shell-history-automation-tool/shellhist/core/__init__.py

"""Core module for shell history loading and parsing."""

import os
import re
from dataclasses import dataclass, field
from datetime import datetime
from typing import Optional, TextIO


@dataclass
class HistoryEntry:
    """Represents a single entry in shell history."""
    command: str
    timestamp: Optional[datetime] = None
    line_number: int = 0
    shell_type: str = "unknown"

    def __hash__(self) -> int:
        return hash(self.command)

    def __eq__(self, other: object) -> bool:
        if isinstance(other, HistoryEntry):
            return self.command == other.command
        return False


@dataclass
class HistoryStore:
    """Stores and manages shell history entries."""
    entries: list[HistoryEntry] = field(default_factory=list)
    command_frequency: dict[str, int] = field(default_factory=dict)
    
    def add_entry(self, entry: HistoryEntry) -> None:
        """Add a history entry to the store."""
        self.entries.append(entry)
        self.command_frequency[entry.command] = self.command_frequency.get(entry.command, 0) + 1
    
    def get_frequency(self, command: str) -> int:
        """Get the frequency of a command."""
        return self.command_frequency.get(command, 0)
    
    def get_most_frequent(self, limit: int = 10) -> list[tuple[str, int]]:
        """Get the most frequently used commands."""
        sorted_freq = sorted(self.command_frequency.items(), key=lambda x: x[1], reverse=True)
        return sorted_freq[:limit]
    
    def get_unique_commands(self) -> list[str]:
        """Get list of unique commands."""
        return list(self.command_frequency.keys())


class HistoryLoader:
    """Loads and parses shell history from various shell formats."""
    
    BASH_TIMESTAMP_PATTERN = re.compile(r'^#(\d{10})$')
    ZSH_TIMESTAMP_PATTERN = re.compile(r'^: (\d{10}):(\d+);(.*)$')
    
    def __init__(self, history_path: Optional[str] = None):
        """Initialize the history loader.
        
        Args:
            history_path: Optional path to history file. If not provided,
                         will attempt to detect from environment.
        """
        self.history_path = history_path
        self.shell_type = self._detect_shell()
    
    def _detect_shell(self) -> str:
        """Detect the shell type from environment."""
        shell = os.environ.get("SHELL", "")
        if "zsh" in shell:
            return "zsh"
        elif "bash" in shell:
            return "bash"
        return "bash"
    
    def _get_default_history_path(self) -> Optional[str]:
        """Get the default history file path based on shell type."""
        home = os.path.expanduser("~")
        if self.shell_type == "zsh":
            return os.environ.get("HISTFILE", f"{home}/.zsh_history")
        return os.environ.get("HISTFILE", f"{home}/.bash_history")
    
    def load(self) -> HistoryStore:
        """Load history from the configured path.
        
        Returns:
            HistoryStore containing all parsed entries.
        """
        path = self.history_path or self._get_default_history_path()
        
        if not path or not os.path.exists(path):
            raise FileNotFoundError(
                f"History file not found: {path}. "
                "Set HISTFILE environment variable or use --history flag."
            )
        
        store = HistoryStore()
        
        with open(path, "r", encoding="utf-8", errors="replace") as f:
            if self.shell_type == "zsh":
                self._parse_zsh_history(f, store)
            else:
                self._parse_bash_history(f, store)
        
        return store
    
    def _parse_bash_history(self, file: TextIO, store: HistoryStore) -> None:
        """Parse bash history format."""
        line_number = 0
        pending_timestamp: Optional[datetime] = None
        
        for line in file:
            line = line.rstrip("\n")
            line_number += 1
            
            timestamp_match = self.BASH_TIMESTAMP_PATTERN.match(line)
            if timestamp_match:
                ts = int(timestamp_match.group(1))
                pending_timestamp = datetime.fromtimestamp(ts)
                continue
            
            if line and not line.startswith("#"):
                command = line
                entry = HistoryEntry(
                    command=command,
                    timestamp=pending_timestamp,
                    line_number=line_number,
                    shell_type="bash"
                )
                store.add_entry(entry)
                pending_timestamp = None
    
    def _parse_zsh_history(self, file: TextIO, store: HistoryStore) -> None:
        """Parse zsh history format."""
        line_number = 0
        
        for line in file:
            line = line.rstrip("\n")
            line_number += 1
            
            timestamp_match = self.ZSH_TIMESTAMP_PATTERN.match(line)
            if timestamp_match:
                ts = int(timestamp_match.group(1))
                command = timestamp_match.group(3)
                timestamp = datetime.fromtimestamp(ts)
                
                entry = HistoryEntry(
                    command=command,
                    timestamp=timestamp,
                    line_number=line_number,
                    shell_type="zsh"
                )
                store.add_entry(entry)
            elif line:
                entry = HistoryEntry(
                    command=line,
                    timestamp=None,
                    line_number=line_number,
                    shell_type="zsh"
                )
                store.add_entry(entry)
    
    @staticmethod
    def from_file(path: str) -> HistoryStore:
        """Convenience method to load history from a specific file.
        
        Args:
            path: Path to history file.
            
        Returns:
            HistoryStore containing parsed entries.
        """
        loader = HistoryLoader(history_path=path)
        return loader.load()