Merge pull request #394 from anthrotype/psLib-py23

make t1Lib and psLib compatible with python 3

Lib/fontTools/misc/eexec.py

@@ -19,19 +19,35 @@ def _encryptChar(plain, R):
 
 
 def decrypt(cipherstring, R):
+	r"""
+	>>> testStr = b"\0\0asdadads asds\265"
+	>>> decryptedStr, R = decrypt(testStr, 12321)
+	>>> decryptedStr == b'0d\nh\x15\xe8\xc4\xb2\x15\x1d\x108\x1a<6\xa1'
+	True
+	>>> R == 36142
+	True
+	"""
 	plainList = []
 	for cipher in cipherstring:
 		plain, R = _decryptChar(cipher, R)
 		plainList.append(plain)
-	plainstring = strjoin(plainList)
+	plainstring = bytesjoin(plainList)
 	return plainstring, int(R)
 
 def encrypt(plainstring, R):
+	r"""
+	>>> testStr = b'0d\nh\x15\xe8\xc4\xb2\x15\x1d\x108\x1a<6\xa1'
+	>>> encryptedStr, R = encrypt(testStr, 12321)
+	>>> encryptedStr == b"\0\0asdadads asds\265"
+	True
+	>>> R == 36142
+	True
+	"""
 	cipherList = []
 	for plain in plainstring:
 		cipher, R = _encryptChar(plain, R)
 		cipherList.append(cipher)
-	cipherstring = strjoin(cipherList)
+	cipherstring = bytesjoin(cipherList)
 	return cipherstring, int(R)
 
 
@@ -41,15 +57,11 @@ def hexString(s):
 
 def deHexString(h):
 	import binascii
-	h = strjoin(h.split())
+	h = bytesjoin(h.split())
 	return binascii.unhexlify(h)
 
 
-def _test():
-	testStr = "\0\0asdadads asds\265"
-	print(decrypt, decrypt(testStr, 12321))
-	print(encrypt, encrypt(testStr, 12321))
-
-
 if __name__ == "__main__":
-	_test()
+	import sys
+	import doctest
+	sys.exit(doctest.testmod().failed)

Lib/fontTools/misc/psLib.py

@@ -7,15 +7,15 @@ import collections
 from string import whitespace
 
 
-ps_special = '()<>[]{}%'	# / is one too, but we take care of that one differently
+ps_special = b'()<>[]{}%'	# / is one too, but we take care of that one differently
 
-skipwhiteRE = re.compile("[%s]*" % whitespace)
-endofthingPat = "[^][(){}<>/%%%s]*" % whitespace
+skipwhiteRE = re.compile(bytesjoin([b"[", whitespace, b"]*"]))
+endofthingPat = bytesjoin([b"[^][(){}<>/%", whitespace, b"]*"])
 endofthingRE = re.compile(endofthingPat)
-commentRE = re.compile("%[^\n\r]*")
+commentRE = re.compile(b"%[^\n\r]*")
 
 # XXX This not entirely correct as it doesn't allow *nested* embedded parens:
-stringPat = r"""
+stringPat = br"""
 	\(
 		(
 			(
@@ -29,22 +29,44 @@ stringPat = r"""
 		[^()]*
 	\)
 """
-stringPat = "".join(stringPat.split())
+stringPat = b"".join(stringPat.split())
 stringRE = re.compile(stringPat)
 
-hexstringRE = re.compile("<[%s0-9A-Fa-f]*>" % whitespace)
+hexstringRE = re.compile(bytesjoin([b"<[", whitespace, b"0-9A-Fa-f]*>"]))
 
