Guido van Rossum | b51eaa1 | 1997-03-07 00:21:55 +0000 | [diff] [blame] | 1 | """Tokenization help for Python programs. |
Guido van Rossum | 4d8e859 | 1992-01-01 19:34:47 +0000 | [diff] [blame] | 2 | |
Florent Xicluna | 43e4ea1 | 2010-09-03 19:54:02 +0000 | [diff] [blame] | 3 | tokenize(readline) is a generator that breaks a stream of bytes into |
| 4 | Python tokens. It decodes the bytes according to PEP-0263 for |
| 5 | determining source file encoding. |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 6 | |
Florent Xicluna | 43e4ea1 | 2010-09-03 19:54:02 +0000 | [diff] [blame] | 7 | It accepts a readline-like method which is called repeatedly to get the |
| 8 | next line of input (or b"" for EOF). It generates 5-tuples with these |
| 9 | members: |
Tim Peters | 4efb6e9 | 2001-06-29 23:51:08 +0000 | [diff] [blame] | 10 | |
| 11 | the token type (see token.py) |
| 12 | the token (a string) |
| 13 | the starting (row, column) indices of the token (a 2-tuple of ints) |
| 14 | the ending (row, column) indices of the token (a 2-tuple of ints) |
| 15 | the original line (string) |
| 16 | |
| 17 | It is designed to match the working of the Python tokenizer exactly, except |
| 18 | that it produces COMMENT tokens for comments and gives type OP for all |
Florent Xicluna | 43e4ea1 | 2010-09-03 19:54:02 +0000 | [diff] [blame] | 19 | operators. Additionally, all token lists start with an ENCODING token |
| 20 | which tells you which encoding was used to decode the bytes stream. |
| 21 | """ |
Guido van Rossum | b51eaa1 | 1997-03-07 00:21:55 +0000 | [diff] [blame] | 22 | |
Ka-Ping Yee | 244c593 | 2001-03-01 13:56:40 +0000 | [diff] [blame] | 23 | __author__ = 'Ka-Ping Yee <ping@lfw.org>' |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 24 | __credits__ = ('GvR, ESR, Tim Peters, Thomas Wouters, Fred Drake, ' |
| 25 | 'Skip Montanaro, Raymond Hettinger, Trent Nelson, ' |
| 26 | 'Michael Foord') |
Serhiy Storchaka | cf4a2f2 | 2015-03-11 17:18:03 +0200 | [diff] [blame] | 27 | from builtins import open as _builtin_open |
Benjamin Peterson | 433f32c | 2008-12-12 01:25:05 +0000 | [diff] [blame] | 28 | from codecs import lookup, BOM_UTF8 |
Raymond Hettinger | 3fb79c7 | 2010-09-09 07:15:18 +0000 | [diff] [blame] | 29 | import collections |
Victor Stinner | 58c0752 | 2010-11-09 01:08:59 +0000 | [diff] [blame] | 30 | from io import TextIOWrapper |
Terry Jan Reedy | 5b8d2c3 | 2014-02-17 23:12:16 -0500 | [diff] [blame] | 31 | from itertools import chain |
| 32 | import re |
| 33 | import sys |
| 34 | from token import * |
| 35 | |
Serhiy Storchaka | e431d3c | 2016-03-20 23:36:29 +0200 | [diff] [blame] | 36 | cookie_re = re.compile(r'^[ \t\f]*#.*?coding[:=][ \t]*([-\w.]+)', re.ASCII) |
Serhiy Storchaka | 768c16c | 2014-01-09 18:36:09 +0200 | [diff] [blame] | 37 | blank_re = re.compile(br'^[ \t\f]*(?:[#\r\n]|$)', re.ASCII) |
Guido van Rossum | 4d8e859 | 1992-01-01 19:34:47 +0000 | [diff] [blame] | 38 | |
Skip Montanaro | 40fc160 | 2001-03-01 04:27:19 +0000 | [diff] [blame] | 39 | import token |
Alexander Belopolsky | b9d10d0 | 2010-11-11 14:07:41 +0000 | [diff] [blame] | 40 | __all__ = token.__all__ + ["COMMENT", "tokenize", "detect_encoding", |
