Temporarily revert PR #1035 bamidei/split_g_l_y_f_to_one_per_file

The split-glyf-to-one-glyph-per-file feature is only partially implemented,
as it was discussed here:

https://github.com/fonttools/fonttools/issues/153#issuecomment-346677171

I need to cut a bugfix release today, but I don't want to ship this as is.
I prefer to temporarily rever, tag a release from master branch, then
revert it again to its present state so that bamidei can complete his
work.

---

Revert "[glyf] make splitGlyphs output more compact"

This reverts commit d08d635a93289982b028aacd3b9039fbe742f3e8.

Revert "more whitespace"

This reverts commit bd030f61c6b360fd360632bbc6c19abb057c9a24.

Revert "minor whitespace"

This reverts commit f2a8c787b13b92ae0e0f61477e584316144747c7.

Revert "Merge pull request #1035 from bamidei/split_g_l_y_f_to_one_per_file"

This reverts commit 17b89d9dde7691dbbb6815efba92eff446af62ac, reversing
changes made to b8482d9666f08d3603ee93d3ca52931550f715fc.

Lib/fontTools/misc/xmlReader.py

@@ -17,8 +17,7 @@ BUFSIZE = 0x4000
 
 class XMLReader(object):
 
-	def __init__(self, fileOrPath, ttFont, progress=None, quiet=None, contentOnly=False):
+	def __init__(self, fileOrPath, ttFont, progress=None, quiet=None):
-  
 		if fileOrPath == '-':
 			fileOrPath = sys.stdin
 		if not hasattr(fileOrPath, "read"):
@@ -36,7 +35,6 @@ class XMLReader(object):
 			self.quiet = quiet
 		self.root = None
 		self.contentStack = []
-		self.contentOnly = contentOnly
 		self.stackSize = 0
 
 	def read(self, rootless=False):
@@ -75,24 +73,8 @@ class XMLReader(object):
 			parser.Parse(chunk, 0)
 
 	def _startElementHandler(self, name, attrs):
-		if self.stackSize == 1 and self.contentOnly:
-			# We already know the table we're parsing, skip
-			# parsing the table tag and continue to
-			# stack '2' which begins parsing content
-			self.contentStack.append([])
-			self.stackSize = 2
-			return
 		stackSize = self.stackSize
 		self.stackSize = stackSize + 1
-		subFile = attrs.get("src")
-		if subFile is not None:
-			if hasattr(self.file, 'name'):
-				# if file has a name, get its parent directory
-				dirname = os.path.dirname(self.file.name)
-			else:
-				# else fall back to using the current working directory
-				dirname = os.getcwd()
-			subFile = os.path.join(dirname, subFile)
 		if not stackSize:
 			if name != "ttFont":
 				raise TTXParseError("illegal root tag: %s" % name)
@@ -103,7 +85,15 @@ class XMLReader(object):
 				self.ttFont.sfntVersion = sfntVersion
 			self.contentStack.append([])
 		elif stackSize == 1:
+			subFile = attrs.get("src")
 			if subFile is not None:
+				if hasattr(self.file, 'name'):
+					# if file has a name, get its parent directory
+					dirname = os.path.dirname(self.file.name)
+				else:
+					# else fall back to using the current working directory
+					dirname = os.getcwd()
+				subFile = os.path.join(dirname, subFile)
 				subReader = XMLReader(subFile, self.ttFont, self.progress)
 				subReader.read()
 				self.contentStack.append([])
@@ -129,11 +119,6 @@ class XMLReader(object):
 				self.currentTable = tableClass(tag)
 				self.ttFont[tag] = self.currentTable
 			self.contentStack.append([])
-		elif stackSize == 2 and subFile is not None:
-			subReader = XMLReader(subFile, self.ttFont, self.progress, contentOnly=True)
-			subReader.read()
-			self.contentStack.append([])
-			self.root = subReader.root
 		elif stackSize == 2:
 			self.contentStack.append([])
 			self.root = (name, attrs, self.contentStack[-1])
@@ -149,7 +134,6 @@ class XMLReader(object):
 	def _endElementHandler(self, name):
 		self.stackSize = self.stackSize - 1
 		del self.contentStack[-1]
-		if not self.contentOnly:
 		if self.stackSize == 1:
 			self.root = None
 		elif self.stackSize == 2:

