version 1.7.1 the resultant file could include
header and footer even if the no other files went in.

With 1.7.1, class :-

   src/main/org/apache/tools/ant/taskdefs/Concat.java

checks for zero source and returns.

I would like to have a file produced if I specify a header or footer,
but I can see this might cause other people problems if they are
expecting no file at all to be produced in this situation, although this
was the result with 1.7.0.

Should I be checking that a file is produced out of concat, instead of
assuming there will be one (my next step is a copy which fails if the
cancat target is not created)?

The following patch would make it work for me, but is this a reasonable
change to suggest?

Index: src/main/org/apache/tools/ant/taskdefs/Concat.java
===================================================================
--- src/main/org/apache/tools/ant/taskdefs/Concat.java  (revision 704405)
+++ src/main/org/apache/tools/ant/taskdefs/Concat.java  (working copy)
@@ -733,7 +733,7 @@
             log(destinationFile + " is up-to-date.", Project.MSG_VERBOSE);
             return;
         }
-        if (c.size() == 0) {
+        if (c.size() == 0 && header == null && footer == null) {
             return;
         }
         OutputStream out;

-- 
Mark

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

Reply via email to