| 41 | "NL", "untokenize", "ENCODING", "TokenInfo"] |
Skip Montanaro | 40fc160 | 2001-03-01 04:27:19 +0000 | [diff] [blame] | 42 | del token |
| 43 | |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 44 | COMMENT = N_TOKENS |
| 45 | tok_name[COMMENT] = 'COMMENT' |
Guido van Rossum | a90c78b | 1998-04-03 16:05:38 +0000 | [diff] [blame] | 46 | NL = N_TOKENS + 1 |
| 47 | tok_name[NL] = 'NL' |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 48 | ENCODING = N_TOKENS + 2 |
| 49 | tok_name[ENCODING] = 'ENCODING' |
| 50 | N_TOKENS += 3 |
Meador Inge | 00c7f85 | 2012-01-19 00:44:45 -0600 | [diff] [blame] | 51 | EXACT_TOKEN_TYPES = { |
| 52 | '(': LPAR, |
| 53 | ')': RPAR, |
| 54 | '[': LSQB, |
| 55 | ']': RSQB, |
| 56 | ':': COLON, |
| 57 | ',': COMMA, |
| 58 | ';': SEMI, |
| 59 | '+': PLUS, |
| 60 | '-': MINUS, |
| 61 | '*': STAR, |
| 62 | '/': SLASH, |
| 63 | '|': VBAR, |
| 64 | '&': AMPER, |
| 65 | '<': LESS, |
| 66 | '>': GREATER, |
| 67 | '=': EQUAL, |
| 68 | '.': DOT, |
| 69 | '%': PERCENT, |
| 70 | '{': LBRACE, |
| 71 | '}': RBRACE, |
| 72 | '==': EQEQUAL, |
| 73 | '!=': NOTEQUAL, |
| 74 | '<=': LESSEQUAL, |
| 75 | '>=': GREATEREQUAL, |
| 76 | '~': TILDE, |
| 77 | '^': CIRCUMFLEX, |
| 78 | '<<': LEFTSHIFT, |
| 79 | '>>': RIGHTSHIFT, |
| 80 | '**': DOUBLESTAR, |
| 81 | '+=': PLUSEQUAL, |
| 82 | '-=': MINEQUAL, |
| 83 | '*=': STAREQUAL, |
| 84 | '/=': SLASHEQUAL, |
| 85 | '%=': PERCENTEQUAL, |
| 86 | '&=': AMPEREQUAL, |
| 87 | '|=': VBAREQUAL, |
| 88 | '^=': CIRCUMFLEXEQUAL, |
| 89 | '<<=': LEFTSHIFTEQUAL, |
| 90 | '>>=': RIGHTSHIFTEQUAL, |
| 91 | '**=': DOUBLESTAREQUAL, |
| 92 | '//': DOUBLESLASH, |
| 93 | '//=': DOUBLESLASHEQUAL, |
Benjamin Peterson | d51374e | 2014-04-09 23:55:56 -0400 | [diff] [blame] | 94 | '@': AT, |
| 95 | '@=': ATEQUAL, |
Meador Inge | 00c7f85 | 2012-01-19 00:44:45 -0600 | [diff] [blame] | 96 | } |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 97 | |
Raymond Hettinger | 3fb79c7 | 2010-09-09 07:15:18 +0000 | [diff] [blame] | 98 | class TokenInfo(collections.namedtuple('TokenInfo', 'type string start end line')): |
Raymond Hettinger | aa17a7f | 2009-04-29 14:21:25 +0000 | [diff] [blame] | 99 | def __repr__(self): |
Raymond Hettinger | a0e7940 | 2010-09-09 08:29:05 +0000 | [diff] [blame] | 100 | annotated_type = '%d (%s)' % (self.type, tok_name[self.type]) |
| 101 | return ('TokenInfo(type=%s, string=%r, start=%r, end=%r, line=%r)' % |
| 102 | self._replace(type=annotated_type)) |
Raymond Hettinger | aa17a7f | 2009-04-29 14:21:25 +0000 | [diff] [blame] | 103 | |
Meador Inge | 00c7f85 | 2012-01-19 00:44:45 -0600 | [diff] [blame] | 104 | @property |
| 105 | def exact_type(self): |
| 106 | if self.type == OP and self.string in EXACT_TOKEN_TYPES: |
| 107 | return EXACT_TOKEN_TYPES[self.string] |
| 108 | else: |
| 109 | return self.type |
| 110 | |
Eric S. Raymond | b08b2d3 | 2001-02-09 11:10:16 +0000 | [diff] [blame] | 111 | def group(*choices): return '(' + '|'.join(choices) + ')' |
Guido van Rossum | 68468eb | 2003-02-27 20:14:51 +0000 | [diff] [blame] | 112 | def any(*choices): return group(*choices) + '*' |
| 113 | def maybe(*choices): return group(*choices) + '?' |
Guido van Rossum | 4d8e859 | 1992-01-01 19:34:47 +0000 | [diff] [blame] | 114 | |
Antoine Pitrou | fd03645 | 2008-08-19 17:56:33 +0000 | [diff] [blame] | 115 | # Note: we use unicode matching for names ("\w") but ascii matching for |
| 116 | # number literals. |
Guido van Rossum | 3b63177 | 1997-10-27 20:44:15 +0000 | [diff] [blame] | 117 | Whitespace = r'[ \f\t]*' |
| 118 | Comment = r'#[^\r\n]*' |
| 119 | Ignore = Whitespace + any(r'\\\r?\n' + Whitespace) + maybe(Comment) |
Benjamin Peterson | 33856de | 2010-08-30 14:41:20 +0000 | [diff] [blame] | 120 | Name = r'\w+' |
Guido van Rossum | 4d8e859 | 1992-01-01 19:34:47 +0000 | [diff] [blame] | 121 | |
Antoine Pitrou | fd03645 | 2008-08-19 17:56:33 +0000 | [diff] [blame] | 122 | Hexnumber = r'0[xX][0-9a-fA-F]+' |
Georg Brandl | fceab5a | 2008-01-19 20:08:23 +0000 | [diff] [blame] | 123 | Binnumber = r'0[bB][01]+' |
| 124 | Octnumber = r'0[oO][0-7]+' |
Antoine Pitrou | fd03645 | 2008-08-19 17:56:33 +0000 | [diff] [blame] | 125 | Decnumber = r'(?:0+|[1-9][0-9]*)' |
Guido van Rossum | cd16bf6 | 2007-06-13 18:07:49 +0000 | [diff] [blame] | 126 | Intnumber = group(Hexnumber, Binnumber, Octnumber, Decnumber) |
Antoine Pitrou | fd03645 | 2008-08-19 17:56:33 +0000 | [diff] [blame] | 127 | Exponent = r'[eE][-+]?[0-9]+' |
| 128 | Pointfloat = group(r'[0-9]+\.[0-9]*', r'\.[0-9]+') + maybe(Exponent) |
| 129 | Expfloat = r'[0-9]+' + Exponent |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 130 | Floatnumber = group(Pointfloat, Expfloat) |
Antoine Pitrou | fd03645 | 2008-08-19 17:56:33 +0000 | [diff] [blame] | 131 | Imagnumber = group(r'[0-9]+[jJ]', Floatnumber + r'[jJ]') |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 132 | Number = group(Imagnumber, Floatnumber, Intnumber) |
Guido van Rossum | 4d8e859 | 1992-01-01 19:34:47 +0000 | [diff] [blame] | 133 | |
Christian Heimes | 0b3847d | 2012-06-20 11:17:58 +0200 | [diff] [blame] | 134 | StringPrefix = r'(?:[bB][rR]?|[rR][bB]?|[uU])?' |
Armin Ronacher | c0eaeca | 2012-03-04 13:07:57 +0000 | [diff] [blame] | 135 | |
Tim Peters | de49583 | 2000-10-07 05:09:39 +0000 | [diff] [blame] | 136 | # Tail end of ' string. |
| 137 | Single = r"[^'\\]*(?:\\.[^'\\]*)*'" |
| 138 | # Tail end of " string. |
| 139 | Double = r'[^"\\]*(?:\\.[^"\\]*)*"' |
| 140 | # Tail end of ''' string. |
| 141 | Single3 = r"[^'\\]*(?:(?:\\.|'(?!''))[^'\\]*)*'''" |
| 142 | # Tail end of """ string. |
| 143 | Double3 = r'[^"\\]*(?:(?:\\.|"(?!""))[^"\\]*)*"""' |
Armin Ronacher | c0eaeca | 2012-03-04 13:07:57 +0000 | [diff] [blame] | 144 | Triple = group(StringPrefix + "'''", StringPrefix + '"""') |
Tim Peters | de49583 | 2000-10-07 05:09:39 +0000 | [diff] [blame] | 145 | # Single-line ' or " string. |
Armin Ronacher | c0eaeca | 2012-03-04 13:07:57 +0000 | [diff] [blame] | 146 | String = group(StringPrefix + r"'[^\n'\\]*(?:\\.[^\n'\\]*)*'", |
| 147 | StringPrefix + r'"[^\n"\\]*(?:\\.[^\n"\\]*)*"') |
Guido van Rossum | 4d8e859 | 1992-01-01 19:34:47 +0000 | [diff] [blame] | 148 | |
Tim Peters | de49583 | 2000-10-07 05:09:39 +0000 | [diff] [blame] | 149 | # Because of leftmost-then-longest match semantics, be sure to put the |
