We have many files which contain only CR for line endings. When using the
code, below, under Ant 1.4.1, I get the expected results, under both W2k and
Linux--CR becomes CRLF under W2k and CR becomes LF under Linux,
respectively. Using Ant 1.5beta, though, when I have two consecutive CRs,
the (non-LF) character following the second CR gets eaten up.
Code:
<target name="fix_preproc_eol" depends="init">
<echo message="Normalizing EOL in ${preproc_dir}/com/wrq ..."/>
<fixcrlf srcdir="${preproc_dir}/com/wrq"
includes="**/*.java,**/*.properties"
/>
</target>
Sample source:
package com.wrq.filexfer.indxfer.ncs.EBCDICToSJIS;
public class EBCDICToSJIS59 implements ...
{
Output under Ant 1.5:
package com.wrq.filexfer.indxfer.ncs.EBCDICToSJIS;
ublic class EBCDICToSJIS59 implements ...
The docs mention that the sequence CR-CR-LF is regarded as a single EOL, but
there is no LF in the sequences in my files.
Thanks,
Ed Mack
WRQ, Inc.
We specialize in integration software and services that let you quickly
adapt your host-intensive environment to meet new business needs.
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
