Taking inspiration from the old "6 lines" barcode scanner I've written
a pretty simple script to scan and add books straight to your Google
Library using the Google Book Search API exposed by the SL4A
gdata.books module.

http://craigloftus.net/blog/2010/07/29/google-books-android-script/

Its by no means pretty code, but it has made adding books pretty
efficient for me.

One word of warning... this may not work via your 3G connection as
some providers, e.g. Three, filter traffic in some way, perhaps by
user agent.

Cheers,
Craig

Craig wrote on 29/07/10 13:17:

Nice!

I just scanned a random selection of books from my desk ..

http://books.google.com/books?uid=10072237719234546165

.. and it worked great (until I exited the scanner without scanning a
book ;-)

sjb

It is doesn't really have any modern conveniences like error
handling... although I have updated the code  in the blog post to fix
the error you report.

I also have yet to come across a book that it fails to scan, so who
knows what it will do then!

Craig

On 30 July 2010 09:11, sjb <ott...@gmail.com> wrote:

Craig Loftus wrote on 30/07/10 12:08:

That's what beta testers are for ;-)

Coincidentally, the app that I wrote to talk to Google docs is now
working, as opposed to returning the 302 response I was getting
previously. I didn't change anything at my end, so maybe Google was
having some problems previously.

Anyhoo, using Google docs as an online OCR works fairly well, with this
deliberately mediocre original ..

http://www.ottaky.com/IMAG0056.jpg

.. exported as this PDF ..

http://www.ottaky.com/pdf_out.pdf

I think it's possible to export as a plain text file, so it may be
possible to knock up a halfway decent photo -> text OCR in a few lines
of code.

That'll give me something to do this afternoon(!)

sjb

sjb wrote on 30/07/10 13:37:

Turns out that cameraInteractiveCapturePicture() will only take photos
at a resolution of 640x384 with no way to increase the resolution in the
camera settings.

Boo!

sjb

sjb wrote on 30/07/10 14:04:

Also, gdata.docs.service.SUPPORTED_FILETYPES doesn't include JPE?G. It
does accept PNG, but converting JPEGs to PNGs is a can of worms I don't
want to open.

So, let's assume that you've taken a high resolution photo of some text
and uploaded it to Google Docs. Providing Google was able to OCR it, you
can now run this ..

http://groups.google.com/group/android-scripting/web/sl4a_gdocs.py

.. choose the image you uploaded and, if you're very lucky, get the text
back.

Should also work for other document types.

sjb

For those people not really interested in typing it in, I've created a
qrcode block for the program as you've posted it on your blog which you can
find at http://www.beresourceful.net/~rusty/images/bookscan.png

Presumptions: you have ASE with the Python interpreter installed, or SL4A
once that is released. Barcode Scanner is installed as well.

Instructions for people using ASE, Open ASE, Menu, Add, From Barcode, scan
in the image.

You should now have the program in your list of apps as 'bookscan.py'. Edit
this app and change the email and password as appropriate, Save and start
using.

-Rusty

On Fri, Jul 30, 2010 at 6:08 AM, Craig Loftus <craiglof...@googlemail.com>wrote: