[GitHub] [tika] theit commented on pull request #1345: TIKA-3948 -- migrate from javax -> jakarta

2023-09-20 Thread via GitHub


theit commented on PR #1345:
URL: https://github.com/apache/tika/pull/1345#issuecomment-1727711669

   > Finally I was able to narrow down this behaviour: This obviously depends 
on the JVM I'm using... Eclipse was configured to use OpenJDK 11 (11.0.20.1 to 
be precisely) whereas on the command-line OpenJDK 20 (20.0.2) was active. 
Switching this to also Java 11 makes the error disappear.
   > 
   > But: When I use Java 20 in Eclipse to execute the JUnit tests in 
tika-server-core, 8 of 75 (with 7 skipped) now fail as they did on the 
command-line. Still investigating...
   
   Also no problems with Java 17. Still searching for the cause...


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: dev-unsubscr...@tika.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org



[GitHub] [tika] theit commented on pull request #1345: TIKA-3948 -- migrate from javax -> jakarta

2023-09-20 Thread via GitHub


theit commented on PR #1345:
URL: https://github.com/apache/tika/pull/1345#issuecomment-1727137134

   > > ```
   > > 4. Running `mvn verify` fails in Tika server core:
   > > ```
   > > 
   > > 
   > > 
   > >   
   > > 
   > > 
   > >   
   > > 
   > > 
   > > 
   > >   
   > > ```
   > > [ERROR] Tests run: 4, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 
0.054 s <<< FAILURE! -- in org.apache.tika.server.core.StackTraceTest
   > > [ERROR] org.apache.tika.server.core.StackTraceTest.testEmptyParser -- 
Time elapsed: 0.006 s <<< FAILURE!
   > > org.opentest4j.AssertionFailedError: bad type: /tika ==> expected: <200> 
but was: <500>
   > >  at 
org.junit.jupiter.api.AssertionFailureBuilder.build(AssertionFailureBuilder.java:151)
   > > (...)
   > > ```
   > 
   > No idea. Let me see if I can reproduce this locally and figure out what's 
going on. Thank you.
   
   Finally I was able to narrow down this behaviour: This obviously depends on 
the JVM I'm using...
   Eclipse was configured to use OpenJDK 11 (11.0.20.1 to be precisely) whereas 
on the command-line OpenJDK 20 (20.0.2) was active. Switching this to also Java 
11 makes the error disappear.
   
   But: When I use Java 20 in Eclipse to execute the JUnit tests in 
tika-server-core, 8 of 75 (with 7 skipped) now fail as they did on the 
command-line. Still investigating...


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: dev-unsubscr...@tika.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org



[GitHub] [tika] theit commented on pull request #1345: TIKA-3948 -- migrate from javax -> jakarta

2023-09-18 Thread via GitHub


theit commented on PR #1345:
URL: https://github.com/apache/tika/pull/1345#issuecomment-1723437367

   > Separately, we're doing a better job of moving most dependencies into 
dependencyManagement and using boms. Should we get rid of the version numbers 
in the elements and just control the versions via the boms where possible?
   
   +1
   IMHO it would also be nice to get rid of properties whose only purpose is to 
define a version that is then later used only in the dependency management 
section. Example: `jackson.version`
   
   Another example is `log4j2.version`, defined in `tika-parent/pom.xml`. It's 
used in a couple of other pom.xml files, but unnecessary there because the 
corresponding dependencies are already referenced/defined via the import BOM.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: dev-unsubscr...@tika.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org



[GitHub] [tika] theit commented on pull request #1345: TIKA-3948 -- migrate from javax -> jakarta

2023-09-18 Thread via GitHub


theit commented on PR #1345:
URL: https://github.com/apache/tika/pull/1345#issuecomment-1722932587

   I was working locally for myself on the migration of Tika from javax to 
jakarta before I found out this PR and now have a few comments/hints, 
especially WRT to `tika-parent/pom.xml`:
   
   1. The different CXF dependencies could be replaced by the import BOM.
   2. The same mechanism could be used for SLF4J that also offers a BOM; no 
need to manage three dependencies.
   3. On my tests I had to add an exclusion for `xml-apis:xml-apis` in the 
dependency manangement section for `xerces:xercesImpl`. Otherwise I get tons of 
errors in my IDE (Eclipse) because the package `org.xml.sax` is accessible from 
more than one module: ``, `java.xml`. Haven't you seen them too?
   4. Running `mvn verify` fails in Tika server core:
   ```
   [ERROR] Tests run: 4, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 
0.054 s <<< FAILURE! -- in org.apache.tika.server.core.StackTraceTest
   [ERROR] org.apache.tika.server.core.StackTraceTest.testEmptyParser -- Time 
elapsed: 0.006 s <<< FAILURE!
   org.opentest4j.AssertionFailedError: bad type: /tika ==> expected: <200> but 
was: <500>
at 
org.junit.jupiter.api.AssertionFailureBuilder.build(AssertionFailureBuilder.java:151)
   (...)
   ```
   However, executing the (single) test in Eclipse, this works. Also when I let 
Eclipse run all JUnit tests in that module. Do you have an idea what is causing 
this?


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: dev-unsubscr...@tika.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org