swarmauri_parser_textblob 0.10.0.dev1

TextBlob Parser for Swarmauri.

---

Swarmauri Parser TextBlob

TextBlob-backed parsers for Swarmauri that split text into sentences or extract noun phrases. Ships two components: TextBlobSentenceParser and TextBlobNounParser.

Features

• Sentence parser returns a Document per sentence with metadata identifying the parser.
• Noun phrase parser returns the original text plus metadata['noun_phrases'] containing the phrases discovered by TextBlob.
• Auto-downloads required NLTK corpora (punkt_tab) during initialization.

Prerequisites

• Python 3.10 or newer.
• TextBlob and its NLTK dependencies (installed automatically).
• Internet access on first run so NLTK can download tokenizer data (or pre-download via python -m textblob.download_corpora).

Installation

# pip
pip install swarmauri_parser_textblob

# poetry
poetry add swarmauri_parser_textblob

# uv (pyproject-based projects)
uv add swarmauri_parser_textblob

Sentence Parsing

from swarmauri_parser_textblob import TextBlobSentenceParser

parser = TextBlobSentenceParser()
text = "One more large chapula please. It should be extra spicy!"

sentences = parser.parse(text)
for doc in sentences:
    print(doc.content)

Noun Phrase Extraction

from swarmauri_parser_textblob import TextBlobNounParser

parser = TextBlobNounParser()
docs = parser.parse("One more large chapula please.")

for doc in docs:
    print(doc.content)
    print(doc.metadata["noun_phrases"])

Tips

• TextBlob uses simple heuristics—it works well for general English text but may struggle with domain-specific jargon.
• Download corpora once in CI/CD or container builds (python -m textblob.download_corpora) to avoid runtime downloads.
• Combine sentence and noun parsers to build structured representations of documents before vectorization or downstream NLP tasks.

Want to help?

If you want to contribute to swarmauri-sdk, read up on our guidelines for contributing that will help you get started.