[MJDEPS-22] dependenciesToAnalyzeIncludes and dependenciesToAnalyzeExcludes don't work as expected

• Type: Bug
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 3.1.1
• Environment: Windows 10, JDK 1.8.0_181
• Assignee: Unassigned
• Reporter: Jean-François Lecomte
• Created: 07/Feb/19 16:04
• Updated: 07/Feb/19 16:04

Description

I would like to analyze all dependencies, but those with a groupId starting with "rhc."
I'm using the following configuration:

                       <configuration>
                            <dependenciesToAnalyzeIncludes>
                                <include>*:*</include>
                            </dependenciesToAnalyzeIncludes>
                            <dependenciesToAnalyzeExcludes>
                                <exclude>rhc.*:*</exclude>
                            </dependenciesToAnalyzeExcludes>
                        </configuration>

I expect the resulting command to look like:

cmd.exe /X /C ""C:\Program Files\Java\jdk1.8.0_181\bin\jdeps.exe" -cp {All rhc.* dependencies} -P [myPath]\target\classes [All other dependencies]

Instead all the dependencies appear in the classpath (-cp) section and only the classes folder is left at the end.

Looking at the code, I see that the method getDependenciesToAnalyze in AbstractJDepsMojo does the following:

                if ( includes.matchesPatternStart( versionlessKey, true ) 
                    && !excludes.matchesPatternStart( versionlessKey, true ) )
                {
                    jdepsClasses.add( artifact.getFile().toPath() );
                }

I don't know what "matchesPatternStart" is suppose to do, but the following:

MatchPatterns.from("org.foo.*:*").matchesPatternStart("com.google.guava:guava", true)

returns true, while I would expect a mismatch here.
Calling matches instead of matchesPatternStart seems to fix the issue.