media/metadata.py

import media.CDDB
import time
import subprocess
from media import common
import math
from PIL import Image
import os
import logging
import sys


try:
    from config import APP_NAME as ROOT_LOGGER_NAME
except ImportError:
    ROOT_LOGGER_NAME = 'root'
logger = logging.getLogger(ROOT_LOGGER_NAME).getChild(__name__)

# make module usable without discid dependency
try:
    import discid
except ModuleNotFoundError:
    logger.warning("Python module discid not installed")
except OSError:
    logger.exception("You might install python3-libdiscid")

__KEY_CAMERA_VENDOR__ = 'camera_vendor'
__KEY_CAMERA_MODEL__ = 'camera_model'


def get_media_data(full_path, user_callback=None):
    #
    ft = common.get_filetype(full_path)
    #
    if ft == common.FILETYPE_AUDIO:
        return get_audio_data(full_path)
    elif ft == common.FILETYPE_IMAGE:
        return get_image_data(full_path)
    elif ft == common.FILETYPE_VIDEO:
        return get_video_data(full_path)
    elif ft == common.FILETYPE_DISC:
        return get_disc_data(full_path, user_callback)
    else:
        logger.warning('Filetype not known: %s', full_path)


def get_audio_data(full_path):
    conv_key_dict = {}
    conv_key_dict['album'] = (str, common.KEY_ALBUM)
    conv_key_dict['TAG:album'] = (str, common.KEY_ALBUM)
    conv_key_dict['TAG:artist'] = (str, common.KEY_ARTIST)
    conv_key_dict['artist'] = (str, common.KEY_ARTIST)
    conv_key_dict['bit_rate'] = (__int_conv__, common.KEY_BITRATE)
    conv_key_dict['duration'] = (float, common.KEY_DURATION)
    conv_key_dict['TAG:genre'] = (str, common.KEY_GENRE)
    conv_key_dict['genre'] = (str, common.KEY_GENRE)
    conv_key_dict['TAG:title'] = (str, common.KEY_TITLE)
    conv_key_dict['title'] = (str, common.KEY_TITLE)
    conv_key_dict['TAG:track'] = (__int_conv__, common.KEY_TRACK)
    conv_key_dict['track'] = (__int_conv__, common.KEY_TRACK)
    conv_key_dict['TAG:date'] = (__int_conv__, common.KEY_YEAR)
    conv_key_dict['date'] = (__int_conv__, common.KEY_YEAR)
    return __adapt__data__(__get_xxprobe_data__(full_path, conv_key_dict), full_path)


def get_disc_data(full_path, user_callback):
    #
    # Read Information from CDDB database
    #
    did = media.CDDB.discid()
    if did is None:
        logger.error("Could not determine disc id...")
        sys.exit(1)
    q = media.CDDB.query(did)
    if q is None:
        data = {
            common.KEY_ARTIST: None,
            common.KEY_ALBUM: None,
            common.KEY_YEAR: None,
            common.KEY_GENRE: None
        }
        for i in range(0, int(did.split('+')[1])):
            data["track_%02d" % (i + 1)] = None
        data = user_callback(common.CALLBACK_MAN_INPUT, data)
        return media.CDDB.my_disc_metadata(**data)

    if len(q) == 1:
        # Only one entry
        did = tuple(q.keys())[0]
    else:
        # multiple entries (choose)
        if user_callback is None:
            logger.warning("No usercallback to handle multiple cddb choices...")
            sys.exit(1)
        did = user_callback(common.CALLBACK_CDDB_CHOICE, q)

    return media.CDDB.cddb(did)
    """
    musicbrainzngs.set_useragent("pyrip", "0.1", "your@mail")
    disc = discid.read()
    disc_id = disc.id
    disc_data = {}
    try:
        result = musicbrainzngs.get_releases_by_discid(disc_id, includes=["artists", "recordings"])
    except musicbrainzngs.ResponseError:
        logger.exception("disc not found or bad response")
        sys.exit(1)
    else:
        disc_data[common.KEY_ARTIST] = result["disc"]["release-list"][0]["artist-credit-phrase"]
        disc_data[common.KEY_ALBUM] = result["disc"]["release-list"][0]["title"]
        disc_data[common.KEY_YEAR] = int(result["disc"]["release-list"][0]["date"][:4])
        data_copy = dict(disc_data)
        disc_data["id"] = result["disc"]["release-list"][0]["id"]
        disc_data["tracks"] = []
        # get tracklist
        for entry in result["disc"]["release-list"][0]["medium-list"][0]["track-list"]:
            track = dict(data_copy)
            track[common.KEY_TITLE] = entry["recording"]["title"]
            track[common.KEY_TRACK] = int(entry['number'])
            disc_data["tracks"].append(track)
    return disc_data
    """


