aboutsummaryrefslogtreecommitdiff
path: root/Lib/fontTools/misc/sstruct.py
blob: ba1f878860964e9eb2691815a0666f2a6e35c04c (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
"""sstruct.py -- SuperStruct

Higher level layer on top of the struct module, enabling to
bind names to struct elements. The interface is similar to
struct, except the objects passed and returned are not tuples
(or argument lists), but dictionaries or instances.

Just like struct, we use fmt strings to describe a data
structure, except we use one line per element. Lines are
separated by newlines or semi-colons. Each line contains
either one of the special struct characters ('@', '=', '<',
'>' or '!') or a 'name:formatchar' combo (eg. 'myFloat:f').
Repetitions, like the struct module offers them are not useful
in this context, except for fixed length strings  (eg. 'myInt:5h'
is not allowed but 'myString:5s' is). The 'x' fmt character
(pad byte) is treated as 'special', since it is by definition
anonymous. Extra whitespace is allowed everywhere.

The sstruct module offers one feature that the "normal" struct
module doesn't: support for fixed point numbers. These are spelled
as "n.mF", where n is the number of bits before the point, and m
the number of bits after the point. Fixed point numbers get
converted to floats.

pack(fmt, object):
	'object' is either a dictionary or an instance (or actually
	anything that has a __dict__ attribute). If it is a dictionary,
	its keys are used for names. If it is an instance, it's
	attributes are used to grab struct elements from. Returns
	a string containing the data.

unpack(fmt, data, object=None)
	If 'object' is omitted (or None), a new dictionary will be
	returned. If 'object' is a dictionary, it will be used to add
	struct elements to. If it is an instance (or in fact anything
	that has a __dict__ attribute), an attribute will be added for
	each struct element. In the latter two cases, 'object' itself
	is returned.

unpack2(fmt, data, object=None)
	Convenience function. Same as unpack, except data may be longer
	than needed. The returned value is a tuple: (object, leftoverdata).

calcsize(fmt)
	like struct.calcsize(), but uses our own fmt strings:
	it returns the size of the data in bytes.
"""

from fontTools.misc.py23 import tobytes, tostr
from fontTools.misc.fixedTools import fixedToFloat as fi2fl, floatToFixed as fl2fi
import struct
import re

__version__ = "1.2"
__copyright__ = "Copyright 1998, Just van Rossum <just@letterror.com>"


class Error(Exception):
	pass

def pack(fmt, obj):
	formatstring, names, fixes = getformat(fmt)
	elements = []
	if not isinstance(obj, dict):
		obj = obj.__dict__
	for name in names:
		value = obj[name]
		if name in fixes:
			# fixed point conversion
			value = fl2fi(value, fixes[name])
		elif isinstance(value, str):
			value = tobytes(value)
		elements.append(value)
	data = struct.pack(*(formatstring,) + tuple(elements))
	return data

def unpack(fmt, data, obj=None):
	if obj is None:
		obj = {}
	data = tobytes(data)
	formatstring, names, fixes = getformat(fmt)
	if isinstance(obj, dict):
		d = obj
	else:
		d = obj.__dict__
	elements = struct.unpack(formatstring, data)
	for i in range(len(names)):
		name = names[i]
		value = elements[i]
		if name in fixes:
			# fixed point conversion
			value = fi2fl(value, fixes[name])
		elif isinstance(value, bytes):
			try:
				value = tostr(value)
			except UnicodeDecodeError:
				pass
		d[name] = value
	return obj

def unpack2(fmt, data, obj=None):
	length = calcsize(fmt)
	return unpack(fmt, data[:length], obj), data[length:]

def calcsize(fmt):
	formatstring, names, fixes = getformat(fmt)
	return struct.calcsize(formatstring)


# matches "name:formatchar" (whitespace is allowed)
_elementRE = re.compile(
		r"\s*"							# whitespace
		r"([A-Za-z_][A-Za-z_0-9]*)"		# name (python identifier)
		r"\s*:\s*"						# whitespace : whitespace
		r"([cbBhHiIlLqQfd]|[0-9]+[ps]|"	# formatchar...
			r"([0-9]+)\.([0-9]+)(F))"	# ...formatchar
		r"\s*"							# whitespace
		r"(#.*)?$"						# [comment] + end of string
	)

# matches the special struct fmt chars and 'x' (pad byte)
_extraRE = re.compile(r"\s*([x@=<>!])\s*(#.*)?$")

# matches an "empty" string, possibly containing whitespace and/or a comment
_emptyRE = re.compile(r"\s*(#.*)?$")

_fixedpointmappings = {
		8: "b",
		16: "h",
		32: "l"}

_formatcache = {}

def getformat(fmt):
	fmt = tostr(fmt, encoding="ascii")
	try:
		formatstring, names, fixes = _formatcache[fmt]
	except KeyError:
		lines = re.split("[\n;]", fmt)
		formatstring = ""
		names = []
		fixes = {}
		for line in lines:
			if _emptyRE.match(line):
				continue
			m = _extraRE.match(line)
			if m:
				formatchar = m.group(1)
				if formatchar != 'x' and formatstring:
					raise Error("a special fmt char must be first")
			else:
				m = _elementRE.match(line)
				if not m:
					raise Error("syntax error in fmt: '%s'" % line)
				name = m.group(1)
				names.append(name)
				formatchar = m.group(2)
				if m.group(3):
					# fixed point
					before = int(m.group(3))
					after = int(m.group(4))
					bits = before + after
					if bits not in [8, 16, 32]:
						raise Error("fixed point must be 8, 16 or 32 bits long")
					formatchar = _fixedpointmappings[bits]
					assert m.group(5) == "F"
					fixes[name] = after
			formatstring = formatstring + formatchar
		_formatcache[fmt] = formatstring, names, fixes
	return formatstring, names, fixes

def _test():
	fmt = """
		# comments are allowed
		>  # big endian (see documentation for struct)
		# empty lines are allowed:

		ashort: h
		along: l
		abyte: b	# a byte
		achar: c
		astr: 5s
		afloat: f; adouble: d	# multiple "statements" are allowed
		afixed: 16.16F
	"""

	print('size:', calcsize(fmt))

	class foo(object):
		pass

	i = foo()

	i.ashort = 0x7fff
	i.along = 0x7fffffff
	i.abyte = 0x7f
	i.achar = "a"
	i.astr = "12345"
	i.afloat = 0.5
	i.adouble = 0.5
	i.afixed = 1.5

	data = pack(fmt, i)
	print('data:', repr(data))
	print(unpack(fmt, data))
	i2 = foo()
	unpack(fmt, data, i2)
	print(vars(i2))

if __name__ == "__main__":
	_test()