deps

#!/usr/bin/env python3

import argparse
import os
import re
import subprocess
import sys
import urllib.request
import urllib.error

############################
# Built-in, Known Corrections, Exclusions
############################

BUILTIN_MODULES = {
    'abc', 'aifc', 'argparse', 'array', 'ast', 'asynchat', 'asyncio', 'asyncore', 'atexit',
    'audioop', 'base64', 'bdb', 'binascii', 'binhex', 'bisect', 'builtins', 'bz2', 'calendar',
    'cgi', 'cgitb', 'chunk', 'cmath', 'cmd', 'code', 'codecs', 'codeop', 'collections', 'colorsys',
    'compileall', 'concurrent', 'configparser', 'contextlib', 'copy', 'copyreg', 'crypt', 'csv',
    'ctypes', 'curses', 'dataclasses', 'datetime', 'dbm', 'decimal', 'difflib', 'dis', 'distutils',
    'doctest', 'dummy_threading', 'email', 'encodings', 'ensurepip', 'enum', 'errno', 'faulthandler',
    'fcntl', 'filecmp', 'fileinput', 'fnmatch', 'formatter', 'fractions', 'ftplib', 'functools',
    'gc', 'getopt', 'getpass', 'gettext', 'glob', 'gzip', 'hashlib', 'heapq', 'hmac', 'html', 'http',
    'imaplib', 'imghdr', 'imp', 'importlib', 'inspect', 'io', 'ipaddress', 'itertools', 'json',
    'keyword', 'lib2to3', 'linecache', 'locale', 'logging', 'lzma', 'mailbox', 'mailcap', 'marshal',
    'math', 'mimetypes', 'modulefinder', 'multiprocessing', 'netrc', 'nntplib', 'numbers', 'operator',
    'optparse', 'os', 'ossaudiodev', 'parser', 'pathlib', 'pdb', 'pickle', 'pickletools', 'pipes',
    'pkgutil', 'platform', 'plistlib', 'poplib', 'posix', 'pprint', 'profile', 'pstats', 'pty',
    'pwd', 'py_compile', 'pyclbr', 'pydoc', 'queue', 'quopri', 'random', 're', 'readline',
    'reprlib', 'resource', 'rlcompleter', 'runpy', 'sched', 'secrets', 'select', 'selectors', 'shelve',
    'shlex', 'shutil', 'signal', 'site', 'smtpd', 'smtplib', 'sndhdr', 'socket', 'socketserver',
    'spwd', 'sqlite3', 'ssl', 'stat', 'statistics', 'string', 'stringprep', 'struct', 'subprocess',
    'sunau', 'symtable', 'sys', 'sysconfig', 'syslog', 'tabnanny', 'tarfile', 'telnetlib', 'tempfile',
    'termios', 'test', 'textwrap', 'threading', 'time', 'timeit', 'token', 'tokenize', 'trace',
    'traceback', 'tracemalloc', 'tty', 'turtle', 'types', 'typing', 'unicodedata', 'unittest',
    'urllib', 'uu', 'uuid', 'venv', 'warnings', 'wave', 'weakref', 'webbrowser', 'xdrlib', 'xml',
    'xmlrpc', 'zipapp', 'zipfile', 'zipimport', 'zlib'
}

KNOWN_CORRECTIONS = {
    'dateutil': 'python-dateutil',
    'dotenv': 'python-dotenv',
    'docx': 'python-docx',
    'tesseract': 'pytesseract',
    'magic': 'python-magic',
    'multipart': 'python-multipart',
    'newspaper': 'newspaper3k',
    'srtm': 'elevation',
    'yaml': 'pyyaml',
    'zoneinfo': 'backports.zoneinfo'
}

