Upload pdf_extractor.py with huggingface_hub

pdf_extractor.py

@@ -0,0 +1,74 @@
+"""
+PDF text extraction for user-uploaded datasheets.
+Uses PyMuPDF (fitz) for robust PDF parsing.
+"""
+
+from __future__ import annotations
+
+import logging
+from pathlib import Path
+
+logger = logging.getLogger(__name__)
+
+
+def extract_text_from_pdf(file_path: str | Path) -> str:
+    """
+    Extract all text from a PDF file.
+    Returns concatenated text from all pages.
+    """
+    try:
+        import fitz  # PyMuPDF
+    except ImportError:
+        logger.error("PyMuPDF not installed. Run: pip install pymupdf")
+        return ""
+
+    try:
+        doc = fitz.open(str(file_path))
+        pages_text = []
+        for page_num, page in enumerate(doc):
+            text = page.get_text()
+            if text.strip():
+                pages_text.append(f"--- Page {page_num + 1} ---\n{text}")
+        doc.close()
+
+        full_text = "\n".join(pages_text)
+        logger.info(
+            "Extracted %d chars from %d pages of %s",
+            len(full_text), len(pages_text), file_path,
+        )
+        return full_text
+
+    except Exception as exc:
+        logger.error("PDF extraction failed for %s: %s", file_path, exc)
+        return ""
+
+
+def extract_text_from_bytes(file_bytes: bytes, filename: str = "upload.pdf") -> str:
+    """
+    Extract text from PDF bytes (for Gradio file upload handling).
+    """
+    try:
+        import fitz
+    except ImportError:
+        logger.error("PyMuPDF not installed.")
+        return ""
+
+    try:
+        doc = fitz.open(stream=file_bytes, filetype="pdf")
+        pages_text = []
+        for page_num, page in enumerate(doc):
+            text = page.get_text()
+            if text.strip():
+                pages_text.append(f"--- Page {page_num + 1} ---\n{text}")
+        doc.close()
+
+        full_text = "\n".join(pages_text)
+        logger.info(
+            "Extracted %d chars from %d pages of %s",
+            len(full_text), len(pages_text), filename,
+        )
+        return full_text
+
+    except Exception as exc:
+        logger.error("PDF bytes extraction failed: %s", exc)
+        return ""