"""
Provide an encoder for a font specification configuration: the encoder is fed
with Unicode characters one by one and determines the needed font switches
between the preceding and the current character.
"""
import re
import xml.dom.minidom
def _indent(string, width=2):
"""Indent the <string> lines by <width> blank characters."""
istr = ' ' * width
s = istr + istr.join(string.splitlines(1))
return s
00020 class UnicodeInterval:
"""Unicode codepoint interval, including all codepoints between its minimum
and maximum boundary.
For any Unicode codepoint it can be queried if it belongs to the interval.
"""
_re_codepoint = re.compile(r'^[Uu]\+?([0-9A-Fa-f]+)$')
def __init__(self):
self._min_boundary = 0
self._max_boundary = 0
00036 def __str__(self):
"""Dump the instance's data attributes."""
string = 'UnicodeInterval: [' + str(self._min_boundary)
if self._max_boundary != self._min_boundary:
string += ',' + str(self._max_boundary)
string += ']'
return string
00044 def _unicode_to_ordinal(self, codepoint):
"""Return the ordinal of the specified codepoint."""
match = self._re_codepoint.match(codepoint)
if match:
return int(match.group(1), 16)
else:
raise RuntimeError, 'Not a unicode codepoint: ' + codepoint
00052 def from_char(self, char):
"""Interval for a single character"""
self._min_boundary = ord(char)
self._max_boundary = self._min_boundary
return self
00058 def from_codepoint(self, codepoint):
"""Interval for a single character defined as unicode string."""
self._min_boundary = self._unicode_to_ordinal(codepoint)
self._max_boundary = self._min_boundary
return self
00064 def from_interval(self, codepoint1, codepoint2):
"""Interval from a unicode range."""
self._min_boundary = self._unicode_to_ordinal(codepoint1)
self._max_boundary = self._unicode_to_ordinal(codepoint2)
if self._min_boundary > self._max_boundary:
self._min_boundary, self._max_boundary = \
self._max_boundary, self._min_boundary
return self
00073 def match(self, char):
"""
Determine whether the specified character is contained in this
instance's interval.
"""
return (ord(char) >= self._min_boundary
and ord(char) <= self._max_boundary)
00082 class FontSpec:
"""
Font specification, consisting of one or several unicode character
intervals and of fonts to select for those characters. The object
fully defines the fonts to switch to.
"""
transition_types = ['enter', 'inter', 'exit']
_re_interval = re.compile(r'^([Uu][0-9A-Fa-f]+)-([Uu][0-9A-Fa-f]+)$')
_re_codepoint = re.compile(r'^([Uu][0-9A-Fa-f]+)$')
00096 def __init__(self, intervals=None, subfont_first=False):
"""Create a font specification from the specified codepoint intervals.
The other data attributes will be set by the caller later.
"""
self.type = ""
self.id = None
self.refmode = None
self.transitions = {}
self.fontspecs = [self]
self.subfont_first = subfont_first
for type in self.transition_types:
self.transitions[type] = {}
if not(intervals):
self._intervals = []
return
try:
self._intervals = list(intervals)
except TypeError:
self._intervals = [intervals]
def fromnode(self, node):
range = node.getAttribute('range')
charset = node.getAttribute('charset')
id = node.getAttribute('id')
refmode = node.getAttribute('refmode')
self.type = node.getAttribute('type')
if (range):
self._intervals = self._parse_range(range)
elif (charset):
intervals = []
for char in charset:
intervals.append(UnicodeInterval().from_char(char))
self._intervals = intervals
if (id):
self.id = id
if (refmode):
self.refmode = refmode
for transition_type in self.transition_types:
self._parse_transitions(node, transition_type)
00143 def _parse_range(self, range):
"""Parse the specified /fonts/fontspec@range attribute to a
UnicodeInterval list.
"""
print range
intervals = []
chunks = range.split()
for chunk in chunks:
match = self._re_interval.match(chunk)
if match:
urange = UnicodeInterval().from_interval(match.group(1),
match.group(2))
intervals.append(urange)
else:
match = self._re_codepoint.match(chunk)
if match:
intervals.append(
UnicodeInterval().from_codepoint(match.group(1)))
else:
raise RuntimeError, 'Unable to parse range: "' + range + '"'
return intervals
00166 def _parse_transitions(self, node, transition_type):
"""Evaluate the font elements of the specified fontspec element for the
specified transition type (enter, inter or exit).
"""
fontlist = self.transitions[transition_type]
for dom_transition in node.getElementsByTagName(transition_type):
for dom_font in dom_transition.getElementsByTagName('font'):
font = ''
types = dom_font.getAttribute("type")
types = types.split()
for dom_child in dom_font.childNodes:
if dom_child.nodeType == dom_child.TEXT_NODE:
font += dom_child.nodeValue
if (font):
for type in types:
fontlist[type] = font
00184 def _switch_to(self, fonts):
"""
Return a string with the XeTeX font switching commands for the
specified font types.
