pointblank 0.18.0__py3-none-any.whl → 0.20.0__py3-none-any.whl

pointblank/locales/__init__.py

@@ -0,0 +1,1476 @@
+"""
+Country-based data generation for synthetic test data.
+
+This module provides country-specific data generation without external dependencies.
+It supports generating realistic names, addresses, emails, and other data types
+with proper localization based on ISO 3166-1 country codes.
+"""
+
+from __future__ import annotations
+
+import json
+import random
+import unicodedata
+from dataclasses import dataclass, field
+from importlib.resources import files
+from typing import TYPE_CHECKING, Any
+
+if TYPE_CHECKING:
+    pass
+
+__all__ = [
+    "LocaleRegistry",
+    "LocaleGenerator",
+    "get_generator",
+    "COUNTRY_CODE_MAP",
+    "COUNTRIES_WITH_FULL_DATA",
+]
+
+
+# ISO 3166-1 country code mappings
+# Maps alpha-2 (2-letter) and alpha-3 (3-letter) codes to internal data directory names
+COUNTRY_CODE_MAP: dict[str, str] = {
+    # United States
+    "US": "US",
+    "USA": "US",
+    # United Kingdom
+    "GB": "GB",
+    "GBR": "GB",
+    "UK": "GB",  # Common alias
+    # Ireland
+    "IE": "IE",
+    "IRL": "IE",
+    # Iceland
+    "IS": "IS",
+    "ISL": "IS",
+    # Australia
+    "AU": "AU",
+    "AUS": "AU",
+    # Argentina
+    "AR": "AR",
+    "ARG": "AR",
+    # Canada
+    "CA": "CA",
+    "CAN": "CA",
+    # Germany
+    "DE": "DE",
+    "DEU": "DE",
+    # Greece
+    "GR": "GR",
+    "GRC": "GR",
+    # Austria
+    "AT": "AT",
+    "AUT": "AT",
+    # Switzerland
+    "CH": "CH",
+    "CHE": "CH",
+    # Chile
+    "CL": "CL",
+    "CHL": "CL",
+    # France
+    "FR": "FR",
+    "FRA": "FR",
+    # Spain
+    "ES": "ES",
+    "ESP": "ES",
+    # Mexico
+    "MX": "MX",
+    "MEX": "MX",
+    # Malta
+    "MT": "MT",
+    "MLT": "MT",
+    # Portugal
+    "PT": "PT",
+    "PRT": "PT",
+    # Brazil
+    "BR": "BR",
+    "BRA": "BR",
+    # India
+    "IN": "IN",
+    "IND": "IN",
+    # Italy
+    "IT": "IT",
+    "ITA": "IT",
+    # Netherlands
+    "NL": "NL",
+    "NLD": "NL",
+    # Belgium
+    "BE": "BE",
+    "BEL": "BE",
+    # Bulgaria
+    "BG": "BG",
+    "BGR": "BG",
+    # Poland
+    "PL": "PL",
+    "POL": "PL",
+    # Romania
+    "RO": "RO",
+    "ROU": "RO",
+    # Russia
+    "RU": "RU",
+    "RUS": "RU",
+    # Slovakia
+    "SK": "SK",
+    "SVK": "SK",
+    # Slovenia
+    "SI": "SI",
+    "SVN": "SI",
+    # Japan
+    "JP": "JP",
+    "JPN": "JP",
+    # South Korea
+    "KR": "KR",
+    "KOR": "KR",
+    # Latvia
+    "LV": "LV",
+    "LVA": "LV",
+    # Lithuania
+    "LT": "LT",
+    "LTU": "LT",
+    # Luxembourg
+    "LU": "LU",
+    "LUX": "LU",
+    # China
+    "CN": "CN",
+    "CHN": "CN",
+    # Colombia
+    "CO": "CO",
+    "COL": "CO",
+    # Cyprus
+    "CY": "CY",
+    "CYP": "CY",
+    # Czech Republic
+    "CZ": "CZ",
+    "CZE": "CZ",
+    # Estonia
+    "EE": "EE",
+    "EST": "EE",
+    # Hong Kong
+    "HK": "HK",
+    "HKG": "HK",
+    # Croatia
+    "HR": "HR",
+    "HRV": "HR",
+    # Hungary
+    "HU": "HU",
+    "HUN": "HU",
+    # Indonesia
+    "ID": "ID",
+    "IDN": "ID",
+    # Taiwan
+    "TW": "TW",
+    "TWN": "TW",
+    # Turkey
+    "TR": "TR",
+    "TUR": "TR",
+    # New Zealand
+    "NZ": "NZ",
+    "NZL": "NZ",
+    # Philippines
+    "PH": "PH",
+    "PHL": "PH",
+}
+
+# Countries that have complete locale data files
+# These are the ISO alpha-2 codes for countries with full address, company,
+# internet, misc, person, and text JSON files in the data directory
+COUNTRIES_WITH_FULL_DATA: list[str] = [
+    "US",  # United States
+    "AR",  # Argentina
+    "AT",  # Austria
+    "AU",  # Australia
+    "BE",  # Belgium
+    "BG",  # Bulgaria
+    "BR",  # Brazil
+    "CA",  # Canada
+    "CH",  # Switzerland
+    "CL",  # Chile
+    "CN",  # China
+    "CO",  # Colombia
+    "CY",  # Cyprus
+    "CZ",  # Czech Republic
+    "DE",  # Germany
+    "DK",  # Denmark
+    "EE",  # Estonia
+    "ES",  # Spain
+    "FI",  # Finland
+    "FR",  # France
+    "GB",  # United Kingdom
+    "GR",  # Greece
+    "HK",  # Hong Kong
+    "HR",  # Croatia
+    "HU",  # Hungary
+    "ID",  # Indonesia
+    "IE",  # Ireland
+    "IN",  # India
+    "IS",  # Iceland
+    "IT",  # Italy
+    "JP",  # Japan
