> For the complete documentation index, see [llms.txt](https://docs.flowiseai.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.flowiseai.com/integrations/langchain/document-loaders/microsoft-word.md).

# Microsoft Word

<figure><img src="/files/EaS3bHpmn4BTnznGeIjH" alt="" width="234"><figcaption></figcaption></figure>

Microsoft Word is a word processing software for creating and editing text documents. This module provides functionality to load and process Word documents using officeparser.

This module provides a sophisticated Word document loader that can:

* Load Word documents
* Extract text content
* Split content into sections
* Handle page numbering
* Process metadata per section
* Support multiple section formats
* Handle various section separators

## Inputs

### Required Parameters

* **Word File**: The Word file(s) to process (.doc, .docx)

### Optional Parameters

* **Text Splitter**: A text splitter to process the extracted content
* **Additional Metadata**: JSON object with additional metadata
* **Omit Metadata Keys**: Comma-separated list of metadata keys to omit

## Outputs

* **Document**: Array of document objects containing metadata and pageContent
* **Text**: Concatenated string from pageContent of documents

## Features

* Text extraction
* Section separation
* Metadata handling
* Error handling
* Memory-efficient processing
* Heuristic section detection
* Content filtering

## Section Detection Methods

### Pattern Recognition

The loader attempts to identify sections using common patterns:

* "Page X" markers
* "Section X" markers
* "Chapter X" markers
* Numbered sections (e.g., "1. ", "2. ")
* ALL CAPS headings
* Long underscore separators
* Long dash separators

### Fallback Mechanisms

If pattern recognition fails:

1. Split by multiple newlines
2. Split by double newlines
3. Treat content as single section

## Document Structure

Each document contains:

* **pageContent**: Extracted text content from the section
* **metadata**:
  * documentType: "word"
  * pageNumber: Sequential section number
  * Additional custom metadata

## Content Processing

* Empty sections are filtered out
* Leading/trailing whitespace removed
* Minimum content length validation
* Reasonable section count validation

## Metadata Attributes

Default attributes include:

* documentType: Type of document (string)
* pageCount: Number of pages/sections (number)
* Custom metadata from input

## Notes

* Uses officeparser for extraction
* Handles various document formats
* Intelligent section detection
* Content validation
* Memory-efficient processing
* Error handling for invalid files
* Flexible output formats
* Robust fallback mechanisms