EXCLUDED_NAMES = {'models', 'data', 'convert', 'example', 'tests'}

############################
# Environment & Installation
############################

def run_command(command):
    process = subprocess.Popen(command, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
    stdout, stderr = process.communicate()
    return process.returncode, stdout.decode(), stderr.decode()

def which(cmd):
    """
    Check if `cmd` is on PATH. Returns True if found, else False.
    """
    for pth in os.environ["PATH"].split(os.pathsep):
        cmd_path = os.path.join(pth, cmd)
        if os.path.isfile(cmd_path) and os.access(cmd_path, os.X_OK):
            return True
    return False

def in_conda_env():
    """
    Returns True if we appear to be in a conda environment,
    typically indicated by CONDA_DEFAULT_ENV or other variables.
    """
    return "CONDA_DEFAULT_ENV" in os.environ

# We'll detect once at runtime (if in a conda env and skip_conda=False):
# we either pick 'mamba' if available, else 'conda' if available, else None
PREFERRED_CONDA_TOOL = None

def detect_conda_tool(skip_conda=False):
    """
    Decide which tool to use for conda-based installation:
    1) If skip_conda is True or not in a conda env -> return None
    2) If mamba is installed, return 'mamba'
    3) Else if conda is installed, return 'conda'
    4) Else return None
    """
    if skip_conda or not in_conda_env():
        return None
    if which("mamba"):
        return "mamba"
    elif which("conda"):
        return "conda"
    return None

def is_package_installed(package, skip_conda=False):
    """
    Checks if 'package' is installed with the chosen conda tool or pip.
    """
    conda_tool = detect_conda_tool(skip_conda)
    if conda_tool == "mamba":
        returncode, stdout, _ = run_command(["mamba", "list"])
        if returncode == 0:
            pattern = rf"^{re.escape(package)}\s"
            if re.search(pattern, stdout, re.MULTILINE):
                return True
    elif conda_tool == "conda":
        returncode, stdout, _ = run_command(["conda", "list"])
        if returncode == 0:
            pattern = rf"^{re.escape(package)}\s"
            if re.search(pattern, stdout, re.MULTILINE):
                return True

    # Fall back to pip
    returncode, stdout, _ = run_command(["pip", "list"])
    pattern = rf"^{re.escape(package)}\s"
    return re.search(pattern, stdout, re.MULTILINE) is not None

def install_package(package, skip_conda=False):
    """
    Installs 'package'.
      1) Decide once if we can use 'mamba' or 'conda' (if skip_conda=False and in conda env).
      2) Try that conda tool for installation
      3) If that fails or not found, fallback to pip
    """
    if is_package_installed(package, skip_conda=skip_conda):
        print(f"Package '{package}' is already installed.")
        return

    conda_tool = detect_conda_tool(skip_conda)

    if conda_tool == "mamba":
        print(f"Installing '{package}' with mamba...")
        returncode, _, _ = run_command(["mamba", "install", "-y", "-c", "conda-forge", package])
        if returncode == 0:
            print(f"Successfully installed '{package}' via mamba.")
            return
        print(f"mamba failed for '{package}'. Falling back to pip...")

    elif conda_tool == "conda":
        print(f"Installing '{package}' with conda...")
        returncode, _, _ = run_command(["conda", "install", "-y", "-c", "conda-forge", package])
        if returncode == 0:
            print(f"Successfully installed '{package}' via conda.")
            return
        print(f"conda failed for '{package}'. Falling back to pip...")

    # fallback: pip
    print(f"Installing '{package}' with pip...")
    returncode, _, _ = run_command(["pip", "install", package])
    if returncode != 0:
        print(f"Failed to install package '{package}'.")
    else:
        print(f"Successfully installed '{package}' via pip.")

