def setFilename(self, filename, append=True):
"""
Use a file to store all messages. The
UTF-8 encoding will be used. Write an informative
message if the file can't be created.
@param filename: C{L{string}}
"""
# Look if file already exists or not
filename = os.path.expanduser(filename)
filename = os.path.realpath(filename)
append = os.access(filename, os.F_OK)
# Create log file (or open it in append mode, if it already exists)
try:
import codecs
if append:
self.__file = codecs.open(filename, "a", "utf-8")
else:
self.__file = codecs.open(filename, "w", "utf-8")
self._writeIntoFile(_("Starting Hachoir"))
except IOError, err:
if err.errno == 2:
self.__file = None
self.info(
_("[Log] setFilename(%s) fails: no such file") % filename)
else:
raise
def useHeader(self, header):
self.width = header["width"].value
self.height = header["height"].value
# Read number of colors and pixel format
if "/palette/size" in header:
nb_colors = header["/palette/size"].value // 3
else:
nb_colors = None
if not header["has_palette"].value:
if header["has_alpha"].value:
self.pixel_format = _("RGBA")
else:
self.pixel_format = _("RGB")
elif "/transparency" in header:
self.pixel_format = _("Color index with transparency")
if nb_colors:
nb_colors -= 1
else:
self.pixel_format = _("Color index")
self.bits_per_pixel = pngBitsPerPixel(header)
if nb_colors:
self.nb_colors = nb_colors
# Read compression, timestamp, etc.
self.compression = header["compression"].display
def getHachoirOptions(parser):
"""
Create an option group (type optparse.OptionGroup) of Hachoir
library options.
"""
def setLogFilename(*args):
log.setFilename(args[2])
common = OptionGroup(parser, _("Hachoir library"), \
"Configure Hachoir library")
common.add_option("--verbose",
help=_("Verbose mode"),
default=False,
action="store_true")
common.add_option("--log",
help=_("Write log in a file"),
type="string",
action="callback",
callback=setLogFilename)
common.add_option("--quiet",
help=_("Quiet mode (don't display warning)"),
default=False,
action="store_true")
common.add_option("--debug",
help=_("Debug mode"),
default=False,
action="store_true")
return common
def useHeader(self, header):
self.width = header["width"].value
self.height = header["height"].value
# Read number of colors and pixel format
if "/palette/size" in header:
nb_colors = header["/palette/size"].value // 3
else:
nb_colors = None
if not header["has_palette"].value:
if header["has_alpha"].value:
self.pixel_format = _("RGBA")
else:
self.pixel_format = _("RGB")
elif "/transparency" in header:
self.pixel_format = _("Color index with transparency")
if nb_colors:
nb_colors -= 1
else:
self.pixel_format = _("Color index")
self.bits_per_pixel = pngBitsPerPixel(header)
if nb_colors:
self.nb_colors = nb_colors
# Read compression, timestamp, etc.
self.compression = header["compression"].display
def setFilename(self, filename, append=True):
"""
Use a file to store all messages. The
UTF-8 encoding will be used. Write an informative
message if the file can't be created.
@param filename: C{L{string}}
"""
# Look if file already exists or not
filename = os.path.expanduser(filename)
filename = os.path.realpath(filename)
append = os.access(filename, os.F_OK)
# Create log file (or open it in append mode, if it already exists)
try:
import codecs
if append:
self.__file = codecs.open(filename, "a", "utf-8")
else:
self.__file = codecs.open(filename, "w", "utf-8")
self._writeIntoFile(_("Starting Hachoir"))
except IOError, err:
if err.errno == 2:
self.__file = None
self.info(_("[Log] setFilename(%s) fails: no such file") % filename)
else:
raise
def __init__(self, size, address, got=None):
self.size = size
self.address = address
self.got = got
if self.got is not None:
msg = _("Can't read %u bits at address %u (got %u bits)") % (self.size, self.address, self.got)
else:
msg = _("Can't read %u bits at address %u") % (self.size, self.address)
InputStreamError.__init__(self, msg)
def __init__(self, size, address, got=None):
self.size = size
self.address = address
self.got = got
if self.got is not None:
msg = _("Can't read %u bits at address %u (got %u bits)") % (
self.size, self.address, self.got)
else:
msg = _("Can't read %u bits at address %u") % (self.size,
self.address)
InputStreamError.__init__(self, msg)
def extract(self, gif):
self.useScreen(gif["/screen"])
if self.has("bits_per_pixel"):
self.nb_colors = (1 << self.get('bits_per_pixel'))
self.compression = _("LZW")
self.format_version = "GIF version %s" % gif["version"].value
for comments in gif.array("comments"):
for comment in gif.array(comments.name + "/comment"):
self.comment = comment.value
if "graphic_ctl/has_transp" in gif and gif["graphic_ctl/has_transp"].value:
self.pixel_format = _("Color index with transparency")
else:
self.pixel_format = _("Color index")
def extract(self, tar):
max_nb = maxNbFile(self)
for index, field in enumerate(tar.array("file")):
if max_nb is not None and max_nb <= index:
self.warning("TAR archive contains many files, but only first %s files are processed" % max_nb)
break
meta = Metadata(self)
self.extractFile(field, meta)
if meta.has("filename"):
title = _('File "%s"') % meta.getText('filename')
else:
title = _("File")
