i18n-key-sync/i18n_key_sync/locale.py

"""Locale file parsing for JSON and YAML format."""

import json
from pathlib import Path
from typing import Any, Dict, List, Optional, Set


class LocaleParser:
    """Parse and manage locale files in JSON and YAML formats."""

    def __init__(self):
        """Initialize the locale parser."""
        self._cache: Dict[str, Dict[str, Any]] = {}

    def parse_file(self, file_path: Path | str) -> Dict[str, Any]:
        """Parse a single locale file.

        Args:
            file_path: Path to the locale file.

        Returns:
            Dictionary of locale keys and values.

        Raises:
            ValueError: If the file format is not supported.
            FileNotFoundError: If the file doesn't exist.

        """
        path = Path(file_path)
        path_str = str(path)
        if path_str in self._cache:
            return self._cache[path_str]

        content = path.read_text(encoding="utf-8")
        suffix = path.suffix.lower()

        if suffix == ".json":
            data = json.loads(content)
        elif suffix in [".yaml", ".yml"]:
            import yaml

            data = yaml.safe_load(content)
        else:
            raise ValueError(f"Unsupported locale file format: {suffix}")

        if not isinstance(data, dict):
            data = {}

        self._cache[str(file_path)] = data
        return data

    def get_all_keys(self, data: Dict[str, Any], prefix: str = "") -> Set[str]:
        """Get all keys from nested dictionary, including nested keys as dot notation.

        Args:
            data: Nested dictionary of locale data.
            prefix: Prefix for nested keys.

        Returns:
            Set of all keys in dot notation.

        """
        keys = set()
        for key, value in data.items():
            full_key = f"{prefix}.{key}" if prefix else key
            if isinstance(value, dict):
                keys.update(self.get_all_keys(value, full_key))
            else:
                keys.add(full_key)
        return keys

    def flatten_keys(self, data: Dict[str, Any], prefix: str = "") -> Dict[str, Any]:
        """Flatten nested dictionary to dot notation.

        Args:
            data: Nested dictionary of locale data.
            prefix: Prefix for nested keys.

        Returns:
            Flattened dictionary with dot notation keys.

        """
        flattened = {}
        for key, value in data.items():
            full_key = f"{prefix}.{key}" if prefix else key
            if isinstance(value, dict):
                flattened.update(self.flatten_keys(value, full_key))
            else:
                flattened[full_key] = value
        return flattened

    def unflatten_keys(self, data: Dict[str, Any]) -> Dict[str, Any]:
        """Unflatten dot notation keys to nested dictionary.

        Args:
            data: Flattened dictionary with dot notation keys.

        Returns:
            Nested dictionary.

        """
        nested: Dict[str, Any] = {}
        for key, value in data.items():
            parts = key.split(".")
            current = nested
            for part in parts[:-1]:
                if part not in current:
                    current[part] = {}
                current = current[part]
            current[parts[-1]] = value
        return nested

    def parse_directory(
        self,
        dir_path: str,
        locales: Optional[List[str]] = None,
    ) -> Dict[str, Set[str]]:
        """Parse all locale files in a directory.

        Args:
            dir_path: Path to the locale directory.
            locales: Optional list of locale names to include.

        Returns:
            Dictionary mapping locale names to their key sets.

        """
        locale_dir = Path(dir_path)
        results: Dict[str, Set[str]] = {}

        if not locale_dir.exists():
            return results

        for locale_file in locale_dir.iterdir():
            if not locale_file.is_file():
                continue

            locale_name = locale_file.stem
            if locales and locale_name not in locales:
                continue

            try:
                data = self.parse_file(locale_file)
                keys = self.get_all_keys(data)
                results[locale_name] = keys
            except (ValueError, json.JSONDecodeError) as e:
                if locales:
                    raise ValueError(f"Error parsing {locale_file}: {e}")

        return results

    def write_file(
        self,
        file_path: Path,
        data: Dict[str, Any],
        format: str = "json",
    ) -> None:
        """Write locale data to a file.

        Args:
            file_path: Path to the output file.
            data: Dictionary of locale data.
            format: Output format (json or yaml).

        """
        content: str
        if format == "json":
            content = json.dumps(data, indent=2, ensure_ascii=False)
        elif format in ["yaml", "yml"]:
            import yaml

            content = yaml.dump(data, default_flow_style=False, allow_unicode=True)
        else:
            raise ValueError(f"Unsupported format: {format}")

        file_path.write_text(content, encoding="utf-8")

    def add_key(
        self,
        data: Dict[str, Any],
        key: str,
        value: Any,
    ) -> Dict[str, Any]:
        """Add a key to locale data, creating nested structure if needed.

        Args:
            data: Existing locale data.
            key: Key in dot notation.
            value: Value to set.

        Returns:
            Updated locale data.

        """
        parts = key.split(".")
        current = data
        for part in parts[:-1]:
            if part not in current:
                current[part] = {}
            current = current[part]
        current[parts[-1]] = value
        return data

    def get_value(self, data: Dict[str, Any], key: str) -> Optional[Any]:
        """Get a value from locale data using dot notation.

        Args:
            data: Locale data dictionary.
            key: Key in dot notation.

        Returns:
            The value if found, None otherwise.

        """
        parts = key.split(".")
        current = data
        for part in parts:
            if isinstance(current, dict) and part in current:
                current = current[part]
            else:
                return None
        return current