Pattern File Recognizer

A Microsoft Presidio recognizer that identifies patterns from text and CSV files using fast keyword matching.

Installation

pip install pattern-file-recognizer

Usage

from presidio_analyzer import AnalyzerEngine
from pattern_file_recognizer import PatternFileRecognizer

# Create recognizer with pattern files
recognizer = PatternFileRecognizer(
    patterns_files=['customers.txt', ('products.csv', 'product_name')],
    allowlist_files=['common_words.txt'],
    entity_type='INTERNAL_NAME',
    csv_column='name',        # default column for CSV files
)

# Add to Presidio Analyzer
analyzer = AnalyzerEngine()
analyzer.registry.add_recognizer(recognizer)

# Analyze text
text = "Working with Acme Corp on a project."
results = analyzer.analyze(text=text, language="en")

File Formats

Plain Text (.txt)

• One pattern per line
• Blank lines ignored
• Lines starting with # are comments
• Leading/trailing whitespace stripped

# Customer names
Acme Corp
Widget Inc

CSV (.csv)

• Standard CSV with header row
• Specify column via csv_column parameter or tuple override

name,region
Acme Corp,West
Widget Inc,East

Parameters

Parameter	Type	Required	Default	Description
patterns_files	list	Yes	-	List of pattern file paths
entity_type	str	Yes	-	Entity type for matches
allowlist_files	list	No	[]	Files with patterns to exclude
csv_column	str	No	"name"	Default column for CSV files
case_sensitive	bool	No	False	Case-sensitive matching