 class PSTokenError(Exception): pass
 class PSError(Exception): pass
 
 
-# StringIO.StringIO is only available in Python 2. The PSTokenizer class
-# attemps to access private attributes from the latter which are not available
-# in other file(-like) objects. Therefore, io.BytesIO (or io.StringIO) can't
-# be used as ready replacements. For now we must drop Python3 support in psLib,
-# and consequently in t1Lib, until we rewrite them for py23.
-# See: https://github.com/behdad/fonttools/issues/391
-class PSTokenizer(StringIO):
+class PSTokenizer(object):
+
+	def __init__(self, buf=b''):
+		# Force self.buf to be a byte string
+		buf = tobytes(buf)
+		self.buf = buf
+		self.len = len(buf)
+		self.pos = 0
+		self.closed = False
+
+	def read(self, n=-1):
+		"""Read at most 'n' bytes from the buffer, or less if the read
+		hits EOF before obtaining 'n' bytes.
+		If 'n' is negative or omitted, read all data until EOF is reached.
+		"""
+		if self.closed:
+			raise ValueError("I/O operation on closed file")
+		if n is None or n < 0:
+			newpos = self.len
+		else:
+			newpos = min(self.pos+n, self.len)
+		r = self.buf[self.pos:newpos]
+		self.pos = newpos
+		return r
+
+	def close(self):
+		if not self.closed:
+			self.closed = True
+			del self.buf, self.pos
 
 	def getnexttoken(self,
 			# localize some stuff, for performance
@@ -53,32 +75,30 @@ class PSTokenizer(StringIO):
 			stringmatch=stringRE.match,
 			hexstringmatch=hexstringRE.match,
 			commentmatch=commentRE.match,
-			endmatch=endofthingRE.match,
-			whitematch=skipwhiteRE.match):
+			endmatch=endofthingRE.match):
 
-		_, nextpos = whitematch(self.buf, self.pos).span()
-		self.pos = nextpos
+		self.skipwhite()
 		if self.pos >= self.len:
 			return None, None
 		pos = self.pos
 		buf = self.buf
-		char = buf[pos]
+		char = bytechr(byteord(buf[pos]))
 		if char in ps_special:
-			if char in '{}[]':
+			if char in b'{}[]':
 				tokentype = 'do_special'
 				token = char
-			elif char == '%':
+			elif char == b'%':
 				tokentype = 'do_comment'
 				_, nextpos = commentmatch(buf, pos).span()
 				token = buf[pos:nextpos]
-			elif char == '(':
+			elif char == b'(':
 				tokentype = 'do_string'
 				m = stringmatch(buf, pos)
 				if m is None:
 					raise PSTokenError('bad string at character %d' % pos)
 				_, nextpos = m.span()
 				token = buf[pos:nextpos]
-			elif char == '<':
+			elif char == b'<':
 				tokentype = 'do_hexstring'
 				m = hexstringmatch(buf, pos)
 				if m is None:
@@ -88,7 +108,7 @@ class PSTokenizer(StringIO):
 			else:
 				raise PSTokenError('bad token at character %d' % pos)
 		else:
-			if char == '/':
+			if char == b'/':
 				tokentype = 'do_literal'
 				m = endmatch(buf, pos+1)
 			else:
@@ -99,6 +119,7 @@ class PSTokenizer(StringIO):
 			_, nextpos = m.span()
 			token = buf[pos:nextpos]
 		self.pos = pos + len(token)
+		token = tostr(token, encoding='ascii')
 		return tokentype, token
 
 	def skipwhite(self, whitematch=skipwhiteRE.match):
@@ -107,7 +128,6 @@ class PSTokenizer(StringIO):
 
 	def starteexec(self):
 		self.pos = self.pos + 1
-		#self.skipwhite()
 		self.dirtybuf = self.buf[self.pos:]
 		self.buf, R = eexec.decrypt(self.dirtybuf, 55665)
 		self.len = len(self.buf)
@@ -119,11 +139,6 @@ class PSTokenizer(StringIO):
 		self.buf = self.dirtybuf
 		del self.dirtybuf
 
-	def flush(self):
-		if self.buflist:
-			self.buf = self.buf + "".join(self.buflist)
-			self.buflist = []
-
 
 class PSInterpreter(PSOperators):
 
