Datasets:
Tasks:
Text Generation
Formats:
parquet
Languages:
English
Size:
10K - 100K
ArXiv:
Tags:
code
DOI:
License:
Update README.md
Browse files
README.md
CHANGED
@@ -429,12 +429,14 @@ configs:
|
|
429 |
# LCA (Bug Localization)
|
430 |
This is the data for **Bug Localization** benchmark as part of LCA.
|
431 |
The bug localization problem can be formulated as follows: given an issue with a bug description and a repository snapshot in a state where the bug is reproducible, identify the files within the repository that need to be modified to address the reported bug.
|
|
|
432 |
The dataset provides all required components for evaluation of **Bug Localization** approaches in real project-level large-scale data collected from GitHub, including:
|
433 |
-
* Bug
|
434 |
-
* Repository content with the commit
|
435 |
* List of files that should be changed in order to solve the bug
|
436 |
* Other additional data and metrics, that can be useful in approach development
|
437 |
|
|
|
438 |
The following sections will describe the utilities around the dataset as well as detailed dataset content description.
|
439 |
|
440 |
## How-to
|
@@ -494,9 +496,8 @@ TODO: some overall structure or repo
|
|
494 |
|
495 |
### Bug localization data
|
496 |
|
497 |
-
|
498 |
-
|
499 |
-
Each example has the following fields:
|
500 |
|
501 |
| **Field** | **Description** |
|
502 |
|:------------------:|:----------------------------------------:|
|
@@ -507,27 +508,47 @@ Each example has the following fields:
|
|
507 |
| `comment_url` | GitHub link to comment with pull request to issue reference <br> `https://github.com/{repo_owner}/{repo_name}/pull/{pull_id}#issuecomment-{comment_id}`. |
|
508 |
| `issue_title` | Issue title. |
|
509 |
| `issue_body` | Issue body. |
|
510 |
-
| `base_sha` | Pull request base
|
511 |
-
| `head_sha` | Pull request head
|
512 |
-
| `diff_url` | Pull request diff
|
513 |
| `diff` | Pull request diff content. |
|
514 |
-
| `
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
515 |
| `changed_files_exts` | Dict from changed files extension to count. |
|
516 |
| `changed_files_count` | Number of changed files. |
|
517 |
| `java_changed_files_count` | Number of changed `.java` files. |
|
518 |
| `kt_changed_files_count` | Number of changed `.kt` files. |
|
519 |
| `py_changed_files_count` | Number of changed `.py` files. |
|
520 |
| `code_changed_files_count` | Number of changed `.java`, `.kt` or `.py` files. |
|
521 |
-
| `
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
522 |
| `issue_lines_count`* | Number of text lines in issue body (separated by \\n). |
|
523 |
| `issue_links_count` | Number of links (\[...\](...)) presented in issue body. |
|
524 |
| `issue_code_blocks_count` | Number of code blocks (\`\`\`...\`\`\`) presented in issue body. |
|
525 |
-
|
526 |
-
|
527 |
-
|
528 |
-
|
529 |
-
| `repo_license` | Licenses assigned to the repository. |
|
530 |
-
|
531 |
### Repos data
|
532 |
|
533 |
TODO: describe repos data as `.zip` archives with list of repos metadata
|
|
|
429 |
# LCA (Bug Localization)
|
430 |
This is the data for **Bug Localization** benchmark as part of LCA.
|
431 |
The bug localization problem can be formulated as follows: given an issue with a bug description and a repository snapshot in a state where the bug is reproducible, identify the files within the repository that need to be modified to address the reported bug.
|
432 |
+
|
433 |
The dataset provides all required components for evaluation of **Bug Localization** approaches in real project-level large-scale data collected from GitHub, including:
|
434 |
+
* Bug issue description
|
435 |
+
* Repository content with the commit SHA where the bug is reproducible
|
436 |
* List of files that should be changed in order to solve the bug
|
437 |
* Other additional data and metrics, that can be useful in approach development
|
438 |
|
439 |
+
The collected dataset was carefully filtered, inhances with useful metrics and, what is more, manually labeled, which assures the data quality and provides a golden subset of good examples for evaluation.