self.addGroup(field.name, meta, title)
def extract(self, gif):
self.useScreen(gif["/screen"])
if self.has("bits_per_pixel"):
self.nb_colors = 1 << self.get("bits_per_pixel")
self.compression = _("LZW")
self.format_version = "GIF version %s" % gif["version"].value
for comments in gif.array("comments"):
for comment in gif.array(comments.name + "/comment"):
self.comment = comment.value
if "graphic_ctl/has_transp" in gif and gif["graphic_ctl/has_transp"].value:
self.pixel_format = _("Color index with transparency")
else:
self.pixel_format = _("Color index")
def extract(self, tar):
max_nb = maxNbFile(self)
for index, field in enumerate(tar.array("file")):
if max_nb is not None and max_nb <= index:
self.warning("TAR archive contains many files, but only first %s files are processed" % max_nb)
break
meta = Metadata(self)
self.extractFile(field, meta)
if meta.has("filename"):
title = _('File "%s"') % meta.getText('filename')
else:
title = _("File")
self.addGroup(field.name, meta, title)
def useFile(self, field):
meta = Metadata(self)
meta.filename = field["filename"].value
meta.file_size = field["filesize"].value
meta.creation_date = field["timestamp"].value
attr = field["attributes"].value
if attr != "(none)":
meta.file_attr = attr
if meta.has("filename"):
title = _("File \"%s\"") % meta.getText('filename')
else:
title = _("File")
self.addGroup(field.name, meta, title)
def useFile(self, field):
meta = Metadata(self)
meta.filename = field["filename"].value
meta.file_size = field["filesize"].value
meta.creation_date = field["timestamp"].value
attr = field["attributes"].value
if attr != "(none)":
meta.file_attr = attr
if meta.has("filename"):
title = _("File \"%s\"") % meta.getText('filename')
else:
title = _("File")
self.addGroup(field.name, meta, title)
def __init__(self, stream, **args):
validate = args.pop("validate", False)
self._mime_type = None
while validate:
nbits = self.getParserTags()["min_size"]
if stream.sizeGe(nbits):
res = self.validate()
if res is True:
break
res = makeUnicode(res)
else:
res = _("stream is smaller than %s.%s bytes" % divmod(nbits, 8))
raise ValidateError(res or _("no reason given"))
self._autofix = True
def startOfFrame(self, sof):
# Set compression method
key = sof["../type"].value
self.compression = "JPEG (%s)" % JpegChunk.START_OF_FRAME[key]
# Read image size and bits/pixel
self.width = sof["width"].value
self.height = sof["height"].value
nb_components = sof["nr_components"].value
self.bits_per_pixel = 8 * nb_components
if nb_components == 3:
self.pixel_format = _("YCbCr")
elif nb_components == 1:
self.pixel_format = _("Grayscale")
self.nb_colors = 256
def __init__(self, stream, **args):
validate = args.pop("validate", False)
self._mime_type = None
while validate:
nbits = self.getParserTags()["min_size"]
if stream.sizeGe(nbits):
res = self.validate()
if res is True:
break
res = makeUnicode(res)
else:
res = _("stream is smaller than %s.%s bytes" %
divmod(nbits, 8))
raise ValidateError(res or _("no reason given"))
self._autofix = True
def extractAVI(self, headers):
audio_index = 1
for stream in headers.array("stream"):
if "stream_hdr/stream_type" not in stream:
continue
stream_type = stream["stream_hdr/stream_type"].value
if stream_type == "vids":
if "stream_hdr" in stream:
meta = Metadata(self)
self.extractAVIVideo(stream["stream_hdr"], meta)
self.addGroup("video", meta, "Video stream")
elif stream_type == "auds":
if "stream_fmt" in stream:
meta = Metadata(self)
self.extractAVIAudio(stream["stream_fmt"], meta)
self.addGroup("audio[%u]" % audio_index, meta,
"Audio stream")
audio_index += 1
if "avi_hdr" in headers:
self.useAviHeader(headers["avi_hdr"])
# Compute global bit rate
if self.has("duration") and "/movie/size" in headers:
self.bit_rate = float(
headers["/movie/size"].value) * 8 / timedelta2seconds(
self.get('duration'))
# Video has index?
if "/index" in headers:
self.comment = _("Has audio/video index (%s)") \
% humanFilesize(headers["/index"].size/8)
def computeBitrate(self, frame):
bit_rate = frame.getBitRate() # may returns None on error
if not bit_rate:
return
self.bit_rate = (bit_rate, _("%s (constant)") % humanBitRate(bit_rate))
self.duration = timedelta(seconds=float(frame["/frames"].size) /
bit_rate)
def extract(self, icon):
for index, header in enumerate(icon.array("icon_header")):
image = Metadata(self)
# Read size and colors from header
image.width = header["width"].value
image.height = header["height"].value
bpp = header["bpp"].value
nb_colors = header["nb_color"].value
if nb_colors != 0:
image.nb_colors = nb_colors
if bpp == 0 and nb_colors in self.color_to_bpp:
bpp = self.color_to_bpp[nb_colors]
elif bpp == 0:
bpp = 8
image.bits_per_pixel = bpp
image.setHeader(_("Icon #%u (%sx%s)")
% (1+index, image.get("width", "?"), image.get("height", "?")))
# Read compression from data (if available)
key = "icon_data[%u]/header/codec" % index
if key in icon:
image.compression = icon[key].display
key = "icon_data[%u]/pixels" % index
if key in icon:
computeComprRate(image, icon[key].size)
# Store new image
self.addGroup("image[%u]" % index, image)
def __delitem__(self, index):
"""
Delete item at position index. May raise IndexError.
