expose AltoPdf as REST API

[jorgeveamurguia]

return xml with text content and all images in base64 format

format XML as :
"+ content +""+ base64imagesxml +"

[coveralls]

Coverage decreased (-0.06%) to 37.742% when pulling 1460d16 on jveamurguia:master into b2bd435 on kermitt2:master.

[lfoppiano]

I'm not sure we want to expose an internal / transportation format at high level such at the API. You write a separate service, that uses grobid-core as a dependency and exposes the information you need or use the information at the programmatic level.

The ALTO format chained with images in a base64 format seems, IMHO, a bit an ad-hoc solution.

Could you add some justification of why would you need such service in grobid?

[jorgeveamurguia]

It is possible that it is poorly implemented. No new service is necessary. It is better to add a parameter and optionally return image linked to the document. You're right.

I am using Grobid to extract text from PDF documents. But, I would also like to extract images too. So I have implemented a new method and I have exposed it to take image (in BASE64 format).

[kermitt2]

Hello @jorgeveamurguia

You can get the embedded images converted into .png by using the batch command. For instance:

> java -Xmx4G -jar grobid-core/build/libs/grobid-core-0.6.0-SNAPSHOT-onejar.jar -gH grobid-home -dIn ~/test/in0/ -dOut ~/test/out0 -exe processFullText 

There is a web service that is doing the same, returning everything in a big zip file, processFulltextAssetDocument, still usable but deprecated.

But if you are simply interested in raw text and images, you'd better use command lines with https://github.com/kermitt2/pdfalto

[jorgeveamurguia]

Hello @kermitt2
I want to extract images and texts. I want text to be related to image.
Position of text and position of image in text is important to me,

[lfoppiano]

I close this, as it has been documented here