Python code coverage for Lib/lib2to3/pgen2/tokenize.py

#	count	content
1	n/a	# Copyright (c) 2001, 2002, 2003, 2004, 2005, 2006 Python Software Foundation.
2	n/a	# All rights reserved.
3	n/a
4	n/a	"""Tokenization help for Python programs.
5	n/a
6	n/a	generate_tokens(readline) is a generator that breaks a stream of
7	n/a	text into Python tokens. It accepts a readline-like method which is called
8	n/a	repeatedly to get the next line of input (or "" for EOF). It generates
9	n/a	5-tuples with these members:
10	n/a
11	n/a	the token type (see token.py)
12	n/a	the token (a string)
13	n/a	the starting (row, column) indices of the token (a 2-tuple of ints)
14	n/a	the ending (row, column) indices of the token (a 2-tuple of ints)
15	n/a	the original line (string)
16	n/a
17	n/a	It is designed to match the working of the Python tokenizer exactly, except
18	n/a	that it produces COMMENT tokens for comments and gives type OP for all
19	n/a	operators
20	n/a
21	n/a	Older entry points
22	n/a	tokenize_loop(readline, tokeneater)
23	n/a	tokenize(readline, tokeneater=printtoken)
24	n/a	are the same, except instead of generating tokens, tokeneater is a callback
25	n/a	function to which the 5 fields described above are passed as 5 arguments,
26	n/a	each time a new token is found."""
27	n/a
28	n/a	__author__ = 'Ka-Ping Yee <ping@lfw.org>'
29	n/a	__credits__ = \
30	n/a	'GvR, ESR, Tim Peters, Thomas Wouters, Fred Drake, Skip Montanaro'
31	n/a
32	n/a	import string, re
33	n/a	from codecs import BOM_UTF8, lookup
34	n/a	from lib2to3.pgen2.token import *
35	n/a
36	n/a	from . import token
37	n/a	__all__ = [x for x in dir(token) if x[0] != '_'] + ["tokenize",
38	n/a	"generate_tokens", "untokenize"]
39	n/a	del token
40	n/a
41	n/a	try:
42	n/a	bytes
43	n/a	except NameError:
44	n/a	# Support bytes type in Python <= 2.5, so 2to3 turns itself into
45	n/a	# valid Python 3 code.
46	n/a	bytes = str
47	n/a
48	n/a	def group(*choices): return '(' + '\|'.join(choices) + ')'
49	n/a	def any(choices): return group(choices) + '*'
50	n/a	def maybe(choices): return group(choices) + '?'
51	n/a
52	n/a	Whitespace = r'[ \f\t]*'
53	n/a	Comment = r'#[^\r\n]*'
54	n/a	Ignore = Whitespace + any(r'\\\r?\n' + Whitespace) + maybe(Comment)
55	n/a	Name = r'[a-zA-Z_]\w*'
56	n/a
57	n/a	Binnumber = r'0[bB][01]*'
58	n/a	Hexnumber = r'0[xX][\da-fA-F]*[lL]?'
59	n/a	Octnumber = r'0[oO]?[0-7]*[lL]?'
60	n/a	Decnumber = r'[1-9]\d*[lL]?'
61	n/a	Intnumber = group(Binnumber, Hexnumber, Octnumber, Decnumber)
62	n/a	Exponent = r'[eE][-+]?\d+'
63	n/a	Pointfloat = group(r'\d+\.\d*', r'\.\d+') + maybe(Exponent)
64	n/a	Expfloat = r'\d+' + Exponent
65	n/a	Floatnumber = group(Pointfloat, Expfloat)
66	n/a	Imagnumber = group(r'\d+[jJ]', Floatnumber + r'[jJ]')
67	n/a	Number = group(Imagnumber, Floatnumber, Intnumber)
68	n/a
69	n/a	# Tail end of ' string.
70	n/a	Single = r"[^'\\](?:\\.[^'\\])*'"
71	n/a	# Tail end of " string.