Lib/fontTools/ttLib/__init__.py

@@ -247,8 +247,7 @@ class TTFont(object):
 
 	def saveXML(self, fileOrPath, progress=None, quiet=None,
 			tables=None, skipTables=None, splitTables=False, disassembleInstructions=True,
-			splitGlyphs=False, bitmapGlyphDataFormat='raw', newlinestr=None):
+			bitmapGlyphDataFormat='raw', newlinestr=None):
-
 		"""Export the font as TTX (an XML-based text file), or as a series of text
 		files when splitTables is true. In the latter case, the 'fileOrPath'
 		argument should be a path to a directory.
@@ -291,7 +290,7 @@ class TTFont(object):
 
 		if not splitTables:
 			writer.newline()
-		if splitTables or splitGlyphs:
+		else:
 			# 'fileOrPath' must now be a path
 			path, ext = os.path.splitext(fileOrPath)
 			fileNameTemplate = path + ".%s" + ext
@@ -300,11 +299,8 @@ class TTFont(object):
 			if progress:
 				progress.set(i)
 			tag = tables[i]
-			if splitTables or (splitGlyphs and tag == 'glyf'):
-				tablePath = fileNameTemplate % tagToIdentifier(tag)
-			else:
-				tablePath = None
 			if splitTables:
+				tablePath = fileNameTemplate % tagToIdentifier(tag)
 				tableWriter = xmlWriter.XMLWriter(tablePath, idlefunc=idlefunc,
 						newlinestr=newlinestr)
 				tableWriter.begintag("ttFont", ttLibVersion=version)
@@ -314,7 +310,7 @@ class TTFont(object):
 				writer.newline()
 			else:
 				tableWriter = writer
-			self._tableToXML(tableWriter, tag, progress, splitGlyphs=splitGlyphs)
+			self._tableToXML(tableWriter, tag, progress)
 			if splitTables:
 				tableWriter.endtag("ttFont")
 				tableWriter.newline()
@@ -328,7 +324,7 @@ class TTFont(object):
 		if not hasattr(fileOrPath, "write") and fileOrPath != "-":
 			writer.close()
 
-	def _tableToXML(self, writer, tag, progress, splitGlyphs=False, quiet=None):
+	def _tableToXML(self, writer, tag, progress, quiet=None):
 		if quiet is not None:
 			deprecateArgument("quiet", "configure logging instead")
 		if tag in self:
@@ -350,9 +346,7 @@ class TTFont(object):
 			attrs['raw'] = True
 		writer.begintag(xmlTag, **attrs)
 		writer.newline()
-		if tag == "glyf":
+		if tag in ("glyf", "CFF "):
-			table.toXML(writer, self, progress, splitGlyphs)
-		elif tag == "CFF ":
 			table.toXML(writer, self, progress)
 		else:
 			table.toXML(writer, self)
@@ -879,8 +873,8 @@ def _escapechar(c):
 		return hex(byteord(c))[2:]
 
 
-def nameToIdentifier(name):
+def tagToIdentifier(tag):
-	"""Convert a name to a valid (but UGLY) python identifier,
+	"""Convert a table tag to a valid (but UGLY) python identifier,
 	as well as a filename that's guaranteed to be unique even on a
 	caseless file system. Each character is mapped to two characters.
 	Lowercase letters get an underscore before the letter, uppercase
@@ -893,25 +887,19 @@ def nameToIdentifier(name):
 		'OS/2' -> 'O_S_2f_2'
 	"""
 	import re
-	while len(name) > 1 and name[-1] == ' ':
+	tag = Tag(tag)
-		name = name[:-1]
+	if tag == "GlyphOrder":
+		return tag
+	assert len(tag) == 4, "tag should be 4 characters long"
+	while len(tag) > 1 and tag[-1] == ' ':
+		tag = tag[:-1]
 	ident = ""
-	for c in name:
+	for c in tag:
 		ident = ident + _escapechar(c)
 	if re.match("[0-9]", ident):
 		ident = "_" + ident
 	return ident
 
