Mobile devices and barcodes

[Bernard Devlin]

I was intrigued by this topic, so did a little research myself.  First of
all, I'm not particularly inclined to jump through Apple's hoops when it
comes to installing applications on a phone I own.  So, I only looked into
doing this for Android.

I came across this 6 line piece of code for reading barcodes on Android:
http://www.mattcutts.com/blog/android-barcode-scanner/

This requires that the Android device has the Android Scripting Environment
installed.
http://google-opensource.blogspot.com/2009/06/introducing-android-scripting.html

Here's information on installing the ASE within the Android emulator:
http://www.akeric.com/blog/?p=879

It's even possible to run scripts within ASE by snapshotting them as QR
barcodes (the script is encoded in the QR barcode):
http://www.submergedspaceman.com/blog/2010/02/16/android-scripting-environment-is-incredible-part-i/

How to access this from within Livecode (given that shell() is inoperable
on Android/iPhone)?  It _might_ work by using the "launch URL" command,
giving a shell script that is recognised by ASE.

Of course, this solution is not going to be of use to those wanting to
distribute a barcode scanner Livecode app to the general public.  However,
in a specific client environment it may be possible to concoct a solution.

Concocting this solution (indeed, investigating that it is possible) is
probably less complicated than writing an image-processor/barcode reader
oneself.  Furthermore, if it is possible, it would open up a world of
possibilities on Android.

Here's a list of tutorials on other things that can be done within the ASE
(It's pretty impressive):
http://code.google.com/p/android-scripting/wiki/Tutorials

Hope that is of some use/interest to someone.

Bernard


On Wed, Dec 7, 2011 at 6:00 PM, stephen barncard <
stephenREVOLUTION2 at barncard.com> wrote:

> The code you want would be like a OCR application, and would require a bit
> of special processing to work.
> Hi contrast, sharpening, THEN the character recognition, which should work
> at different angles.
>
> My point is this is not trivial code.
>
>