72	n/a	Double = r'[^"\\](?:\\.[^"\\])*"'
73	n/a	# Tail end of ''' string.
74	n/a	Single3 = r"[^'\\](?:(?:\\.\|'(?!''))[^'\\])*'''"
75	n/a	# Tail end of """ string.
76	n/a	Double3 = r'[^"\\](?:(?:\\.\|"(?!""))[^"\\])*"""'
77	n/a	Triple = group("[ubUB]?[rR]?'''", '[ubUB]?[rR]?"""')
78	n/a	# Single-line ' or " string.
79	n/a	String = group(r"[uU]?[rR]?'[^\n'\\](?:\\.[^\n'\\])*'",
80	n/a	r'[uU]?[rR]?"[^\n"\\](?:\\.[^\n"\\])*"')
81	n/a
82	n/a	# Because of leftmost-then-longest match semantics, be sure to put the
83	n/a	# longest operators first (e.g., if = came before ==, == would get
84	n/a	# recognized as two instances of =).
85	n/a	Operator = group(r"\\=?", r">>=?", r"<<=?", r"<>", r"!=",
86	n/a	r"//=?", r"->",
87	n/a	r"[+\-*/%&@\|^=<>]=?",
88	n/a	r"~")
89	n/a
90	n/a	Bracket = '[][(){}]'
91	n/a	Special = group(r'\r?\n', r'[:;.,`@]')
92	n/a	Funny = group(Operator, Bracket, Special)
93	n/a
94	n/a	PlainToken = group(Number, Funny, String, Name)
95	n/a	Token = Ignore + PlainToken
96	n/a
97	n/a	# First (or only) line of ' or " string.
98	n/a	ContStr = group(r"[uUbB]?[rR]?'[^\n'\\](?:\\.[^\n'\\])*" +
99	n/a	group("'", r'\\\r?\n'),
100	n/a	r'[uUbB]?[rR]?"[^\n"\\](?:\\.[^\n"\\])*' +
101	n/a	group('"', r'\\\r?\n'))
102	n/a	PseudoExtras = group(r'\\\r?\n', Comment, Triple)
103	n/a	PseudoToken = Whitespace + group(PseudoExtras, Number, Funny, ContStr, Name)
104	n/a
105	n/a	tokenprog, pseudoprog, single3prog, double3prog = list(map(
106	n/a	re.compile, (Token, PseudoToken, Single3, Double3)))
107	n/a	endprogs = {"'": re.compile(Single), '"': re.compile(Double),
108	n/a	"'''": single3prog, '"""': double3prog,
109	n/a	"r'''": single3prog, 'r"""': double3prog,
110	n/a	"u'''": single3prog, 'u"""': double3prog,
111	n/a	"b'''": single3prog, 'b"""': double3prog,
112	n/a	"ur'''": single3prog, 'ur"""': double3prog,
113	n/a	"br'''": single3prog, 'br"""': double3prog,
114	n/a	"R'''": single3prog, 'R"""': double3prog,
115	n/a	"U'''": single3prog, 'U"""': double3prog,
116	n/a	"B'''": single3prog, 'B"""': double3prog,
117	n/a	"uR'''": single3prog, 'uR"""': double3prog,
118	n/a	"Ur'''": single3prog, 'Ur"""': double3prog,
119	n/a	"UR'''": single3prog, 'UR"""': double3prog,
120	n/a	"bR'''": single3prog, 'bR"""': double3prog,
121	n/a	"Br'''": single3prog, 'Br"""': double3prog,
122	n/a	"BR'''": single3prog, 'BR"""': double3prog,
123	n/a	'r': None, 'R': None,
124	n/a	'u': None, 'U': None,
125	n/a	'b': None, 'B': None}
126	n/a
127	n/a	triple_quoted = {}
128	n/a	for t in ("'''", '"""',
129	n/a	"r'''", 'r"""', "R'''", 'R"""',
130	n/a	"u'''", 'u"""', "U'''", 'U"""',
131	n/a	"b'''", 'b"""', "B'''", 'B"""',
132	n/a	"ur'''", 'ur"""', "Ur'''", 'Ur"""',
133	n/a	"uR'''", 'uR"""', "UR'''", 'UR"""',
134	n/a	"br'''", 'br"""', "Br'''", 'Br"""',
135	n/a	"bR'''", 'bR"""', "BR'''", 'BR"""',):
136	n/a	triple_quoted[t] = t
137	n/a	single_quoted = {}
138	n/a	for t in ("'", '"',
139	n/a	"r'", 'r"', "R'", 'R"',
140	n/a	"u'", 'u"', "U'", 'U"',
141	n/a	"b'", 'b"', "B'", 'B"',
142	n/a	"ur'", 'ur"', "Ur'", 'Ur"',
143	n/a	"uR'", 'uR"', "UR'", 'UR"',
144	n/a	"br'", 'br"', "Br'", 'Br"',
