Ryan created ARCHETYPE-616:
------------------------------
Summary: Filelist scanning does not use substitute property in
directory name
Key: ARCHETYPE-616
URL: https://issues.apache.org/jira/browse/ARCHETYPE-616
Project: Maven Archetype
Issue Type: Bug
Affects Versions: 3.2.0
Reporter: Ryan
When creating an archetype, I would like directories to be copied based on a
property *p*.
eg. given a directory structure
root
— a
------ pom.xml
— b
------ pom.xml
if the *p* property is passed as "a" Only the "a" directory is copied, "b" is
not copied.
My archetype-metadata.xml looks like the following:
{quote}<archetype-descriptor>
<fileSets>
<fileSet filtered="true" packaged="false" encoding="UTF-8">
<directory>__p__</directory>
<includes>
<include>***/***</include>
</includes>
</fileSet>
</fileSets>
<requiredProperties>
<requiredProperty key="p">
<defaultValue>a</defaultValue>
</requiredProperty>
</requiredProperties>
</archetype-descriptor>
{quote}
And I can see in the debug logs that the directory name is being substituted
correctly
[DEBUG] Replacing property 'applicationFramework' in file path
'__applicationFramework__//' with value 'a'.
[DEBUG] Final interpolated file path: 'a//'
*However the file search itself is not using the substituted directory name,
leading to it finding no files*
[DEBUG] Processing fileset __p__ (Filtered-Flat) [**/** -- ] -> 0:
[DEBUG] Processed 0 files.
I am very confident this is not an issue with my archetype-metadata.xml because
when I replace "__p__" with either "a" or "b" it searches, finds and copies the
files in that directory successfully.
*Expected behaviour:*
Property is substituted in directory path and that same directory is scanned
for files, as is stated in the docs
*Actual behaviour:*
Property is substituted, but the scan takes place in the non-substituted
___p___ directory. This may be a feature if the directory was named "__p__" but
both situations should be covered
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