############################
# Parsing Python Imports
############################

def process_requirements_file(file_path):
    packages = set()
    with open(file_path, 'r') as file:
        for line in file:
            line = line.strip()
            if line and not line.startswith('#'):
                packages.add(line)
    return packages

def process_python_file(file_path):
    """
    Return a set of external imports (not built-in or excluded).
    Applies known corrections to recognized package names.
    """
    imports = set()
    with open(file_path, 'r', encoding='utf-8', errors='ignore') as f:
        content = f.read()

    for line in content.split('\n'):
        line = line.strip()
        if line.startswith(('import ', 'from ')) and not line.startswith('#'):
            if line.startswith('import '):
                modules = line.replace('import ', '').split(',')
                for mod in modules:
                    mod = re.sub(r'\s+as\s+\w+', '', mod).split('.')[0].strip()
                    if mod and not mod.isupper() and mod not in EXCLUDED_NAMES and mod not in BUILTIN_MODULES:
                        imports.add(KNOWN_CORRECTIONS.get(mod, mod))
            elif line.startswith('from '):
                mod = line.split(' ')[1].split('.')[0].strip()
                if mod and not mod.isupper() and mod not in EXCLUDED_NAMES and mod not in BUILTIN_MODULES:
                    imports.add(KNOWN_CORRECTIONS.get(mod, mod))

    return imports

def find_imports_in_path(path, recurse=False):
    """
    Finds Python imports in the specified path. If path is a file, parse that file;
    if path is a dir, parse .py files in that dir. Recurse subdirs if 'recurse=True'.
    """
    imports = set()
    if not os.path.exists(path):
        print(f"Warning: Path does not exist: {path}")
        return imports

    if os.path.isfile(path):
        if path.endswith('.py'):
            imports.update(process_python_file(path))
        else:
            print(f"Skipping non-Python file: {path}")
        return imports

    # Directory:
    if recurse:
        for root, _, filenames in os.walk(path):
            for fn in filenames:
                if fn.endswith('.py'):
                    imports.update(process_python_file(os.path.join(root, fn)))
    else:
        for fn in os.listdir(path):
            fullpath = os.path.join(path, fn)
            if os.path.isfile(fullpath) and fn.endswith('.py'):
                imports.update(process_python_file(fullpath))

    return imports

############################
# PyPI Availability Check
############################

def check_library_on_pypi(library):
    """
    Returns True if 'library' is on PyPI, else False.
    Using urllib to avoid external dependencies.
    """
    url = f"https://pypi.org/pypi/{library}/json"
    try:
        with urllib.request.urlopen(url, timeout=5) as resp:
            return (resp.status == 200)  # 200 => available
    except (urllib.error.URLError, urllib.error.HTTPError, ValueError):
        return False

############################
# Writing to requirements/missing
############################

def append_to_file(line, filename):
    """
    Append 'line' to 'filename' only if it's not already in there.
    """
    if not os.path.isfile(filename):
        with open(filename, 'w') as f:
            f.write(line + '\n')
        return

    with open(filename, 'r') as f:
        lines = {l.strip() for l in f.readlines() if l.strip()}
    if line not in lines:
        with open(filename, 'a') as f:
            f.write(line + '\n')

############################
# Subcommand: ls
############################

def subcmd_ls(parsed_args):
    """
    Gathers imports, displays them, then writes them to requirements.txt or missing-packages.txt
    just like the original import_finder script did.
    """
    path = parsed_args.path
    recurse = parsed_args.recurse

    imports = find_imports_in_path(path, recurse=recurse)
    if not imports:
        print("No Python imports found (or none that require external packages).")
        return

    print("Imports found:")
    for imp in sorted(imports):
        print(f" - {imp}")

    # Now we replicate the logic of import_finder:
    # If on PyPI => requirements.txt; else => missing-packages.txt
    for lib in sorted(imports):
        if check_library_on_pypi(lib):
            append_to_file(lib, 'requirements.txt')
        else:
            append_to_file(lib, 'missing-packages.txt')

    print("\nWrote results to requirements.txt (PyPI-available) and missing-packages.txt (unavailable).")

############################
# Subcommand: install
############################

def subcmd_install(parsed_args):
    """
    If the user typed no direct packages/scripts or only used '-r' for recursion with no other args,
    we gather imports from the current dir, check PyPI availability, and install them with conda/mamba/pip
    (unless --no-conda is given).
    