def get_video_data(full_path):
    conv_key_dict = {}
    conv_key_dict['creation_time'] = (__vid_datetime_conv__, common.KEY_TIME)
    conv_key_dict['TAG:creation_time'] = (__vid_datetime_conv__, common.KEY_TIME)
    conv_key_dict['bit_rate'] = (__int_conv__, common.KEY_BITRATE)
    conv_key_dict['duration'] = (float, common.KEY_DURATION)
    conv_key_dict['height'] = (__int_conv__, common.KEY_HEIGHT)
    conv_key_dict['width'] = (__int_conv__, common.KEY_WIDTH)
    conv_key_dict['display_aspect_ratio'] = (__ratio_conv__, common.KEY_RATIO)
    return __adapt__data__(__get_xxprobe_data__(full_path, conv_key_dict), full_path)


def get_image_data(full_path):
    return __adapt__data__(__get_exif_data__(full_path), full_path)


def __adapt__data__(data, full_path):
    data[common.KEY_SIZE] = os.path.getsize(full_path)
    # Join Camera Vendor and Camera Model
    if __KEY_CAMERA_MODEL__ in data and __KEY_CAMERA_VENDOR__ in data:
        model = data.pop(__KEY_CAMERA_MODEL__)
        vendor = data.pop(__KEY_CAMERA_VENDOR__)
        data[common.KEY_CAMERA] = '%s: %s' % (vendor, model)
    # Add time if not exists
    if common.KEY_TIME not in data:
        if common.KEY_YEAR in data and common.KEY_TRACK in data:
            if data[common.KEY_YEAR] != 0:  # ignore year 0 - must be wrong
                # Use a date where track 1 is the newest in the given year
                minute = int(data[common.KEY_TRACK] / 60)
                second = (data[common.KEY_TRACK] - 60 * minute) % 60
                #
                data[common.KEY_TIME] = int(time.mktime((data[common.KEY_YEAR], 1, 1, 0, 59 - minute, 59 - second, 0, 0, 0)))
                data[common.KEY_TIME_IS_SUBSTITUTION] = True
        else:
            data[common.KEY_TIME] = int(os.path.getmtime(full_path))
            data[common.KEY_TIME_IS_SUBSTITUTION] = True
    return data


def __get_xxprobe_data__(full_path, conv_key_dict):
    def _ffprobe_command(full_path):
        return ['ffprobe', '-v', 'quiet', '-show_format', '-show_streams', full_path]

    def _avprobe_command(full_path):
        return ['avprobe', '-v', 'quiet', '-show_format', '-show_streams', full_path]

    try:
        xxprobe_text = subprocess.check_output(_avprobe_command(full_path))
    except FileNotFoundError:
        try:
            xxprobe_text = subprocess.check_output(_ffprobe_command(full_path))
        except FileNotFoundError:
            logger.warning('ffprobe and avprobe seem to be not installed')
            return {}
    #
    rv = {}
    for line in xxprobe_text.decode('utf-8').splitlines():
        try:
            key, val = [snippet.strip() for snippet in line.split('=')]
        except ValueError:
            continue
        else:
            if key in conv_key_dict:
                tp, name = conv_key_dict[key]
                try:
                    rv[name] = tp(val)
                except ValueError:
                    logger.log(logging.WARNING if val else logger.INFO, 'Can\'t convert %s (%s) for %s', repr(val), name, name)
    return rv


