Python Docx

Python-docx: Working with Word Documents

Overview

The python-docx library allows you to create, read, and update Microsoft Word (.docx) files programmatically. It's essential for automating document generation, template filling, and Word document manipulation in Python.

Installation

pip install python-docx

Core Concepts

Document Structure

• Document: Root object representing the entire .docx file
• Paragraphs: Text blocks in the document
• Runs: Individual text segments within a paragraph (can have different formatting)
• Tables: Structured data with rows/cells
• Sections: Document sections with different headers/footers

Common Operations

Loading and Saving Documents

from docx import Document

# Open existing document
doc = Document('template.docx')

# Create new document
doc = Document()

# Save document
doc.save('output.docx')

Working with Paragraphs and Text

# Access all paragraphs
for para in doc.paragraphs:
    print(para.text)

# Add new paragraph
new_para = doc.add_paragraph('Text here')

# Add text with formatting
run = new_para.add_run('Bold text')
run.bold = True

Finding and Replacing Text

# Search through all paragraphs and runs
for para in doc.paragraphs:
    for run in para.runs:
        if '{{PLACEHOLDER}}' in run.text:
            run.text = run.text.replace('{{PLACEHOLDER}}', 'replacement')

# Note: Text can be split across multiple runs!
# A safer approach:
full_text = ''.join(run.text for run in para.runs)
if '{{PLACEHOLDER}}' in full_text:
    # Clear all runs and recreate with replacement
    for run in para.runs:
        run.text = ''
    para.text = full_text.replace('{{PLACEHOLDER}}', 'replacement')

Working with Tables

# Access table rows and cells
for table in doc.tables:
    for row in table.rows:
        for cell in row.cells:
            print(cell.text)
            # Replace text in cell
            cell.text = cell.text.replace('{{PLACEHOLDER}}', 'value')

Handling Multiple Runs

The trickiest part of python-docx is that text in a paragraph can be split across multiple "runs" (especially in templates). For placeholder replacement:

def replace_text_in_paragraph(para, old_text, new_text):
    """Replace text handling split runs"""
    if old_text in para.text:
        # Clear existing runs
        for run in para.runs:
            run.text = ''
        # Add new text as single run
        para.text = new_text

Important Considerations

1. Text Fragmentation: When you open a .docx file (especially templates), the same logical text might be split across multiple run objects due to formatting. Always check the full paragraph text before manipulating individual runs.

1. Preserve Formatting: When replacing text, consider whether you want to preserve the original formatting or apply new formatting.

1. Iterating Safely: If modifying document structure while iterating, make copies of lists first:

``python paragraphs = list(doc.paragraphs) # Create copy for para in paragraphs: # Safe to modify now ``

1. Headers/Footers: Accessed via doc.sections[0].header and doc.sections[0].footer

Example: Template Filling

from docx import Document

def fill_template(template_path, replacements, output_path):
    doc = Document(template_path)

    # Replace in paragraphs
    for para in doc.paragraphs:
        for key, value in replacements.items():
            if key in para.text:
                para.text = para.text.replace(key, value)

    # Replace in tables
    for table in doc.tables:
        for row in table.rows:
            for cell in row.cells:
                for key, value in replacements.items():
                    if key in cell.text:
                        cell.text = cell.text.replace(key, value)

    doc.save(output_path)

Debugging Tips

• Always print para.text (full paragraph) to see actual content vs individual run.text
• Check len(para.runs) to see if text is fragmented
• Use doc.element.xml to inspect underlying XML structure if needed

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: cxcscmu
• Source: cxcscmu/SkillLearnBench
• License: MIT
• Homepage: https://cxcscmu.github.io/SkillLearnBench

Install and usage instructions live in the source repository linked above.