145	n/a	"bR'", 'bR"', "BR'", 'BR"', ):
146	n/a	single_quoted[t] = t
147	n/a
148	n/a	tabsize = 8
149	n/a
150	n/a	class TokenError(Exception): pass
151	n/a
152	n/a	class StopTokenizing(Exception): pass
153	n/a
154	n/a	def printtoken(type, token, xxx_todo_changeme, xxx_todo_changeme1, line): # for testing
155	n/a	(srow, scol) = xxx_todo_changeme
156	n/a	(erow, ecol) = xxx_todo_changeme1
157	n/a	print("%d,%d-%d,%d:\t%s\t%s" % \
158	n/a	(srow, scol, erow, ecol, tok_name[type], repr(token)))
159	n/a
160	n/a	def tokenize(readline, tokeneater=printtoken):
161	n/a	"""
162	n/a	The tokenize() function accepts two parameters: one representing the
163	n/a	input stream, and one providing an output mechanism for tokenize().
164	n/a
165	n/a	The first parameter, readline, must be a callable object which provides
166	n/a	the same interface as the readline() method of built-in file objects.
167	n/a	Each call to the function should return one line of input as a string.
168	n/a
169	n/a	The second parameter, tokeneater, must also be a callable object. It is
170	n/a	called once for each token, with five arguments, corresponding to the
171	n/a	tuples generated by generate_tokens().
172	n/a	"""
173	n/a	try:
174	n/a	tokenize_loop(readline, tokeneater)
175	n/a	except StopTokenizing:
176	n/a	pass
177	n/a
178	n/a	# backwards compatible interface
179	n/a	def tokenize_loop(readline, tokeneater):
180	n/a	for token_info in generate_tokens(readline):
181	n/a	tokeneater(*token_info)
182	n/a
183	n/a	class Untokenizer:
184	n/a
185	n/a	def __init__(self):
186	n/a	self.tokens = []
187	n/a	self.prev_row = 1
188	n/a	self.prev_col = 0
189	n/a
190	n/a	def add_whitespace(self, start):
191	n/a	row, col = start
192	n/a	assert row <= self.prev_row
193	n/a	col_offset = col - self.prev_col
194	n/a	if col_offset:
195	n/a	self.tokens.append(" " * col_offset)
196	n/a
197	n/a	def untokenize(self, iterable):
198	n/a	for t in iterable:
199	n/a	if len(t) == 2:
200	n/a	self.compat(t, iterable)
201	n/a	break
202	n/a	tok_type, token, start, end, line = t
203	n/a	self.add_whitespace(start)
204	n/a	self.tokens.append(token)
205	n/a	self.prev_row, self.prev_col = end
206	n/a	if tok_type in (NEWLINE, NL):
207	n/a	self.prev_row += 1
208	n/a	self.prev_col = 0
209	n/a	return "".join(self.tokens)
210	n/a
211	n/a	def compat(self, token, iterable):
212	n/a	startline = False
213	n/a	indents = []
214	n/a	toks_append = self.tokens.append
215	n/a	toknum, tokval = token
216	n/a	if toknum in (NAME, NUMBER):
217	n/a	tokval += ' '
218	n/a	if toknum in (NEWLINE, NL):
219	n/a	startline = True
220	n/a	for tok in iterable:
221	n/a	toknum, tokval = tok[:2]
222	n/a
223	n/a	if toknum in (NAME, NUMBER, ASYNC, AWAIT):
224	n/a	tokval += ' '
225	n/a
226	n/a	if toknum == INDENT:
227	n/a	indents.append(tokval)
228	n/a	continue
229	n/a	elif toknum == DEDENT:
230	n/a	indents.pop()
231	n/a	continue
232	n/a	elif toknum in (NEWLINE, NL):
233	n/a	startline = True
234	n/a	elif startline and indents:
235	n/a	toks_append(indents[-1])
236	n/a	startline = False
237	n/a	toks_append(tokval)
238	n/a
239	n/a	cookie_re = re.compile(r'^[ \t\f]#.?coding[:=][ \t]*([-\w.]+)', re.ASCII)
