[Feature]: If page has text, force OCR and rasterize page

[mikejokic]

Describe the proposed feature

My multi-page pdf pages contains a mix of digital and scanned images that can be OCR'd.

In this case, redo-ocr is the current option to blank out certain areas of the page from OCR while avoiding rasterizing. However, it is possible for these results to be poorer than force-ocr on that same page. Force-ocr however rastertizes the page and can cause increase in file size. Force-ocr also enables this setting for all pages regardless of whether it may be the best option (if page has no text but only images, normal workflow w/o full page raster is the best option)

So, can we detect if page has text and then only force-ocr? If page has no text, ocrmypdf proceeds as normal and only overlays text layer on top of original page. If page has text, force-ocr and rasterize.

Sorry if this is already possible, but seems like force-ocr applies to all pages if enabled. @jbarlow83 Thank you for the assist!

[mikejokic]

bumping this @jbarlow83 I know you must be busy - but is there a way to selectively apply force-ocr to pages that have text, and proceed normally for pages that are clean?