Compare/review PDF vs OCR output

[interrogator]

Law history project is just one of many possible projects where the input data is the output of an OCR process.

For such corpora, ideally, buzzword provides an interface that displays the original PDF beside its plaintext. Users should be able to submit changes (i.e. corrections) and/or add metadata tags to the text. In terms of implementation, the text displayed on the right should use a good high-level lib for this (martor right now). Needed features:

• Corpus model updated to handle a parallel 'PDF corpus'
• the compare/<slug> endpoint, possibly with page/pages query params for starting at a certain place
• Navigation between pages. Essentially (<<first <prev next> last>>) [ goto ] is good enough for now.
• Admin interface for viewing/accepting submitted corrections
• Email service to tell users the status of their changes/tell admins that there are new suggestions

This should be built with knowledge that the explorer and compare interfaces will eventually be linked via #40

[interrogator]

No email yet, but everything else done