240	n/a	blank_re = re.compile(br'^[ \t\f]*(?:[#\r\n]\|$)', re.ASCII)
241	n/a
242	n/a	def _get_normal_name(orig_enc):
243	n/a	"""Imitates get_normal_name in tokenizer.c."""
244	n/a	# Only care about the first 12 characters.
245	n/a	enc = orig_enc[:12].lower().replace("_", "-")
246	n/a	if enc == "utf-8" or enc.startswith("utf-8-"):
247	n/a	return "utf-8"
248	n/a	if enc in ("latin-1", "iso-8859-1", "iso-latin-1") or \
249	n/a	enc.startswith(("latin-1-", "iso-8859-1-", "iso-latin-1-")):
250	n/a	return "iso-8859-1"
251	n/a	return orig_enc
252	n/a
253	n/a	def detect_encoding(readline):
254	n/a	"""
255	n/a	The detect_encoding() function is used to detect the encoding that should
256	n/a	be used to decode a Python source file. It requires one argument, readline,
257	n/a	in the same way as the tokenize() generator.
258	n/a
259	n/a	It will call readline a maximum of twice, and return the encoding used
260	n/a	(as a string) and a list of any lines (left as bytes) it has read
261	n/a	in.
262	n/a
263	n/a	It detects the encoding from the presence of a utf-8 bom or an encoding
264	n/a	cookie as specified in pep-0263. If both a bom and a cookie are present, but
265	n/a	disagree, a SyntaxError will be raised. If the encoding cookie is an invalid
266	n/a	charset, raise a SyntaxError. Note that if a utf-8 bom is found,
267	n/a	'utf-8-sig' is returned.
268	n/a
269	n/a	If no encoding is specified, then the default of 'utf-8' will be returned.
270	n/a	"""
271	n/a	bom_found = False
272	n/a	encoding = None
273	n/a	default = 'utf-8'
274	n/a	def read_or_stop():
275	n/a	try:
276	n/a	return readline()
277	n/a	except StopIteration:
278	n/a	return bytes()
279	n/a
280	n/a	def find_cookie(line):
281	n/a	try:
282	n/a	line_string = line.decode('ascii')
283	n/a	except UnicodeDecodeError:
284	n/a	return None
285	n/a	match = cookie_re.match(line_string)
286	n/a	if not match:
287	n/a	return None
288	n/a	encoding = _get_normal_name(match.group(1))
289	n/a	try:
290	n/a	codec = lookup(encoding)
291	n/a	except LookupError:
292	n/a	# This behaviour mimics the Python interpreter
293	n/a	raise SyntaxError("unknown encoding: " + encoding)
294	n/a
295	n/a	if bom_found:
296	n/a	if codec.name != 'utf-8':
297	n/a	# This behaviour mimics the Python interpreter
298	n/a	raise SyntaxError('encoding problem: utf-8')
299	n/a	encoding += '-sig'
300	n/a	return encoding
301	n/a
302	n/a	first = read_or_stop()
303	n/a	if first.startswith(BOM_UTF8):
304	n/a	bom_found = True
305	n/a	first = first[3:]
306	n/a	default = 'utf-8-sig'
307	n/a	if not first:
308	n/a	return default, []
309	n/a
310	n/a	encoding = find_cookie(first)
311	n/a	if encoding:
312	n/a	return encoding, [first]
313	n/a	if not blank_re.match(first):
314	n/a	return default, [first]
315	n/a
316	n/a	second = read_or_stop()
317	n/a	if not second:
318	n/a	return default, [first]
319	n/a
320	n/a	encoding = find_cookie(second)
321	n/a	if encoding:
322	n/a	return encoding, [first, second]
323	n/a
324	n/a	return default, [first, second]
325	n/a
326	n/a	def untokenize(iterable):
327	n/a	"""Transform tokens back into Python source code.
