def create_sfnt(self, text_item):
get_table = partial(self.qt_hack.get_sfnt_table, text_item)
try:
ans = Font(Sfnt(get_table))
except UnsupportedFont as e:
raise UnsupportedFont('The font %s is not a valid sfnt. Error: %s'%(
text_item.font().family(), e))
glyph_map = self.qt_hack.get_glyph_map(text_item)
gm = {}
for uc, glyph_id in enumerate(glyph_map):
if glyph_id not in gm:
gm[glyph_id] = unichr(uc)
ans.full_glyph_map = gm
return ans
def mergeable(fonts):
has_type0 = False
for font in fonts:
if font['Subtype'] == 'Type0':
has_type0 = True
if not font['Encoding'] or not font['Encoding'].startswith('Identity-'):
return False
else:
if not font['Data']:
return False
try:
sfnt = Sfnt(font['Data'])
except UnsupportedFont:
return False
font['sfnt'] = sfnt
if b'glyf' not in sfnt:
return False
return has_type0
def create_sfnt(self, text_item):
get_table = partial(self.qt_hack.get_sfnt_table, text_item)
try:
ans = Font(Sfnt(get_table))
except UnsupportedFont as e:
raise UnsupportedFont(
'The font %s is not a valid sfnt. Error: %s' %
(text_item.font().family(), e))
glyph_map = self.qt_hack.get_glyph_map(text_item)
gm = {}
ans.ignore_glyphs = set()
for uc, glyph_id in enumerate(glyph_map):
if glyph_id not in gm:
gm[glyph_id] = codepoint_to_chr(uc)
if uc in (0xad, 0x200b):
ans.ignore_glyphs.add(glyph_id)
ans.full_glyph_map = gm
return ans
def subset_fonts(pdf_doc, log):
all_fonts = pdf_doc.list_fonts(True)
for font in all_fonts:
if font['Subtype'] != 'Type0' and font['Data']:
try:
sfnt = Sfnt(font['Data'])
except UnsupportedFont:
continue
if b'glyf' not in sfnt:
continue
num, gen = font['Reference']
glyphs = all_glyph_ids_in_w_arrays(
(font['W'] or (), font['W2'] or ()), as_set=True)
pdf_subset(sfnt, glyphs)
data = sfnt()[0]
log('Subset embedded font from: {} to {}'.format(
human_readable(len(font['Data'])), human_readable(len(data))))
pdf_doc.replace_font_data(data, num, gen)
def subset(raw, individual_chars, ranges=(), warnings=None):
warn = partial(do_warn, warnings)
chars = set(map(ord, individual_chars))
for r in ranges:
chars |= set(xrange(ord(r[0]), ord(r[1]) + 1))
# Always add the space character for ease of use from the command line
if ord(' ') not in chars:
chars.add(ord(' '))
sfnt = Sfnt(raw)
old_sizes = sfnt.sizes()
# Remove the Digital Signature table since it is useless in a subset
# font anyway
sfnt.pop(b'DSIG', None)
# Remove non core tables as they aren't likely to be used by renderers
# anyway
core_tables = {
b'cmap', b'hhea', b'head', b'hmtx', b'maxp', b'name', b'OS/2', b'post',
b'cvt ', b'fpgm', b'glyf', b'loca', b'prep', b'CFF ', b'VORG', b'EBDT',
b'EBLC', b'EBSC', b'BASE', b'GSUB', b'GPOS', b'GDEF', b'JSTF', b'gasp',
b'hdmx', b'kern', b'LTSH', b'PCLT', b'VDMX', b'vhea', b'vmtx', b'MATH'
}
for tag in list(sfnt):
if tag not in core_tables:
del sfnt[tag]
try:
cmap = sfnt[b'cmap']
except KeyError:
raise UnsupportedFont('This font has no cmap table')
# Get mapping of chars to glyph ids for all specified chars
character_map = cmap.get_character_map(chars)
extra_glyphs = set()
if b'GSUB' in sfnt:
# Parse all substitution rules to ensure that glyphs that can be
# substituted for the specified set of glyphs are not removed
gsub = sfnt[b'GSUB']
try:
gsub.decompile()
extra_glyphs = gsub.all_substitutions(character_map.itervalues())
except UnsupportedFont as e:
warn('Usupported GSUB table: %s' % e)
except Exception as e:
warn('Failed to decompile GSUB table:', traceback.format_exc())
if b'loca' in sfnt and b'glyf' in sfnt:
# TrueType Outlines
subset_truetype(sfnt, character_map, extra_glyphs)
elif b'CFF ' in sfnt:
# PostScript Outlines
subset_postscript(sfnt, character_map, extra_glyphs)
else:
raise UnsupportedFont('This font does not contain TrueType '
'or PostScript outlines')
# Restrict the cmap table to only contain entries for the resolved glyphs
cmap.set_character_map(character_map)
