rrwrite-extract-figures-tables

name: rrwrite-extract-figures-tables description: Extract figures and tables from analyzed repository and generate supplementary visualizations allowed_tools: - Read - Write - Bash - Glob - Grep fork_mode: fork

RRWrite: Figure and Table Extraction

Extract existing figures/tables from the analyzed repository and generate supplementary analysis visualizations. Creates manifests with priority metadata.

Overview

This skill performs two extraction tasks:

1. Repository Extraction (Priority 1): Copy existing figures/tables from the analyzed code repository
2. Supplementary Generation (Priority 2): Generate analysis visualizations from repository data

Prerequisites

Before running this skill, ensure:

• ✓ Repository analysis completed (/rrwrite-analyze-repository)
• ✓ data_tables/ directory exists with analysis outputs
• ✓ Repository path is known

Inputs

Required arguments:

• --target_dir: Manuscript output directory (e.g., manuscript/project_v1)
• --repo_path: Path to analyzed repository

Optional arguments:

• --extract: What to extract: figures, tables, or figures,tables (default: both)
• --generate: What to generate: figures, tables, or figures,tables (default: both)

Execution

# Run extraction
python scripts/rrwrite-extract-figures-tables.py \
    --repo-path {repo_path} \
    --manuscript-dir {target_dir} \
    --extract figures,tables \
    --generate figures \
    --verbose

Output Structure

Creates the following directory structure:

{target_dir}/
├── figures/
│   ├── from_repo/              # Priority 1: Original repository figures
│   │   ├── workflow_diagram.pdf
│   │   ├── results_plot.png
│   │   └── README.md           # Auto-generated index
│   ├── generated/              # Priority 2: Analysis visualizations
│   │   ├── repository_composition.png
│   │   ├── repository_composition.pdf
│   │   ├── file_size_distribution.png
│   │   └── research_topics.png
│   └── figure_manifest.json    # Metadata with priorities
├── tables/
│   ├── from_repo/              # Priority 1: Original data tables
│   │   ├── experimental_data.csv
│   │   └── benchmark_results.tsv
│   ├── generated/              # Priority 2: Analysis tables
│   │   ├── repository_statistics.tsv
│   │   └── file_inventory.tsv
│   └── table_manifest.json

Manifest Format

Figure Manifest (figures/figure_manifest.json)

{
  "version": "1.0",
  "created_at": "2025-01-15T10:30:00",
  "total_figures": 8,
  "figures_from_repo": 5,
  "figures_generated": 3,
  "figures": [
    {
      "id": "fig_repo_001",
      "path": "figures/from_repo/workflow_diagram.pdf",
      "source": "from_repo",
      "priority": 1,
      "original_path": "docs/figures/workflow.pdf",
      "recommended_sections": ["methods", "introduction"],
      "default_caption": "Workflow diagram showing analysis pipeline",
      "generating_script": "scripts/create_workflow.py"
    },
    {
      "id": "fig_gen_001",
      "path": "figures/generated/repository_composition.png",
      "source": "generated",
      "priority": 2,
      "recommended_sections": ["methods", "results"],
      "default_caption": "Repository composition by file category"
    }
  ]
}

State Tracking

Updates StateManager with extraction results:

from pathlib import Path
import sys
sys.path.insert(0, 'scripts')
from rrwrite_state_manager import StateManager

state = StateManager(output_dir="{target_dir}")

# Mark stage as in-progress
state.update_stage("figure_table_extraction", "in_progress")

# After extraction, mark as completed
state.update_figure_table_extraction(
    figures_from_repo=5,
    figures_generated=3,
    tables_from_repo=2,
    tables_generated=4,
    figure_manifest_path="figures/figure_manifest.json",
    table_manifest_path="tables/table_manifest.json",
    scripts_parsed=12
)

Validation

After extraction, validate manifests:

# Validate against schema
python scripts/rrwrite_manifest_generator.py \
    --manuscript-dir {target_dir} \
    --validate \
    --schemas-dir schemas

Using Extracted Figures in Sections

When drafting sections, query manifests for available figures:

from pathlib import Path
from rrwrite_manifest_generator import ManifestGenerator

generator = ManifestGenerator(Path("{target_dir}"))

# Get figures for specific section (prioritizes repo figures)
results_figures = generator.get_figures_for_section("results")

for fig in results_figures:
    print(f"Priority {fig['priority']}: {fig['path']}")
    print(f"Caption: {fig['default_caption']}")

Exclusion Rules

The extractor automatically excludes:

• Thumbnails and icons (patterns: *thumb*, *icon*, *logo*, *badge*)
• Build artifacts (build/, dist/, node_modules/, __pycache__/)
• Version control (.git/, .ipynb_checkpoints/)
• Oversized files (>10MB for figures, >5MB for tables)

Troubleshooting

No figures/tables found

Cause: Repository may not contain figure files or extraction patterns don't match

Solution: Check repository for figure files manually:

find {repo_path} -name "*.png" -o -name "*.pdf" -o -name "*.svg"

Missing generated figures

Cause: Repository analysis didn't create data_tables/ directory

Solution: Re-run repository analysis:

/rrwrite-analyze-repository --repo_path {repo_path} --output_dir {target_dir}

Manifest validation fails

Cause: Schema files missing or malformed manifest

Solution: Check schemas exist in schemas/ directory:

ls schemas/figure_manifest_schema.json
ls schemas/table_manifest_schema.json

Integration with Workflow

This stage runs after research and before drafting:

1. Repository Analysis → generates data_tables/
2. Planning → creates outline
3. Assessment → selects journal
4. Literature Research → finds citations
5. Figure/Table Extraction ← THIS STAGE
6. Section Drafting → uses manifests to include figures/tables
7. Assembly → embeds all figures/tables in final manuscript
8. Critique

Next Steps

After successful extraction:

1. Review figures/from_repo/README.md to verify detected figures
2. Check manifest files for correct priority assignments
3. Proceed to section drafting: /rrwrite-draft-section --section introduction

Expected Duration

• Small repositories (<50 figures): 30-60 seconds
• Medium repositories (50-200 figures): 1-2 minutes
• Large repositories (>200 figures): 2-4 minutes

Success Criteria

✓ Figures extracted and copied to figures/from_repo/ ✓ Generated figures created in figures/generated/ ✓ Tables extracted to tables/from_repo/ ✓ Manifests created with valid JSON schema ✓ StateManager updated with extraction counts ✓ Priority metadata correctly assigned (1=repo, 2=generated)