humanmint 0.1.11

Clean, functional data processing for human-centric applications. Normalize and standardize names, emails, phones, departments, and job titles with a single unified API.

HumanMint

Clean, normalized contact data in one line of code.

Standardize names, emails, phones, addresses, departments, job titles, and organizations with intelligent parsing and fuzzy matching.

from humanmint import mint

result = mint(
    name="Dr. John Q. Smith, PhD",
    email="[e-mail]",
    phone="[phone] ext 456",
    department="001 - Public Works Dept",
    title="Chief of Police"
)

print(result.name_str)          # "John Q Smith"
print(result.email_str)         # "[e-mail]"
print(result.phone_str)         # "[phone]"
print(result.department_str)    # "Public Works"
print(result.title_str)         # "police chief"

Why HumanMint?

Real-world contact data is messy:

• Names with titles: "Dr. Jane Smith, PhD"
• Inconsistent formatting: "[e-mail]" vs "[e-mail]"
• Phone number variations: "[phone] x101" vs "202.555.0101"
• Departments with noise: "000171 - Public Works 202-555-0150 ext 200"
• Abbreviated titles: "Sr. Water Engr."

HumanMint handles all of this with zero configuration.

Installation

pip install humanmint

Key Features

• Names: Parse, normalize, infer gender, detect nicknames, strip titles
• Emails: Validate, normalize, detect free providers (Gmail, Yahoo, etc.)
• Phones: Format (E.164), extract extensions, validate, detect type (mobile/landline)
• Departments: Canonicalize, categorize, fuzzy match (23K+ dept names → 64 categories)
• Titles: Standardize, match against curated list (100K+ job titles), confidence scores
• Addresses: Parse US postal addresses (street, city, state, ZIP)
• Organizations: Normalize agency/org names
• Comparison: compare(result_a, result_b) for deduplication with 0-100 similarity scores
• Batch: Parallel processing with bulk(records, workers=4) for high throughput
• Export: JSON, CSV, Parquet, SQL with flatten option for direct database import

Quick Examples

Field Accessor Reference

All fields provide three access patterns:

Pattern	Example	Description
Dict access	result.title["canonical"]	Access specific processing stages
Property	result.title_str	Shorthand for canonical/standardized form
Full dict	result.title	All stages: raw, normalized, canonical, is_valid

Common _str properties: name_str, email_str, phone_str, department_str, title_str

Accessing title fields

result = mint(title="Chief of Police")

# Dict access - different processing stages
result.title["raw"]         # "Chief of Police" (original input)
result.title["normalized"]  # "Chief of Police" (cleaned)
result.title["canonical"]   # "police chief" (standardized form)
result.title["is_valid"]    # True

# Shorthand properties
result.title_str            # "police chief" (same as canonical)
result.title_normalized     # "Chief of Police"

Comparing records

from humanmint import compare

r1 = mint(name="John Smith", email="[e-mail]")
r2 = mint(name="Jon Smith", email="[e-mail]")

score = compare(r1, r2)  # Returns 0-100 similarity score
# Typically: >85 = likely duplicate, >70 = similar, <50 = different

Batch processing

from humanmint import bulk

records = [
    {"name": "Alice", "email": "[e-mail]"},
    {"name": "Bob", "email": "[e-mail]"},
]

results = bulk(records, workers=4, progress=True)

Performance

Dataset	Time	Per Record	Throughput
1,000	561 ms	0.56 ms	1,783 rec/sec
10,000	3.1 s	0.31 ms	3,178 rec/sec
50,000	14.0 s	0.28 ms	3,576 rec/sec

Documentation

• API Reference — Full function documentation
• Use Cases — Real-world examples (Government contacts, HR, Salesforce, etc.)
• Fields Guide — Access all returned fields
• Advanced — Custom weights, overrides, batch export

CLI

humanmint clean input.csv output.csv --name-col name --email-col email

Testing

pytest -q unittests

License

MIT