add_cloze

#!/usr/bin/env python3
"""
Add Anki cloze deletions of increasing size to a text.
"""
import argparse
import logging

logging.basicConfig(level=logging.INFO)

CHUNK_MULTIPLIER = 2
INITIAL_CHUNK_SIZE = 2


def read_lines_from_file(filename):
    with open(filename, "r") as file:
        lines = [l.strip() for l in file]
        while not lines[-1]:
            lines.pop()
        return lines


def start_cloze(line, cloze_index):
    return "{{c" + str(cloze_index) + "::" + line


def finish_cloze(line):
    return line + "}}"


def encloze(line, cloze_index):
    return finish_cloze(start_cloze(line, cloze_index))


def create_nested_cloze_chunks(lines):
    """
    Partially overlapping cloze deletions are not supported, so bigger chunks
    must contain smaller chunks. Guaranteed if bigger chunk sizes are
    multiples of smaller chunk sizes.

    See
    "There is no support for partial overlaps" in
    https://docs.ankiweb.net/editing.html?highlight=cloze#cloze-deletion.
    """
    cloze_index = 1

    added_chunks = set()
    # Distinguishing case chunk_size == 1 to skip empty lines. We assume
    # there's at most one consecutive empty line.
    for i in range(len(lines)):
        if not lines[i]:
            continue
        lines[i] = encloze(lines[i], cloze_index)
        added_chunks.add((i, i))
        cloze_index += 1

    chunk_size = 2
    while chunk_size < len(lines):
        for start in range(0, len(lines), chunk_size):
            end = min(start + chunk_size, len(lines)) - 1
            if (start, end) in added_chunks:
                logging.debug(f"{start}, {end} already in.")
                continue
            logging.debug(f"{start}, {end}, {cloze_index} added.")
            logging.debug(f"Begin: {lines[start]}")
            logging.debug(f"End: {lines[end]}")
            added_chunks.add((start, end))
            lines[start] = start_cloze(lines[start], cloze_index)
            lines[end] = finish_cloze(lines[end])
            cloze_index += 1
        chunk_size *= CHUNK_MULTIPLIER


def write_lines_to_file(filename, lines, overwrite=False):
    cloze_file = "cloze_" + filename
    mode = "w" if overwrite else "x"
    try:
        with open(cloze_file, mode) as file:
            file.write("\n".join(lines))
    except FileExistsError as e:
        logging.error("File already exists: %s", e)


def add_cloze_deletion(filename, overwrite):
    lines = read_lines_from_file(filename)
    create_nested_cloze_chunks(lines)
    write_lines_to_file(filename, lines, overwrite)


if __name__ == "__main__":
    parser = argparse.ArgumentParser(
        description="Add cloze deletion tags to chunks of text in a file."
    )
    parser.add_argument("filename", help="Input filename.")
    parser.add_argument(
        "-f",
        "--force",
        action="store_true",
        default=False,
        help="Overwrite existing file.",
    )

    args = parser.parse_args()
    add_cloze_deletion(args.filename, overwrite=args.force)