mirror of
https://github.com/Alfresco/alfresco-transform-core.git
synced 2025-08-14 17:58:27 +00:00
7ca8a483ad
Adding extra error messages to the t-config checking, for the case where a pipeline specifies source and target mimetypes that cannot be provided by the step transformers, so that it will be clearer that the pipeline t-config is wrong. In the case of the AI rendition tests the AI-transform t-config has a pipeline that uses libreoffice as a step transformer, to transform some source to text/plain before asking AWS_AI to process it into the final mimetype. However libreoffice does not convert to text/plain. What is happening is that the request was still being sent to the all-in-one t-engine that contains libreoffice, and it workout that it should be using the tika transformer. As a result the pipeline works by accident. The size check that was commented out (uncommented now) was just finding out that libreoffice was unable to do the transform and was reporting it. officeToComprehendPiiEntityTypesViaText is the pipeline with the error.