|
440 |
The following sections will describe the utilities around the dataset as well as detailed dataset content description.
|
441 |
|
442 |
## How-to
|
|
|
496 |
|
497 |
### Bug localization data
|
498 |
|
499 |
+
Each data point contains main subset of field as well as additional metrics calculated upon them.
|
500 |
+
The core fields are:
|
|
|
501 |
|
502 |
| **Field** | **Description** |
|
503 |
|:------------------:|:----------------------------------------:|
|
|
|
508 |
| `comment_url` | GitHub link to comment with pull request to issue reference <br> `https://github.com/{repo_owner}/{repo_name}/pull/{pull_id}#issuecomment-{comment_id}`. |
|
509 |
| `issue_title` | Issue title. |
|
510 |
| `issue_body` | Issue body. |
|
511 |
+
| `base_sha` | Pull request base SHA. |
|
512 |
+
| `head_sha` | Pull request head SHA. |
|
513 |
+
| `diff_url` | Pull request diff link between base and head SHA <br> `https://github.com/{repo_owner}/{repo_name}/compare/{base_sha}...{head_sha}`. |
|
514 |
| `diff` | Pull request diff content. |
|
515 |
+
| `pull_create_at` | Data of pull request creation in format yyyy-mm-ddThh:mm:ssZ. |
|
516 |
+
| `repo_stars` | Number of repo stars. |
|
517 |
+
| `changed_files`* | List of changed files parsed from diff. |
|
518 |
+
| `repo_language` | Main programming language used in the repository. |
|
519 |
+
| `repo_languages` | All programming languages used in the repository. |
|
520 |
+
| `repo_license` | Licenses assigned to the repository. |
|
521 |
+
|
522 |
+
* Excluding test files that do not contain bug causes, rather changed in order to add test for proving that bug is gone.
|
523 |
+
|
524 |
+
The metrics fields are:
|
525 |
+
| **Field** | **Description** |
|
526 |
+
|:------------------:|:----------------------------------------:|
|
527 |
+
|
528 |
| `changed_files_exts` | Dict from changed files extension to count. |
|
529 |
| `changed_files_count` | Number of changed files. |
|
530 |
| `java_changed_files_count` | Number of changed `.java` files. |
|
531 |
| `kt_changed_files_count` | Number of changed `.kt` files. |
|
532 |
| `py_changed_files_count` | Number of changed `.py` files. |
|
533 |
| `code_changed_files_count` | Number of changed `.java`, `.kt` or `.py` files. |
|
534 |
+
| `repo_symbols_count`* | Number of symbols in repository. |
|
535 |
+
| `repo_tokens_count`* | Number of tokens** in files in repository. |
|
536 |
+
| `repo_lines_count`* | Number of lines in files in repository. |
|
537 |
+
| `repo_files_without_tests_count`* | Number of files in repository. |
|
538 |
+
| `changed_symbols_count`* | Number of symbols in changed lines in diff. |
|
539 |
+
| `changed_tokens_count`* | Number of tokens** in changed lines in diff. |
|
540 |
+
| `changed_lines_count`* | Number of changed lines in diff (including added and deleted). |
|
541 |
+
| `changed_files_without_tests_count`* | Number of files in diff. |
|
542 |
+
| `issue_symbols_count`* | Number of symbols in issue body. |
|
543 |
+
| `issue_words_count`* | Number of words in issue body (separated by space symbols). |
|
544 |
+
| `issue_tokens_count`* | Number of tokens** in issue body. |
|
545 |
| `issue_lines_count`* | Number of text lines in issue body (separated by \\n). |
|
546 |
| `issue_links_count` | Number of links (\[...\](...)) presented in issue body. |
|
547 |
| `issue_code_blocks_count` | Number of code blocks (\`\`\`...\`\`\`) presented in issue body. |
|
548 |
+
|
549 |
+
* Excluding test files that do not contain bug causes, rather changed in order to add test for proving that bug is gone. \
|
550 |
+
** Using GPT-4 tokenizer via ticktoken.
|
551 |
+
|
|
|
|
|
552 |
### Repos data
|
553 |
|
554 |
TODO: describe repos data as `.zip` archives with list of repos metadata
|