| 150 | # longest operators first (e.g., if = came before ==, == would get |
| 151 | # recognized as two instances of =). |
Guido van Rossum | b053cd8 | 2006-08-24 03:53:23 +0000 | [diff] [blame] | 152 | Operator = group(r"\*\*=?", r">>=?", r"<<=?", r"!=", |
Neal Norwitz | c150536 | 2006-12-28 06:47:50 +0000 | [diff] [blame] | 153 | r"//=?", r"->", |
Benjamin Peterson | d51374e | 2014-04-09 23:55:56 -0400 | [diff] [blame] | 154 | r"[+\-*/%&@|^=<>]=?", |
Tim Peters | de49583 | 2000-10-07 05:09:39 +0000 | [diff] [blame] | 155 | r"~") |
Thomas Wouters | e1519a1 | 2000-08-24 21:44:52 +0000 | [diff] [blame] | 156 | |
Guido van Rossum | 4d8e859 | 1992-01-01 19:34:47 +0000 | [diff] [blame] | 157 | Bracket = '[][(){}]' |
Georg Brandl | dde0028 | 2007-03-18 19:01:53 +0000 | [diff] [blame] | 158 | Special = group(r'\r?\n', r'\.\.\.', r'[:;.,@]') |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 159 | Funny = group(Operator, Bracket, Special) |
Guido van Rossum | 4d8e859 | 1992-01-01 19:34:47 +0000 | [diff] [blame] | 160 | |
Guido van Rossum | 3b63177 | 1997-10-27 20:44:15 +0000 | [diff] [blame] | 161 | PlainToken = group(Number, Funny, String, Name) |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 162 | Token = Ignore + PlainToken |
Guido van Rossum | 4d8e859 | 1992-01-01 19:34:47 +0000 | [diff] [blame] | 163 | |
Tim Peters | de49583 | 2000-10-07 05:09:39 +0000 | [diff] [blame] | 164 | # First (or only) line of ' or " string. |
Armin Ronacher | c0eaeca | 2012-03-04 13:07:57 +0000 | [diff] [blame] | 165 | ContStr = group(StringPrefix + r"'[^\n'\\]*(?:\\.[^\n'\\]*)*" + |
Ka-Ping Yee | 1ff08b1 | 2001-01-15 22:04:30 +0000 | [diff] [blame] | 166 | group("'", r'\\\r?\n'), |
Armin Ronacher | c0eaeca | 2012-03-04 13:07:57 +0000 | [diff] [blame] | 167 | StringPrefix + r'"[^\n"\\]*(?:\\.[^\n"\\]*)*' + |
Ka-Ping Yee | 1ff08b1 | 2001-01-15 22:04:30 +0000 | [diff] [blame] | 168 | group('"', r'\\\r?\n')) |
Ezio Melotti | 2cc3b4b | 2012-11-03 17:38:43 +0200 | [diff] [blame] | 169 | PseudoExtras = group(r'\\\r?\n|\Z', Comment, Triple) |
Guido van Rossum | 3b63177 | 1997-10-27 20:44:15 +0000 | [diff] [blame] | 170 | PseudoToken = Whitespace + group(PseudoExtras, Number, Funny, ContStr, Name) |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 171 | |
Benjamin Peterson | 33856de | 2010-08-30 14:41:20 +0000 | [diff] [blame] | 172 | def _compile(expr): |
| 173 | return re.compile(expr, re.UNICODE) |
| 174 | |
Antoine Pitrou | 10a99b0 | 2011-10-11 15:45:56 +0200 | [diff] [blame] | 175 | endpats = {"'": Single, '"': Double, |
| 176 | "'''": Single3, '"""': Double3, |
| 177 | "r'''": Single3, 'r"""': Double3, |
| 178 | "b'''": Single3, 'b"""': Double3, |
Antoine Pitrou | 10a99b0 | 2011-10-11 15:45:56 +0200 | [diff] [blame] | 179 | "R'''": Single3, 'R"""': Double3, |
| 180 | "B'''": Single3, 'B"""': Double3, |
Armin Ronacher | c0eaeca | 2012-03-04 13:07:57 +0000 | [diff] [blame] | 181 | "br'''": Single3, 'br"""': Double3, |
Antoine Pitrou | 10a99b0 | 2011-10-11 15:45:56 +0200 | [diff] [blame] | 182 | "bR'''": Single3, 'bR"""': Double3, |
| 183 | "Br'''": Single3, 'Br"""': Double3, |
| 184 | "BR'''": Single3, 'BR"""': Double3, |
