🐫 CAMEL: The first and the best multi-agent framework. Finding the Scaling Law of Agents. https://www.camel-ai.org
Python agentai-societiesartificial-intelligencecommunicative-aicooperative-aideep-learninglarge-language-modelsmulti-agent-systemsnatural-language-processing
<a id="camel.loaders.markitdown"></a>
<a id="camel.loaders.markitdown.MarkItDownLoader"></a>
## MarkItDownLoader
```python
class MarkItDownLoader:
```
MarkitDown convert various file types into Markdown format.
Supported Input Formats:
- PDF
- Microsoft Office documents:
- Word (.doc, .docx)
- Excel (.xls, .xlsx)
- PowerPoint (.ppt, .pptx)
- EPUB
- HTML
- Images (with EXIF metadata and OCR support)
- Audio files (with EXIF metadata and speech transcription)
- Text-based formats:
- CSV
- JSON
- XML
- ZIP archives (iterates over contents)
- YouTube URLs (via transcript extraction)
<a id="camel.loaders.markitdown.MarkItDownLoader.__init__"></a>
### __init__
```python
def __init__(
self,
llm_client: Optional[object] = None,
llm_model: Optional[str] = None
):
```
Initializes the Converter.
**Parameters:**
- **llm_client** (Optional[object]): Optional client for LLM integration. (default: :obj:`None`)
- **llm_model** (Optional[str]): Optional model name for the LLM. (default: :obj:`None`)
<a id="camel.loaders.markitdown.MarkItDownLoader._validate_format"></a>
### _validate_format
```python
def _validate_format(self, file_path: str):
```
Validates if the file format is supported.
**Parameters:**
- **file_path** (str): Path to the input file.
**Returns:**
bool: True if the format is supported, False otherwise.
<a id="camel.loaders.markitdown.MarkItDownLoader.convert_file"></a>
### convert_file
```python
def convert_file(self, file_path: str):
```
Converts the given file to Markdown format.
**Parameters:**
- **file_path** (str): Path to the input file.
**Returns:**
str: Converted Markdown text.
<a id="camel.loaders.markitdown.MarkItDownLoader.convert_files"></a>
### convert_files
```python
def convert_files(
self,
file_paths: List[str],
parallel: bool = False,
skip_failed: bool = False
):
```
Converts multiple files to Markdown format.
**Parameters:**
- **file_paths** (List[str]): List of file paths to convert.
- **parallel** (bool): Whether to process files in parallel. (default: :obj:`False`)
- **skip_failed** (bool): Whether to skip failed files instead of including error messages. (default: :obj:`False`)
**Returns:**
Dict[str, str]: Dictionary mapping file paths to their
converted Markdown text.