Add TTF subsetting to backend_pdf.py.

svn path=/trunk/matplotlib/; revision=3493

Author: mdboom

lib/matplotlib/backends/backend_pdf.py

@@ -15,6 +15,7 @@
 from cStringIO import StringIO
 from datetime import datetime
 from math import ceil, cos, floor, pi, sin
+import sets
 
 from matplotlib import __version__, rcParams, agg, get_data_path
 from matplotlib._pylab_helpers import Gcf
@@ -29,6 +30,7 @@
 from matplotlib.mathtext import math_parse_s_pdf
 from matplotlib.numerix import Float32, UInt8, fromstring, arange, infinity, isnan, asarray
 from matplotlib.transforms import Bbox
+from matplotlib import ttconv
 
 # Overview
 #
@@ -453,7 +455,8 @@ def writeFonts(self):
             if filename.endswith('.afm'):
                 fontdictObject = self._write_afm_font(filename)
             else:
-                fontdictObject = self.embedTTF(filename)
+                fontdictObject = self.embedTTF(
+                    filename, self.used_characters[filename])
             fonts[Fx] = fontdictObject
             #print >>sys.stderr, filename
         self.writeObject(self.fontObject, fonts)
@@ -471,10 +474,11 @@ def _write_afm_font(self, filename):
         self.writeObject(fontdictObject, fontdict)
         return fontdictObject
 
-    def embedTTF(self, filename):
+    def embedTTF(self, filename, characters):
         """Embed the TTF font from the named file into the document."""
 
         font = FT2Font(str(filename))
+        fonttype = rcParams['pdf.fonttype']
 
         def cvt(length, upe=font.units_per_EM, nearest=True):
             "Convert font coordinates to PDF glyph coordinates"
@@ -515,21 +519,36 @@ def get_char_width(charcode):
 
         firstchar, lastchar = 0, 255
         widths = [ get_char_width(charcode) for charcode in range(firstchar, lastchar+1) ]
+        font_bbox = [ cvt(x, nearest=False) for x in font.bbox ]
 
         widthsObject = self.reserveObject('font widths')
         fontdescObject = self.reserveObject('font descriptor')
         # TODO: "WinAnsiEncoding" could become a parameter of PdfFile. The PDF encoding
         # "WinAnsiEncoding" matches the Python enconding "cp1252" used in method
         # RendererPdf.draw_text and RendererPdf.get_text_width_height to encode Unicode strings.
         fontdict = { 'Type': Name('Font'),
-                     'Subtype': Name('TrueType'),
-                     'Encoding': Name('WinAnsiEncoding'),
                      'BaseFont': ps_name,
                      'FirstChar': firstchar,
                      'LastChar': lastchar,
                      'Widths': widthsObject,
                      'FontDescriptor': fontdescObject }
 
+        if fonttype == 3:
+            charprocsObject = self.reserveObject('character procs')
+            differencesArray = []
+            fontdict['Subtype'] = Name('Type3')
+            fontdict['Name'] = ps_name
+            fontdict['FontBBox'] = font_bbox
+            fontdict['FontMatrix'] = [ .001, 0, 0, .001, 0, 0 ]
+            fontdict['CharProcs'] = charprocsObject
+            fontdict['Encoding'] = {
+                'Type': Name('Encoding'), 
+                'Differences': differencesArray}
+        elif fonttype == 42:
+            fontdict['Subtype'] = Name('TrueType')
+            fontdict['Encoding'] = Name('WinAnsiEncoding'),
+
+            
         flags = 0
         symbolic = False #ps_name.name in ('Cmsy10', 'Cmmi10', 'Cmex10')
         if ff & FIXED_WIDTH: flags |= 1 << 0
@@ -551,11 +570,13 @@ def get_char_width(charcode):
             'CapHeight': cvt(pclt['capHeight'], nearest=False),
             'XHeight': cvt(pclt['xHeight']),
             'ItalicAngle': post['italicAngle'][1], # ???
-            'FontFile2': self.reserveObject('font file'),
             'MaxWidth': max(widths),
             'StemV': 0 # ???
         }
 
+        if fonttype == 42:
+            descriptor['FontFile2'] = self.reserveObject('font file')
+
         # Other FontDescriptor keys include:
         # /FontFamily /Times (optional)
         # /FontStretch /Normal (optional)
@@ -584,24 +605,57 @@ def get_char_width(charcode):
 
             widths = [ firstchar, widths ]
 