328	n/a
329	n/a	Each element returned by the iterable must be a token sequence
330	n/a	with at least two elements, a token number and token value. If
331	n/a	only two tokens are passed, the resulting output is poor.
332	n/a
333	n/a	Round-trip invariant for full input:
334	n/a	Untokenized source will match input source exactly
335	n/a
336	n/a	Round-trip invariant for limited intput:
337	n/a	# Output text will tokenize the back to the input
338	n/a	t1 = [tok[:2] for tok in generate_tokens(f.readline)]
339	n/a	newcode = untokenize(t1)
340	n/a	readline = iter(newcode.splitlines(1)).next
341	n/a	t2 = [tok[:2] for tokin generate_tokens(readline)]
342	n/a	assert t1 == t2
343	n/a	"""
344	n/a	ut = Untokenizer()
345	n/a	return ut.untokenize(iterable)
346	n/a
347	n/a	def generate_tokens(readline):
348	n/a	"""
349	n/a	The generate_tokens() generator requires one argument, readline, which
350	n/a	must be a callable object which provides the same interface as the
351	n/a	readline() method of built-in file objects. Each call to the function
352	n/a	should return one line of input as a string. Alternately, readline
353	n/a	can be a callable function terminating with StopIteration:
354	n/a	readline = open(myfile).next # Example of alternate readline
355	n/a
356	n/a	The generator produces 5-tuples with these members: the token type; the
357	n/a	token string; a 2-tuple (srow, scol) of ints specifying the row and
358	n/a	column where the token begins in the source; a 2-tuple (erow, ecol) of
359	n/a	ints specifying the row and column where the token ends in the source;
360	n/a	and the line on which the token was found. The line passed is the
361	n/a	logical line; continuation lines are included.
362	n/a	"""
363	n/a	lnum = parenlev = continued = 0
364	n/a	namechars, numchars = string.ascii_letters + '_', '0123456789'
365	n/a	contstr, needcont = '', 0
366	n/a	contline = None
367	n/a	indents = [0]
368	n/a
369	n/a	# 'stashed' and 'async_*' are used for async/await parsing
370	n/a	stashed = None
371	n/a	async_def = False
372	n/a	async_def_indent = 0
373	n/a	async_def_nl = False
374	n/a
375	n/a	while 1: # loop over lines in stream
376	n/a	try:
377	n/a	line = readline()
378	n/a	except StopIteration:
379	n/a	line = ''
380	n/a	lnum = lnum + 1
381	n/a	pos, max = 0, len(line)
382	n/a
383	n/a	if contstr: # continued string
384	n/a	if not line:
385	n/a	raise TokenError("EOF in multi-line string", strstart)
386	n/a	endmatch = endprog.match(line)
387	n/a	if endmatch:
388	n/a	pos = end = endmatch.end(0)
389	n/a	yield (STRING, contstr + line[:end],
390	n/a	strstart, (lnum, end), contline + line)
391	n/a	contstr, needcont = '', 0
392	n/a	contline = None
393	n/a	elif needcont and line[-2:] != '\\\n' and line[-3:] != '\\\r\n':
394	n/a	yield (ERRORTOKEN, contstr + line,
395	n/a	strstart, (lnum, len(line)), contline)
396	n/a	contstr = ''
397	n/a	contline = None
