icscat

#!/usr/bin/bash

# Given several ICS files, concatenate the events.
# Use the calendar-level metadata (e.g. calendar name, source URL, timezone
# definitions) from the *last* calendar specified.
#
# Warning: This is pretty fragile text editing.  Make sure TZIDs still match,
# and the file layout is the same (events are the last thing in the file), etc.
#
# Usage: icscat part1.ics part2.ics ... partN.ics > out.ics

# Use the preamble of the last file.
# Preamble is from start to (but not including) the first BEGIN:VEVENT.
sed -nre '/BEGIN:VEVENT/ q; p' "${!#}"

for ics; do # for each positional argument
    # Print from the first BEGIN:VEVENT to (but not including) END:VCALENDAR
    sed -nre '/BEGIN:VEVENT/,$ { /END:VCALENDAR/ q; p }' "$ics"
done

printf 'END:VCALENDAR\r\n'