if b'kern' in sfnt:
try:
sfnt[b'kern'].restrict_to_glyphs(
frozenset(character_map.itervalues()))
except UnsupportedFont as e:
warn('kern table unsupported, ignoring: %s' % e)
except Exception as e:
warn('Subsetting of kern table failed, ignoring:',
traceback.format_exc())
raw, new_sizes = sfnt()
return raw, old_sizes, new_sizes
cmap = self.cmap.get_character_map(chars)
glyph_ids = (cmap[c] for c in chars)
pixel_size_x = stretch * pixel_size
xscale = pixel_size_x / self.units_per_em
return tuple(i * xscale for i in self.glyph_widths(glyph_ids))
def glyph_widths(self, glyph_ids):
last = len(self._advance_widths)
return tuple(self._advance_widths[i if i < last else -1]
for i in glyph_ids)
def width(self, string, pixel_size=12.0, stretch=1.0):
'The width of the string at the specified pixel size and stretch, in pixels'
return sum(self.advance_widths(string, pixel_size, stretch))
if __name__ == '__main__':
import sys
from calibre.utils.fonts.sfnt.container import Sfnt
with open(sys.argv[-1], 'rb') as f:
raw = f.read()
sfnt = Sfnt(raw)
m = FontMetrics(sfnt)
print('Ascent:', m.pdf_ascent)
print('Descent:', m.pdf_descent)
print('PDF BBox:', m.pdf_bbox)
print('CapHeight:', m.pdf_capheight)
print('AvgWidth:', m.pdf_avg_width)
print('ItalicAngle', m.post.italic_angle)
print('StemV', m.pdf_stemv)
def load_font(stream_or_path):
raw = stream_or_path
if hasattr(raw, 'read'):
raw = raw.read()
from calibre.utils.fonts.sfnt.container import Sfnt
return Sfnt(raw)
def subset(raw, individual_chars, ranges=(), warnings=None):
warn = partial(do_warn, warnings)
chars = set(map(ord, individual_chars))
for r in ranges:
chars |= set(xrange(ord(r[0]), ord(r[1])+1))
# Always add the space character for ease of use from the command line
if ord(' ') not in chars:
chars.add(ord(' '))
sfnt = Sfnt(raw)
old_sizes = sfnt.sizes()
# Remove the Digital Signature table since it is useless in a subset
# font anyway
sfnt.pop(b'DSIG', None)
# Remove non core tables as they aren't likely to be used by renderers
# anyway
core_tables = {b'cmap', b'hhea', b'head', b'hmtx', b'maxp', b'name',
b'OS/2', b'post', b'cvt ', b'fpgm', b'glyf', b'loca', b'prep',
b'CFF ', b'VORG', b'EBDT', b'EBLC', b'EBSC', b'BASE', b'GSUB',
b'GPOS', b'GDEF', b'JSTF', b'gasp', b'hdmx', b'kern', b'LTSH',
b'PCLT', b'VDMX', b'vhea', b'vmtx', b'MATH'}
for tag in list(sfnt):
if tag not in core_tables:
del sfnt[tag]
try:
cmap = sfnt[b'cmap']
except KeyError:
raise UnsupportedFont('This font has no cmap table')
# Get mapping of chars to glyph ids for all specified chars
character_map = cmap.get_character_map(chars)
extra_glyphs = set()
if b'GSUB' in sfnt:
# Parse all substitution rules to ensure that glyphs that can be
# substituted for the specified set of glyphs are not removed
gsub = sfnt[b'GSUB']
try:
gsub.decompile()
extra_glyphs = gsub.all_substitutions(character_map.itervalues())
except UnsupportedFont as e:
warn('Usupported GSUB table: %s'%e)
except Exception as e:
warn('Failed to decompile GSUB table:', traceback.format_exc())
if b'loca' in sfnt and b'glyf' in sfnt:
# TrueType Outlines
subset_truetype(sfnt, character_map, extra_glyphs)
elif b'CFF ' in sfnt:
# PostScript Outlines
subset_postscript(sfnt, character_map, extra_glyphs)
else:
raise UnsupportedFont('This font does not contain TrueType '
'or PostScript outlines')
# Restrict the cmap table to only contain entries for the resolved glyphs
cmap.set_character_map(character_map)
if b'kern' in sfnt:
try:
sfnt[b'kern'].restrict_to_glyphs(frozenset(character_map.itervalues()))
except UnsupportedFont as e:
warn('kern table unsupported, ignoring: %s'%e)
except Exception as e:
warn('Subsetting of kern table failed, ignoring:',
traceback.format_exc())
raw, new_sizes = sfnt()
return raw, old_sizes, new_sizes