@@ -163,7 +178,6 @@ class PSInterpreter(PSOperators):
 		try:
 			while 1:
 				tokentype, token = getnexttoken()
-				#print token
 				if not token:
 					break
 				if tokentype:
@@ -345,12 +359,3 @@ def suckfont(data):
 		rawfont = fontdir[fontNames[0]]
 	interpreter.close()
 	return unpack_item(rawfont)
-
-
-if __name__ == "__main__":
-	import EasyDialogs
-	path = EasyDialogs.AskFileForOpen()
-	if path:
-		from fontTools import t1Lib
-		data, kind = t1Lib.read(path)
-		font = suckfont(data)

Lib/fontTools/t1Lib/__init__.py

@@ -13,7 +13,7 @@ write(path, data, kind='OTHER', dohex=False)
 	'kind' can be one of 'LWFN', 'PFB' or 'OTHER'; it defaults to 'OTHER'.
 	'dohex' is a flag which determines whether the eexec encrypted
 	part should be written as hexadecimal or binary, but only if kind
-	is 'LWFN' or 'PFB'.
+	is 'OTHER'.
 """
 from __future__ import print_function, division, absolute_import
 from fontTools.misc.py23 import *
@@ -36,7 +36,6 @@ except ImportError:
 	haveMacSupport = 0
 else:
 	haveMacSupport = 1
-	import MacOS
 
 
 class T1Error(Exception): pass
@@ -56,8 +55,8 @@ class T1Font(object):
 		else:
 			pass # XXX
 
-	def saveAs(self, path, type):
-		write(path, self.getData(), type)
+	def saveAs(self, path, type, dohex=False):
+		write(path, self.getData(), type, dohex)
 
 	def getData(self):
 		# XXX Todo: if the data has been converted to Python object,
@@ -266,7 +265,7 @@ def writeOther(path, data, dohex=False):
 			if code == 2 and dohex:
 				while chunk:
 					f.write(eexec.hexString(chunk[:hexlinelen]))
-					f.write('\r')
+					f.write(b'\r')
 					chunk = chunk[hexlinelen:]
 			else:
 				f.write(chunk)
@@ -276,13 +275,13 @@ def writeOther(path, data, dohex=False):
 
 # decryption tools
 
-EEXECBEGIN = "currentfile eexec"
-EEXECEND = '0' * 64
-EEXECINTERNALEND = "currentfile closefile"
-EEXECBEGINMARKER = "%-- eexec start\r"
-EEXECENDMARKER = "%-- eexec end\r"
+EEXECBEGIN = b"currentfile eexec"
+EEXECEND = b'0' * 64
+EEXECINTERNALEND = b"currentfile closefile"
+EEXECBEGINMARKER = b"%-- eexec start\r"
+EEXECENDMARKER = b"%-- eexec end\r"
 
-_ishexRE = re.compile('[0-9A-Fa-f]*$')
+_ishexRE = re.compile(b'[0-9A-Fa-f]*$')
 
 def isHex(text):
 	return _ishexRE.match(text) is not None
@@ -300,7 +299,7 @@ def decryptType1(data):
 			if decrypted[-len(EEXECINTERNALEND)-1:-1] != EEXECINTERNALEND \
 					and decrypted[-len(EEXECINTERNALEND)-2:-2] != EEXECINTERNALEND:
 				raise T1Error("invalid end of eexec part")
-			decrypted = decrypted[:-len(EEXECINTERNALEND)-2] + '\r'
+			decrypted = decrypted[:-len(EEXECINTERNALEND)-2] + b'\r'
 			data.append(EEXECBEGINMARKER + decrypted + EEXECENDMARKER)
 		else:
 			if chunk[-len(EEXECBEGIN)-1:-1] == EEXECBEGIN:
@@ -333,7 +332,7 @@ def findEncryptedChunks(data):
 	return chunks
 
 def deHexString(hexstring):
-	return eexec.deHexString(strjoin(hexstring.split()))
+	return eexec.deHexString(bytesjoin(hexstring.split()))
 
 
 # Type 1 assertion
@@ -357,7 +356,7 @@ def assertType1(data):
 # pfb helpers
 
 def longToString(long):
-	s = ""
+	s = b""
 	for i in range(4):
 		s += bytechr((long & (0xff << (i * 8))) >> i * 8)
 	return s

