Replica Package
Paper: A Topic Guided Pointer-Generator Model for Generating Natural Language Code Summaries
Describe
We provide data and files used in paper as follows:
The Sequential Data for Dataset:
| File List | Brief Description |
|---|---|
| ToPNN-dataset/train/ | sequential data for original training data (i.e., class, method, code topics, code summary) |
| ToPNN-dataset/valid/ | sequential data for original validation data (i.e., class, method, code topics, code summary) |
| ToPNN-dataset/test/ | sequential data for original test data (i.e., class, method, code topics, code summary) |
The Packed Data for Dataset:
| File List | Brief Description |
|---|---|
| ToPNN-dataset/pkl/dats.tok | vocabulary for code tokens |
| ToPNN-dataset/pkl/smls.tok | vocabulary for code topics |
| ToPNN-dataset/pkl/coms.tok | vocabulary for code summary |
| ToPNN-dataset/pkl/dataset.pkl | packed data for method, code topics, code summary |
Source Code of Our Model
| File List | Brief Description |
|---|---|
| ToPNN-sourcecode/ToPNN-full | ToPNN model |