You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tika.apache.org by "Uwe Schindler (JIRA)" <ji...@apache.org> on 2008/11/15 19:27:50 UTC

[jira] Created: (TIKA-173) Creating of a binary release that does not bundle all JARS in one big one

Creating of a binary release that does not bundle all JARS in one big one
-------------------------------------------------------------------------

                 Key: TIKA-173
                 URL: https://issues.apache.org/jira/browse/TIKA-173
             Project: Tika
          Issue Type: Wish
    Affects Versions: 0.2-incubating
            Reporter: Uwe Schindler


The maven code for building the project works good. For me it is a little bit annoying that you are not able (as end user who will just download a binary release for inclusion into his project) to include the JAR and all dependent JARs into your project. The standalone JAR is good for starting TIKA as GUI or for a simple command line, but when including it into other projects, where some of the libraries used by TIKA are also used, you have duplicate class entries in your classpath and depending on the position of the standalone.jar you may overwrite your local classes by old versions.

It would be good to have a maven build task/step, that creates a lib directory for the binary release containing TIKA's JAR file and all dependant libs as separate files. Then a user is possible to only add the missing libs to his own project and e.g. replace by newer versions of PDFBox, POI, XALAN, XERCES and do not pollute his classpath.

By the way: TIKA is a Java 1.5 Project, the use of xml-apis.jar is not good. JAXP is already included in Java, so i would remove this from the dependencies.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


[jira] Commented: (TIKA-173) Creating of a binary release that does not bundle all JARS in one big one

Posted by "Uwe Schindler (JIRA)" <ji...@apache.org>.
    [ https://issues.apache.org/jira/browse/TIKA-173?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12647909#action_12647909 ] 

Uwe Schindler commented on TIKA-173:
------------------------------------

Thank you,

"mvn dependency:copy-dependencies" was what I missed. I have seen both JAR files (the TIKA only and standalone), but I needed a build with collecting all dependencies in the target.

I opened this issue, because I wanted to be sure that in future (when TIKA gets released also in binary version), dependencies are in their original JAR files bundled together with TIKA and not only mixed together in the standalone JAR.

For quick testing of your build, the standalone jar is really good, but not for re-using the library in own projects.

Thanks!

> Creating of a binary release that does not bundle all JARS in one big one
> -------------------------------------------------------------------------
>
>                 Key: TIKA-173
>                 URL: https://issues.apache.org/jira/browse/TIKA-173
>             Project: Tika
>          Issue Type: Wish
>    Affects Versions: 0.2-incubating
>            Reporter: Uwe Schindler
>
> The maven code for building the project works good. For me it is a little bit annoying that you are not able (as end user who will just download a binary release for inclusion into his project) to include the JAR and all dependent JARs into your project. The standalone JAR is good for starting TIKA as GUI or for a simple command line, but when including it into other projects, where some of the libraries used by TIKA are also used, you have duplicate class entries in your classpath and depending on the position of the standalone.jar you may overwrite your local classes by old versions.
> It would be good to have a maven build task/step, that creates a lib directory for the binary release containing TIKA's JAR file and all dependant libs as separate files. Then a user is possible to only add the missing libs to his own project and e.g. replace by newer versions of PDFBox, POI, XALAN, XERCES and do not pollute his classpath.
> By the way: TIKA is a Java 1.5 Project, the use of xml-apis.jar is not good. JAXP is already included in Java, so i would remove this from the dependencies.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


[jira] Resolved: (TIKA-173) Creating of a binary release that does not bundle all JARS in one big one

Posted by "Jukka Zitting (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/jira/browse/TIKA-173?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jukka Zitting resolved TIKA-173.
--------------------------------

    Resolution: Invalid

The main build artifact already is what you're looking for. The "standalone" jar is just an additional build artifact, targeted for cases where you do want all the dependencies in a single package. Resolving this issue as Invalid.

The xml-apis library is a dependency of Xerces that's a dependency of NekoHtml. You may want to open a separate issue for better dealing with that dependency.

> Creating of a binary release that does not bundle all JARS in one big one
> -------------------------------------------------------------------------
>
>                 Key: TIKA-173
>                 URL: https://issues.apache.org/jira/browse/TIKA-173
>             Project: Tika
>          Issue Type: Wish
>    Affects Versions: 0.2-incubating
>            Reporter: Uwe Schindler
>
> The maven code for building the project works good. For me it is a little bit annoying that you are not able (as end user who will just download a binary release for inclusion into his project) to include the JAR and all dependent JARs into your project. The standalone JAR is good for starting TIKA as GUI or for a simple command line, but when including it into other projects, where some of the libraries used by TIKA are also used, you have duplicate class entries in your classpath and depending on the position of the standalone.jar you may overwrite your local classes by old versions.
> It would be good to have a maven build task/step, that creates a lib directory for the binary release containing TIKA's JAR file and all dependant libs as separate files. Then a user is possible to only add the missing libs to his own project and e.g. replace by newer versions of PDFBox, POI, XALAN, XERCES and do not pollute his classpath.
> By the way: TIKA is a Java 1.5 Project, the use of xml-apis.jar is not good. JAXP is already included in Java, so i would remove this from the dependencies.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


[jira] Commented: (TIKA-173) Creating of a binary release that does not bundle all JARS in one big one

Posted by "Jukka Zitting (JIRA)" <ji...@apache.org>.
    [ https://issues.apache.org/jira/browse/TIKA-173?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12647907#action_12647907 ] 

Jukka Zitting commented on TIKA-173:
------------------------------------

Also, as for the Maven goal that creates a directory with all the dependency jars, run "mvn dependency:copy-dependencies" and look at "target/dependency".

> Creating of a binary release that does not bundle all JARS in one big one
> -------------------------------------------------------------------------
>
>                 Key: TIKA-173
>                 URL: https://issues.apache.org/jira/browse/TIKA-173
>             Project: Tika
>          Issue Type: Wish
>    Affects Versions: 0.2-incubating
>            Reporter: Uwe Schindler
>
> The maven code for building the project works good. For me it is a little bit annoying that you are not able (as end user who will just download a binary release for inclusion into his project) to include the JAR and all dependent JARs into your project. The standalone JAR is good for starting TIKA as GUI or for a simple command line, but when including it into other projects, where some of the libraries used by TIKA are also used, you have duplicate class entries in your classpath and depending on the position of the standalone.jar you may overwrite your local classes by old versions.
> It would be good to have a maven build task/step, that creates a lib directory for the binary release containing TIKA's JAR file and all dependant libs as separate files. Then a user is possible to only add the missing libs to his own project and e.g. replace by newer versions of PDFBox, POI, XALAN, XERCES and do not pollute his classpath.
> By the way: TIKA is a Java 1.5 Project, the use of xml-apis.jar is not good. JAXP is already included in Java, so i would remove this from the dependencies.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.