-def tagToIdentifier(tag):
-	"""This performs the same conversion which nameToIdentifiier does
-	with the additional assertion that the source tag is 4 characters
-	long which is criteria for a valid tag name.
-	"""
-	if tag == "GlyphOrder":
-		return tag
-	ret = nameToIdentifier(tag)
-	assert len(tag) == 4, "tag should be 4 characters long"
-	return ret
 
 def identifierToTag(ident):
 	"""the opposite of tagToIdentifier()"""

Lib/fontTools/ttLib/tables/_g_l_y_f.py

@@ -5,7 +5,6 @@ from collections import namedtuple
 from fontTools.misc.py23 import *
 from fontTools.misc import sstruct
 from fontTools import ttLib
-from fontTools import version
 from fontTools.misc.textTools import safeEval, pad
 from fontTools.misc.arrayTools import calcBounds, calcIntBounds, pointInRect
 from fontTools.misc.bezierTools import calcQuadraticBounds
@@ -17,17 +16,10 @@ import sys
 import struct
 import array
 import logging
-import os
+
-from fontTools.misc import xmlWriter
-from fontTools.ttLib import nameToIdentifier
 
 log = logging.getLogger(__name__)
 
-# We compute the version the same as is computed in ttlib/__init__
-# so that we can write 'ttLibVersion' attribute of the glyf TTX files
-# when glyf is written to separate files.
-version = ".".join(version.split('.')[:2])
-
 #
 # The Apple and MS rasterizers behave differently for
 # scaled composite components: one does scale first and then translate
@@ -118,14 +110,10 @@ class table__g_l_y_f(DefaultTable.DefaultTable):
 			ttFont['maxp'].numGlyphs = len(self.glyphs)
 		return data
 
-	def toXML(self, writer, ttFont, progress=None, splitGlyphs=False):
+	def toXML(self, writer, ttFont, progress=None):
-		notice = (
-			"The xMin, yMin, xMax and yMax values\n"
-			"will be recalculated by the compiler.")
-		glyphNames = ttFont.getGlyphNames()
-		if not splitGlyphs:
 		writer.newline()
-			writer.comment(notice)
+		glyphNames = ttFont.getGlyphNames()
+		writer.comment("The xMin, yMin, xMax and yMax values\nwill be recalculated by the compiler.")
 		writer.newline()
 		writer.newline()
 		counter = 0
@@ -138,43 +126,20 @@ class table__g_l_y_f(DefaultTable.DefaultTable):
 			counter = counter + 1
 			glyph = self[glyphName]
 			if glyph.numberOfContours:
-				if splitGlyphs:
+				writer.begintag('TTGlyph', [
-					path, ext = os.path.splitext(writer.file.name)
-					fileNameTemplate = path + ".%s" + ext
-					glyphPath = fileNameTemplate % nameToIdentifier(glyphName)
-					glyphWriter = xmlWriter.XMLWriter(
-						glyphPath, idlefunc=writer.idlefunc,
-						newlinestr=writer.newlinestr)
-					glyphWriter.begintag("ttFont", ttLibVersion=version)
-					glyphWriter.newline()
-					glyphWriter.begintag("glyf")
-					glyphWriter.newline()
-					glyphWriter.comment(notice)
-					glyphWriter.newline()
-					writer.simpletag("TTGlyph", src=os.path.basename(glyphPath))
-				else:
-					glyphWriter = writer
-				glyphWriter.begintag('TTGlyph', [
 						("name", glyphName),
 						("xMin", glyph.xMin),
 						("yMin", glyph.yMin),
 						("xMax", glyph.xMax),
 						("yMax", glyph.yMax),
 						])
-				glyphWriter.newline()
+				writer.newline()
-				glyph.toXML(glyphWriter, ttFont)
+				glyph.toXML(writer, ttFont)
-				glyphWriter.endtag('TTGlyph')
+				writer.endtag('TTGlyph')
-				glyphWriter.newline()
+				writer.newline()
-				if splitGlyphs:
-					glyphWriter.endtag("glyf")
-					glyphWriter.newline()
-					glyphWriter.endtag("ttFont")
-					glyphWriter.newline()
-					glyphWriter.close()
 			else:
 				writer.simpletag('TTGlyph', name=glyphName)
 				writer.comment("contains no outline data")
-				if not splitGlyphs:
 				writer.newline()
 			writer.newline()
 