>>> d=Dict( ((6, 'six'), (9, 'neuf'), (4, 'quatre')) )
>>> del d[1]
>>> d
{6: 'six', 4: 'quatre'}
"""
if index < 0:
index += len(self._value_list)
if not (0 <= index < len(self._value_list)):
raise IndexError(_("list assignment index out of range (%s/%s)")
% (index, len(self._value_list)))
del self._value_list[index]
del self._key_list[index]
# First loop which may alter self._index
for key, item_index in self._index.iteritems():
if item_index == index:
del self._index[key]
break
# Second loop update indexes
for key, item_index in self._index.iteritems():
if index < item_index:
self._index[key] -= 1
def extract(self, icon):
for index, header in enumerate(icon.array("icon_header")):
image = Metadata(self)
# Read size and colors from header
image.width = header["width"].value
image.height = header["height"].value
bpp = header["bpp"].value
nb_colors = header["nb_color"].value
if nb_colors != 0:
image.nb_colors = nb_colors
if bpp == 0 and nb_colors in self.color_to_bpp:
bpp = self.color_to_bpp[nb_colors]
elif bpp == 0:
bpp = 8
image.bits_per_pixel = bpp
image.setHeader(_("Icon #%u (%sx%s)") % (1 + index, image.get("width", "?"), image.get("height", "?")))
# Read compression from data (if available)
key = "icon_data[%u]/header/codec" % index
if key in icon:
image.compression = icon[key].display
key = "icon_data[%u]/pixels" % index
if key in icon:
computeComprRate(image, icon[key].size)
# Store new image
self.addGroup("image[%u]" % index, image)
def extractAVI(self, headers):
audio_index = 1
for stream in headers.array("stream"):
if "stream_hdr/stream_type" not in stream:
continue
stream_type = stream["stream_hdr/stream_type"].value
if stream_type == "vids":
if "stream_hdr" in stream:
meta = Metadata(self)
self.extractAVIVideo(stream["stream_hdr"], meta)
self.addGroup("video", meta, "Video stream")
elif stream_type == "auds":
if "stream_fmt" in stream:
meta = Metadata(self)
self.extractAVIAudio(stream["stream_fmt"], meta)
self.addGroup("audio[%u]" % audio_index, meta, "Audio stream")
audio_index += 1
if "avi_hdr" in headers:
self.useAviHeader(headers["avi_hdr"])
# Compute global bit rate
if self.has("duration") and "/movie/size" in headers:
self.bit_rate = float(headers["/movie/size"].value) * 8 / timedelta2seconds(self.get("duration"))
# Video has index?
if "/index" in headers:
self.comment = _("Has audio/video index (%s)") % humanFilesize(headers["/index"].size / 8)
def computeVariableBitrate(self, mp3):
if self.quality <= QUALITY_FAST:
return
count = 0
if QUALITY_BEST <= self.quality:
self.warning(
"Process all MPEG audio frames to compute exact duration")
max_count = None
else:
max_count = 500 * self.quality
total_bit_rate = 0.0
for index, frame in enumerate(mp3.array("frames/frame")):
if index < 3:
continue
bit_rate = frame.getBitRate()
if bit_rate:
total_bit_rate += float(bit_rate)
count += 1
if max_count and max_count <= count:
break
if not count:
return
bit_rate = total_bit_rate / count
self.bit_rate = (bit_rate,
_("%s (Variable bit rate)") % humanBitRate(bit_rate))
duration = timedelta(seconds=float(mp3["frames"].size) / bit_rate)
self.duration = duration
def displayStat(self, stat):
"""
Display statistics to stdout:
- best time (minimum)
- average time (arithmetic average)
- worst time (maximum)
- total time (sum)
Use arithmetic avertage instead of geometric average because
geometric fails if any value is zero (returns zero) and also
because floating point multiplication lose precision with many
values.
"""
average = stat.getSum() / len(stat)
values = (stat.getMin(), average, stat.getMax(), stat.getSum())
values = tuple(self.formatTime(value) for value in values)
print _("Benchmark: best=%s average=%s worst=%s total=%s") \
% values
def displayStat(self, stat):
"""
Display statistics to stdout:
- best time (minimum)
- average time (arithmetic average)
- worst time (maximum)
- total time (sum)
Use arithmetic avertage instead of geometric average because
geometric fails if any value is zero (returns zero) and also
because floating point multiplication lose precision with many
values.
"""
average = stat.getSum() / len(stat)
values = (stat.getMin(), average, stat.getMax(), stat.getSum())
values = tuple(self.formatTime(value) for value in values)
print _("Benchmark: best=%s average=%s worst=%s total=%s") \
% values
def append(self, key, value):
"""
Append new value
"""
if key in self._index:
raise UniqKeyError(_("Key '%s' already exists") % key)
self._index[key] = len(self._value_list)
self._key_list.append(key)
self._value_list.append(value)
def getHachoirOptions(parser):
"""
Create an option group (type optparse.OptionGroup) of Hachoir
library options.
"""
def setLogFilename(*args):
log.setFilename(args[2])
common = OptionGroup(parser, _("Hachoir library"), \
"Configure Hachoir library")
common.add_option("--verbose", help=_("Verbose mode"),
default=False, action="store_true")
common.add_option("--log", help=_("Write log in a file"),
type="string", action="callback", callback=setLogFilename)
common.add_option("--quiet", help=_("Quiet mode (don't display warning)"),
default=False, action="store_true")
common.add_option("--debug", help=_("Debug mode"),
default=False, action="store_true")
return common
def __setattr__(self, key, value):
"""
Add a new value to data with name 'key'. Skip duplicates.
