Difference between revisions of "Ideas for Google Summer of Code/Python library"

From Apertium
Jump to navigation Jump to search
 
(One intermediate revision by one other user not shown)
Line 1: Line 1:
Either of these provide meaningful signal. So pick one (or both)!


==Coding challenge 1==


Working installation of [https://github.com/apertium/apertium-python apertium] via a <code>setup.py</code> file in a Windows environment.
==Coding challenge==

'''Note:''' You have to parse and run the mode files with the binaries - you can't use the apertium Bash helper script. It can be done similar to [https://github.com/apertium/apertium-simpleton/blob/master/src/simpleton.cpp#L67 simpleton.cpp#L67] or [https://github.com/apertium/apertium-omegat-native/blob/master/src/com/tinodidriksen/omegat/apertiumnative/ApertiumNative.java#L210 ApertiumNative.java#L210] + [https://github.com/apertium/apertium-omegat-native/blob/master/src/com/tinodidriksen/omegat/apertiumnative/ApertiumNative.java#L64 ApertiumNative.java#L64]

==Coding challenge 2==

Develop a proof-of-concept SWIG API for lttoolbox, it should let you do of the following:


* Proof-of-concept SWIG API for lttoolbox, it should let you do the following:
*# Create a transducer
*# Create a transducer
*# Make <tt>show()</tt> return a list of tuples (in, out, in, out)
*# Make <tt>show()</tt> return a list of tuples (in, out, in, out)

Latest revision as of 21:12, 19 March 2019

Either of these provide meaningful signal. So pick one (or both)!

Coding challenge 1[edit]

Working installation of apertium via a setup.py file in a Windows environment.

Note: You have to parse and run the mode files with the binaries - you can't use the apertium Bash helper script. It can be done similar to simpleton.cpp#L67 or ApertiumNative.java#L210 + ApertiumNative.java#L64

Coding challenge 2[edit]

Develop a proof-of-concept SWIG API for lttoolbox, it should let you do of the following:

    1. Create a transducer
    2. Make show() return a list of tuples (in, out, in, out)
    3. Perform morphological analysis using FSTProcessor
  • Improve the SWIG API with Python wrapper functions that let you do
    1. t = Transducer([(0,"a",1), (1,"b",2)], [2]) where [2] is the list of finals and 0 is the initial state