| # CodeReality-1T Data Directory | |
| ## Location | |
| The complete 3TB dataset is located at: | |
| ``` | |
| /mnt/z/CodeReality_Final/unified_dataset/ | |
| ``` | |
| ## Contents | |
| - **52,692 JSONL files** containing 397,475 repositories | |
| - **Total size**: 3.05 TB uncompressed | |
| - **Format**: JSONL (JSON Lines) with complete repository metadata | |
| ## File Structure | |
| Each JSONL file contains repositories with: | |
| - Source code files with full paths | |
| - Git commit history and messages | |
| - Issue tracking data | |
| - Repository metadata (stars, forks, topics) | |
| - License information | |
| - Enhanced Blueprint metadata | |
| ## Usage | |
| To access the data programmatically: | |
| ```python | |
| import json | |
| import os | |
| data_dir = "/mnt/z/CodeReality_Final/unified_dataset" | |
| for filename in os.listdir(data_dir): | |
| if filename.endswith('.jsonl'): | |
| with open(os.path.join(data_dir, filename), 'r') as f: | |
| for line in f: | |
| repo_data = json.loads(line) | |
| # Process repository data | |
| print(repo_data.get('name', 'Unknown')) | |
| ``` | |
| ## Integrity Verification | |
| Use the dataset index file for verification: | |
| ```bash | |
| # Verify against index | |
| python3 -c " | |
| import json | |
| with open('../analysis/dataset_index.json', 'r') as f: | |
| index = json.load(f) | |
| print(f'Total files: {len(index[\"files\"])}') | |
| print(f'Total repositories: {sum(f[\"repository_count\"] for f in index[\"files\"])}') | |
| " | |
| ``` | |
| ## See Also | |
| - Dataset Card: `../docs/DATASET_CARD.md` | |
| - Analysis Results: `../analysis/metrics.json` | |
| - Evaluation Subset: `../eval/subset/` |