def __get_exif_data__(full_path):
    rv = {}
    im = Image.open(full_path)
    try:
        exif = dict(im._getexif().items())
    except AttributeError:
        logger.debug('%s does not have any exif information', full_path)
    else:
        conv_key_dict = {}
        # IMAGE
        conv_key_dict[0x9003] = (__datetime_conv__, common.KEY_TIME)
        conv_key_dict[0x8822] = (__exposure_program_conv__, common.KEY_EXPOSURE_PROGRAM)
        conv_key_dict[0x829A] = (__num_denum_conv__, common.KEY_EXPOSURE_TIME)
        conv_key_dict[0x9209] = (__flash_conv__, common.KEY_FLASH)
        conv_key_dict[0x829D] = (__num_denum_conv__, common.KEY_APERTURE)
        conv_key_dict[0x920A] = (__num_denum_conv__, common.KEY_FOCAL_LENGTH)
        conv_key_dict[0x8825] = (__gps_conv__, common.KEY_GPS)
        conv_key_dict[0xA003] = (__int_conv__, common.KEY_HEIGHT)
        conv_key_dict[0x8827] = (__int_conv__, common.KEY_ISO)
        conv_key_dict[0x010F] = (str, __KEY_CAMERA_VENDOR__)
        conv_key_dict[0x0110] = (str, __KEY_CAMERA_MODEL__)
        conv_key_dict[0x0112] = (__int_conv__, common.KEY_ORIENTATION)
        conv_key_dict[0xA002] = (__int_conv__, common.KEY_WIDTH)
        for key in conv_key_dict:
            if key in exif:
                tp, name = conv_key_dict[key]
                raw_value = exif[key]
                logger.debug("Converting %s out of %s", name, repr(raw_value))
                value = tp(raw_value)
                if value is not None:
                    rv[name] = value
    return rv


# TODO: Join datetime converter __datetime_conv__ and __vid_datetime_conv_
def __datetime_conv__(dt):
    format_string = "%Y:%m:%d %H:%M:%S"
    return int(time.mktime(time.strptime(dt, format_string)))


def __vid_datetime_conv__(dt):
    try:
        dt = dt[:dt.index('.')]
    except ValueError:
        pass  # time string seems to have no '.'
    dt = dt.replace('T', ' ').replace('/', '').replace('\\', '')
    if len(dt) == 16:
        dt += ':00'
    format_string = '%Y-%m-%d %H:%M:%S'
    return int(time.mktime(time.strptime(dt, format_string)))


def __exposure_program_conv__(n):
    return {
        0: 'Unidentified',
        1: 'Manual',
        2: 'Program Normal',
        3: 'Aperture Priority',
        4: 'Shutter Priority',
        5: 'Program Creative',
        6: 'Program Action',
        7: 'Portrait Mode',
        8: 'Landscape Mode'
    }.get(n, None)


def __flash_conv__(n):
    return {
        0: 'No',
        1: 'Fired',
        5: 'Fired (?)',  # no return sensed
        7: 'Fired (!)',  # return sensed
        9: 'Fill Fired',
        13: 'Fill Fired (?)',
        15: 'Fill Fired (!)',
        16: 'Off',
        24: 'Auto Off',
        25: 'Auto Fired',
        29: 'Auto Fired (?)',
        31: 'Auto Fired (!)',
        32: 'Not Available'
    }.get(n, None)


def __int_conv__(value):
    try:
        return int(value)
    except ValueError:
        for c in ['.', '/', '-']:
            p = value.find(c)
            if p >= 0:
                value = value[:p]
    try:
        return int(value)
    except ValueError:
        return None


def __num_denum_conv__(data):
    try:
        return float(data)
    except TypeError:
        num, denum = data
        return num / denum


def __gps_conv__(data):
    def lat_lon_cal(lon_or_lat):
        lon_lat = 0.
        fac = 1.
        for data in lon_or_lat:
            try:
                lon_lat += float(data[0]) / float(data[1]) * fac
            except TypeError:
                lon_lat += data * fac
            except ZeroDivisionError:
                return 0.
            fac *= 1. / 60.
        if math.isnan(lon_lat):
            return 0.
        return lon_lat
    try:
        lon = lat_lon_cal(data[0x0004])
        lat = lat_lon_cal(data[0x0002])
        if lon != 0 or lat != 0:    # do not use lon and lat equal 0, caused by motorola gps weakness
            return {'lon': lon, 'lat': lat}
    except KeyError:
        logger.warning('GPS data extraction failed for %s', repr(data))


def __ratio_conv__(ratio):
    ratio = ratio.replace('\\', '')
    num, denum = ratio.split(':')
    return float(num) / float(denum)