"""
# Invalid key?
if key not in self.__data:
raise KeyError(_("%s has no metadata '%s'") % (self.__class__.__name__, key))
# Skip duplicates
self.__data[key].add(value)
def __setattr__(self, key, value):
"""
Add a new value to data with name 'key'. Skip duplicates.
"""
# Invalid key?
if key not in self.__data:
raise KeyError(
_("%s has no metadata '%s'") % (self.__class__.__name__, key))
# Skip duplicates
self.__data[key].add(value)
def extract(self, pcx):
self.width = 1 + pcx["xmax"].value
self.height = 1 + pcx["ymax"].value
self.width_dpi = pcx["horiz_dpi"].value
self.height_dpi = pcx["vert_dpi"].value
self.bits_per_pixel = pcx["bpp"].value
if 1 <= pcx["bpp"].value <= 8:
self.nb_colors = 2 ** pcx["bpp"].value
self.compression = _("Run-length encoding (RLE)")
self.format_version = "PCX: %s" % pcx["version"].display
if "image_data" in pcx:
computeComprRate(self, pcx["image_data"].size)
def extract(self, pcx):
self.width = 1 + pcx["xmax"].value
self.height = 1 + pcx["ymax"].value
self.width_dpi = pcx["horiz_dpi"].value
self.height_dpi = pcx["vert_dpi"].value
self.bits_per_pixel = pcx["bpp"].value
if 1 <= pcx["bpp"].value <= 8:
self.nb_colors = 2 ** pcx["bpp"].value
self.compression = _("Run-length encoding (RLE)")
self.format_version = "PCX: %s" % pcx["version"].display
if "image_data" in pcx:
computeComprRate(self, pcx["image_data"].size)
def insert(self, index, key, value):
"""
Insert an item at specified position index.
>>> d=Dict( ((6, 'six'), (9, 'neuf'), (4, 'quatre')) )
>>> d.insert(1, '40', 'quarante')
>>> d
{6: 'six', '40': 'quarante', 9: 'neuf', 4: 'quatre'}
"""
if key in self:
raise UniqKeyError(_("Insert error: key '%s' ready exists") % key)
_index = index
if index < 0:
index += len(self._value_list)
if not(0 <= index <= len(self._value_list)):
raise IndexError(_("Insert error: index '%s' is invalid") % _index)
for item_key, item_index in self._index.iteritems():
if item_index >= index:
self._index[item_key] += 1
self._index[key] = index
self._key_list.insert(index, key)
self._value_list.insert(index, value)
def humanFilesize(size):
"""
Convert a file size in byte to human natural representation.
It uses the values: 1 KB is 1024 bytes, 1 MB is 1024 KB, etc.
The result is an unicode string.
>>> humanFilesize(1)
u'1 byte'
>>> humanFilesize(790)
u'790 bytes'
>>> humanFilesize(256960)
u'250.9 KB'
"""
if size < 10000:
return ngettext("%u byte", "%u bytes", size) % size
units = [_("KB"), _("MB"), _("GB"), _("TB")]
size = float(size)
divisor = 1024
for unit in units:
size = size / divisor
if size < divisor:
return "%.1f %s" % (size, unit)
return "%u %s" % (size, unit)
def humanFilesize(size):
"""
Convert a file size in byte to human natural representation.
It uses the values: 1 KB is 1024 bytes, 1 MB is 1024 KB, etc.
The result is an unicode string.
>>> humanFilesize(1)
u'1 byte'
>>> humanFilesize(790)
u'790 bytes'
>>> humanFilesize(256960)
u'250.9 KB'
"""
if size < 10000:
return ngettext("%u byte", "%u bytes", size) % size
units = [_("KB"), _("MB"), _("GB"), _("TB")]
size = float(size)
divisor = 1024
for unit in units:
size = size / divisor
if size < divisor:
return "%.1f %s" % (size, unit)
return "%u %s" % (size, unit)
def timestampMac32(value):
"""
Convert an Mac (32-bit) timestamp to string. The format is the number
of seconds since the 1st January 1904 (to 2040). Returns unicode string.
>>> timestampMac32(0)
datetime.datetime(1904, 1, 1, 0, 0)
>>> timestampMac32(2843043290)
datetime.datetime(1994, 2, 2, 14, 14, 50)
"""
if not isinstance(value, (float, int, long)):
raise TypeError("an integer or float is required")
if not(0 <= value <= 4294967295):
return _("invalid Mac timestamp (%s)") % value
return MAC_TIMESTAMP_T0 + timedelta(seconds=value)
def timestampMac32(value):
"""
Convert an Mac (32-bit) timestamp to string. The format is the number
of seconds since the 1st January 1904 (to 2040). Returns unicode string.
>>> timestampMac32(0)
datetime.datetime(1904, 1, 1, 0, 0)
>>> timestampMac32(2843043290)
datetime.datetime(1994, 2, 2, 14, 14, 50)
"""
if not isinstance(value, (float, int, long)):
raise TypeError("an integer or float is required")
if not (0 <= value <= 4294967295):
return _("invalid Mac timestamp (%s)") % value
return MAC_TIMESTAMP_T0 + timedelta(seconds=value)
def timestampWin64(value):
"""
Convert Windows 64-bit timestamp to string. The timestamp format is
a 64-bit number which represents number of 100ns since the
1st January 1601 at 00:00. Result is an unicode string.
