Bruno Thomas

Thank you @Skn0tt for your reply. We're still stuck on this. What we tried: - create a simple [vite](https://vitejs.dev/) app with cron-parser and luxon. It worked ok ; - so...

A workaround that we found to remove the blocker is to remove the dependency on `luxon`. We added the types in dev dependencies: ```json "devDependencies": { // ... "@types/luxon": "^3.3.7",...

normally it is done with `make major/minor/patch` but I'm not sure that it will work with a3/a4 :/ I don't understand what has been done so far in dev branch....

the result of extraction is stored into TikaDocumentSource class : ```java public class TikaDocumentSource { public final Metadata metadata; public final byte[] content; public TikaDocumentSource(final Metadata metadata, final byte[] content)...

maybe needs #1385

there is still an issue about parallelization of artifact tasks. see https://github.com/ICIJ/datashare/pull/1580 see #1598

See how it will perform with actual filtering on the consumers side? To be reopen for optimisation (if tasks are ending in deadletter queues).

See with TaskCancel/TaskCancelled messages to udpate tasks persistence map.

For now we are using in memory documents when indexing them. it seems [in this thread](https://discuss.elastic.co/t/how-to-bulk-load-huge-json-docs-from-files-using-java-client/142607) that we cannot stream the data to ES.

see how to deploy the jar on maven central