pricetag 1.0.0

A pure-Python library for extracting price and currency information from unstructured text

Pricetag

A pure-Python library for extracting price and currency information from unstructured text, with a primary focus on salary data in job postings.

Features

• Zero Dependencies: Uses only Python standard library
• High Performance: Processes 1000+ documents per second
• Comprehensive Pattern Support:
  • Individual amounts: $50,000, $50k, 50k USD
  • Ranges: $50k-$75k, $50,000 to $75,000
  • Hourly/Annual/Monthly rates: $25/hour, $50k/year, $4k/month
  • Contextual terms: six figures, competitive, DOE
• Smart Normalization: Converts all amounts to annual USD for easy comparison
• Confidence Scoring: Each extraction includes a confidence score
• Validation & Sanity Checks: Flags unusual or problematic values

Installation

pip install pricetag

Quick Start

from pricetag import PriceExtractor

# Initialize the extractor
extractor = PriceExtractor()

# Extract prices from text
text = "Senior Engineer position paying $120,000 - $150,000 annually"
results = extractor.extract(text)

# Access the results
for result in results:
    print(f"Value: {result['value']}")
    print(f"Type: {result['type']}")
    print(f"Confidence: {result['confidence']}")
    print(f"Annual value: {result['normalized_annual']}")

Configuration Options

extractor = PriceExtractor(
    min_confidence=0.5,        # Minimum confidence score to include
    include_contextual=True,   # Extract terms like "six figures"
    normalize_to_annual=True,  # Convert to annual amounts
    min_salary=10000,          # Minimum reasonable salary
    max_salary=10000000,       # Maximum reasonable salary
    assume_hours_per_year=2080,# For hourly→annual conversion
    fast_mode=False,           # Enable for better performance
    max_results_per_text=None  # Limit number of results
)

Examples

Basic Salary Extraction

text = "The position offers $75,000 per year with benefits"
results = extractor.extract(text)
# Returns: [{'value': 75000.0, 'type': 'annual', ...}]

Hourly Rate with Normalization

text = "Paying $45/hour for senior developers"
results = extractor.extract(text)
# Returns: [{'value': 45.0, 'type': 'hourly', 'normalized_annual': 93600.0, ...}]

Salary Range

text = "Salary range: $60k-$80k depending on experience"
results = extractor.extract(text)
# Returns: [{'value': (60000.0, 80000.0), 'is_range': True, ...}]

Contextual Terms

text = "Looking for someone with 5+ years experience, six figures"
results = extractor.extract(text)
# Returns: [{'value': (100000, 999999), 'type': 'unknown', 'confidence': 0.7, ...}]

Multiple Prices

text = "Base: $100k, Bonus: up to $30k, Equity: 0.5%"
results = extractor.extract(text)
# Returns multiple results with appropriate flags

Batch Processing

texts = [
    "Salary: $50,000",
    "Rate: $30/hour", 
    "Competitive pay"
]
results_batch = extractor.extract_batch(texts)

Output Format

Each extraction returns a PriceResult dictionary:

{
    'value': float | tuple[float, float],  # Single value or (min, max)
    'raw_text': str,                       # Original matched text
    'position': tuple[int, int],           # Character positions
    'type': str,                           # 'hourly', 'annual', 'monthly', etc.
    'confidence': float,                   # 0.0 to 1.0
    'normalized_annual': float | tuple,    # Annual USD amount
    'currency': str,                       # Always 'USD' in v1
    'is_range': bool,                      # True for ranges
    'flags': list[str]                     # Validation flags
}

Validation Flags

• invalid_range: Max less than min
• below_minimum: Below configured threshold
• above_maximum: Above configured threshold
• unreasonable_hourly_rate: Outside $7-$500/hour
• potential_inconsistency: Large discrepancy with other prices
• ambiguous_type: Unclear if hourly/annual
• approximate: Estimated from "approximately"
• requires_market_data: Needs external data (e.g., "competitive")
• experience_dependent: Depends on experience (e.g., "DOE")

Performance

The library is optimized for high-volume processing:

• Pre-compiled regex patterns
• Number parsing cache
• Quick pre-filtering
• Fast mode for bulk processing
• Batch processing support

# Fast mode for high-volume processing
extractor = PriceExtractor(fast_mode=True, max_results_per_text=5)

# Process 1000 documents
texts = ["..." for _ in range(1000)]
results = extractor.extract_batch(texts)  # < 1 second

Testing

Run the test suite:

# Install dev dependencies
pip install -e ".[dev]"

# Run tests
pytest tests/

# Run with coverage
pytest tests/ --cov=pricetag

Limitations

• Currently supports USD only
• Optimized for US salary formats
• Context window limited to surrounding text
• Does not handle equity/stock compensation

Contributing

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

License

MIT License - see LICENSE file for details

Author

Michelle Pellon - [e-mail]

Acknowledgments

Built with pure Python for maximum compatibility and zero dependencies.