Skip to main content

​
Readers Module

The Readers module provides concrete implementations for loading documents from various sources into the knowledge base.

​
Import

from praisonai.adapters import AutoReader, TextReader, MarkItDownReader, DirectoryReader

​
Quick Example

from praisonai.adapters import AutoReader

# AutoReader automatically detects source type
reader = AutoReader()

# Load from file
docs = reader.load("document.pdf")

# Load from directory
docs = reader.load("./docs/")

# Load from URL
docs = reader.load("https://example.com/page.html")

​
Features

  • Automatic source type detection and routing
  • Multiple reader implementations (Text, MarkItDown, Directory, URL, Glob)
  • Lazy loading of optional dependencies
  • Metadata preservation for loaded documents
  • Recursive directory traversal with exclusion patterns

​
Classes

​
AutoReader

Automatic reader that detects source type and routes to the appropriate reader. 
from praisonai.adapters import AutoReader

reader = AutoReader()

# Handles files, directories, URLs, and glob patterns
docs = reader.load("report.pdf")
docs = reader.load("./documents/")
docs = reader.load("https://example.com")
docs = reader.load("*.md")

​
TextReader

Simple text file reader for plain text files. 
from praisonai.adapters import TextReader

reader = TextReader()
docs = reader.load("notes.txt")
Supported Extensions: .txt, .text, .log

​
MarkItDownReader

Document reader using markitdown for rich document conversion. 
from praisonai.adapters import MarkItDownReader

reader = MarkItDownReader()
docs = reader.load("report.pdf")
Supported Extensions: .pdf, .doc, .docx, .ppt, .pptx, .xls, .xlsx, .html, .htm, .md, .markdown, .csv, .json, .xml, .jpg, .jpeg, .png, .gif, .bmp, .tiff, .webp, .mp3, .wav, .ogg, .m4a, .flac
Requires markitdown package: pip install markitdown

​
DirectoryReader

Recursively reads all files in a directory. 
from praisonai.adapters import DirectoryReader

reader = DirectoryReader(
    recursive=True,
    exclude_patterns=["*.pyc", "__pycache__", ".git", "node_modules"]
)
docs = reader.load("./project/")
Parameters:
ParameterTypeDefaultDescription
recursiveboolTrueRecursively traverse subdirectories
exclude_patternsList[str]See belowGlob patterns to exclude
Default Exclusions: *.pyc, __pycache__, .git, .svn, node_modules, *.egg-info, .env, .venv, venv

​
Methods

​
load(source, metadata=None)

Load documents from a source. Parameters:
  • source (str): File path, directory, URL, or glob pattern
  • metadata (dict, optional): Additional metadata to attach to documents
Returns: List[Document] - List of loaded documents

​
can_handle(source)

Check if the reader can handle the given source. Parameters:
  • source (str): Source to check
Returns: bool - True if the reader can handle this source

​
Example: Custom Metadata

from praisonai.adapters import AutoReader

reader = AutoReader()

# Add custom metadata to loaded documents
docs = reader.load(
    "technical_docs/",
    metadata={
        "category": "technical",
        "version": "2.0",
        "author": "engineering"
    }
)

for doc in docs:
    print(f"Source: {doc.metadata['source']}")
    print(f"Category: {doc.metadata['category']}")

​
Example: URL Reading

from praisonai.adapters.readers import URLReader

reader = URLReader()
docs = reader.load("https://docs.python.org/3/tutorial/index.html")

# Content is automatically extracted from HTML
print(docs[0].content[:500])

​
CLI Usage

praisonai knowledge add <source>
Examples: 
# Add a single file
praisonai knowledge add document.pdf

# Add all files in a directory
praisonai knowledge add ./docs/

# Add files matching a pattern
praisonai knowledge add "*.pdf"

# Add from URL
praisonai knowledge add https://example.com/page.html