[
https://issues.apache.org/jira/browse/TIKA-4512?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18029078#comment-18029078
]
ASF GitHub Bot commented on TIKA-4512:
--------------------------------------
tballison commented on PR #2362:
URL: https://github.com/apache/tika/pull/2362#issuecomment-3391530764
This is just the first step. It won't build fully yet.
I'm using: `mvn clean install -am -pl :tika-fetcher-file-system,tika-app
-Dcheckstyle.skip -DskipTests`
I have this running with only the file system fetcher and tika-app.
I'm struggling with how to configure the plugins. I'd like to move away from
our custom serialization stuff. So, now everything depends on jackson-databind,
which adds to dependency bloat but makes the code much cleaner.
For now, I'm putting the `file-system-fetcher.json` config in the plugin
root, and reading it from there. This means that we'll have a bunch of
different *.json configs running around, which I don't like.
The current proof of concept only works when using the .jar plugin, not the
.zip plugin...
The big change here is that we can't have multiple fetchers of the same
class with different configurations like we used to.
However, users can change the configuration of a given fetcher at runtime by
passing in the `FileSystemFetcherConfig` via the ParseContext.
@nddipiazza would you be willing to take a look and maybe chat?
> Experiment with pf4j for tika-pipes
> -----------------------------------
>
> Key: TIKA-4512
> URL: https://issues.apache.org/jira/browse/TIKA-4512
> Project: Tika
> Issue Type: Task
> Reporter: Tim Allison
> Priority: Major
>
> This is proof of concept work based on [~ndipiazza] 's TIKA-4272-docker
> branch.
> I really like a lot of that work.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
