From cef9f4f8042f67d7fb7b01a8633609d5e6cb08fd Mon Sep 17 00:00:00 2001
From: Simon Howard <fraggle@gmail.com>
Date: Fri, 14 Feb 2014 04:39:51 +0000
Subject: [PATCH] scripts: Add music duplicates script.

We have lots of music tracks serving double duty for multiple levels,
but it's hard to tell which are the missing ones at a glance. Use
some heuristics to identify duplicated tracks and print a report on
what we still need.
---
 scripts/music-duplicates.py | 111 ++++++++++++++++++++++++++++++++++++
 1 file changed, 111 insertions(+)
 create mode 100755 scripts/music-duplicates.py

diff --git a/scripts/music-duplicates.py b/scripts/music-duplicates.py
new file mode 100755
index 00000000..5db55772
--- /dev/null
+++ b/scripts/music-duplicates.py
@@ -0,0 +1,111 @@
+#!/usr/bin/env python
+#
+# Find duplicated music tracks and create a summary report of music
+# that the project needs.
+
+from glob import glob
+import os
+import re
+import sys
+
+PHASE1_MATCH_RE = re.compile(r'(e\dm\d)', re.I)
+PHASE2_MATCH_RE = re.compile(r'(map\d\d)', re.I)
+FREEDM_MATCH_RE = re.compile(r'(dm\d\d)', re.I)
+
+DOOM2_TRACKS = (
+	'runnin', 'stalks', 'countd', 'betwee', 'doom', 'the_da', 'shawn',
+	'ddtblu', 'in_cit', 'dead', 'stlks2', 'theda2', 'doom2', 'ddtbl2',
+	'runni2', 'dead2', 'stlks3', 'romero', 'shawn2', 'messag', 'count2',
+	'ddtbl3', 'ampie', 'theda3', 'adrian', 'messg2', 'romer2', 'tense',
+	'shawn3', 'openin', 'evil', 'ultima',
+)
+
+def get_music_tracks():
+	"""Returns a dictionary mapping from MIDI name (subpath of musics/)
+	   to a list of game tracks that use that MIDI."""
+	result = {}
+	musics_path = os.path.join(os.path.dirname(sys.argv[0]), '../musics')
+	for mus in glob('%s/*.mus' % musics_path):
+		try:
+			symlink = os.readlink(mus)
+			if symlink not in result:
+				result[symlink] = []
+			result[symlink].append(os.path.basename(mus))
+		except OSError:
+			pass
+	return result
+
+def doom2_level_for_file(filename):
+	"""Given a filename that may be named like a Doom 2 music name
+	   (eg. d_stalks.mid), get the level number it corresponds to
+	   or 0 if it doesn't match."""
+	filename = os.path.basename(filename)
+	for i, doom2_name in enumerate(DOOM2_TRACKS):
+		if filename.startswith('d_%s.' % doom2_name):
+			return i + 1
+	else:
+		return 0
+
+def get_prime_track(midi, tracks):
+	"""Given a list of tracks that all use the same MIDI, find the
+	   "prime" one (the one that isn't a reuse/duplicate)."""
+	# We have almost all Phase 2 tracks fulfilled. So if the same
+	# track is used in Phase 1 and Phase 2, or Phase 2 and FreeDM,
+	# the Phase 2 track is probably the leader.
+	phase2_tracks = [x for x in tracks if PHASE2_MATCH_RE.search(x)]
+	if len(phase2_tracks) == 1:
+		return phase2_tracks[0]
+
+	level = doom2_level_for_file(midi)
+	if level:
+		for track in phase2_tracks:
+			if ('map%02i' % level) in track:
+				return track
+
+	# If the filename of the MIDI file (symlink target) describes a
+	# level, then that level is the leader.
+	m = PHASE1_MATCH_RE.search(os.path.basename(midi))
+	if m:
+		level = m.group(1)
+		for track in tracks:
+			if level in track:
+				return track
+
+	# We're out of options. Pick the first one in the list.
+	#print 'Warning: Don't know which of %s is the leader for %s.' % (
+	#	tracks, midi)
+	return tracks[0]
+
+def find_missing_tracks(tracks):
+	"""Given a dictionary of tracks, get a list of "missing" tracks."""
+	result = []
+	for midi, tracks in tracks.items():
+		if len(tracks) < 2:
+			continue
+		prime_track = get_prime_track(midi, tracks)
+		result.extend(x for x in tracks if x != prime_track)
+	return result
+
+def tracks_matching_regexp(tracks, regexp):
+	return set([x for x in tracks if regexp.search(x)])
+
+def print_report(title, tracks):
+	if len(tracks) == 0:
+		return
+	print(title)
+	for track in sorted(tracks):
+		print('\t%s' % track.replace('.mus', '').upper())
+	print('')
+
+missing_tracks = set(find_missing_tracks(get_music_tracks()))
+phase1_tracks = tracks_matching_regexp(missing_tracks, PHASE1_MATCH_RE)
+phase2_tracks = tracks_matching_regexp(missing_tracks, PHASE2_MATCH_RE)
+freedm_tracks = tracks_matching_regexp(missing_tracks, FREEDM_MATCH_RE)
+other_tracks = missing_tracks - phase1_tracks - phase2_tracks - freedm_tracks
+
+print('=== Missing tracks (tracks currently using duplicates):\n')
+print_report('Phase 1 tracks:', phase1_tracks)
+print_report('Phase 2 tracks:', phase2_tracks)
+print_report('FreeDM tracks:', freedm_tracks)
+print_report('Other tracks:', other_tracks)
+