+    "KR",  # South Korea
+    "LV",  # Latvia
+    "LT",  # Lithuania
+    "LU",  # Luxembourg
+    "MT",  # Malta
+    "MX",  # Mexico
+    "NL",  # Netherlands
+    "NO",  # Norway
+    "NZ",  # New Zealand
+    "PL",  # Poland
+    "PH",  # Philippines
+    "PT",  # Portugal
+    "RO",  # Romania
+    "RU",  # Russia
+    "SE",  # Sweden
+    "SK",  # Slovakia
+    "SI",  # Slovenia
+    "TR",  # Turkey
+    "TW",  # Taiwan
+]
+
+# Fallback chains for countries (when a country's data is incomplete)
+COUNTRY_FALLBACKS: dict[str, list[str]] = {
+    # English-speaking countries fall back to US
+    "GB": ["GB", "US"],
+    "IE": ["IE", "GB", "US"],
+    "AU": ["AU", "GB", "US"],
+    "CA": ["CA", "US"],
+    # German-speaking countries
+    "DE": ["DE", "US"],
+    "AT": ["AT", "DE", "US"],
+    "CH": ["CH", "DE", "US"],
+    # French-speaking
+    "FR": ["FR", "US"],
+    # Belgian (Dutch/French bilingual)
+    "BE": ["BE", "NL", "FR", "US"],
+    # Scandinavian
+    "DK": ["DK", "DE", "US"],
+    "NO": ["NO", "DK", "DE", "US"],
+    "SE": ["SE", "DK", "DE", "US"],
+    "FI": ["FI", "SE", "US"],
+    # Spanish-speaking
+    "ES": ["ES", "US"],
+    "MX": ["MX", "ES", "US"],
+    # Portuguese-speaking
+    "PT": ["PT", "US"],
+    "BR": ["BR", "PT", "US"],
+    # Other European
+    "IT": ["IT", "US"],
+    "NL": ["NL", "US"],
+    "PL": ["PL", "US"],
+    "RU": ["RU", "US"],
+    # Asian countries
+    "JP": ["JP", "US"],
+    "KR": ["KR", "US"],
+    "CN": ["CN", "US"],
+    "TW": ["TW", "CN", "US"],
+    # Turkey
+    "TR": ["TR", "US"],
+}
+
+
+@dataclass
+class LocaleData:
+    """Container for all locale-specific data."""
+
+    locale: str
+    person: dict[str, Any] = field(default_factory=dict)
+    address: dict[str, Any] = field(default_factory=dict)
+    company: dict[str, Any] = field(default_factory=dict)
+    internet: dict[str, Any] = field(default_factory=dict)
+    text: dict[str, Any] = field(default_factory=dict)
+    misc: dict[str, Any] = field(default_factory=dict)
+
+
+# Transliteration map for special characters (umlauts add 'e', others simplified)
+_TRANSLITERATION_MAP: dict[str, str] = {
+    # German umlauts -> add 'e'
+    "ä": "ae",
+    "ö": "oe",
+    "ü": "ue",
+    "Ä": "Ae",
+    "Ö": "Oe",
+    "Ü": "Ue",
+    "ß": "ss",
+    # Scandinavian
+    "å": "aa",
+    "Å": "Aa",
+    "ø": "oe",
+    "Ø": "Oe",
+    "æ": "ae",
+    "Æ": "Ae",
+    # French/Spanish/Portuguese/Italian accents
+    "à": "a",
+    "á": "a",
+    "â": "a",
+    "ã": "a",
+    "À": "A",
+    "Á": "A",
+    "Â": "A",
+    "Ã": "A",
+    "è": "e",
+    "é": "e",
+    "ê": "e",
+    "ë": "e",
+    "È": "E",
+    "É": "E",
+    "Ê": "E",
+    "Ë": "E",
+    "ì": "i",
+    "í": "i",
+    "î": "i",
+    "ï": "i",
+    "Ì": "I",
+    "Í": "I",
+    "Î": "I",
+    "Ï": "I",
+    "ò": "o",
+    "ó": "o",
+    "ô": "o",
+    "õ": "o",
+    "Ò": "O",
+    "Ó": "O",
+    "Ô": "O",
+    "Õ": "O",
+    "ù": "u",
+    "ú": "u",
+    "û": "u",
+    "Ù": "U",
+    "Ú": "U",
+    "Û": "U",
+    "ñ": "n",
+    "Ñ": "N",
+    "ç": "c",
+    "Ç": "C",
+    "ý": "y",
+    "ÿ": "y",
+    "Ý": "Y",
+    # Eastern European
+    "ł": "l",
+    "Ł": "L",
+    "ń": "n",
+    "Ń": "N",
+    "ś": "s",
+    "Ś": "S",
+    "ź": "z",
+    "Ź": "Z",
+    "ż": "z",
+    "Ż": "Z",
+    "ć": "c",
+    "Ć": "C",
+    "ě": "e",
+    "Ě": "E",
+    "š": "s",
+    "Š": "S",
+    "č": "c",
+    "Č": "C",
+    "ř": "r",
+    "Ř": "R",
+    "ž": "z",
+    "Ž": "Z",
+    "ů": "u",
+    "Ů": "U",
+    "ď": "d",
+    "Ď": "D",
+    "ť": "t",
+    "Ť": "T",
+    "ň": "n",
+    "Ň": "N",
+    # Other
+    "đ": "d",
+    "Đ": "D",
+    "ğ": "g",
+    "Ğ": "G",
+    "ı": "i",
+    "İ": "I",
+    "ş": "s",
+    "Ş": "S",
+    "ț": "t",
+    "Ț": "T",
+    "ă": "a",
+    "Ă": "A",
+}
+
+
+def _transliterate_to_ascii(text: str) -> str:
+    """
+    Transliterate text to ASCII-safe characters for email addresses and usernames.
+
+    Handles German umlauts specially (ü -> ue, ö -> oe, ä -> ae) and converts
+    other accented characters to their base ASCII equivalents.
+
+    Parameters
+    ----------
+    text
+        The text to transliterate.
+
+    Returns
+    -------
+    str
+        ASCII-safe version of the text.
