JobFulltext
This job applies the full-text indexing of the documents.
Parameters
{
"job_fulltext_scan": "all",
"batch_size": 100,
"retry_max": 5,
"force_reindex": false
}
The job_fulltext_scan parameter can take the following values:
- all: applies the indexing on every document.
- filtered: applies the indexing only on the documents matching a GQL predicate. Add a
filter parameter holding the predicate, for instance
standard.status = 'Approved'. - query: applies the indexing only on the documents matching a GQL query. Add a query parameter holding the full query.
Other parameters:
- batch_size: number of documents processed per batch before being stored in the base. If vectorization is enabled, set this parameter to 5 to avoid timeouts.
- nb_threads: number of threads used to process several files at the same time.
- retry_max: maximum number of indexing attempts for a document in error.
- force_reindex: set to
trueto reindex the documents already indexed. - force_reindex_reload_partial: set to
trueto reindex the documents reloaded from MongoDB with a partial full text. The contents reloaded with a complete full text are not impacted. - force_ocr_overwrite: set to
trueto regenerate the markdown produced by the A.I. OCR, even when it is already present on the document.