Le samedi 24 septembre 2016 17:33:31 Karl Heinz Marbaise a écrit : > Hi Hervé, > > On 24/09/16 13:46, Hervé BOUTEMY wrote: > > if you look at > > https://cwiki.apache.org/confluence/display/MAVENOLD/Versioning that lead > > to ComparableVersion [1] implementation: "The current implementation for > > version schemes is rather limited. It only supports 5 properties" (and by > > "current", it meant DefaultArtifactVersion) > > > > ComparableVersion is meant to be more generic, with any number of parts. > > > > And since DefaultArtifactVersion is not used in core (core only compares > > version, does not really need to split in major/minor/...), I never > > investigated how it really splits the version string in parts: having a > > fixed number of part is broken by design, IMHO, then nobody should use > > DefaultArtifactVersion with its broken API. But I never went ahead and > > removed this class from core, to avoid incompatibility: perhaps just > > deprecating this class would be the most simple thing to do. > > So I have marked the class as deprecated... > > But it implements an interface: > > public interface ArtifactVersion > extends Comparable<ArtifactVersion> > > which should be marked as deprecated as well in consequence of the > previous step (my opinion ?) WDYT ? +1 the issue is both in interface and implementation
> > > So one more question: > > What about moving DefaultArtifactVersion to maven-compat module... good idea > and replace it with a better implementation ? if you can improve it, why not > > > What might be done is to create an enhanced version of the interface and > make a better implemtation (based on ComparableVersion or make same > getters to get the parsed information from it...) yes, adding getters in ComparableVersion would be a good idea the only issue is that I fear that nobody will use it, since it would require Maven 3.4 Regards, Hervé > > > > Kind regards > Karl Heinz Marbaise > > > Regards, > > > > Hervé > > > > [1] > > http://maven.apache.org/ref/3-LATEST/maven-artifact/apidocs/org/apache/ma > > ven/artifact/versioning/ComparableVersion.html> > > Le samedi 24 septembre 2016 12:42:02 Karl Heinz Marbaise a écrit : > >> Hi to all, > >> > >> at the moment I'm trying to understand a thing...(motivated by an issue > >> https://github.com/mojohaus/build-helper-maven-plugin/issues/33) > >> > >> Lets say I have the following version: "20.4.00.0-SNAPSHOT" > >> > >> The DefaultArtifactVersion class will parse this version with this > >> result: > >> > >> artifactVersion.getMajorVersion() = "0" > >> artifactVersion.getMinorVersion() = "0" > >> artifactVersion.getIncrementalVersion() = "0" > >> artifactVersion.getBuildNumber() = "0" > >> artifactVersion.getQualifier() = "20.4.00.0-SNAPSHOT" > >> > >> And another version: "20.4.0-SNAPSHOT" and the DefaultArtifactVersion > >> class will parse this version with this result: > >> > >> artifactVersion.getMajorVersion() = "20" > >> artifactVersion.getMinorVersion() = "4" > >> artifactVersion.getIncrementalVersion() = "0" > >> artifactVersion.getBuildNumber() = "0" > >> artifactVersion.getQualifier() = "SNAPSHOT" > >> > >> but If I do a version comparison via: > >> > >> java -jar /usr/local/apache-maven-3.3.9/lib/maven-artifact-3.3.9.jar > >> 20.4.00.0-SNAPSHOT 20.4.0-SNAPSHOT > >> Display parameters as parsed by Maven (in canonical form) and comparison > >> result: > >> 1. 20.4.00.0-SNAPSHOT == 20.4-snapshot > >> > >> 20.4.00.0-SNAPSHOT == 20.4.0-SNAPSHOT > >> > >> 2. 20.4.0-SNAPSHOT == 20.4-snapshot > >> > >> So the question is why is the first one parsed differently by > >> DefaultArtifactVersion) than the second one... > >> > >> Yes I know there is a difference (20.4.00.0...) but from a logical point > >> of view this does not make sense? > >> > >> So it looks like that DefaultArtifactVersion follows hardly the pattern: > >> > >> <majorversion [> . <minorversion [> . <incrementalversion ] ] [> - > >> <buildnumber | qualifier ]> > >> > >> whereas the version comparison goes a different path... > >> > >> > >> Furthermore I have checked some other cases like this: > >> > >> given Version: 1.2.3-01-SNAPSHOT > >> > >> artifactVersion.getMajorVersion() = "1" > >> artifactVersion.getMinorVersion() = "2" > >> artifactVersion.getIncrementalVersion() = "3" > >> artifactVersion.getBuildNumber() = "0" > >> artifactVersion.getQualifier() = "01-SNAPSHOT" > >> > >> given Version: 1.2.3-01 > >> > >> artifactVersion.getMajorVersion() = "1" > >> artifactVersion.getMinorVersion() = "2" > >> artifactVersion.getIncrementalVersion() = "3" > >> artifactVersion.getBuildNumber() = "0" > >> artifactVersion.getQualifier() = "01" > >> > >> Version: 1.2.3-1 > >> > >> artifactVersion.getMajorVersion() = "1" > >> artifactVersion.getMinorVersion() = "2" > >> artifactVersion.getIncrementalVersion() = "3" > >> artifactVersion.getBuildNumber() = "1" > >> artifactVersion.getQualifier() = null > >> > >> So this means a buildnumber is not allowed to be prefixed with a digit > >> "0" ? I think the last two examples look strange to me...I would have > >> expected that a qualifier is something which starts with a character and > >> not with a digit... > >> > >> WDYT ? > >> > >> Kind regards > >> Karl Heinz > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org > For additional commands, e-mail: dev-h...@maven.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org For additional commands, e-mail: dev-h...@maven.apache.org