+    """
+    # First apply our custom transliteration map
+    result = []
+    for char in text:
+        if char in _TRANSLITERATION_MAP:
+            result.append(_TRANSLITERATION_MAP[char])
+        else:
+            result.append(char)
+    text = "".join(result)
+
+    # Then use unicodedata to handle any remaining non-ASCII characters
+    # NFD decomposes characters (e.g., é -> e + combining accent)
+    # We then filter to keep only ASCII characters
+    normalized = unicodedata.normalize("NFD", text)
+    ascii_text = "".join(c for c in normalized if unicodedata.category(c) != "Mn")
+
+    return ascii_text
+
+
+def _normalize_country(country: str) -> str:
+    """
+    Normalize a country code to the standard 2-letter ISO 3166-1 alpha-2 format.
+
+    Parameters
+    ----------
+    country
+        Country code in alpha-2 (US), alpha-3 (USA), or legacy locale format (en_US).
+
+    Returns
+    -------
+    str
+        The normalized 2-letter country code.
+
+    Raises
+    ------
+    ValueError
+        If the country code is not recognized.
+    """
+    # Uppercase and strip whitespace
+    code = country.strip().upper()
+
+    # Handle legacy locale format (en_US, de-DE, etc.)
+    if "_" in code or "-" in code:
+        # Extract country part from locale code
+        parts = code.replace("-", "_").split("_")
+        if len(parts) == 2:
+            code = parts[1]  # Take the country part (US from en_US)
+
+    # Look up in the country code map
+    if code in COUNTRY_CODE_MAP:
+        return COUNTRY_CODE_MAP[code]
+
+    # If already a valid 2-letter code in fallbacks, use it
+    if code in COUNTRY_FALLBACKS:
+        return code
+
+    # Default to US with a warning (or raise an error)
+    raise ValueError(
+        f"Unknown country code: {country!r}. "
+        f"Supported codes: {', '.join(sorted(set(COUNTRY_CODE_MAP.keys())))}"
+    )
+
+
+class LocaleRegistry:
+    """Registry for country data with fallback support."""
+
+    _instance: LocaleRegistry | None = None
+    _cache: dict[str, LocaleData]
+
+    def __new__(cls) -> LocaleRegistry:
+        if cls._instance is None:
+            cls._instance = super().__new__(cls)
+            cls._instance._cache = {}
+        return cls._instance
+
+    def get(self, country: str) -> LocaleData:
+        """
+        Get country data with fallback chain.
+
+        Parameters
+        ----------
+        country
+            Country code (e.g., "US", "DE", "USA", "DEU").
+            Also accepts legacy locale codes like "en_US" for backwards compatibility.
+
+        Returns
+        -------
+        LocaleData
+            The country data, falling back to parent countries if needed.
+        """
+        # Normalize to 2-letter country code
+        country_code = _normalize_country(country)
+
+        if country_code in self._cache:
+            return self._cache[country_code]
+
+        # Get fallback chain
+        fallback_chain = COUNTRY_FALLBACKS.get(country_code, [country_code, "US"])
+        if country_code not in fallback_chain:
+            fallback_chain = [country_code] + fallback_chain
+
+        # Load data with fallback
+        locale_data = self._load_with_fallback(fallback_chain)
+        self._cache[country_code] = locale_data
+        return locale_data
+
+    def _load_with_fallback(self, fallback_chain: list[str]) -> LocaleData:
+        """Load country data, falling back through the chain."""
+        merged_data = LocaleData(locale=fallback_chain[0])
+
+        # Load shared/universal data first (e.g., file extensions, MIME types)
+        shared_data = self._load_country_files("_shared")
+        if shared_data:
+            self._merge_data(merged_data, shared_data)
+
+        # Load in reverse order so more specific countries override
+        for country in reversed(fallback_chain):
+            data = self._load_country_files(country)
+            if data:
+                self._merge_data(merged_data, data)
+
+        return merged_data
+
+    def _load_country_files(self, country: str) -> dict[str, Any] | None:
+        """Load all data files for a country."""
+        try:
+            data_path = files("pointblank.locales.data") / country
+            if not data_path.is_dir():
+                return None
+
+            result: dict[str, Any] = {}
+            for category in ["person", "address", "company", "internet", "text", "misc"]:
+                file_path = data_path / f"{category}.json"
+                try:
+                    content = file_path.read_text(encoding="utf-8")
+                    result[category] = json.loads(content)
+                except (FileNotFoundError, json.JSONDecodeError):
+                    pass
+
+            return result if result else None
+        except (TypeError, FileNotFoundError):
+            return None
+
+    def _merge_data(self, target: LocaleData, source: dict[str, Any]) -> None:
+        """Merge source data into target LocaleData."""
+        for category, data in source.items():
+            if hasattr(target, category):
+                existing = getattr(target, category)
+                if isinstance(existing, dict) and isinstance(data, dict):
+                    existing.update(data)
+                else:
+                    setattr(target, category, data)
+
+    def clear_cache(self) -> None:
+        """Clear the country data cache."""
+        self._cache.clear()
+
+
+class LocaleGenerator:
+    """
+    Generator for country-specific test data.
+
+    This class provides methods to generate realistic data like names, emails,
+    addresses, etc. based on country-specific patterns and data.
+    """
+
+    def __init__(self, country: str = "US", seed: int | None = None):
+        """
+        Initialize the country data generator.
+
+        Parameters
+        ----------
+        country
+            Country code (e.g., "US", "DE", "USA", "DEU").
+            Also accepts legacy locale codes like "en_US" for backwards compatibility.
+        seed
+            Random seed for reproducibility.
+        """
+        self.country_code = _normalize_country(country)
+        self.rng = random.Random(seed)
+        self._registry = LocaleRegistry()
+        self._data = self._registry.get(self.country_code)
+
+    def seed(self, seed: int) -> None:
+        """Set the random seed."""
+        self.rng.seed(seed)
+
+    # =========================================================================
+    # Person
+    # =========================================================================
+
+    _current_person: dict[str, str] | None = None
+    _row_persons: list[dict[str, str]] | None = None
+
+    def _get_person(self, gender: str | None = None) -> dict[str, str]:
+        """Get a coherent person (first_name, last_name, gender) from the data."""