Armin Ronacher | c0eaeca | 2012-03-04 13:07:57 +0000 | [diff] [blame] | 185 | "rb'''": Single3, 'rb"""': Double3, |
| 186 | "Rb'''": Single3, 'Rb"""': Double3, |
| 187 | "rB'''": Single3, 'rB"""': Double3, |
| 188 | "RB'''": Single3, 'RB"""': Double3, |
Armin Ronacher | 6ecf77b | 2012-03-04 12:04:06 +0000 | [diff] [blame] | 189 | "u'''": Single3, 'u"""': Double3, |
Armin Ronacher | 6ecf77b | 2012-03-04 12:04:06 +0000 | [diff] [blame] | 190 | "U'''": Single3, 'U"""': Double3, |
Armin Ronacher | 6ecf77b | 2012-03-04 12:04:06 +0000 | [diff] [blame] | 191 | 'r': None, 'R': None, 'b': None, 'B': None, |
| 192 | 'u': None, 'U': None} |
Guido van Rossum | 4d8e859 | 1992-01-01 19:34:47 +0000 | [diff] [blame] | 193 | |
Guido van Rossum | 9d6897a | 2002-08-24 06:54:19 +0000 | [diff] [blame] | 194 | triple_quoted = {} |
| 195 | for t in ("'''", '"""', |
| 196 | "r'''", 'r"""', "R'''", 'R"""', |
Guido van Rossum | 4fe72f9 | 2007-11-12 17:40:10 +0000 | [diff] [blame] | 197 | "b'''", 'b"""', "B'''", 'B"""', |
| 198 | "br'''", 'br"""', "Br'''", 'Br"""', |
Armin Ronacher | 6ecf77b | 2012-03-04 12:04:06 +0000 | [diff] [blame] | 199 | "bR'''", 'bR"""', "BR'''", 'BR"""', |
Armin Ronacher | c0eaeca | 2012-03-04 13:07:57 +0000 | [diff] [blame] | 200 | "rb'''", 'rb"""', "rB'''", 'rB"""', |
| 201 | "Rb'''", 'Rb"""', "RB'''", 'RB"""', |
Armin Ronacher | 6ecf77b | 2012-03-04 12:04:06 +0000 | [diff] [blame] | 202 | "u'''", 'u"""', "U'''", 'U"""', |
Christian Heimes | 0b3847d | 2012-06-20 11:17:58 +0200 | [diff] [blame] | 203 | ): |
Guido van Rossum | 9d6897a | 2002-08-24 06:54:19 +0000 | [diff] [blame] | 204 | triple_quoted[t] = t |
| 205 | single_quoted = {} |
| 206 | for t in ("'", '"', |
| 207 | "r'", 'r"', "R'", 'R"', |
Guido van Rossum | 4fe72f9 | 2007-11-12 17:40:10 +0000 | [diff] [blame] | 208 | "b'", 'b"', "B'", 'B"', |
| 209 | "br'", 'br"', "Br'", 'Br"', |
Armin Ronacher | 6ecf77b | 2012-03-04 12:04:06 +0000 | [diff] [blame] | 210 | "bR'", 'bR"', "BR'", 'BR"' , |
Armin Ronacher | c0eaeca | 2012-03-04 13:07:57 +0000 | [diff] [blame] | 211 | "rb'", 'rb"', "rB'", 'rB"', |
| 212 | "Rb'", 'Rb"', "RB'", 'RB"' , |
Armin Ronacher | 6ecf77b | 2012-03-04 12:04:06 +0000 | [diff] [blame] | 213 | "u'", 'u"', "U'", 'U"', |
Christian Heimes | 0b3847d | 2012-06-20 11:17:58 +0200 | [diff] [blame] | 214 | ): |
Guido van Rossum | 9d6897a | 2002-08-24 06:54:19 +0000 | [diff] [blame] | 215 | single_quoted[t] = t |
| 216 | |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 217 | tabsize = 8 |
Fred Drake | 9b8d801 | 2000-08-17 04:45:13 +0000 | [diff] [blame] | 218 | |
Ka-Ping Yee | 28c62bb | 2001-03-23 05:22:49 +0000 | [diff] [blame] | 219 | class TokenError(Exception): pass |
| 220 | |
| 221 | class StopTokenizing(Exception): pass |
Fred Drake | 9b8d801 | 2000-08-17 04:45:13 +0000 | [diff] [blame] | 222 | |
Tim Peters | 5ca576e | 2001-06-18 22:08:13 +0000 | [diff] [blame] | 223 | |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 224 | class Untokenizer: |
| 225 | |
| 226 | def __init__(self): |
| 227 | self.tokens = [] |
| 228 | self.prev_row = 1 |