    Otherwise, if the user typed e.g. '-R <reqfile>', or a .py file, or direct package names, we handle them.
    """
    skip_conda = parsed_args.no_conda
    is_recursive = parsed_args.recurse

    # If user typed no leftover arguments or only the recursion flag, we do the "auto-scan & install" mode
    if not parsed_args.packages:
        # Means: "deps install" or "deps install -r"
        imports_found = find_imports_in_path('.', recurse=is_recursive)
        if not imports_found:
            print("No imports found in current directory.")
            return
        # Filter out those that are on PyPI
        to_install = []
        for lib in sorted(imports_found):
            if check_library_on_pypi(lib):
                to_install.append(lib)
            else:
                print(f"Skipping '{lib}' (not found on PyPI).")
        if not to_install:
            print("No PyPI-available packages found to install.")
            return
        print("Installing packages:", ', '.join(to_install))
        for pkg in to_install:
            install_package(pkg, skip_conda=skip_conda)
        return

    # Otherwise, we have leftover items: direct packages, .py files, or "-R" requirements.
    leftover_args = parsed_args.packages
    packages_to_install = set()

    i = 0
    while i < len(leftover_args):
        arg = leftover_args[i]
        if arg == '-R':
            # next arg is a requirements file
            if i + 1 < len(leftover_args):
                req_file = leftover_args[i + 1]
                if os.path.isfile(req_file):
                    pkgs = process_requirements_file(req_file)
                    packages_to_install.update(pkgs)
                else:
                    print(f"Requirements file not found: {req_file}")
                i += 2
            else:
                print("Error: -R requires a file path.")
                return
        elif arg.endswith('.py'):
            # parse imports from that script
            if os.path.isfile(arg):
                pkgs = process_python_file(arg)
                packages_to_install.update(pkgs)
            else:
                print(f"File not found: {arg}")
            i += 1
        else:
            # treat as a direct package name
            packages_to_install.add(arg)
            i += 1

    # Now install them
    for pkg in sorted(packages_to_install):
        install_package(pkg, skip_conda=skip_conda)

############################
# Main
############################

def main():
    parser = argparse.ArgumentParser(description='deps - Manage and inspect Python dependencies.')
    subparsers = parser.add_subparsers(dest='subcommand', required=True)

    # Subcommand: ls
    ls_parser = subparsers.add_parser(
        'ls',
        help="List imports in a file/folder (and write them to requirements.txt/missing-packages.txt)."
    )
    ls_parser.add_argument(
        '-r', '--recurse',
        action='store_true',
        help='Recurse into subfolders.'
    )
    ls_parser.add_argument(
        'path',
        nargs='?',
        default='.',
        help='File or directory to scan (default is current directory).'
    )
    ls_parser.set_defaults(func=subcmd_ls)

    # Subcommand: install
    install_parser = subparsers.add_parser(
        'install',
        help="Install packages or dependencies from .py files / current folder / subfolders."
    )
    install_parser.add_argument(
        '-r', '--recurse',
        action='store_true',
        help="If no packages are specified, scanning current dir for imports will be recursive."
    )
    install_parser.add_argument(
        '--no-conda',
        action='store_true',
        help="Skip using mamba/conda entirely and install only with pip."
    )
    install_parser.add_argument(
        'packages',
        nargs='*',
        help=(
            "Direct package names, .py files, or '-R <reqfile>'. If empty, scans current dir; "
            "if combined with -r, scans recursively. Example usage:\n"
            "  deps install requests flask\n"
            "  deps install script.py\n"
            "  deps install -R requirements.txt\n"
            "  deps install -r   (recursively scan current dir)\n"
        )
    )
    install_parser.set_defaults(func=subcmd_install)

    parsed_args = parser.parse_args()
    parsed_args.func(parsed_args)

if __name__ == "__main__":
    main()