markov-chain-linearizer

#!/usr/bin/env python3
# -*- coding: utf-8 -*-

# Copyright © 2017 Martin Ueding <mu@martin-ueding.de>

import argparse
import os
import re

DIGITS_PATTERN = re.compile(r'(\d{3,4})')

def get_index(path):
    basename = os.path.basename(path)
    m = DIGITS_PATTERN.search(basename)
    if m:
        return int(m.group(1))

    raise RuntimeError('The path `{}` did not contain a four digit number.'.format(path))


def extract_all_indices(paths):
    return {
        get_index(path): path
        for path in paths
    }


def get_new_path(path, new_dirname, new_index):
    new_name = DIGITS_PATTERN.sub('{:04d}'.format(new_index), os.path.basename(path))
    new_path = os.path.join(new_dirname, new_name)
    return new_path


def main():
    options = _parse_args()

    forward = extract_all_indices(options.forward)
    backward = extract_all_indices(options.backward)

    # Get the index of the last backward element. This is going to be the very
    # first element in the joined chain.
    last = max(backward.keys())

    new_midpoint = last - options.common

    renames_forward = [
        (new_midpoint + (index - options.common), path)
        for index, path in sorted(forward.items())
    ]
    renames_backward = [
        (new_midpoint - (index - options.common), path)
        for index, path in sorted(backward.items())
    ]
    renames = renames_forward + renames_backward

    mapping = [
        (path, get_new_path(path, options.target, new_index))
        for new_index, path in renames
    ]

    if options.doit:
        os.makedirs(options.target, exist_ok=True)

    action = 'Symlinked' if options.doit else 'Would symlink'

    for path, new_path in sorted(mapping):
        print('{} {:30s} to {:30s}'.format(action, path, new_path))

        if options.doit:
            os.symlink(path, new_path)

    if not options.doit:
        print()
        print('Run again with option `--doit` to actually create symlinks.')


def _parse_args():
    '''
    Parses the command line arguments.

    :return: Namespace with arguments.
    :rtype: Namespace
    '''
    parser = argparse.ArgumentParser(description='Linearizes two Markov chains into a single one with continous index numbers by creating symlinks.')
    parser.add_argument('--forward', nargs='+', required=True, help='Paths to files from the forward chain')
    parser.add_argument('--backward', nargs='+', required=True, help='Paths to files from the backward chain')
    parser.add_argument('--common', type=int, required=True, help='Index which is common to both parts of the chain')
    parser.add_argument('--target', required=True, help='Target directory for symlinks')
    parser.add_argument('--doit', action='store_true', help='Actually create symlinks, otherwise a dry run is performed only')
    options = parser.parse_args()

    return options


if __name__ == '__main__':
    main()