#!/usr/bin/env python3

"""Report on gaps found in syslog-formatted files."""

import gzip
import sys
import time
import typing


def usage(retval: int) -> None:
    """Output a usage message."""
    if retval == 0:
        write = sys.stdout.write
    else:
        write = sys.stderr.write

    write(f'Usage: {sys.argv[0]} --gap-duration 300 --both -- file1 file2.gz ... file3\n')
    write('    The units on --gap-duration are seconds.\n')
    write('    --both to output the "before gap" and "after gap" lines, separated by a blank line\n')
    sys.exit(retval)


def get_lines() -> typing.Iterator[typing.Tuple[float, str]]:
    """Read and generate lines from sys.argv[1:]."""
    for filename in sys.argv[1:]:
        try:
            if filename.endswith('.gz'):
                with gzip.open(filename, 'rb') as gz_file:
                    for lineno, line in enumerate(gz_file, 1):
                        string = line.decode('iso-8859-1')
                        yield (get_time(string), string)
            else:
                with open(filename, 'rb') as plain_file:
                    for lineno, line in enumerate(plain_file, 1):
                        string = line.decode('iso-8859-1')
                        yield (get_time(string), string)
        except BadTimeError:
            print(f'{sys.argv[0]}: {filename} has an unexpected timestamp at line {lineno}; ignoring', file=sys.stderr)


class BadTimeError(Exception):
    """An exception to raise on a timestamp with an unexpected or invalid format."""

    pass


def get_time(line: str) -> float:
    """Extract the time, in seconds since the epoch, from a single log line."""
    # Jan 12 06:06:59 zareason-limbo colord[1010]: failed to get session [pid 2882]: No data available
    fields = line.split()
    just_time = ' '.join(fields[:3])
    try:
        tm_struct = time.strptime(just_time, '%b %d %H:%M:%S')
    except ValueError:
        raise BadTimeError
    tm_secs = time.mktime(tm_struct)
    return tm_secs


def gap_starts(duration: float) -> typing.Iterator[typing.Tuple[str, str]]:
    """Yield the first line of all adjacent log line pairs that have a gap of more than 300 seconds."""
    lines = list(get_lines())
    lines.sort()
    for first_line, second_line in zip(lines, lines[1:]):
        first_time = first_line[0]
        second_time = second_line[0]
        difference = second_time - first_time
        if difference > duration:
            yield (first_line[1], second_line[1])


def main() -> None:
    """Get the ball rolling."""
    gap_duration = 300.0
    both = False

    if not sys.argv[1:]:
        usage(0)

    while sys.argv[1:]:
        if sys.argv[1] == '--gap-duration':
            gap_duration = float(sys.argv[2])
            del sys.argv[1]
        elif sys.argv[1] == '--both':
            both = True
        elif sys.argv[1] in ('-h', '--help'):
            usage(0)
        elif sys.argv[1] == '--':
            del sys.argv[1]
            break
        elif not sys.argv[1].startswith('--'):
            break
        else:
            print(f'{sys.argv[0]}: unexpected option: {sys.argv[1]}', file=sys.stderr)
            usage(1)
        del sys.argv[1]

    for gap_start, gap_end in gap_starts(gap_duration):
        if both:
            sys.stdout.write(gap_start)
            sys.stdout.write(gap_end)
            sys.stdout.write('\n')
        else:
            sys.stdout.write(gap_start)


main()