+        if fonttype == 3:
+            cmap = font.get_charmap()
+            glyph_ids = []
+            differences = []
+            for c in characters:
+                ccode = ord(c)
+                gind = cmap.get(ccode) or 0
+                glyph_ids.append(gind)
+                differences.append((ccode, font.get_glyph_name(gind)))
+            differences.sort()
+
+            last_c = -256
+            for c, name in differences:
+                if c != last_c + 1:
+                    differencesArray.append(c)
+                differencesArray.append(Name(name))
+                last_c = c
+
+            rawcharprocs = ttconv.get_pdf_charprocs(filename, glyph_ids)
+            charprocs = {}
+            charprocsRef = {}
+            for charname, stream in rawcharprocs.items():
+                charprocObject = self.reserveObject('charProc for %s' % name)
+                self.beginStream(charprocObject.id,
+                                 None,
+                                 {'Length':  len(stream)})
+                self.currentstream.write(stream)                 
+                self.endStream()
+                charprocs[charname] = charprocObject
+            self.writeObject(charprocsObject, charprocs)
+
+        elif fonttype == 42:
+            length1Object = self.reserveObject('decoded length of a font')
+            self.beginStream(descriptor['FontFile2'].id,
+                             self.reserveObject('length of font stream'),
+                             {'Length1': length1Object})
+            fontfile = open(filename, 'rb')
+            length1 = 0
+            while True:
+                data = fontfile.read(4096)
+                if not data: break
+                length1 += len(data)
+                self.currentstream.write(data)
+            fontfile.close()
+            self.endStream()
+            self.writeObject(length1Object, length1)
+
         fontdictObject = self.reserveObject('font dictionary')
-        length1Object = self.reserveObject('decoded length of a font')
         self.writeObject(fontdictObject, fontdict)
         self.writeObject(widthsObject, widths)
         self.writeObject(fontdescObject, descriptor)
-        self.beginStream(descriptor['FontFile2'].id,
-                         self.reserveObject('length of font stream'),
-                         {'Length1': length1Object})
-        fontfile = open(filename, 'rb')
-        length1 = 0
-        while True:
-            data = fontfile.read(4096)
-            if not data: break
-            length1 += len(data)
-            self.currentstream.write(data)
-        fontfile.close()
-        self.endStream()
-        self.writeObject(length1Object, length1)
 
         return fontdictObject
 
@@ -849,6 +903,7 @@ def __init__(self, file):
         self.gc = self.new_gc()
         self.truetype_font_cache = {}
         self.afm_font_cache = {}
+        self.file.used_characters = self.used_characters = {}
 
     def finalize(self):
         self.gc.finalize()
@@ -865,6 +920,16 @@ def check_gc(self, gc, fillcolor=None):
         # Restore gc to avoid unwanted side effects
         gc._fillcolor = orig_fill
 
+    def track_characters(self, font, s):
+        """Keeps track of which characters are required from
+        each font."""
+        if isinstance(font, (str, unicode)):
+            fname = font
+        else:
+            fname = font.fname
+        used_characters = self.used_characters.setdefault(fname, sets.Set())
+        used_characters.update(s)
+
     def draw_arc(self, gcEdge, rgbFace, x, y, width, height,
                  angle1, angle2, rotation):
         """
@@ -1018,7 +1083,7 @@ def _setup_textpos(self, x, y, angle, oldx=0, oldy=0, oldangle=0):
     def draw_mathtext(self, gc, x, y, s, prop, angle):
         # TODO: fix positioning and encoding
         fontsize = prop.get_size_in_points()
-        width, height, pswriter = math_parse_s_pdf(s, 72, fontsize)
+        width, height, pswriter = math_parse_s_pdf(s, 72, fontsize, 0, self.track_characters)
 
         self.check_gc(gc, gc._rgb)
         self.file.output(Op.begin_text)
@@ -1114,6 +1179,7 @@ def draw_text(self, gc, x, y, s, prop, angle, ismath=False):
             y -= b * fontsize / 1000
         else:
             font = self._get_font_ttf(prop)
+            self.track_characters(font, s)
             font.set_text(s, 0.0)
             y += font.get_descent() / 64.0
 