See also durationWin64(). Maximum date is 28 may 60056.
>>> timestampWin64(0)
datetime.datetime(1601, 1, 1, 0, 0)
>>> timestampWin64(127840491566710000)
datetime.datetime(2006, 2, 10, 12, 45, 56, 671000)
"""
try:
return WIN64_TIMESTAMP_T0 + durationWin64(value)
except OverflowError:
raise ValueError(_("date newer than year %s (value=%s)") % (MAXYEAR, value))
def timestampWin64(value):
"""
Convert Windows 64-bit timestamp to string. The timestamp format is
a 64-bit number which represents number of 100ns since the
1st January 1601 at 00:00. Result is an unicode string.
See also durationWin64(). Maximum date is 28 may 60056.
>>> timestampWin64(0)
datetime.datetime(1601, 1, 1, 0, 0)
>>> timestampWin64(127840491566710000)
datetime.datetime(2006, 2, 10, 12, 45, 56, 671000)
"""
try:
return WIN64_TIMESTAMP_T0 + durationWin64(value)
except OverflowError:
raise ValueError(
_("date newer than year %s (value=%s)") % (MAXYEAR, value))
def FileInputStream(filename, real_filename=None, **args):
"""
Create an input stream of a file. filename must be unicode.
real_filename is an optional argument used to specify the real filename,
its type can be 'str' or 'unicode'. Use real_filename when you are
not able to convert filename to real unicode string (ie. you have to
use unicode(name, 'replace') or unicode(name, 'ignore')).
"""
assert isinstance(filename, unicode)
if not real_filename:
real_filename = filename
try:
inputio = open(real_filename, "rb")
except IOError, err:
charset = getTerminalCharset()
errmsg = unicode(str(err), charset)
raise InputStreamError(_("Unable to open file %s: %s") % (filename, errmsg))
def _get(self, index):
if index >= len(self.buffers):
return ''
buf = self.buffers[index]
if buf is None:
raise InputStreamError(_("Error: Buffers too small. Can't seek backward."))
if self.last != index:
next = buf[1]
prev = buf[2]
self.buffers[next][2] = prev
self.buffers[prev][1] = next
first = self.buffers[self.last][1]
buf[1] = first
buf[2] = self.last
self.buffers[first][2] = index
self.buffers[self.last][1] = index
self.last = index
return buf[0]
def FileInputStream(filename, real_filename=None, **args):
"""
Create an input stream of a file. filename must be unicode.
real_filename is an optional argument used to specify the real filename,
its type can be 'str' or 'unicode'. Use real_filename when you are
not able to convert filename to real unicode string (ie. you have to
use unicode(name, 'replace') or unicode(name, 'ignore')).
"""
assert isinstance(filename, unicode)
if not real_filename:
real_filename = filename
try:
inputio = open(real_filename, 'rb')
except IOError, err:
charset = getTerminalCharset()
errmsg = unicode(str(err), charset)
raise InputStreamError(_("Unable to open file %s: %s") % (filename, errmsg))
def __init__(self, input, size=None, **args):
if not hasattr(input, "seek"):
if size is None:
input = InputPipe(input, self._setSize)
else:
input = InputPipe(input)
elif size is None:
try:
input.seek(0, 2)
size = input.tell() * 8
except IOError, err:
if err.errno == ESPIPE:
input = InputPipe(input, self._setSize)
else:
charset = getTerminalCharset()
errmsg = unicode(str(err), charset)
source = args.get("source", "<inputio:%r>" % input)
raise InputStreamError(_("Unable to get size of %s: %s") % (source, errmsg))
def computeQuality(self, jpeg):
# This function is an adaption to Python of ImageMagick code
# to compute JPEG quality using quantization tables
# Read quantization tables
qtlist = []
for dqt in jpeg.array("quantization"):
for qt in dqt.array("content/qt"):
# TODO: Take care of qt["index"].value?
qtlist.append(qt)
if not qtlist:
return
# Compute sum of all coefficients
sumcoeff = 0
for qt in qtlist:
coeff = qt.array("coeff")
for index in xrange(64):
sumcoeff += coeff[index].value
# Choose the right quality table and compute hash value
try:
hashval = qtlist[0]["coeff[2]"].value + qtlist[0]["coeff[53]"].value
if 2 <= len(qtlist):
hashval += qtlist[1]["coeff[0]"].value + qtlist[1][
"coeff[63]"].value
hashtable = QUALITY_HASH_COLOR
sumtable = QUALITY_SUM_COLOR
else:
hashtable = QUALITY_HASH_GRAY
sumtable = QUALITY_SUM_GRAY
except (MissingField, IndexError):
# A coefficient is missing, so don't compute JPEG quality
return
# Find the JPEG quality
for index in xrange(100):
if (hashval >= hashtable[index]) or (sumcoeff >= sumtable[index]):
quality = "%s%%" % (index + 1)
if (hashval > hashtable[index]) or (sumcoeff >
sumtable[index]):
quality += " " + _("(approximate)")
self.comment = "JPEG quality: %s" % quality
return
def createDisplay(self, human=True):
if not human:
if self._raw_value is None:
self._raw_value = GenericString.createValue(self, False)
value = makePrintable(self._raw_value, "ASCII", to_unicode=True)
elif self._charset:
value = makePrintable(self.value, "ISO-8859-1", to_unicode=True)
else:
value = self.value
if config.max_string_length < len(value):
# Truncate string if needed
value = "%s(...)" % value[:config.max_string_length]
if not self._charset or not human:
return makePrintable(value, "ASCII", quote='"', to_unicode=True)
else:
if value:
return '"%s"' % value.replace('"', '\\"')
else:
return _("(empty)")
def timestampUUID60(value):
"""
Convert UUID 60-bit timestamp to string. The timestamp format is
a 60-bit number which represents number of 100ns since the
the 15 October 1582 at 00:00. Result is an unicode string.
