#!/usr/bin/python
# coding: euc-jp

# Requires: CDDB.py http://cddb-py.sourceforge.net/
# Requires: eyeD3 http://eyed3.nicfit.net/

import CDDB, DiscID
import string
import codecs
import sys, os
import re
import subprocess
import struct
from array import array
import eyeD3
from threading import Thread

OUT_DIR = "/share03/music/Music"

BROKEN_CHARSET = "cp949"  # 韓国語
#BROKEN_CHARSET = "cp874"  # タイ語

DIR_NAME = "%(artist)s - %(album)s"

FILE_NAME = "%(track_no)02d_%(track)s"
#FILE_NAME = "%(artist)s - %(track)s"

CDROM = "/dev/cdrom"

MP3_COMMAND = ["gogo", "-silent"]

CDDB.proto = 6

def escape_filename(filename):
    filename = re.sub(r'[\\/:*?"<>|]', '_', filename)
    filename = re.sub(r'[. ]+$', '', filename)

    return filename

def track2filename(i, info, ext):
    track_info = {}
    track_info = dict(category=info["category"],
                      artist=info["artist"],
                      album=info["album"],
                      year=info["year"],
                      track_no=i + 1,
                      track=info["tracks"][i])

    filename = FILE_NAME % track_info
    filename = "%s.%s" % (escape_filename(filename), ext)
    return filename.encode("utf-8")

def fix_broken_korean(str):
    """古い韓国などのCDのcddbを取得すると、euc-krの「c1f6」が「c381c3b6」
       に文字化けしている。これは、c1/f6 を1バイトずつutf-8に変換
       した結果であるので、c381/c3b6 を c1/f6 に変換し、更に、これを
       euc-krからutf-8に変換する。(pythonでは書きにくい・・・)"""

    b_str = array("B", str)

    for c in b_str:
        if c >= 0x80:
            # 最初に出てくる非asciiな文字が0xc0 <= 0xc3の場合のみ
            # ハングルなどの文字化けパターンとして処理する
            if 0xc0 <= c <= 0xc3:
                break
            else:
                return str

    # U+0080...U+07FF のUTF-8 -> Unicode変換を行う
    # 但し、今回は0x80...0xFFの範囲のみとして処理する
    new_array = array("B")
    prev = 0
    in_utf = False
    for c in b_str:
        if 0xc0 <= c <= 0xc3:
            prev = (c & 3)
            in_utf = True
        else:
            if in_utf:
                 new_array.append((c & 0x3f) | (prev << 6))
                 in_utf = False
            else:
                 new_array.append(c)

    try:
        return new_array.tostring().decode(BROKEN_CHARSET).encode("utf-8")

    # ヨーロッパ系言語?
    except UnicodeDecodeError:
        return str


def get_cddb():
    cdrom = DiscID.open()
    disc_id = DiscID.disc_id(cdrom)

    (query_status, query_info) = CDDB.query(disc_id)

    if query_status != 200:
        return None

    query_info["title"] = fix_broken_korean(query_info["title"]).decode("utf-8")

    (artist, album) = map(string.strip, query_info["title"].split(" / ", 1))

    (read_status, read_info) = CDDB.read(query_info['category'], query_info['disc_id'])

    track_titles = []
    for i in range(disc_id[1]):
        title = fix_broken_korean(read_info["TTITLE%s" % (i)]).decode("utf-8")
        track_titles.append(title)

    info = dict(category=query_info["category"],
                artist=artist,
                album=album,
                tracks=track_titles,
                disc_id=disc_id,
                year=read_info["DYEAR"]
               )

    return info

def rip_cd(i, info):
    track = info["tracks"][i]
    print ("Ripping: %02d %s ..." % (i+1, track))
    cmd = ("cdda2wav",
               "-D", CDROM,
               "-H",
               "-q",
               "-t", "%s" % (i + 1),
               track2filename(i, info, "wav"))
    p = subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
    p.wait()
    print ("Ripping: %02d %s (Done)" % (i+1, track))

def wav2mp3(i, info, remove=True):
    track = info["tracks"][i]
    print (" Converting: %02d %s ..." % (i+1, track))
    cmd = MP3_COMMAND + [track2filename(i, info, "wav")]
    p = subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
    p.wait()

    mp3_filename = track2filename(i, info, "mp3")
#    os.chown(mp3_filename, 0, 512)
#    os.chmod(mp3_filename, 0664)
    if remove:
        _remove_wav(i, info)
    print (" Converting: %02d %s (Done)" % (i+1, track))

def _remove_wav(i, info):
    os.remove(track2filename(i, info, "wav"))

def set_id3(info, dir):
    tracks = info["tracks"]
    tracks_len = len(tracks)
    for i in range(tracks_len):
        track = tracks[i]
        filepath = track2filename(i, info, "mp3")
        tag = eyeD3.Tag()
        tag.link(filepath)
        tag.header.setVersion(eyeD3.ID3_V2_3);
        tag.setTextEncoding(eyeD3.UTF_16_ENCODING);

        tag.setAlbum(info["album"])
        tag.setArtist(info["artist"])
        tag.setTitle(track)
        tag.setGenre(info["category"])
        tag.setDate(info["year"])
        tag.setTrackNum((i+1, tracks_len))
        tag.update()

def close_cdrom():
    subprocess.call(("eject", "-t"))

def open_cdrom():
    subprocess.call(("eject",))

def main():
    sys.stdout = codecs.getwriter('utf_8')(sys.stdout)

    close_cdrom()

    info = get_cddb()
    if info == None:
        print "[ERROR] Could not get CDDB."
        open_cdrom()
        sys.exit(2)

    print "%s / %s / %s / %s" % (info["artist"], info["album"],
                                 info["category"], info["year"])

    dir = DIR_NAME % info

    dir = os.path.join(OUT_DIR, escape_filename(dir)).encode("utf-8")
    try:
        os.mkdir(dir)
    except OSError, e:
        if e.errno == 17: # File exists
            print "[ERROR] Directory already exists."
            open_cdrom()
            sys.exit(1)
    except:
        raise

#    os.chown(dir, 0, 512)
#    os.chmod(dir, 04775)
    os.chdir(dir)

    open("disc_id", "w").write("%s\n" % info["disc_id"])

    tracks = info["tracks"]

    for i in range(len(tracks)):
        print ("%02d.%s" % (i+1, tracks[i]))

    threads = []
    for i in range(len(tracks)):
        rip_cd(i, info)

        # convert wave to mp3 in background
        th = Thread(target=wav2mp3, args=(i, info))
        threads.append(th)
        th.start()

    open_cdrom()

    for th in threads:
        th.join()

    set_id3(info, dir)
    print "Done."

if __name__ == '__main__':
    main()