contraction-fix 0.2.0

A fast and efficient library for fixing contractions in text with batch processing support

Contraction Fix

A fast and efficient library for fixing contractions in text. This package provides tools to expand contractions in English text while maintaining high performance and accuracy.

Features

• Fast text processing using precompiled regex patterns
• Batch processing for multiple texts with optimized performance
• Support for standard contractions, informal contractions, and internet slang
• Configurable dictionary usage
• Optimized caching for improved performance
• Preview functionality to see contractions before fixing
• Easy addition and removal of custom contractions
• Thread-safe operations

Installation

pip install contraction-fix

Usage

Basic Usage

from contraction_fix import fix

text = "I can't believe it's not butter!"
fixed_text = fix(text)
print(fixed_text)  # "I cannot believe it is not butter!"

Batch Processing (NEW!)

For processing multiple texts efficiently:

from contraction_fix import fix_batch

texts = [
    "I can't believe it's working!",
    "They're going to the store",
    "We'll see what happens"
]

fixed_texts = fix_batch(texts)
print(fixed_texts)
# Output: ["I cannot believe it is working!", "They are going to the store", "We will see what happens"]

Instantiating ContractionFixer

Start by creating an instance of the ContractionFixer class:

from contraction_fix import ContractionFixer

fixer = ContractionFixer()

Optional Parameters:

• use_informal: bool = True

  • Enables informal contractions like "gonna" → "going to".

  • Set to False to avoid informal style expansions.

• use_slang: bool = True

  • Enables slang contractions like "brb" → "be right back".

  • Set to False for more formal or academic applications.

• cache_size: int = 1024

  • Sets the LRU cache size for memoization. Improves performance when processing repeated inputs.

Example – Disabling slang:

fixer = ContractionFixer(use_slang=False)
print(fixer.fix("brb, idk what's up"))  
# Output: "brb, I don't know what is up"  (brb is skipped because use_slang=False)

Contractions vs. Possessives

The package intelligently differentiates between contractions and possessive forms:

from contraction_fix import fix

text = "I can't find Sarah's keys, and she won't be at her brother's house until it's dark."
fixed_text = fix(text)
print(fixed_text)  # "I cannot find Sarah's keys, and she will not be at her brother's house until it is dark."

Notice how the package:

• Expands contractions: "can't" → "cannot", "won't" → "will not", "it's" → "it is"
• Preserves possessives: "Sarah's" and "brother's" remain unchanged

Advanced Usage

from contraction_fix import ContractionFixer

# Create a custom fixer instance
fixer = ContractionFixer(use_informal=True, use_slang=False)

# Fix single text
text = "I'd like to see y'all tomorrow"
fixed_text = fixer.fix(text)
print(fixed_text)  # "I would like to see you all tomorrow"

# Fix multiple texts efficiently
texts = [
    "I can't believe it's working",
    "They're going home",
    "We'll see what happens"
]
fixed_texts = fixer.fix_batch(texts)
print(fixed_texts)  # ["I cannot believe it is working", "They are going home", "We will see what happens"]

# Preview contractions
matches = fixer.preview(text, context_size=5)
for match in matches:
    print(f"Found '{match.text}' at position {match.start}")
    print(f"Context: '{match.context}'")
    print(f"Will be replaced with: '{match.replacement}'")

# Add custom contraction
fixer.add_contraction("gonna", "going to")

# Remove contraction
fixer.remove_contraction("won't")

Dictionary Types

The package uses three types of dictionaries:

1. Standard Contractions: Common English contractions like "can't", "won't", etc.
2. Informal Contractions: Less formal contractions and patterns like "goin'", "doin'", etc.
3. Internet Slang: Modern internet slang and abbreviations like "lol", "btw", etc.

Performance

The package is optimized for speed through:

• Precompiled regex patterns with cached compilation
• LRU caching of results for repeated inputs
• Efficient dictionary lookups with optimized key ordering
• Batch processing for multiple texts
• Minimal memory usage with frozenset constants
• Thread-safe operations

Batch Processing Performance

When processing multiple texts, use fix_batch() for better performance:

from contraction_fix import fix_batch

# More efficient for multiple texts
texts = ["I can't go", "They're here", "We'll see"]
results = fix_batch(texts)  # Uses shared cache and optimized processing

# Less efficient for multiple texts
results = [fix(text) for text in texts]  # Creates new instances

API Reference

Functions

• fix(text: str, use_informal: bool = True, use_slang: bool = True) -> str
• fix_batch(texts: List[str], use_informal: bool = True, use_slang: bool = True) -> List[str]

Classes

• ContractionFixer(use_informal: bool = True, use_slang: bool = True, cache_size: int = 1024)
  • fix(text: str) -> str
  • fix_batch(texts: List[str]) -> List[str]
  • preview(text: str, context_size: int = 10) -> List[Match]
  • add_contraction(contraction: str, expansion: str) -> None
  • remove_contraction(contraction: str) -> None

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.