A Julia package which can help porting Python code into Julia. It does a syntactical transformation of a Python file at the level of tokens.
What this package is not:
-
It is not a transpiler. Most of the time the output will be an unrunnable mix of Python and Julia.
-
It is not something which allows you to port code without knowing both languages.
-
It does not help you with the design changes you may need when porting the code.
What it does do:
-
It reduces the amount of typing you have to do for a literal port. In particular it eliminates the most repetitive parts of the typing.
-
It automates the error prone step of converting indentation-marked blocks into
end-delimited blocks. -
It retains comments, docstrings, and formatting.
-
It supports adding custom translation rules which are appropriate for your code base.
using Pkg
pkg"add CrudePythonTranslator"
Basic usage is
using CrudePythonTranslator
translate(filename)
See the translate docstring for more options.
A custom translation rule is a function which receives a sequence of
tokens and returns a possibly modified sequence of tokens. To inspect
what tokens you currently have, use the verbose keyword argument to
translate. For examples of translation rules, see the implementation
of the standard translations in the source code. Enter your custom
translations as additional positional arguments of translate, or as
a vector of functions.
Simple translations can often be generated with the simple_rule
function, which does not require detailed understanding of the token
representation.
These are used by default. If some of them do not do what you want, you can disable the default application and add the ones you like.
These translation rules are available with
using CrudePython.Library
but you need to add the ones you want to use to the translate call.
Did you find some missing translation which would be generally applicable? File an issue or make a PR. Providing a short example of an input Python file and a translated Julia file, which could be used as testcase, is particularly appreciated.
This tool is too crude to make use of large language models (GPT-4 etc.) If those turn out to be more useful tools, by all means use them.