>>> timestampUUID60(0)
datetime.datetime(1582, 10, 15, 0, 0)
>>> timestampUUID60(130435676263032368)
datetime.datetime(1996, 2, 14, 5, 13, 46, 303236)
"""
if not isinstance(value, (float, int, long)):
raise TypeError("an integer or float is required")
if value < 0:
raise ValueError("value have to be a positive or nul integer")
try:
return UUID60_TIMESTAMP_T0 + timedelta(microseconds=value/10)
except OverflowError:
raise ValueError(_("timestampUUID60() overflow (value=%s)") % value)
def __init__(self, input, size=None, **args):
if not hasattr(input, "seek"):
if size is None:
input = InputPipe(input, self._setSize)
else:
input = InputPipe(input)
elif size is None:
try:
input.seek(0, 2)
size = input.tell() * 8
except IOError, err:
if err.errno == ESPIPE:
input = InputPipe(input, self._setSize)
else:
charset = getTerminalCharset()
errmsg = unicode(str(err), charset)
source = args.get("source", "<inputio:%r>" % input)
raise InputStreamError(
_("Unable to get size of %s: %s") % (source, errmsg))
def _get(self, index):
if index >= len(self.buffers):
return ''
buf = self.buffers[index]
if buf is None:
raise InputStreamError(
_("Error: Buffers too small. Can't seek backward."))
if self.last != index:
next = buf[1]
prev = buf[2]
self.buffers[next][2] = prev
self.buffers[prev][1] = next
first = self.buffers[self.last][1]
buf[1] = first
buf[2] = self.last
self.buffers[first][2] = index
self.buffers[self.last][1] = index
self.last = index
return buf[0]
def timestampUUID60(value):
"""
Convert UUID 60-bit timestamp to string. The timestamp format is
a 60-bit number which represents number of 100ns since the
the 15 October 1582 at 00:00. Result is an unicode string.
>>> timestampUUID60(0)
datetime.datetime(1582, 10, 15, 0, 0)
>>> timestampUUID60(130435676263032368)
datetime.datetime(1996, 2, 14, 5, 13, 46, 303236)
"""
if not isinstance(value, (float, int, long)):
raise TypeError("an integer or float is required")
if value < 0:
raise ValueError("value have to be a positive or nul integer")
try:
return UUID60_TIMESTAMP_T0 + timedelta(microseconds=value / 10)
except OverflowError:
raise ValueError(_("timestampUUID60() overflow (value=%s)") % value)
def createDisplay(self, human=True):
if not human:
if self._raw_value is None:
self._raw_value = GenericString.createValue(self, False)
value = makePrintable(self._raw_value, "ASCII", to_unicode=True)
elif self._charset:
value = makePrintable(self.value, "ISO-8859-1", to_unicode=True)
else:
value = self.value
if config.max_string_length < len(value):
# Truncate string if needed
value = "%s(...)" % value[:config.max_string_length]
if not self._charset or not human:
return makePrintable(value, "ASCII", quote='"', to_unicode=True)
else:
if value:
return '"%s"' % value.replace('"', '\\"')
else:
return _("(empty)")
def computeQuality(self, jpeg):
# This function is an adaption to Python of ImageMagick code
# to compute JPEG quality using quantization tables
# Read quantization tables
qtlist = []
for dqt in jpeg.array("quantization"):
for qt in dqt.array("content/qt"):
