Batch utility about grammark

[markfullmer]

Hi @yannis1962 , that's definitely possible, though if you're talking large corpus-sized processing, it might be better to do this in a non-web interface. Regardless, I think the main thing would be deciding things like:

• What would be the preferred interface for importing texts?
• How would any given researcher be able to define what symbols/codes would be associated with which items?
• What would be the best output format for the processed texts? A single file? Multiple files?
• What other kind of corpus-type statistics on the processed text should be provided?

I think the first step would be to add an "API" that abstracts the output of a single processed text from what is currently HTML markup, and instead provides the processed text in, say, JSON or XML format. Multiple files could then be concatenated together subsequently.

Would you be able to delineate in more detail answers to some of the questions, above?