Lib/fontTools/t1Lib/t1Lib_test.py

@@ -0,0 +1,86 @@
+from __future__ import print_function, division, absolute_import
+from fontTools.misc.py23 import *
+import unittest
+import os
+from fontTools import t1Lib
+from fontTools.pens.basePen import NullPen
+import random
+
+
+CWD = os.path.abspath(os.path.dirname(__file__))
+DATADIR = os.path.join(CWD, 'testdata')
+PFA = os.path.join(DATADIR, 'TestT1-Regular.pfa')
+PFB = os.path.join(DATADIR, 'TestT1-Regular.pfb')
+
+
+class FindEncryptedChunksTest(unittest.TestCase):
+
+	def test_findEncryptedChunks(self):
+		with open(PFA, "rb") as f:
+			data = f.read()
+		chunks = t1Lib.findEncryptedChunks(data)
+		self.assertEqual(len(chunks), 3)
+		self.assertFalse(chunks[0][0])
+		# the second chunk is encrypted
+		self.assertTrue(chunks[1][0])
+		self.assertFalse(chunks[2][0])
+
+
+class DecryptType1Test(unittest.TestCase):
+
+	def test_decryptType1(self):
+		with open(PFA, "rb") as f:
+			data = f.read()
+		decrypted = t1Lib.decryptType1(data)
+		self.assertNotEqual(decrypted, data)
+
+
+class ReadWriteTest(unittest.TestCase):
+
+	def test_read_pfa_write_pfb(self):
+		font = t1Lib.T1Font(PFA)
+		data = self.write(font, 'PFB')
+		self.assertEqual(font.getData(), data)
+
+	def test_read_pfb_write_pfa(self):
+		font = t1Lib.T1Font(PFB)
+		# 'OTHER' == 'PFA'
+		data = self.write(font, 'OTHER', dohex=True)
+		self.assertEqual(font.getData(), data)
+
+	@staticmethod
+	def write(font, outtype, dohex=False):
+		temp = os.path.join(DATADIR, 'temp.' + outtype.lower())
+		try:
+			font.saveAs(temp, outtype, dohex=dohex)
+			newfont = t1Lib.T1Font(temp)
+			data = newfont.getData()
+		finally:
+			if os.path.exists(temp):
+				os.remove(temp)
+		return data
+
+
+class T1FontTest(unittest.TestCase):
+
+	def test_parse_pfa(self):
+		font = t1Lib.T1Font(PFA)
+		font.parse()
+		self.assertEqual(font['FontName'], 'TestT1-Regular')
+		self.assertTrue('Subrs' in font['Private'])
+
+	def test_parse_pfb(self):
+		font = t1Lib.T1Font(PFB)
+		font.parse()
+		self.assertEqual(font['FontName'], 'TestT1-Regular')
+		self.assertTrue('Subrs' in font['Private'])
+
+	def test_getGlyphSet(self):
+		font = t1Lib.T1Font(PFA)
+		glyphs = font.getGlyphSet()
+		i = random.randrange(len(glyphs))
+		aglyph = list(glyphs.values())[i]
+		self.assertTrue(hasattr(aglyph, 'draw'))
+		self.assertFalse(hasattr(aglyph, 'width'))
+		aglyph.draw(NullPen())
+		self.assertTrue(hasattr(aglyph, 'width'))