+        # If no gender specified, randomly select one (weighted toward male/female)
+        if gender is None:
+            gender = self.rng.choice(["male", "female"])
+
+        return {
+            "first_name": self._generate_first_name(gender),
+            "last_name": self._generate_last_name(),
+            "gender": gender,
+        }
+
+    def _generate_first_name(self, gender: str | None = None) -> str:
+        """Generate a random first name (internal, no caching)."""
+        names = self._data.person.get("first_names", {})
+
+        if gender and gender in names:
+            name_list = names[gender]
+        elif "neutral" in names:
+            # Combine all available names
+            all_names = []
+            for category in ["male", "female", "neutral"]:
+                all_names.extend(names.get(category, []))
+            name_list = all_names if all_names else ["Alex"]
+        else:
+            # Flatten all categories
+            all_names = []
+            for category_names in names.values():
+                if isinstance(category_names, list):
+                    all_names.extend(category_names)
+            name_list = all_names if all_names else ["Alex"]
+
+        return self.rng.choice(name_list)
+
+    def _generate_last_name(self) -> str:
+        """Generate a random last name (internal, no caching)."""
+        names = self._data.person.get("last_names", ["Smith"])
+        return self.rng.choice(names)
+
+    def init_row_persons(self, n_rows: int) -> None:
+        """
+        Pre-generate person data for multiple rows to ensure coherence across columns.
+
+        This should be called before generating a dataset with person-related columns.
+        When active, first_name(), last_name(), name(), email() will use the person
+        for the current row (set via set_row()).
+
+        Parameters
+        ----------
+        n_rows
+            Number of rows to pre-generate persons for.
+        """
+        self._row_persons = [self._get_person() for _ in range(n_rows)]
+
+    def clear_row_persons(self) -> None:
+        """Clear all pre-generated row persons."""
+        self._row_persons = None
+
+    def new_person(self, gender: str | None = None) -> dict[str, str]:
+        """
+        Select a new random person and cache it for coherent generation.
+
+        Call this before generating related person components (first_name, last_name, email)
+        to ensure they all refer to the same person.
+
+        Returns
+        -------
+        dict
+            The selected person with first_name and last_name.
+        """
+        self._current_person = self._get_person(gender)
+        return self._current_person
+
+    def _get_current_person(self) -> dict[str, str]:
+        """Get the current cached person, or select a new one."""
+        # If row persons are active, use those
+        if self._row_persons is not None and self._current_row is not None:
+            return self._row_persons[self._current_row]
+        # Otherwise use single cached person
+        if self._current_person is None:
+            self._current_person = self._get_person()
+        return self._current_person
+
+    def clear_person(self) -> None:
+        """Clear the cached person so the next call will select a new one."""
+        self._current_person = None
+
+    def first_name(self, gender: str | None = None) -> str:
+        """Generate a random first name (coherent with current person context)."""
+        person = self._get_current_person()
+        return person.get("first_name", "Alex")
+
+    def last_name(self) -> str:
+        """Generate a random last name (coherent with current person context)."""
+        person = self._get_current_person()
+        return person.get("last_name", "Smith")
+
+    def name(self, gender: str | None = None) -> str:
+        """Generate a simple full name (first + last, coherent with current person context).
+
+        For names with prefixes (Mr., Ms., Dr., etc.) and occasional suffixes (Jr., III),
+        use name_full() instead.
+        """
+        person = self._get_current_person()
+        first = person.get("first_name", "Alex")
+        last = person.get("last_name", "Smith")
+
+        # Check if locale uses "last first" order (e.g., Japanese)
+        formats = self._data.person.get("name_formats", ["{first_name} {last_name}"])
+        # Use the simplest format (usually first one, which is typically "first last" or "last first")
+        if formats and "{last_name} {first_name}" in formats[0]:
+            return f"{last} {first}"
+        return f"{first} {last}"
+
+    def name_full(self, gender: str | None = None) -> str:
+        """Generate a full name with optional prefix and rare suffix.
+
+        Includes honorific prefixes with realistic frequencies:
+        - Common honorifics (Mr., Ms., Mrs., etc.): ~95% of names
+        - Professional titles (Dr., Prof., Rev., etc.): ~5% of names
+
+        Suffixes (Jr., II, III) appear very rarely (~1 in 2000).
+        """
+        person = self._get_current_person()
+        first = person.get("first_name", "Alex")
+        last = person.get("last_name", "Smith")
+
+        # Get gender for prefix selection (from person context or parameter)
+        person_gender = person.get("gender", "neutral")
+        if gender:
+            person_gender = gender
+
+        # Professional titles are rare (~2-3% of population for Dr., ~0.5% for Prof.)
+        # These should appear infrequently
+        professional_titles = {
+            "Dr.",
+            "Prof.",
+            "Professor",
+            "Rev.",
+            "Pr.",
+            "Prof. Dr.",
+            "Rabbi",
+            "Father",
+            "Sister",
+            "Pastor",
+            "Elder",
+        }
+
+        # Get prefix based on gender from locale data
+        prefixes = self._data.person.get("prefixes", {})
+        prefix_list = prefixes.get(person_gender, prefixes.get("neutral", []))
+
+        # Separate common honorifics and professional titles from locale data
+        locale_common = [p for p in prefix_list if p not in professional_titles]
+        locale_professional = [p for p in prefix_list if p in professional_titles]
+
+        # Select prefix with realistic probabilities
+        # ~95% common honorific, ~5% professional title
+        if locale_professional and self.rng.random() < 0.05:
+            prefix = self.rng.choice(locale_professional)
+        elif locale_common:
+            prefix = self.rng.choice(locale_common)
+        else:
+            # Fallback defaults if no common prefixes in locale
+            fallback = {"male": "Mr.", "female": "Ms.", "neutral": "Mr."}
+            prefix = fallback.get(person_gender, "")
+
+        # Get suffix - very rare (approximately 1/2000 chance)
+        suffix = ""
+        if self.rng.random() < 0.0005:  # 1 in 2000
+            suffixes = self._data.person.get("suffixes", [])
+            # Filter out empty strings
+            suffixes = [s for s in suffixes if s]
+            if suffixes:
+                suffix = self.rng.choice(suffixes)
+
+        # Check if locale uses "last first" order (e.g., Japanese)
+        formats = self._data.person.get("name_formats", ["{first_name} {last_name}"])
+        if formats and "{last_name} {first_name}" in formats[0]:
+            # For "last first" cultures, prefix typically comes before everything
+            parts = [prefix, last, first] if prefix else [last, first]
+        else:
+            parts = [prefix, first, last] if prefix else [first, last]
+
+        if suffix:
+            parts.append(suffix)
+
+        return " ".join(parts)
+
+    # =========================================================================
+    # Address
+    # =========================================================================
+
+    _current_location: dict[str, str] | None = None
+    _row_locations: list[dict[str, str]] | None = None
+    _current_row: int | None = None
+
+    def _get_location(self) -> dict[str, str]:
+        """Get a coherent location (city, state, postcode_prefix) from the data."""