# TODO: Take care of qt["index"].value?
qtlist.append(qt)
if not qtlist:
return
# Compute sum of all coefficients
sumcoeff = 0
for qt in qtlist:
coeff = qt.array("coeff")
for index in xrange(64):
sumcoeff += coeff[index].value
# Choose the right quality table and compute hash value
try:
hashval= qtlist[0]["coeff[2]"].value + qtlist[0]["coeff[53]"].value
if 2 <= len(qtlist):
hashval += qtlist[1]["coeff[0]"].value + qtlist[1]["coeff[63]"].value
hashtable = QUALITY_HASH_COLOR
sumtable = QUALITY_SUM_COLOR
else:
hashtable = QUALITY_HASH_GRAY
sumtable = QUALITY_SUM_GRAY
except (MissingField, IndexError):
# A coefficient is missing, so don't compute JPEG quality
return
# Find the JPEG quality
for index in xrange(100):
if (hashval >= hashtable[index]) or (sumcoeff >= sumtable[index]):
quality = "%s%%" % (index + 1)
if (hashval > hashtable[index]) or (sumcoeff > sumtable[index]):
quality += " " + _("(approximate)")
self.comment = "JPEG quality: %s" % quality
return
def _getValue(self):
try:
value = self.createValue()
except HACHOIR_ERRORS, err:
self.error(_("Unable to create value: %s") % unicode(err))
value = None
class JpegMetadata(RootMetadata):
EXIF_KEY = {
# Exif metadatas
ExifEntry.TAG_CAMERA_MANUFACTURER: "camera_manufacturer",
ExifEntry.TAG_CAMERA_MODEL: "camera_model",
ExifEntry.TAG_ORIENTATION: "image_orientation",
ExifEntry.TAG_EXPOSURE: "camera_exposure",
ExifEntry.TAG_FOCAL: "camera_focal",
ExifEntry.TAG_BRIGHTNESS: "camera_brightness",
ExifEntry.TAG_APERTURE: "camera_aperture",
# Generic metadatas
ExifEntry.TAG_IMG_TITLE: "title",
ExifEntry.TAG_SOFTWARE: "producer",
ExifEntry.TAG_FILE_TIMESTAMP: "creation_date",
ExifEntry.TAG_WIDTH: "width",
ExifEntry.TAG_HEIGHT: "height",
ExifEntry.TAG_USER_COMMENT: "comment",
}
IPTC_KEY = {
80: "author",
90: "city",
101: "country",
116: "copyright",
120: "title",
231: "comment",
}
orientation_name = {
1: _('Horizontal (normal)'),
2: _('Mirrored horizontal'),
3: _('Rotated 180'),
4: _('Mirrored vertical'),
5: _('Mirrored horizontal then rotated 90 counter-clock-wise'),
6: _('Rotated 90 clock-wise'),
7: _('Mirrored horizontal then rotated 90 clock-wise'),
8: _('Rotated 90 counter clock-wise'),
}
def extract(self, jpeg):
if "start_frame/content" in jpeg:
self.startOfFrame(jpeg["start_frame/content"])
elif "start_scan/content/nr_components" in jpeg:
self.bits_per_pixel = 8 * jpeg["start_scan/content/nr_components"].value
if "app0/content" in jpeg:
self.extractAPP0(jpeg["app0/content"])
if "exif/content" in jpeg:
for ifd in jpeg.array("exif/content/ifd"):
for entry in ifd.array("entry"):
self.processIfdEntry(ifd, entry)
self.readGPS(ifd)
if "photoshop/content" in jpeg:
psd = jpeg["photoshop/content"]
if "version/content/reader_name" in psd:
self.producer = psd["version/content/reader_name"].value
if "iptc/content" in psd:
self.parseIPTC(psd["iptc/content"])
for field in jpeg.array("comment"):
if "content/comment" in field:
self.comment = field["content/comment"].value
self.computeQuality(jpeg)
if "data" in jpeg:
computeComprRate(self, jpeg["data"].size)
if not self.has("producer") and "photoshop" in jpeg:
self.producer = u"Adobe Photoshop"
if self.has("compression"):
self.compression = "JPEG"
@fault_tolerant
def startOfFrame(self, sof):
# Set compression method
key = sof["../type"].value
self.compression = "JPEG (%s)" % JpegChunk.START_OF_FRAME[key]
# Read image size and bits/pixel
self.width = sof["width"].value
self.height = sof["height"].value
nb_components = sof["nr_components"].value
self.bits_per_pixel = 8 * nb_components
if nb_components == 3:
self.pixel_format = _("YCbCr")
elif nb_components == 1:
self.pixel_format = _("Grayscale")
self.nb_colors = 256
@fault_tolerant
def computeQuality(self, jpeg):
# This function is an adaption to Python of ImageMagick code
# to compute JPEG quality using quantization tables
# Read quantization tables
qtlist = []
for dqt in jpeg.array("quantization"):
for qt in dqt.array("content/qt"):
# TODO: Take care of qt["index"].value?
qtlist.append(qt)
if not qtlist:
return
# Compute sum of all coefficients
sumcoeff = 0
for qt in qtlist:
coeff = qt.array("coeff")
for index in xrange(64):
sumcoeff += coeff[index].value
# Choose the right quality table and compute hash value
try:
hashval= qtlist[0]["coeff[2]"].value + qtlist[0]["coeff[53]"].value
if 2 <= len(qtlist):
hashval += qtlist[1]["coeff[0]"].value + qtlist[1]["coeff[63]"].value
hashtable = QUALITY_HASH_COLOR
sumtable = QUALITY_SUM_COLOR
else:
hashtable = QUALITY_HASH_GRAY
sumtable = QUALITY_SUM_GRAY
except (MissingField, IndexError):
# A coefficient is missing, so don't compute JPEG quality
return
# Find the JPEG quality
for index in xrange(100):
if (hashval >= hashtable[index]) or (sumcoeff >= sumtable[index]):
quality = "%s%%" % (index + 1)
if (hashval > hashtable[index]) or (sumcoeff > sumtable[index]):
quality += " " + _("(approximate)")
self.comment = "JPEG quality: %s" % quality
return
@fault_tolerant
def extractAPP0(self, app0):
self.format_version = u"JFIF %u.%02u" \
% (app0["ver_maj"].value, app0["ver_min"].value)
if "y_density" in app0:
self.width_dpi = app0["x_density"].value
self.height_dpi = app0["y_density"].value
