add pdf to word conversion tab

app.py

@@ -6,10 +6,13 @@ from PIL import Image
 from loadimg import load_img
 import io
 from markitdown import MarkItDown
+from docx import Document
+import pdfplumber
 
 converter = Converter()
 md = MarkItDown()
 
+
 def convert_file_to_pdf(filename: str) -> str:
     """
     Converts a markdown file to PDF format.
@@ -65,6 +68,30 @@ def convert_file_to_markdown(filename: str) -> str:
     return md.convert(filename.name).text_content
 
 
+def convert_pdf_to_word(filename: str) -> str:
+    """
+    Converts a PDF file to Word format.
+
+    Args:
+        filename: str
+            The path to the PDF file to be converted.
+
+    Returns:
+        str: The file path of the generated Word file.
+    """
+    output_path = filename.name.rsplit(".", 1)[0] + ".docx"
+    doc = Document()
+
+    with pdfplumber.open(filename.name) as pdf:
+        for page in pdf.pages:
+            text = page.extract_text()
+            if text:
+                doc.add_paragraph(text)
+
+    doc.save(output_path)
+    return output_path
+
+
 # Create individual interfaces
 file_to_pdf = gr.Interface(
     fn=convert_file_to_pdf,
@@ -90,10 +117,18 @@ file_to_markdown = gr.Interface(
     description="Convert your files to markdown format",
 )
 
+pdf_to_word = gr.Interface(
+    fn=convert_pdf_to_word,
+    inputs=gr.File(label="Upload PDF file"),
+    outputs=gr.File(label="Converted Word Document"),
+    title="PDF to Word Converter",
+    description="Convert your PDF files to Word format",
+)
+
 # Create tabbed interface
 demo = gr.TabbedInterface(
-    [file_to_pdf, file_to_image, file_to_markdown],
-    ["File to PDF", "File to Image", "File to Markdown"],
+    [file_to_pdf, file_to_image, file_to_markdown, pdf_to_word],
+    ["File to PDF", "File to Image", "File to Markdown", "PDF to Word"],
 )
 
 if __name__ == "__main__":

requirements.txt

@@ -9,4 +9,5 @@ python-pptx
 pdfitdown
 loadimg
 docling
-markitdown
+markitdown
+PyMuPDF