+        locations = self._data.address.get("locations", [])
+        if locations:
+            return self.rng.choice(locations)
+        # Fallback for old-style data
+        return {
+            "city": "Springfield",
+            "state": "State",
+            "state_abbr": "ST",
+            "postcode_prefix": "000",
+        }
+
+    def init_row_locations(self, n_rows: int) -> None:
+        """
+        Pre-generate locations for multiple rows to ensure coherence across columns.
+
+        This should be called before generating a dataset with address-related columns.
+        When active, city(), state(), postcode() etc. will use the location for the
+        current row (set via set_row()).
+
+        Parameters
+        ----------
+        n_rows
+            Number of rows to pre-generate locations for.
+        """
+        self._row_locations = [self._get_location() for _ in range(n_rows)]
+        self._current_row = None
+
+    def set_row(self, row_index: int) -> None:
+        """
+        Set the current row index for location-based generation.
+
+        When row locations are initialized, this sets which row's location to use.
+
+        Parameters
+        ----------
+        row_index
+            The row index (0-based).
+        """
+        self._current_row = row_index
+
+    def clear_row_locations(self) -> None:
+        """Clear all pre-generated row locations."""
+        self._row_locations = None
+        self._current_row = None
+
+    def new_location(self) -> dict[str, str]:
+        """
+        Select a new random location and cache it for coherent address generation.
+
+        Call this before generating related address components (city, state, postcode)
+        to ensure they all refer to the same location.
+
+        Returns
+        -------
+        dict
+            The selected location with city, state, state_abbr, and postcode_prefix.
+        """
+        self._current_location = self._get_location()
+        return self._current_location
+
+    def _get_current_location(self) -> dict[str, str]:
+        """Get the current cached location, or select a new one."""
+        # If row locations are active, use those
+        if self._row_locations is not None and self._current_row is not None:
+            return self._row_locations[self._current_row]
+        # Otherwise use single cached location
+        if self._current_location is None:
+            self._current_location = self._get_location()
+        return self._current_location
+
+    def clear_location(self) -> None:
+        """Clear the cached location so the next call will select a new one."""
+        self._current_location = None
+
+    def city(self) -> str:
+        """Generate a random city name (coherent with current location context).
+
+        Returns the exonym (English name) if available, otherwise the native city name.
+        This allows addresses to use native names while city presets use international names.
+        """
+        location = self._get_current_location()
+        # Prefer exonym (English name) for standalone city preset
+        return location.get("exonym", location.get("city", "Springfield"))
+
+    def _city_native(self) -> str:
+        """Get the native city name (used internally for addresses).
+
+        Always returns the native name, ignoring any exonym.
+        """
+        location = self._get_current_location()
+        return location.get("city", "Springfield")
+
+    def state(self, abbr: bool = False) -> str:
+        """Generate a random state/province name (coherent with current location context)."""
+        location = self._get_current_location()
+        if abbr:
+            return location.get("state_abbr", "ST")
+        return location.get("state", "State")
+
+    def country(self) -> str:
+        """Generate the country name for this locale."""
+        return self._data.address.get("country", "United States")
+
+    def postcode(self) -> str:
+        """Generate a random postal code (coherent with current location context)."""
+        location = self._get_current_location()
+        prefix = location.get("postcode_prefix", "")
+        postcode_format = self._data.address.get("postcode_format", "")
+
+        # If format uses pattern characters (? for letter, # for digit), generate accordingly
+        if "?" in postcode_format or "#" in postcode_format:
+            # Generate the full postcode from the format pattern
+            # Replace ? with random uppercase letter, # with random digit
+            result = []
+            prefix_idx = 0
+            for char in postcode_format:
+                if char == "?":
+                    # Use prefix character if available, otherwise random letter
+                    if prefix_idx < len(prefix) and prefix[prefix_idx].isalpha():
+                        result.append(prefix[prefix_idx])
+                        prefix_idx += 1
+                    else:
+                        result.append(self.rng.choice("ABCDEFGHIJKLMNOPQRSTUVWXYZ"))
+                elif char == "#":
+                    # Use prefix character if available, otherwise random digit
+                    if prefix_idx < len(prefix) and prefix[prefix_idx].isdigit():
+                        result.append(prefix[prefix_idx])
+                        prefix_idx += 1
+                    else:
+                        result.append(str(self.rng.randint(0, 9)))
+                else:
+                    # Keep literal characters (spaces, dashes, etc.)
+                    result.append(char)
+            return "".join(result)
+
+        # Default: append digits to complete the postal code
+        remaining = 5 - len(prefix)
+        suffix = "".join(str(self.rng.randint(0, 9)) for _ in range(remaining))
+        return prefix + suffix
+
+    def street_name(self) -> str:
+        """Generate a random street name.
+
+        If the locale has `streets_by_city`, use city-specific streets.
+        Otherwise, fall back to combining `street_names` and `street_suffixes`.
+        """
+        # Check if locale uses city-specific streets
+        streets_by_city = self._data.address.get("streets_by_city")
+        if streets_by_city:
+            # Get current city from location
+            location = self._get_current_location()
+            city = location.get("city", "")
+            city_streets = streets_by_city.get(city)
+            if city_streets:
+                return self.rng.choice(city_streets)
+
+        # Fall back to old street_names + street_suffixes approach
+        names = self._data.address.get("street_names", ["Main"])
+        suffixes = self._data.address.get("street_suffixes", ["St"])
+        return f"{self.rng.choice(names)} {self.rng.choice(suffixes)}"
+
+    def building_number(self) -> str:
+        """Generate a random building number."""