398	n/a	continue
399	n/a	else:
400	n/a	contstr = contstr + line
401	n/a	contline = contline + line
402	n/a	continue
403	n/a
404	n/a	elif parenlev == 0 and not continued: # new statement
405	n/a	if not line: break
406	n/a	column = 0
407	n/a	while pos < max: # measure leading whitespace
408	n/a	if line[pos] == ' ': column = column + 1
409	n/a	elif line[pos] == '\t': column = (column//tabsize + 1)*tabsize
410	n/a	elif line[pos] == '\f': column = 0
411	n/a	else: break
412	n/a	pos = pos + 1
413	n/a	if pos == max: break
414	n/a
415	n/a	if stashed:
416	n/a	yield stashed
417	n/a	stashed = None
418	n/a
419	n/a	if line[pos] in '#\r\n': # skip comments or blank lines
420	n/a	if line[pos] == '#':
421	n/a	comment_token = line[pos:].rstrip('\r\n')
422	n/a	nl_pos = pos + len(comment_token)
423	n/a	yield (COMMENT, comment_token,
424	n/a	(lnum, pos), (lnum, pos + len(comment_token)), line)
425	n/a	yield (NL, line[nl_pos:],
426	n/a	(lnum, nl_pos), (lnum, len(line)), line)
427	n/a	else:
428	n/a	yield ((NL, COMMENT)[line[pos] == '#'], line[pos:],
429	n/a	(lnum, pos), (lnum, len(line)), line)
430	n/a	continue
431	n/a
432	n/a	if column > indents[-1]: # count indents or dedents
433	n/a	indents.append(column)
434	n/a	yield (INDENT, line[:pos], (lnum, 0), (lnum, pos), line)
435	n/a	while column < indents[-1]:
436	n/a	if column not in indents:
437	n/a	raise IndentationError(
438	n/a	"unindent does not match any outer indentation level",
439	n/a	("<tokenize>", lnum, pos, line))
440	n/a	indents = indents[:-1]
441	n/a
442	n/a	if async_def and async_def_indent >= indents[-1]:
443	n/a	async_def = False
444	n/a	async_def_nl = False
445	n/a	async_def_indent = 0
446	n/a
447	n/a	yield (DEDENT, '', (lnum, pos), (lnum, pos), line)
448	n/a
449	n/a	if async_def and async_def_nl and async_def_indent >= indents[-1]:
450	n/a	async_def = False
451	n/a	async_def_nl = False
452	n/a	async_def_indent = 0
453	n/a
454	n/a	else: # continued statement
455	n/a	if not line:
456	n/a	raise TokenError("EOF in multi-line statement", (lnum, 0))
457	n/a	continued = 0
458	n/a
459	n/a	while pos < max:
460	n/a	pseudomatch = pseudoprog.match(line, pos)
461	n/a	if pseudomatch: # scan for tokens
462	n/a	start, end = pseudomatch.span(1)
463	n/a	spos, epos, pos = (lnum, start), (lnum, end), end
464	n/a	token, initial = line[start:end], line[start]
465	n/a
466	n/a	if initial in numchars or \
467	n/a	(initial == '.' and token != '.'): # ordinary number
468	n/a	yield (NUMBER, token, spos, epos, line)
469	n/a	elif initial in '\r\n':
470	n/a	newline = NEWLINE
471	n/a	if parenlev > 0:
472	n/a	newline = NL
473	n/a	elif async_def:
474	n/a	async_def_nl = True
475	n/a	if stashed:
476	n/a	yield stashed
477	n/a	stashed = None
478	n/a	yield (newline, token, spos, epos, line)
479	n/a
480	n/a	elif initial == '#':
481	n/a	assert not token.endswith("\n")
482	n/a	if stashed:
483	n/a	yield stashed
484	n/a	stashed = None
485	n/a	yield (COMMENT, token, spos, epos, line)
486	n/a	elif token in triple_quoted:
487	n/a	endprog = endprogs[token]
488	n/a	endmatch = endprog.match(line, pos)
489	n/a	if endmatch: # all on one line
490	n/a	pos = endmatch.end(0)
491	n/a	token = line[start:pos]
492	n/a	if stashed:
493	n/a	yield stashed
494	n/a	stashed = None
495	n/a	yield (STRING, token, spos, (lnum, pos), line)
496	n/a	else:
497	n/a	strstart = (lnum, start) # multiple lines
498	n/a	contstr = line[start:]
499	n/a	contline = line
500	n/a	break
501	n/a	elif initial in single_quoted or \
502	n/a	token[:2] in single_quoted or \
503	n/a	token[:3] in single_quoted:
504	n/a	if token[-1] == '\n': # continued string
505	n/a	strstart = (lnum, start)
506	n/a	endprog = (endprogs[initial] or endprogs[token[1]] or
507	n/a	endprogs[token[2]])
508	n/a	contstr, needcont = line[start:], 1
509	n/a	contline = line
510	n/a	break
511	n/a	else: # ordinary string
512	n/a	if stashed:
513	n/a	yield stashed
514	n/a	stashed = None
515	n/a	yield (STRING, token, spos, epos, line)
516	n/a	elif initial in namechars: # ordinary name
517	n/a	if token in ('async', 'await'):
518	n/a	if async_def:
519	n/a	yield (ASYNC if token == 'async' else AWAIT,
520	n/a	token, spos, epos, line)
521	n/a	continue
522	n/a
523	n/a	tok = (NAME, token, spos, epos, line)
524	n/a	if token == 'async' and not stashed:
525	n/a	stashed = tok
526	n/a	continue
527	n/a
528	n/a	if token == 'def':
529	n/a	if (stashed
530	n/a	and stashed[0] == NAME
531	n/a	and stashed[1] == 'async'):
532	n/a
533	n/a	async_def = True
534	n/a	async_def_indent = indents[-1]
535	n/a
536	n/a	yield (ASYNC, stashed[1],
537	n/a	stashed[2], stashed[3],
538	n/a	stashed[4])
539	n/a	stashed = None
540	n/a
541	n/a	if stashed:
542	n/a	yield stashed
543	n/a	stashed = None
544	n/a
545	n/a	yield tok
546	n/a	elif initial == '\\': # continued stmt
547	n/a	# This yield is new; needed for better idempotency:
548	n/a	if stashed:
549	n/a	yield stashed
550	n/a	stashed = None
551	n/a	yield (NL, token, spos, (lnum, pos), line)
552	n/a	continued = 1
553	n/a	else:
554	n/a	if initial in '([{': parenlev = parenlev + 1
555	n/a	elif initial in ')]}': parenlev = parenlev - 1
556	n/a	if stashed:
557	n/a	yield stashed
558	n/a	stashed = None
559	n/a	yield (OP, token, spos, epos, line)
560	n/a	else:
561	n/a	yield (ERRORTOKEN, line[pos],
562	n/a	(lnum, pos), (lnum, pos+1), line)
563	n/a	pos = pos + 1
564	n/a
565	n/a	if stashed:
566	n/a	yield stashed
567	n/a	stashed = None
568	n/a
569	n/a	for indent in indents[1:]: # pop remaining indent levels
570	n/a	yield (DEDENT, '', (lnum, 0), (lnum, 0), '')
571	n/a	yield (ENDMARKER, '', (lnum, 0), (lnum, 0), '')
572	n/a
573	n/a	if __name__ == '__main__': # testing
574	n/a	import sys
575	n/a	if len(sys.argv) > 1: tokenize(open(sys.argv[1]).readline)
576	n/a	else: tokenize(sys.stdin.readline)