"""
s = ''
for type, font in fonts.items():
s += '\set%sfont{%s}' % (type, font)
if s:
s = r"\savefont" + s + r"\restorefont{}"
return s
def enter(self):
s = self._switch_to(self.transitions["enter"])
return s
def exit(self):
s = self._switch_to(self.transitions["exit"])
return s
def interchar(self):
s = self._switch_to(self.transitions["inter"])
return s
00208 def __str__(self):
"""Dump the instance's data attributes."""
string = 'FontSpec:'
string += '\n Id: %s' % self.id
string += '\n Refmode: %s' % self.refmode
for interval in self._intervals:
string += '\n' + _indent(str(interval))
return string
def add_subfont(self, fontspec):
print "%s -> %s" % (self.id, fontspec.id)
if self.subfont_first:
self.fontspecs.insert(-1, fontspec)
else:
self.fontspecs.append(fontspec)
def add_uranges(self, ranges, depth=1):
if depth:
for f in self.fontspecs:
if f != self:
f.add_uranges(ranges)
self._intervals.extend(ranges)
def get_uranges(self):
return self._intervals
def contains(self, char):
for interval in self._intervals:
if interval.match(char):
return True
else:
return False
00243 def match(self, char):
"""Determine whether the font specification matches the specified
object, thereby considering refmode.
"""
fontspec = None
for fontspec in self.fontspecs:
if fontspec.contains(char):
try:
print "%s has '%s'" % (fontspec.id, str(char))
except:
print "%s has '%s'" % (fontspec.id, ord(char))
return fontspec
return None
00259 class DefaultFontSpec(FontSpec):
"""
The default fontspec gives priority to its children, and
contains any character.
"""
def __init__(self):
FontSpec.__init__(self, subfont_first=True)
def contains(self, char):
return True
00271 class FontSpecConfig:
"""
This object parses an XML fontspec configuration file and build the
resulting fontspec tree, the root fontspec being the default font
to apply.
The fontspec configuration file defines the fonts to apply in order
of precedence (and for some Unicode ranges) and the font levels (or
subfonts) thanks to the 'refmode' attribute that links a font to its
parent.
"""
00283 def __init__(self, conf_file):
"""Create a font specification configuration from the specified file
(file name or file-like object).
"""
self.fontspecs = []
self.fontnames = {}
self.default_fontspec = DefaultFontSpec()
dom_document = xml.dom.minidom.parse(conf_file)
for dom_fontspec in dom_document.getElementsByTagName('fontspec'):
default = dom_fontspec.getAttribute('default')
if default:
print "has default"
fontspec = self.default_fontspec
else:
fontspec = FontSpec()
fontspec.fromnode(dom_fontspec)
if fontspec != self.default_fontspec:
self.fontspecs.append(fontspec)
if fontspec.id:
self.fontnames[fontspec.id] = fontspec
dom_document.unlink()
self.build_tree()
00312 def build_tree(self):
"""
Build the fontspec tree, the root node being the default font
to apply. The fontspecs without a refmode (i.e. not being
explicitely a subfont) are direct children of the default font.
"""
to_ignore = []
for fontspec in self.fontspecs:
if fontspec.type == "ignore":
to_ignore.append(fontspec)
continue
if not(fontspec.refmode):
f = self.default_fontspec
else:
f = self.fontnames.get(fontspec.refmode, None)
if (f):
f.add_subfont(fontspec)
else:
raise ValueError("wrong fontspec tree")
for f in to_ignore:
self.default_fontspec.add_uranges(f.get_uranges())
00338 def __str__(self):
"""Dump the instance's data attributes."""
string = 'FontSpecConfig:'
string += '\n Fontspec list:'
for fontspec in self.fontspecs:
string += '\n' + _indent(str(fontspec), 4)
return string
00347 class FontSpecEncoder:
"""
Encoder working with font specifications: it is fed
with Unicode characters one by one and it inserts the needed font switches
between the preceding and the current character.
"""
00354 def __init__(self, configuration):
"""
Create a font specification encoder from the specified configuration
file (file name or file-like object).
"""
self._conf = FontSpecConfig(configuration)
self._cur_fontspec = None
self._ref_stack = [self._conf.default_fontspec]
00363 def _switch_to(self, fontspec):
"""
Insert the transition string, according to the newly selected
fontspec and the previously selected fontspec
"""
s = ""
if fontspec == self._cur_fontspec:
return fontspec.interchar()
if self._cur_fontspec:
s += self._cur_fontspec.exit()
self._cur_fontspec = fontspec
s += fontspec.enter()
return s
00382 def _encode(self, char):
"""
Select the fontspec matching the specified <char>, and switch to
this font as current font.
The principle to find out the fontspec is to:
- find from the current font level a matching font
(the current font leaf or the direct font children)
- if no font is found try with the parent font, and so on,
up to the default root font (that must exist).
"""
fontspec = self._cur_fontspec or self._conf.default_fontspec
fontspec = fontspec.match(char)
while not(fontspec):
self._ref_stack.pop()
fontspec = self._ref_stack[-1].match(char)
if fontspec != self._ref_stack[-1]:
self._ref_stack.append(fontspec)
return self._switch_to(fontspec)
def ignorechars(self, charset):
"Characters to ignore in font selection (maintain the current one)"
intervals = [ UnicodeInterval().from_char(c) for c in charset ]
self._conf.default_fontspec.add_uranges(intervals)
00410 def encode(self, char):
"""
Return a string consisting of the specified character prepended by
all necessary font switching commands.
"""
return (self._encode(char), char)
00417 def stop(self):
"""
Cleanly exit from the current fontspec
"""
if self._cur_fontspec:
s = self._cur_fontspec.exit()
self._cur_fontspec = None
return s