+        return str(self.rng.randint(1, 9999))
+
+    def address(self) -> str:
+        """Generate a full coherent address (city, state, postcode are consistent)."""
+        # Only select a new location if row locations are not active
+        # This ensures coherence with other address-related columns (city, state, etc.)
+        using_row_context = self._row_locations is not None and self._current_row is not None
+        if not using_row_context:
+            self.new_location()
+
+        formats = self._data.address.get(
+            "address_formats",
+            ["{building_number} {street}, {city}, {state} {postcode}"],
+        )
+        fmt = self.rng.choice(formats)
+
+        result = fmt.format(
+            building_number=self.building_number(),
+            street=self.street_name(),
+            city=self._city_native(),  # Use native name in addresses
+            state=self.state(abbr=False),
+            state_abbr=self.state(abbr=True),
+            postcode=self.postcode(),
+            country=self.country(),
+            unit=str(self.rng.randint(1, 999)),
+        )
+
+        # Clear location after generating full address (only if we set it)
+        if not using_row_context:
+            self.clear_location()
+        return result
+
+    def phone_number(self) -> str:
+        """Generate a phone number with area code matching the current location's state."""
+        location = self._get_current_location()
+        state = location.get("state", "California")
+
+        # Get area codes for this state
+        area_codes = self._data.address.get("phone_area_codes", {})
+        state_codes = area_codes.get(state, ["555"])  # 555 is fictional fallback
+        area_code = self.rng.choice(state_codes)
+
+        # Generate the rest of the number
+        exchange = str(self.rng.randint(200, 999))  # Exchange can't start with 0 or 1
+        subscriber = str(self.rng.randint(0, 9999)).zfill(4)
+
+        return f"({area_code}) {exchange}-{subscriber}"
+
+    def latitude(self) -> str:
+        """Generate a random latitude (bounded by current location if available)."""
+        location = self._get_current_location()
+        lat_min = location.get("lat_min", -90)
+        lat_max = location.get("lat_max", 90)
+        return f"{self.rng.uniform(lat_min, lat_max):.6f}"
+
+    def longitude(self) -> str:
+        """Generate a random longitude (bounded by current location if available)."""
+        location = self._get_current_location()
+        lon_min = location.get("lon_min", -180)
+        lon_max = location.get("lon_max", 180)
+        return f"{self.rng.uniform(lon_min, lon_max):.6f}"
+
+    # =========================================================================
+    # Company
+    # =========================================================================
+
+    def company(self) -> str:
+        """Generate a random company name.
+
+        Has a ~15% chance to return a well-known company name, with preference
+        for companies that have offices in the current city (if location context is active).
+        Otherwise generates a fictional company name.
+        """
+        # 15% chance to use a well-known company
+        if self.rng.random() < 0.15:
+            well_known = self._data.company.get("well_known_companies", [])
+            if well_known:
+                # Get current city if location context is active
+                current_city = None
+                if self._row_locations is not None and self._current_row is not None:
+                    current_city = self._row_locations[self._current_row].get("city")
+
+                # Collect all companies, preferring those in the current city
+                city_companies = []
+                all_companies = []
+
+                for company in well_known:
+                    name = company.get("name") if isinstance(company, dict) else company
+                    cities = company.get("cities", []) if isinstance(company, dict) else []
+                    all_companies.append(name)
+                    if current_city and current_city in cities:
+                        city_companies.append(name)
+
+                # 70% chance to use city-relevant company if available
+                if city_companies and self.rng.random() < 0.7:
+                    return self.rng.choice(city_companies)
+                elif all_companies:
+                    return self.rng.choice(all_companies)
+
+        # Generate a fictional company name
+        formats = self._data.company.get("formats", ["{last_name} {suffix}"])
+        fmt = self.rng.choice(formats)
+
+        suffixes = self._data.company.get("suffixes", ["Inc", "LLC", "Corp"])
+        adjectives = self._data.company.get("adjectives", ["Global", "Advanced"])
+        nouns = self._data.company.get("nouns", ["Solutions", "Systems"])
+
+        # Count how many {last_name} placeholders are in the format
+        # and generate distinct last names for each
+        last_name_count = fmt.count("{last_name}")
+        if last_name_count <= 1:
+            company_last_name = self._generate_last_name()
+            return fmt.format(
+                last_name=company_last_name,
+                suffix=self.rng.choice(suffixes),
+                adjective=self.rng.choice(adjectives),
+                noun=self.rng.choice(nouns),
+            )
+        else:
+            # Generate distinct last names for formats like "{last_name} and {last_name}"
+            last_names = []
+            for _ in range(last_name_count):
+                new_name = self._generate_last_name()
+                # Ensure we don't repeat the same name
+                attempts = 0
+                while new_name in last_names and attempts < 10:
+                    new_name = self._generate_last_name()
+                    attempts += 1
+                last_names.append(new_name)
+
+            # Replace placeholders one at a time
+            result = fmt
+            for name in last_names:
+                result = result.replace("{last_name}", name, 1)
+
+            return result.format(
+                suffix=self.rng.choice(suffixes),
+                adjective=self.rng.choice(adjectives),
+                noun=self.rng.choice(nouns),
+            )
+
+    def job(self) -> str:
+        """Generate a random job title."""
+        jobs = self._data.company.get("jobs", ["Manager"])
+        return self.rng.choice(jobs)
+
+    def catch_phrase(self) -> str:
+        """Generate a random business catch phrase."""
+        adjectives = self._data.company.get("catch_phrase_adjectives", ["Innovative", "Dynamic"])
+        nouns = self._data.company.get("catch_phrase_nouns", ["solutions", "paradigms"])
+        verbs = self._data.company.get("catch_phrase_verbs", ["deliver", "leverage"])
+        return (
+            f"{self.rng.choice(adjectives)} {self.rng.choice(nouns)} that {self.rng.choice(verbs)}"
+        )
+
+    # =========================================================================
+    # Internet
+    # =========================================================================
+
+    def email(self) -> str:
+        """Generate a random email address (coherent with current person context)."""
