polytext

Doc Utils

A Python package for document conversion and text extraction.

Features

• Convert various document formats (DOCX, ODT, PPT, etc.) to PDF
• Extract text from PDF, Markdown, IMAGE, and audio files
• Support for both local files and S3/GCS cloud storage
• Multiple PDF parsing backends (PyPDF, PyMuPDF)
• Transcribe audio & video files (local or cloud) to text/markdown
• Extract YouTube video transcripts
• Extract text from URLs

Installation

# Library only – assumes system requirements are already present
pip install polytext

Heads-up: Polytext’s PDF generator relies on [WeasyPrint] under the hood.
The PyPI wheel contains only Python code; you still need WeasyPrint’s native libraries (Pango, Cairo, GDK-PixBuf, HarfBuzz, Fontconfig) installed at the OS level.

System requirements

Requirement	Notes	macOS (Homebrew)	Ubuntu / Debian
Python	Supported on 3.11 – 3.13 WeasyPrint still requires its native libraries	brew install python@3.11	sudo apt install python3.11
WeasyPrint – native stack	installs Pango, Cairo, etc.	brew install weasyprint	sudo apt install weasyprint
LibreOffice	used for Office → PDF conversion	brew install --cask libreoffice	sudo apt install libreoffice

Usage

Converting Documents to PDF

from polytext import convert_to_pdf, ConversionError

try:
    # Convert a document to PDF
    pdf_path = convert_to_pdf('input.docx', 'output.pdf')
    print(f"PDF saved to: {pdf_path}")
except ConversionError as e:
    print(f"Conversion failed: {e}")

Features that require the API key for Google Gemini are:

• audio
• video
• image
• youtube

from polytext.loader.base import BaseLoader

llm_api_key = "your_google_gemini_api_key"  # Set your Google Gemini API key here

# Instantiate the loader 
loader = BaseLoader(llm_api_key=llm_api_key)

Text or Markdown Extraction

from polytext.loader.base import BaseLoader

markdown_output = False # Change if you want to extract text as markdown
source = "local" # Change to "cloud" if you want to extract from cloud storage (s3 or GCS)

# Instantiate the loader (optionally set markdown_output, llm_api_key, etc.)
loader = BaseLoader(markdown_output=markdown_output, source=source)

# Extract text from a local file
result = loader.get_text(input_list=["/path/to/document.docx"])
print(result["text"])
# Extract text from cloud file
result = loader.get_text(input_list=["s3://your-bucket/path/to/document.docx"])
print(result["text"])

# Extract text from a markdown file (local)
result = loader.get_text(input_list=["/path/to/document.md"])
print(result["text"])
# Extract text from cloud file
result = loader.get_text(input_list=["s3://your-bucket/path/to/document.md"])
print(result["text"])

# Extract text from an audio file (local)
result = loader.get_text(input_list=["/path/to/audio.mp3"])
print(result["text"])
# Extract text from cloud file
result = loader.get_text(input_list=["s3://your-bucket/path/to/audio.mp3"])
print(result["text"])

# Extract text from a video file (local)
result = loader.get_text(input_list=["/path/to/video.mp4"])
print(result["text"])
# Extract text from cloud file
result = loader.get_text(input_list=["s3://your-bucket/path/to/video.mp4"])
print(result["text"])

# Extract text from Image (local)
result = loader.get_text(input_list=["/path/to/image.jpg"])
print(result["text"])
# Extract text from cloud file
result = loader.get_text(input_list=["s3://your-bucket/path/to/image.jpg"])
print(result["text"])

# Extract transcript from a YouTube video
result = loader.get_text(input_list=["https://www.youtube.com/watch?v=xxxx"])
print(result["text"])

# Extract text from a URL
result = loader.get_text(input_list=["https://www.domain-name.com/path"])
print(result["text"])

S3 authentication

By default, Polytext uses the standard boto3 credential chain when loading s3:// inputs (environment variables, AWS profiles, IAM roles, and other boto3-supported providers).

For runtimes that need to assume an AWS role through Google OIDC, STS web identity authentication can be enabled explicitly:

from polytext.loader.base import BaseLoader

loader = BaseLoader(
    aws_auth_mode="sts_web_identity",
    aws_role_arn="arn:aws:iam::111122223333:role/ExampleRole",
    aws_region="eu-central-1",
    aws_role_session_name="polytext-session",
    gcp_id_token_audience="example-gcp-audience",
)

The same configuration can also come from environment variables:

POLYTEXT_AWS_AUTH_MODE=sts_web_identity
AWS_ROLE_ARN=arn:aws:iam::111122223333:role/ExampleRole
AWS_REGION=eu-central-1
AWS_ROLE_SESSION_NAME=polytext-session
GCP_ID_TOKEN_AUDIENCE=example-gcp-audience
GOOGLE_APPLICATION_CREDENTIALS=/absolute/path/to/service_account.json

Polytext uses the temporary STS credentials only to create the S3 client. It does not export them to os.environ and does not reset boto3's global session.

License

MIT Licence