Include Dictionary metric

[M3ssman]

Description

Include Dictionary metric as textual OCR metric.

Example Implementation: https://github.com/ulb-sachsen-anhalt/ocr-pipeline/blob/master/lib/ocr_step.py#L337 ff

[M3ssman]

Please add parameter for

• language (i.e., from metadata ger => de-DE, fra => fr, en => en-GB )
• DictLT => call to languagetool (default connection: localhost:8010/v2/check, default container name digital-eval-languagetool )
• DictXX => call to XX (default connection param)

Dictionary-endpoint (connection URL, i.e. like for current language-tool implementation, or local call to spellchecker like hunspell) is currently set as default on module level.

Please add communication for unavailable endpoints (i.e., service is down, container not running, tool not installed, etc.)

Please add communication for unsupported languages related to specific endpoint (cf. ulb-ocr-odem) .
Recent languages seem to include beside German and French also Arabic and Persian.

Please review test data and fixtures from ocr-pipeline