@@ -1131,7 +1197,8 @@ def get_text_width_height(self, s, prop, ismath):
 
         if ismath:
             fontsize = prop.get_size_in_points()
-            w, h, pswriter = math_parse_s_pdf(s, 72, fontsize)
+            w, h, pswriter = math_parse_s_pdf(
+                s, 72, fontsize, 0, self.track_characters)
 
         elif rcParams['pdf.use14corefonts']:
             font = self._get_font_afm(prop)

lib/matplotlib/backends/backend_ps.py

@@ -20,7 +20,7 @@ def _fn_name(): return sys._getframe(1).f_code.co_name
 
 from matplotlib.font_manager import fontManager
 from matplotlib.ft2font import FT2Font, KERNING_UNFITTED, KERNING_DEFAULT, KERNING_UNSCALED
-from matplotlib.ttf2ps import convert_ttf_to_ps
+from matplotlib.ttconv import convert_ttf_to_ps
 from matplotlib.mathtext import math_parse_s_ps
 from matplotlib.text import Text
 
@@ -1033,8 +1033,7 @@ def print_figure(self, outfile, dpi=72, facecolor='w', edgecolor='w',
                     cmap = font.get_charmap()
                     glyph_ids = []
                     for c in chars:
-                        ccode = ord(c)
-                        gind = cmap.get(ccode) or 0
+                        gind = cmap.get(ord(c)) or 0
                         glyph_ids.append(gind)
                     convert_ttf_to_ps(font_filename, fh, rcParams['ps.fonttype'], glyph_ids)
             print >>fh, "end"

lib/matplotlib/mathtext.py

@@ -818,6 +818,8 @@ def render(self, ox, oy, font, sym, fontsize, dpi):
         fontname, metrics, glyphname, offset = \
                 self._get_info(font, sym, fontsize, dpi)
         filename, num = self._get_filename_and_num(font, sym, fontsize, dpi)
+        if self.character_tracker:
+            self.character_tracker(filename, unichr(num))
         if fontname.lower() == 'cmex10':
             oy += offset - 512/2048.*10.
 
@@ -1567,7 +1569,7 @@ def __call__(self, s, dpi, fontsize, angle=0, character_tracker=None):
                 #self.font_object = MyUnicodeFonts(output='PS')
                 Element.fonts = self.font_object
         elif self.output == 'PDF':
-            self.font_object = BakomaPDFFonts()
+            self.font_object = BakomaPDFFonts(character_tracker)
             Element.fonts = self.font_object
 
         handler.clear()

lib/matplotlib/mpl-data/matplotlibrc

@@ -258,10 +258,12 @@ numerix      : numpy  # numpy, Numeric or numarray
                                           # xpdf intended for production of publication quality files,
                                           # but requires ghostscript, xpdf and ps2eps
 #ps.distiller.res  : 6000      # dpi
+#ps.fonttype       : 3         # Output Type 3 (Type3) or Type 42 (TrueType)
 
 # pdf backend params
 #pdf.compression   : 6 # integer from 0 to 9
                        # 0 disables compression (good for debugging)
+#pdf.fonttype       : 3         # Output Type 3 (Type3) or Type 42 (TrueType)
 
 # svg backend params
 #svg.image_inline : True       # write raster image data directly into the svg file

lib/matplotlib/rcsetup.py

@@ -53,12 +53,20 @@ def validate_int(s):
     except ValueError:
         raise ValueError('Could not convert "%s" to int' % s)
 
-def validate_psfonttype(s):
-    'confirm that this is a Postscript font type that we know how to convert to'
-    fonttype = validate_int(s)
-    if fonttype not in (3, 42):
-        raise ValueError('Supported Postscript font types are 3 and 42')
-    return fonttype
+def validate_fonttype(s):
+    'confirm that this is a Postscript of PDF font type that we know how to convert to'
+    fonttypes = { 'type3':    3,
+                  'truetype': 42 }
+    try:
+        fonttype = validate_int(s)
+    except ValueError:
+        if s.lower() in fonttypes.keys():
+            return fonttypes[s.lower()]
+        raise ValueError('Supported Postscript/PDF font types are %s' % fonttypes.keys())
+    else:
+        if fonttype not in fonttypes.values():
+            raise ValueError('Supported Postscript/PDF font types are %s' % fonttypes.values())
+        return fonttype
 