+        # Get person data - uses cached person if available
+        person = self._get_current_person()
+        first = person.get("first_name", "user").lower()
+        last = person.get("last_name", "name").lower()
+        domains = self._data.internet.get("free_email_domains", ["gmail.com", "outlook.com"])
+
+        # Transliterate to ASCII for valid email addresses
+        first = _transliterate_to_ascii(first)
+        last = _transliterate_to_ascii(last)
+
+        # Clean names for email (remove non-alphanumeric)
+        first = "".join(c for c in first if c.isalnum())
+        last = "".join(c for c in last if c.isalnum())
+
+        # Various realistic email patterns
+        patterns = [
+            f"{first}.{last}",  # john.smith
+            f"{first}{last}",  # johnsmith
+            f"{first}_{last}",  # john_smith
+            f"{first[0]}{last}",  # jsmith
+            f"{first}{self.rng.randint(1, 999)}",  # john123
+            f"{first[0]}{last}{self.rng.randint(1, 99)}",  # jsmith42
+            f"{first}.{last}{self.rng.randint(1, 99)}",  # john.smith99
+            f"{first[0]}_{last}",  # j_smith
+        ]
+
+        return f"{self.rng.choice(patterns)}@{self.rng.choice(domains)}"
+
+    def user_name(self) -> str:
+        """Generate a random username (coherent with current person context)."""
+        # Get person data - uses cached person if available
+        person = self._get_current_person()
+        first = person.get("first_name", "user").lower()
+        last = person.get("last_name", "name").lower()
+
+        # Transliterate to ASCII for valid usernames
+        first = _transliterate_to_ascii(first)
+        last = _transliterate_to_ascii(last)
+
+        # Clean names
+        first = "".join(c for c in first if c.isalnum())
+        last = "".join(c for c in last if c.isalnum())
+
+        patterns = [
+            f"{first}{last}",
+            f"{first}_{last}",
+            f"{first}{self.rng.randint(1, 999)}",
+            f"{first[0]}{last}{self.rng.randint(1, 99)}",
+        ]
+
+        return self.rng.choice(patterns)
+
+    def password(self, length: int = 12) -> str:
+        """Generate a random password."""
+        import string
+
+        chars = string.ascii_letters + string.digits + "!@#$%^&*"
+        return "".join(self.rng.choice(chars) for _ in range(length))
+
+    def url(self) -> str:
+        """Generate a random URL."""
+        protocols = ["https://"]
+        tlds = self._data.internet.get("tlds", ["com", "org", "net"])
+        words = self._data.text.get("words", ["example", "test", "sample"])
+
+        domain = self.rng.choice(words).lower()
+        domain = "".join(c for c in domain if c.isalnum())
+
+        return f"{self.rng.choice(protocols)}www.{domain}.{self.rng.choice(tlds)}"
+
+    def domain_name(self) -> str:
+        """Generate a random domain name."""
+        tlds = self._data.internet.get("tlds", ["com", "org", "net"])
+        words = self._data.text.get("words", ["example", "test", "sample"])
+
+        domain = self.rng.choice(words).lower()
+        domain = "".join(c for c in domain if c.isalnum())
+
+        return f"{domain}.{self.rng.choice(tlds)}"
+
+    def ipv4(self) -> str:
+        """Generate a random IPv4 address."""
+        return ".".join(str(self.rng.randint(0, 255)) for _ in range(4))
+
+    def ipv6(self) -> str:
+        """Generate a random IPv6 address."""
+        return ":".join(f"{self.rng.randint(0, 65535):04x}" for _ in range(8))
+
+    # =========================================================================
+    # Text
+    # =========================================================================
+
+    def word(self) -> str:
+        """Generate a random word."""
+        words = self._data.text.get("words", ["lorem", "ipsum", "dolor"])
+        return self.rng.choice(words)
+
+    def sentence(self, num_words: int | None = None) -> str:
+        """Generate a random sentence."""
+        if num_words is None:
+            num_words = self.rng.randint(5, 15)
+
+        words = [self.word() for _ in range(num_words)]
+        words[0] = words[0].capitalize()
+        return " ".join(words) + "."
+
+    def paragraph(self, num_sentences: int | None = None) -> str:
+        """Generate a random paragraph."""
+        if num_sentences is None:
+            num_sentences = self.rng.randint(3, 7)
+
+        return " ".join(self.sentence() for _ in range(num_sentences))
+
+    def text(self, max_chars: int = 200) -> str:
+        """Generate random text up to max_chars."""
+        result = []
+        current_length = 0
+
+        while current_length < max_chars:
+            sentence = self.sentence()
+            if current_length + len(sentence) + 1 > max_chars:
+                break
+            result.append(sentence)
+            current_length += len(sentence) + 1
+
+        return " ".join(result) if result else self.sentence()[:max_chars]
+
+    # =========================================================================
+    # Financial
+    # =========================================================================
+
+    def credit_card_number(self) -> str:
+        """Generate a random credit card number (not valid for transactions)."""
+        # Generate a 16-digit number with valid Luhn checksum
+        prefix = self.rng.choice(["4", "5", "37", "6011"])  # Visa, MC, Amex, Discover
+        length = 15 if prefix == "37" else 16
+
+        # Generate digits (minus check digit)
+        digits = list(prefix)
+        while len(digits) < length - 1:
+            digits.append(str(self.rng.randint(0, 9)))
+
+        # Calculate Luhn check digit
+        check_digit = self._luhn_checksum(digits)
+        digits.append(str(check_digit))
+
+        return "".join(digits)
+
+    def _luhn_checksum(self, digits: list[str]) -> int:
+        """Calculate Luhn check digit for a partial card number.
+
+        The check digit is appended to make the full number pass the Luhn algorithm.
+        We process from right to left, doubling every second digit starting from
+        the rightmost digit of the partial number (since the check digit will be
+        at position 0 and won't be doubled).
+        """
+        nums = [int(d) for d in digits]
+        total = 0
+        for i, d in enumerate(reversed(nums)):
+            if i % 2 == 0:  # These positions get doubled (check digit at pos 0 won't be)
+                d = d * 2
+                if d > 9:
+                    d -= 9
+            total += d
+        return (10 - (total % 10)) % 10
+
+    def iban(self) -> str:
+        """Generate a random IBAN."""
