OCR-to-RAG pipeline for life science screenshots

PaddleOCR with preprocessing, scispaCy NER, figure/table detection,
citation extraction, and chunked Markdown output with frontmatter.
Includes watch mode and notebook reprocessing.

src/ocr_pipeline/cli.py

@@ -0,0 +1,163 @@
+from __future__ import annotations
+
+import sys
+from pathlib import Path
+
+import typer
+from rich.console import Console
+from rich.table import Table
+
+from ocr_pipeline.config import settings
+from ocr_pipeline.pipeline import OCRPipeline
+from ocr_pipeline.utils.logging import get_logger, setup_logging
+from ocr_pipeline.utils.migrate import migrate_notebook
+
+app = typer.Typer(
+    name="ocr-pipeline",
+    help="OCR pipeline for screenshots -> RAG-ready Markdown",
+    add_completion=False,
+)
+console = Console()
+logger = get_logger(__name__)
+
+
+@app.callback()
+def callback(
+    config: Path | None = typer.Option(None, "--config", "-c", help="Config file path"),
+    log_level: str = typer.Option("INFO", "--log-level", "-l", help="Log level"),
+    log_file: Path | None = typer.Option(None, "--log-file", help="Log file path"),
+):
+    """OCR Pipeline - Extract text from screenshots for RAG"""
+    setup_logging(log_level, log_file)
+
+    if config:
+        settings.update_from_yaml(config)
+
+
+@app.command()
+def run(
+    input_dir: Path | None = typer.Option(None, "--input-dir", "-i", help="Input directory"),
+    output_dir: Path | None = typer.Option(None, "--output-dir", "-o", help="Output directory"),
+    workers: int | None = typer.Option(
+        None, "--workers", "-w", help="Number of worker processes"
+    ),
+    engine: str | None = typer.Option(
+        None, "--engine", "-e", help="OCR engine (paddleocr/tesseract/auto)"
+    ),
+):
+    """Run OCR pipeline on all screenshots"""
+    if input_dir:
+        settings.input.paths = [str(input_dir)]
+    if output_dir:
+        settings.output.base_directory = str(output_dir)
+    if workers:
+        settings.processing.workers = workers
+    if engine:
+        settings.ocr.engine = engine
+
+    pipeline = OCRPipeline()
+    result = pipeline.process_all()
+
+    table = Table(title="Pipeline Results")
+    table.add_column("Metric", style="cyan")
+    table.add_column("Value", style="green")
+
+    table.add_row("Total Images", str(result.total_images))
+    table.add_row("Successful", str(result.successful))
+    table.add_row("Failed", str(result.failed))
+    table.add_row("Chunks Created", str(result.chunks_created))
+    table.add_row("Output Files", str(len(result.output_files)))
+
+    console.print(table)
+
+    if result.errors:
+        console.print("\n[red]Errors:[/red]")
+        for err in result.errors:
+            console.print(f"  {err['path']}: {err['error']}")
+
+    if result.failed > 0:
+        sys.exit(1)
+
+
+@app.command()
+def watch():
+    """Watch for new screenshots and process automatically"""
+    pipeline = OCRPipeline()
+    console.print("[green]Starting watch mode...[/green]")
+    console.print("Press Ctrl+C to stop")
+    try:
+        pipeline.watch()
+    except KeyboardInterrupt:
+        console.print("\n[yellow]Stopping watcher...[/yellow]")
+
+
+@app.command()
+def reprocess(
+    notebook: Path = typer.Argument(..., help="Path to Jupyter notebook"),
+):
+    """Reprocess screenshots from a Jupyter notebook"""
+    pipeline = OCRPipeline()
+    console.print(f"[green]Migrating from notebook: {notebook}[/green]")
+    result = migrate_notebook(notebook, pipeline)
+
+    table = Table(title="Migration Results")
+    table.add_column("Metric", style="cyan")
+    table.add_column("Value", style="green")
+
+    for key, value in result.items():
+        if key != "errors":
+            table.add_row(key.capitalize(), str(value))
+
+    console.print(table)
+
+    if result.get("errors"):
+        console.print("\n[red]Errors:[/red]")
+        for err in result["errors"]:
+            console.print(f"  {err['path']}: {err['error']}")
+
+
+@app.command()
+def status():
+    """Show pipeline status and statistics"""
+    pipeline = OCRPipeline()
+    stats = pipeline.get_stats()
+
+    table = Table(title="Pipeline Status")
+    table.add_column("Metric", style="cyan")
+    table.add_column("Value", style="green")
+
+    for key, value in stats.items():
+        table.add_row(key.replace("_", " ").title(), str(value))
+
+    console.print(table)
+
+
+@app.command()
+def stats(days: int = typer.Option(30, "--days", "-d", help="Number of days to analyze")):
+    """Show processing statistics"""
+    pipeline = OCRPipeline()
+    stats = pipeline.get_stats()
+
+    table = Table(title=f"Statistics (last {days} days)")
+    table.add_column("Metric", style="cyan")
+    table.add_column("Value", style="green")
+
+    for key, value in stats.items():
+        table.add_row(key.replace("_", " ").title(), str(value))
+
+    console.print(table)
+
+
+@app.command()
+def config():
+    """Show current configuration"""
+    console.print("[cyan]Current Configuration:[/cyan]")
+    console.print(settings.model_dump_json(indent=2))
+
+
+def main():
+    app()
+
+
+if __name__ == "__main__":
+    main()