Slice text output to columns and convert to equivalent CSV

slice

#!/usr/bin/env python3
"""Slice ascii output along columns of empty space into a table.

Any vertical column of whitespace spanning the entire height of the input
generates a column separator.

Ideas:
- We could enhance this to detect 2+ spaces in the header field names as the
  only legitimate places for separation (to avoid false positives).
- Handle tabs.
"""

__copyright__ = "Copyright (C) 2023 Martin Blais. All Rights Reserved."
__author__ = "Martin Blais <blais@furius.ca>"

import argparse
import csv
import itertools
import re
import sys


def slice_line(line, spans):
    for start, end in spans[:-1]:
        yield line[start:end].strip()
    start, end = spans[-1]
    yield line[start:end].strip()


def main():
    parser = argparse.ArgumentParser(description=__doc__.strip())
    parser.add_argument("filename", nargs="?", help="Filename")
    parser.add_argument(
        "-n",
        "--no-header",
        action="store_false",
        default=True,
        dest="header",
        help="Ensure header is present or merge detected columns.",
    )
    args = parser.parse_args()

    infile = sys.stdin if args.filename in {None, "-"} else open(args.filename, "r")
    lines = infile.readlines()

    chars = []
    for line in lines:
        diff_len = len(line) - len(chars)
        if diff_len > 0:
            for _ in range(diff_len):
                chars.append(" ")
        chars = [
            (p if c == " " else "x") for (p, c) in itertools.zip_longest(chars, line)
        ]

    spans = [match.span() for match in re.finditer("x+", "".join(chars))]

    if args.header:
        new_spans = []
        spaniter = zip(spans, slice_line(lines[0], spans))
        (start, end), _ = next(spaniter)
        for span, field in spaniter:
            if field:
                new_spans.append((start, end))
                start, end = span
            else:
                _, end = span
        new_spans.append((start, end))
        spans = new_spans

    writer = csv.writer(sys.stdout)
    for line in lines:
        writer.writerow(slice_line(line, spans))


if __name__ == "__main__":
    main()