Merge pull request #27 from ks6088ts-labs/feature/issue-26_extract-da…

…ta-from-pdf

add a CLI to extract data from PDF

README.md

@@ -52,3 +52,29 @@ To publish the docker image to Docker Hub, you need to set the following secrets
 gh secret set DOCKERHUB_USERNAME --body $DOCKERHUB_USERNAME
 gh secret set DOCKERHUB_TOKEN --body $DOCKERHUB_TOKEN
 ```
+
+## scripts
+
+### pdf_cli.py
+
+```shell
+# help
+poetry run python scripts/pdf_cli.py --help
+
+# Convert a PDF file to a Markdown file.
+poetry run python scripts/pdf_cli.py pdf2md \
+    --in-pdf "./datasets/sample.pdf" \
+    --out-md "./datasets/sample.md" \
+    --verbose
+
+# Dump the table of contents (TOC) of a PDF file.
+poetry run python scripts/pdf_cli.py toc \
+    --in-pdf "./datasets/sample.pdf" \
+    --verbose
+
+# Dump tables of the specified page of a PDF file.
+poetry run python scripts/pdf_cli.py tables \
+    --in-pdf "./datasets/sample.pdf" \
+    --page-number 123 \
+    --verbose
+```

scripts/pdf_cli.py

@@ -0,0 +1,76 @@
+import logging
+import pathlib
+
+import pymupdf
+import pymupdf4llm
+import typer
+
+app = typer.Typer(
+    add_completion=False,
+)
+
+logger = logging.getLogger(__name__)
+
+
+@app.command(
+    help="Convert a PDF file to a Markdown file.",
+)
+def pdf2md(
+    in_pdf: str = typer.Option(..., help="Path to the input PDF file."),
+    out_md: str = typer.Option(..., help="Path to the output Markdown file."),
+    verbose: bool = typer.Option(True, help="Verbose mode."),
+):
+    if verbose:
+        logging.basicConfig(level=logging.DEBUG)
+    try:
+        logger.info(f"Converting {in_pdf} to {out_md}")
+        md_text = pymupdf4llm.to_markdown(
+            doc=in_pdf,
+        )
+        pathlib.Path(out_md).write_bytes(md_text.encode())
+    except Exception as e:
+        logger.error(e)
+
+
+@app.command(
+    help="Dump the table of contents (TOC) of a PDF file.",
+)
+def toc(
+    in_pdf: str = typer.Option(..., help="Path to the input PDF file."),
+    verbose: bool = typer.Option(True, help="Verbose mode."),
+):
+    if verbose:
+        logging.basicConfig(level=logging.DEBUG)
+    try:
+        doc = pymupdf.open(in_pdf)
+        # https://pymupdf.readthedocs.io/en/latest/document.html#Document.get_toc
+        toc = doc.get_toc()
+        for i, (level, title, page) in enumerate(toc):
+            logger.info(f"{i}: {level}, {title}, {page}")
+    except Exception as e:
+        logger.error(e)
+
+
+@app.command(
+    help="Dump tables of the specified page of a PDF file.",
+)
+def tables(
+    in_pdf: str = typer.Option(..., help="Path to the input PDF file."),
+    page_number: int = typer.Option(0, help="Page number."),
+    verbose: bool = typer.Option(True, help="Verbose mode."),
+):
+    if verbose:
+        logging.basicConfig(level=logging.DEBUG)
+    try:
+        doc = pymupdf.open(in_pdf)
+        page = doc[page_number - 1]  # 0-based index
+        # https://pymupdf.readthedocs.io/en/latest/page.html#Page.find_tables
+        tables = page.find_tables()
+        for i, table in enumerate(tables):
+            logger.info(f"Table {i}: {table.to_markdown()}")
+    except Exception as e:
+        logger.error(e)
+
+
+if __name__ == "__main__":
+    app()