Pseudogen
A tool to automatically generate pseudo-code from source code.
Installation
Requirements
$ sudo apt-get install git libboost-all-dev autoconf automake autotools-dev libtool zlib1g-dev cmake
$ pip install nltk
for Mac OS X users:
GIZA++
is written for Linux, so you may need to do some modifications to install.
http://catherinegasnier.blogspot.jp/2014/04/install-giza-107-on-mac-osx-1092.html
$ git clone https://github.com/delihiros/pseudogen.git
$ cd pseudogen
$ ./tool_setup.sh
Usage
Download and extract corpus from annotated Django source code.
$ mkdir data
$ cd data
$ wget -O- http://ahclab.naist.jp/pseudogen/en-django.tar.gz | tar zxvf -
$ mv en-django/all.* .
$ ./../train-pseudogen.sh -p all.code -e all.anno
$ ./../run-pseudogen.sh -f tune/travatar.ini
$ # input Python code you want to translate
$ # in some environments, you may need to press Ctrl+D few times in order to start tranlating
How does Pseudogen work?
Papers
Tools Used
GIZA++
to make alignmentTravatar
to train Tree-to-String machine translation modelmteval
to evaluate