 validate_backend = ValidateInStrings('backend',[
     'Agg2', 'Agg', 'Aqt', 'Cairo', 'CocoaAgg', 'EMF', 'GD', 'GDK',
@@ -416,11 +424,12 @@ def __call__(self, s):
     'ps.useafm'          : [False, validate_bool],  # Set PYTHONINSPECT
     'ps.usedistiller'    : [False, validate_ps_distiller], # use ghostscript or xpdf to distill ps output
     'ps.distiller.res'   : [6000, validate_int],     # dpi
-    'ps.fonttype'        : [3, validate_psfonttype], # 3 or 42
+    'ps.fonttype'        : [3, validate_fonttype], # 3 (Type3) or 42 (Truetype)
     'pdf.compression'    : [6, validate_int],        # compression level from 0 to 9; 0 to disable
     'pdf.inheritcolor'   : [False, validate_bool],   # ignore any color-setting commands from the frontend
     'pdf.use14corefonts' : [False, validate_bool],  # use only the 14 PDF core fonts
                                                     # embedded in every PDF viewing application
+    'pdf.fonttype'      : [3, validate_fonttype],  # 3 (Type3) or 42 (Truetype)
     'svg.image_inline'  : [True, validate_bool],    # write raster image data directly into the svg file
     'svg.image_noscale' : [False, validate_bool],  # suppress scaling of raster data embedded in SVG
     'plugins.directory' : ['.matplotlib_plugins', str], # where plugin directory is locate

matplotlibrc.template

@@ -258,11 +258,12 @@ numerix      : %(numerix)s  # numpy, Numeric or numarray
                                           # xpdf intended for production of publication quality files,
                                           # but requires ghostscript, xpdf and ps2eps
 #ps.distiller.res  : 6000      # dpi
-#ps.fonttype       : 3         # Output Type 3 (raw Postscript) or Type 42 (embedded TrueType)
+#ps.fonttype       : 3         # Output Type 3 (Type3) or Type 42 (TrueType)
 
 # pdf backend params
 #pdf.compression   : 6 # integer from 0 to 9
                        # 0 disables compression (good for debugging)
+#pdf.fonttype       : 3         # Output Type 3 (Type3) or Type 42 (TrueType)
 
 # svg backend params
 #svg.image_inline : True       # write raster image data directly into the svg file

setup.py

@@ -82,7 +82,7 @@
 from setupext import build_agg, build_gtkagg, build_tkagg, build_wxagg,\
      build_ft2font, build_image, build_windowing, build_transforms, \
      build_contour, build_nxutils, build_enthought, build_swigagg, build_gdk, \
-     build_subprocess, build_isnan, build_ttf2ps
+     build_subprocess, build_isnan, build_ttconv
 import distutils.sysconfig
 
 for line in file('lib/matplotlib/__init__.py').readlines():
@@ -120,7 +120,7 @@
 
 # these are not optional
 BUILD_FT2FONT = 1
-BUILD_TTF2PS  = 1
+BUILD_TTCONV  = 1
 BUILD_CONTOUR = 1
 BUILD_NXUTILS = 1
 
@@ -270,8 +270,8 @@ def havegtk():
 if BUILD_FT2FONT:
     build_ft2font(ext_modules, packages)
 
-if BUILD_TTF2PS:
-    build_ttf2ps(ext_modules, packages)
+if BUILD_TTCONV:
+    build_ttconv(ext_modules, packages)
 
 if BUILD_WINDOWING and sys.platform=='win32':
    build_windowing(ext_modules, packages)

setupext.py

@@ -73,7 +73,7 @@
 
 BUILT_AGG       = False
 BUILT_FT2FONT   = False
-BUILT_TTF2PS    = False
+BUILT_TTCONV    = False
 BUILT_GTKAGG    = False
 BUILT_IMAGE     = False
 BUILT_TKAGG     = False
@@ -501,17 +501,17 @@ def build_ft2font(ext_modules, packages):
     ext_modules.append(module)
     BUILT_FT2FONT = True
 
-def build_ttf2ps(ext_modules, packages):
-    global BUILT_TTF2PS
-    if BUILT_TTF2PS: return # only build it if you you haven't already
-    deps = ['src/_ttf2ps.cpp', 'ttconv/pprdrv_tt.cpp', 'ttconv/pprdrv_tt2.cpp', 'ttconv/ttutil.cpp']
+def build_ttconv(ext_modules, packages):
+    global BUILT_TTCONV
+    if BUILT_TTCONV: return # only build it if you you haven't already
+    deps = ['src/_ttconv.cpp', 'ttconv/pprdrv_tt.cpp', 'ttconv/pprdrv_tt2.cpp', 'ttconv/ttutil.cpp']
     deps.extend(glob.glob('CXX/*.cxx'))
     deps.extend(glob.glob('CXX/*.c'))
 
-    module = Extension('matplotlib.ttf2ps', deps)
+    module = Extension('matplotlib.ttconv', deps)
     add_base_flags(module)
     ext_modules.append(module)
-    BUILT_TTF2PS = True
+    BUILT_TTCONV = True
 
 def build_gtkagg(ext_modules, packages, numerix):
     global BUILT_GTKAGG