Skip to content

Create doc2code #31

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 4 commits into
base: master
Choose a base branch
from
Open
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 15 additions & 0 deletions _requests_for_research/doc2code
Original file line number Diff line number Diff line change
@@ -0,0 +1,15 @@
---
title: 'Doc2Code'
summary: "Given a docstring, generate the source code of a program that satisfies program description in the docstring. Or alternatively, given source code, generate docstrings that describe the function of the source code. If you're feeling extra ambitious, acquire/generate test cases for each triplet and then generate the source code of programs that satisfy test cases as well."
difficulty: 3 # out of 3
---


<p> The (extremely) ambitious goal of this request is to solve the problem
of generating code from docstrings, or to alternatively solve the problem of generating docstrings from code.
It is outside the reach of current machine learning algorithms.
However, <a href="https://github.com/Avmb">Antonio Valerio Miceli Barone</a>
has collected <a href="https://github.com/EdinburghNLP/code-docstring-corpus">150,370 triples of function declarations,
docstrings and bodies</a>. Refer to the paper https://arxiv.org/abs/1707.02275 for a baseline and an explanation behind
the motivation of the dataset. If you're feeling extra ambitious, first figure out a way to acquire/generate test cases
for each of the triplets, and then use those test cases to verify that the generated code is functionally correct. </p>