@fault_tolerant
def processIfdEntry(self, ifd, entry):
# Skip unknown tags
tag = entry["tag"].value
if tag not in self.EXIF_KEY:
return
key = self.EXIF_KEY[tag]
if key in ("width", "height") and self.has(key):
# EXIF "valid size" are sometimes not updated when the image is scaled
# so we just ignore it
return
# Read value
if "value" in entry:
value = entry["value"].value
else:
value = ifd["value_%s" % entry.name].value
# Convert value to string
if tag == ExifEntry.TAG_ORIENTATION:
value = self.orientation_name.get(value, value)
elif tag == ExifEntry.TAG_EXPOSURE:
if not value:
return
if isinstance(value, float):
value = (value, u"1/%g" % (1/value))
elif entry["type"].value in (ExifEntry.TYPE_RATIONAL, ExifEntry.TYPE_SIGNED_RATIONAL):
value = (value, u"%.3g" % value)
# Store information
setattr(self, key, value)
@fault_tolerant
def readGPS(self, ifd):
# Read latitude and longitude
latitude_ref = None
longitude_ref = None
latitude = None
longitude = None
altitude_ref = 1
altitude = None
timestamp = None
datestamp = None
for entry in ifd.array("entry"):
tag = entry["tag"].value
if tag == ExifEntry.TAG_GPS_LATITUDE_REF:
if entry["value"].value == "N":
latitude_ref = 1
else:
latitude_ref = -1
elif tag == ExifEntry.TAG_GPS_LONGITUDE_REF:
if entry["value"].value == "E":
longitude_ref = 1
else:
longitude_ref = -1
elif tag == ExifEntry.TAG_GPS_ALTITUDE_REF:
if entry["value"].value == 1:
altitude_ref = -1
else:
altitude_ref = 1
elif tag == ExifEntry.TAG_GPS_LATITUDE:
latitude = [ifd["value_%s[%u]" % (entry.name, index)].value for index in xrange(3)]
elif tag == ExifEntry.TAG_GPS_LONGITUDE:
longitude = [ifd["value_%s[%u]" % (entry.name, index)].value for index in xrange(3)]
elif tag == ExifEntry.TAG_GPS_ALTITUDE:
altitude = ifd["value_%s" % entry.name].value
elif tag == ExifEntry.TAG_GPS_DATESTAMP:
datestamp = ifd["value_%s" % entry.name].value
elif tag == ExifEntry.TAG_GPS_TIMESTAMP:
items = [ifd["value_%s[%u]" % (entry.name, index)].value for index in xrange(3)]
items = map(int, items)
items = map(str, items)
timestamp = ":".join(items)
if latitude_ref and latitude:
value = deg2float(*latitude)
if latitude_ref < 0:
value = -value
self.latitude = value
if longitude and longitude_ref:
value = deg2float(*longitude)
if longitude_ref < 0:
value = -value
self.longitude = value
if altitude:
value = altitude
if altitude_ref < 0:
value = -value
self.altitude = value
if datestamp:
if timestamp:
datestamp += " " + timestamp
self.creation_date = datestamp
def parseIPTC(self, iptc):
datestr = hourstr = None
for field in iptc:
# Skip incomplete field
if "tag" not in field or "content" not in field:
continue
# Get value
value = field["content"].value
if isinstance(value, (str, unicode)):
value = value.replace("\r", " ")
value = value.replace("\n", " ")
# Skip unknown tag
tag = field["tag"].value
if tag == 55:
datestr = value
continue
if tag == 60:
hourstr = value
continue
if tag not in self.IPTC_KEY:
if tag != 0:
self.warning("Skip IPTC key %s: %s" % (
field["tag"].display, makeUnicode(value)))
continue
setattr(self, self.IPTC_KEY[tag], value)
if datestr and hourstr:
try:
year = int(datestr[0:4])
month = int(datestr[4:6])
day = int(datestr[6:8])
hour = int(hourstr[0:2])
min = int(hourstr[2:4])
sec = int(hourstr[4:6])
self.creation_date = datetime(year, month, day, hour, min, sec)
except ValueError:
pass
for key, value in self.parser_args.iteritems():
setattr(parser_obj, key, value)
return parser_obj
except ValidateError, err:
res = unicode(err)
if fallback and self.fallback:
fb = parser
except HACHOIR_ERRORS, err:
res = unicode(err)
if warn:
if parser == self.other:
warn = info
warn(_("Skip parser '%s': %s") % (parser.__name__, res))
fallback = False
if self.use_fallback and fb:
warning(_("Force use of parser '%s'") % fb.__name__)
return fb(stream)
def guessParser(stream):
return QueryParser(stream.tags).parse(stream)
def createParser(filename, real_filename=None, tags=None):
"""
Create a parser from a file or returns None on error.
Options:
- filename (unicode): Input file name ;
- real_filename (str|unicode): Real file name.
"""
from lib.hachoir_core.i18n import _, ngettext
NB_CHANNEL_NAME = {1: _("mono"), 2: _("stereo")}
def humanAudioChannel(value):
return NB_CHANNEL_NAME.get(value, unicode(value))
def humanFrameRate(value):
if isinstance(value, (int, long, float)):
return _("%.1f fps") % value
else:
return value
def humanComprRate(rate):
return u"%.1fx" % rate
def humanAltitude(value):
return ngettext("%.1f meter", "%.1f meters", value) % value
def humanPixelSize(value):
return ngettext("%s pixel", "%s pixels", value) % value
def humanDPI(value):
return u"%s DPI" % value
def humanFrameRate(value):
if isinstance(value, (int, long, float)):
return _("%.1f fps") % value
else:
return value
def processMovieHeader(self, hdr):
self.creation_date = hdr["creation_date"].value
self.last_modification = hdr["lastmod_date"].value
self.duration = timedelta(seconds=float(hdr["duration"].value) / hdr["time_scale"].value)
self.comment = _("Play speed: %.1f%%") % (hdr["play_speed"].value*100)
self.comment = _("User volume: %.1f%%") % (float(hdr["volume"].value)*100//255)