Lib/fontTools/ttx.py

@@ -38,9 +38,6 @@ usage: ttx [options] inputfile1 [... inputfileN]
        to the individual table dumps. This file can be used as
        input to ttx, as long as the table files are in the
        same directory.
-    -g Split glyf table: Save the glyf data into separate TTX files
-       per glyph and write a small TTX for the glyf table which
-       contains references to the individual TTGlyph elements.
     -i Do NOT disassemble TT instructions: when this option is given,
        all TrueType programs (glyph programs, the font program and the
        pre-program) will be written to the TTX file as hex data
@@ -113,7 +110,6 @@ class Options(object):
 	verbose = False
 	quiet = False
 	splitTables = False
-	splitGlyphs = False
 	disassembleInstructions = True
 	mergeFile = None
 	recalcBBoxes = True
@@ -164,8 +160,6 @@ class Options(object):
 				self.skipTables.append(value)
 			elif option == "-s":
 				self.splitTables = True
-			elif option == "-g":
-				self.splitGlyphs = True
 			elif option == "-i":
 				self.disassembleInstructions = False
 			elif option == "-z":
@@ -261,7 +255,6 @@ def ttDump(input, output, options):
 			tables=options.onlyTables,
 			skipTables=options.skipTables,
 			splitTables=options.splitTables,
-			splitGlyphs=options.splitGlyphs,
 			disassembleInstructions=options.disassembleInstructions,
 			bitmapGlyphDataFormat=options.bitmapGlyphDataFormat,
 			newlinestr=options.newlinestr)
@@ -325,7 +318,7 @@ def guessFileType(fileName):
 
 
 def parseOptions(args):
-	rawOptions, files = getopt.getopt(args, "ld:o:fvqht:x:sgim:z:baey:",
+	rawOptions, files = getopt.getopt(args, "ld:o:fvqht:x:sim:z:baey:",
 			['unicodedata=', "recalc-timestamp", 'flavor=', 'version',
 			 'with-zopfli', 'newline='])
 

Tests/misc/xmlReader_test.py

@@ -142,48 +142,7 @@ class TestXMLReader(unittest.TestCase):
 		self.assertTrue(reader.file.closed)
 		os.remove(tmp.name)
 
-	def test_read_sub_file(self):
-		# Verifies that sub-file content is able to be read to a table.
-		expectedContent = u'testContent'
-		expectedNameID = '1'
-		expectedPlatform = '3'
-		expectedLangId = '0x409'
 
-		with tempfile.NamedTemporaryFile(delete=False) as tmp:
-			subFileData = (
-				'<ttFont ttLibVersion="3.15">'
-					'<name>'
-						'<namerecord nameID="%s" platformID="%s" platEncID="1" langID="%s">'
-							'%s'
-						'</namerecord>'
-					'</name>'
-				'</ttFont>'
-			) % (expectedNameID, expectedPlatform, expectedLangId, expectedContent)
-			tmp.write(subFileData.encode("utf-8"))
-
-		with tempfile.NamedTemporaryFile(delete=False) as tmp2:
-			fileData = (
-				'<ttFont ttLibVersion="3.15">'
-					'<name>'
-						'<namerecord src="%s"/>'
-					'</name>'
-				'</ttFont>'
-			) % tmp.name
-			tmp2.write(fileData.encode('utf-8'))
-
-		ttf = TTFont()
-		with open(tmp2.name, "rb") as f:
-			reader = XMLReader(f, ttf)
-			reader.read()
-			reader.close()
-			nameTable = ttf['name']
-			self.assertTrue(int(expectedNameID) == nameTable.names[0].nameID)
-			self.assertTrue(int(expectedLangId, 16) == nameTable.names[0].langID)
-			self.assertTrue(int(expectedPlatform) == nameTable.names[0].platformID)
-			self.assertEqual(expectedContent, nameTable.names[0].string.decode(nameTable.names[0].getEncoding()))
-
-		os.remove(tmp.name)
-		os.remove(tmp2.name)
 
 if __name__ == '__main__':
 	import sys