Lib/fontTools/t1Lib/testdata/TestT1-Regular.pfa

@@ -0,0 +1,60 @@
+%!FontType1-1.1: TestT1-Regular 1.0
+%%BeginResource: font TestT1-Regular
+12 dict dup begin
+/FontType 1 def
+/FontName /TestT1-Regular def
+/FontInfo 14 dict dup begin
+/version (1.0) def
+/Notice (Test T1 is not a trademark of FontTools.) def
+/Copyright (Copyright c 2015 by FontTools. No rights reserved.) def
+/FullName (Test T1) def
+/FamilyName (Test T1) def
+/Weight (Regular) def
+/ItalicAngle 0.000000 def
+/isFixedPitch false def
+/UnderlinePosition -75.000000 def
+/UnderlineThickness 50.000000 def
+/FSType 0 def
+end def
+/PaintType 0 def
+/FontMatrix [0.001 0 0 0.001 0 0] def
+/Encoding 256 array
+0 1 255 {1 index exch /.notdef put} for
+def
+/FontBBox {50.000000 0.000000 668.000000 750.000000} def
+end
+currentfile eexec bab431ea06bb0a1031e1aa11919e714ac1ac5197cb08b39a4d7e746fca0af12d89ac0ebd1bc11ab1
+b3887b922efcec739534242d2fd22e7c30e3edce24b93798627e1ac3387816a8c4b84d76047dada8
+28b2ad27c5603046fecbc2a97adc5c37a68912324d2d435f2ee0ccc38df10ba1271a1c9af8897a6d
+6e425cd7d18fd6bd64c2adadb74365bc101a850841669886291e158cbfa7f204b3fe0ba49ffe0c80
+4f6795d32eb770c5fcd38a3879c06a4bb87b2d3ab100d8c2b5f89e9be99248575575025c66381446
+e4d9183674880aef57fb2032a1e00431133b16f6d758de7c3d0c48a0fada1d40034742a69fb3a6f9
+450d2251e659158a04697cbfa70907346d27d37ef683284385c44a1b5089bd29b4629b6483122dc8
+cbce7327bdc33dd30e6fcdb346c0ddaf433a5ac740423aa35639b2386673832f5ae8cc380e9703ba
+d3369533bfa85af9f56a090c9d97f5fc26ed102c07b647137e83632be51a65a532bd26430b59a31c
+3cb037ded351c1d4e944733feb30a3e6f81c1a7b74ac4e0eadbe705412d47991c246e8820876bbc6
+1f6a3e264ae6b2ad4b864b0d7abee289308bea26eb15d00d2b9103861386e0a5f1802ba06f916810
+62110d2b1c3641806f78eea365614f440b580185e84bac6f87bee36108d95174c786600cf0e9dc4a
+5545d1a84cfe8392115c0b7027c17fd460481d21f684af32204085690946327bfded992852645149
+8d44150d2495bd2efe0db6a450c6e28d0a52ca234e252129d5095596b0d8de096682d2eb00bc8320
+f257fd653b05a22eab7a193ccc315a6ee274a03ff1fdf443b310157a02656ca4b06c581dca8ced72
+c6ddcab26eb856ad1093452c587438b7f8408c1311e19254955914612c09828fd4d4fc2b8b0406ea
+2ee38348a8bdab88a77b8033366b2e469834c01b7bd73207b7c67756937c7a9232947fde2e0ea327
+7b7d610e601b91389ccbcdd813c87db5333c0c723e48d3ef69285f246327978ce68ae9081076a227
+1a962a2a10e2b1147ec40b0f6553a00c8b329118569d16fe04a4fa195caf1b04c52c9a562b72e0cd
+e411d747af796b9d2fb086ed927efb0e5fc9f50aa18aaf4949cba0de0805210620a19eec4319dfef
+a74d9d13d16f8ad793323a231347e6b40022a1100c1e064b8679c1da63a26dfb217a6037096ad796
+320da5a9d0526eed51d7d64d3223e285c1a8c70780c59ecc9dd9bc90a0f84ffa038834918cebe247
+f6e8fa4ca0654019196388f2df008e63bc32c8e5e686dbb69193b7749638c22b389fb1f090fbb007
+fdb8a6ee4e4b29e123fe1652fe72239bd2c8
+0000000000000000000000000000000000000000000000000000000000000000
+0000000000000000000000000000000000000000000000000000000000000000
+0000000000000000000000000000000000000000000000000000000000000000
+0000000000000000000000000000000000000000000000000000000000000000
+0000000000000000000000000000000000000000000000000000000000000000
+0000000000000000000000000000000000000000000000000000000000000000
+0000000000000000000000000000000000000000000000000000000000000000
+0000000000000000000000000000000000000000000000000000000000000000
+cleartomark
+%%EndResource
+%%EOF