+        # Simplified - generates a plausible-looking IBAN
+        country = self._data.address.get("country_code", "US")
+        if country == "US":
+            # US doesn't use IBAN, use DE as example
+            country = "DE"
+
+        check_digits = f"{self.rng.randint(10, 99)}"
+        bank_code = "".join(str(self.rng.randint(0, 9)) for _ in range(8))
+        account = "".join(str(self.rng.randint(0, 9)) for _ in range(10))
+
+        return f"{country}{check_digits}{bank_code}{account}"
+
+    def currency_code(self) -> str:
+        """Generate a random currency code."""
+        codes = self._data.misc.get("currency_codes", ["USD", "EUR", "GBP", "JPY", "CNY"])
+        return self.rng.choice(codes)
+
+    # =========================================================================
+    # Identifiers
+    # =========================================================================
+
+    def uuid4(self) -> str:
+        """Generate a random UUID4."""
+        # Use our RNG to generate deterministic UUIDs
+        hex_chars = "0123456789abcdef"
+        parts = [
+            "".join(self.rng.choice(hex_chars) for _ in range(8)),
+            "".join(self.rng.choice(hex_chars) for _ in range(4)),
+            "4" + "".join(self.rng.choice(hex_chars) for _ in range(3)),  # Version 4
+            self.rng.choice("89ab")
+            + "".join(self.rng.choice(hex_chars) for _ in range(3)),  # Variant
+            "".join(self.rng.choice(hex_chars) for _ in range(12)),
+        ]
+        return "-".join(parts)
+
+    def ssn(self) -> str:
+        """Generate a random SSN-like identifier."""
+        # US format: XXX-XX-XXXX
+        fmt = self._data.misc.get("ssn_format", "###-##-####")
+        return self._generate_from_format(fmt)
+
+    def license_plate(self) -> str:
+        """Generate a random license plate."""
+        fmt = self._data.misc.get("license_plate_format", "???-####")
+        return self._generate_from_format(fmt)
+
+    # =========================================================================
+    # Date/Time (string representations)
+    # =========================================================================
+
+    def date_this_year(self) -> str:
+        """Generate a random date from this year as ISO string."""
+        from datetime import date, timedelta
+
+        today = date.today()
+        start = date(today.year, 1, 1)
+        days = (today - start).days
+        random_date = start + timedelta(days=self.rng.randint(0, max(days, 1)))
+        return random_date.isoformat()
+
+    def date_this_decade(self) -> str:
+        """Generate a random date from this decade as ISO string."""
+        from datetime import date, timedelta
+
+        today = date.today()
+        decade_start = (today.year // 10) * 10
+        start = date(decade_start, 1, 1)
+        days = (today - start).days
+        random_date = start + timedelta(days=self.rng.randint(0, max(days, 1)))
+        return random_date.isoformat()
+
+    def time(self) -> str:
+        """Generate a random time as string."""
+        hour = self.rng.randint(0, 23)
+        minute = self.rng.randint(0, 59)
+        second = self.rng.randint(0, 59)
+        return f"{hour:02d}:{minute:02d}:{second:02d}"
+
+    # =========================================================================
+    # Misc
+    # =========================================================================
+
+    def color_name(self) -> str:
+        """Generate a random color name."""
+        colors = self._data.misc.get(
+            "colors",
+            [
+                "Red",
+                "Blue",
+                "Green",
+                "Yellow",
+                "Purple",
+                "Orange",
+                "Pink",
+                "Brown",
+                "Black",
+                "White",
+                "Gray",
+                "Cyan",
+                "Magenta",
+            ],
+        )
+        return self.rng.choice(colors)
+
+    def file_name(self) -> str:
+        """Generate a random file name."""
+        words = self._data.text.get("words", ["document", "file", "report"])
+        extensions = self._data.misc.get("file_extensions", ["txt", "pdf", "doc", "xlsx"])
+        word = self.rng.choice(words).lower()
+        word = "".join(c for c in word if c.isalnum())
+        return f"{word}.{self.rng.choice(extensions)}"
+
+    def file_extension(self) -> str:
+        """Generate a random file extension."""
+        extensions = self._data.misc.get(
+            "file_extensions", ["txt", "pdf", "doc", "xlsx", "png", "jpg"]
+        )
+        return self.rng.choice(extensions)
+
+    def mime_type(self) -> str:
+        """Generate a random MIME type."""
+        mime_types = self._data.misc.get(
+            "mime_types",
+            [
+                "text/plain",
+                "text/html",
+                "application/json",
+                "application/pdf",
+                "image/png",
+                "image/jpeg",
+            ],
+        )
+        return self.rng.choice(mime_types)
+
+    # =========================================================================
+    # Utilities
+    # =========================================================================
+
+    def _generate_from_format(self, fmt: str) -> str:
+        """
+        Generate a string from a format pattern.
+
+        Patterns:
+        - # = digit (0-9)
+        - ? = uppercase letter (A-Z)
+        - * = alphanumeric
+        """
+        result = []
+        for char in fmt:
+            if char == "#":
+                result.append(str(self.rng.randint(0, 9)))
+            elif char == "?":
+                result.append(self.rng.choice("ABCDEFGHIJKLMNOPQRSTUVWXYZ"))
+            elif char == "*":
+                result.append(self.rng.choice("ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789"))
+            else:
+                result.append(char)
+        return "".join(result)
+
+
+# Module-level convenience function
+_default_registry = LocaleRegistry()
+
+
+def get_generator(country: str = "US", seed: int | None = None) -> LocaleGenerator:
+    """
+    Get a country data generator instance.
+
+    Parameters
+    ----------
+    country
+        Country code (e.g., "US", "DE", "USA", "DEU").
+        Also accepts legacy locale codes like "en_US" for backwards compatibility.
+    seed
+        Random seed for reproducibility.
+
+    Returns
+    -------
+    LocaleGenerator
+        A generator configured for the specified country.
+    """
+    return LocaleGenerator(country=country, seed=seed)