| 229 | self.prev_col = 0 |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 230 | self.encoding = None |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 231 | |
| 232 | def add_whitespace(self, start): |
| 233 | row, col = start |
Terry Jan Reedy | 5e6db31 | 2014-02-17 16:45:48 -0500 | [diff] [blame] | 234 | if row < self.prev_row or row == self.prev_row and col < self.prev_col: |
| 235 | raise ValueError("start ({},{}) precedes previous end ({},{})" |
| 236 | .format(row, col, self.prev_row, self.prev_col)) |
Terry Jan Reedy | 9dc3a36 | 2014-02-23 23:33:08 -0500 | [diff] [blame] | 237 | row_offset = row - self.prev_row |
Terry Jan Reedy | f106f8f | 2014-02-23 23:39:57 -0500 | [diff] [blame] | 238 | if row_offset: |
Terry Jan Reedy | 9dc3a36 | 2014-02-23 23:33:08 -0500 | [diff] [blame] | 239 | self.tokens.append("\\\n" * row_offset) |
| 240 | self.prev_col = 0 |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 241 | col_offset = col - self.prev_col |
| 242 | if col_offset: |
| 243 | self.tokens.append(" " * col_offset) |
| 244 | |
| 245 | def untokenize(self, iterable): |
Terry Jan Reedy | 5b8d2c3 | 2014-02-17 23:12:16 -0500 | [diff] [blame] | 246 | it = iter(iterable) |
Dingyuan Wang | e411b66 | 2015-06-22 10:01:12 +0800 | [diff] [blame] | 247 | indents = [] |
| 248 | startline = False |
Terry Jan Reedy | 5b8d2c3 | 2014-02-17 23:12:16 -0500 | [diff] [blame] | 249 | for t in it: |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 250 | if len(t) == 2: |
Terry Jan Reedy | 5b8d2c3 | 2014-02-17 23:12:16 -0500 | [diff] [blame] | 251 | self.compat(t, it) |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 252 | break |
| 253 | tok_type, token, start, end, line = t |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 254 | if tok_type == ENCODING: |
| 255 | self.encoding = token |
| 256 | continue |
Terry Jan Reedy | 9dc3a36 | 2014-02-23 23:33:08 -0500 | [diff] [blame] | 257 | if tok_type == ENDMARKER: |
| 258 | break |
Dingyuan Wang | e411b66 | 2015-06-22 10:01:12 +0800 | [diff] [blame] | 259 | if tok_type == INDENT: |
| 260 | indents.append(token) |
| 261 | continue |
| 262 | elif tok_type == DEDENT: |
| 263 | indents.pop() |
| 264 | self.prev_row, self.prev_col = end |
| 265 | continue |
| 266 | elif tok_type in (NEWLINE, NL): |
| 267 | startline = True |
| 268 | elif startline and indents: |
| 269 | indent = indents[-1] |
| 270 | if start[1] >= len(indent): |
| 271 | self.tokens.append(indent) |
| 272 | self.prev_col = len(indent) |
| 273 | startline = False |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 274 | self.add_whitespace(start) |
| 275 | self.tokens.append(token) |
| 276 | self.prev_row, self.prev_col = end |
| 277 | if tok_type in (NEWLINE, NL): |
| 278 | self.prev_row += 1 |
| 279 | self.prev_col = 0 |
| 280 | return "".join(self.tokens) |
| 281 | |
| 282 | def compat(self, token, iterable): |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 283 | indents = [] |
| 284 | toks_append = self.tokens.append |
Terry Jan Reedy | 5b8d2c3 | 2014-02-17 23:12:16 -0500 | [diff] [blame] | 285 | startline = token[0] in (NEWLINE, NL) |
Christian Heimes | ba4af49 | 2008-03-28 00:55:15 +0000 | [diff] [blame] | 286 | prevstring = False |
Terry Jan Reedy | 5b8d2c3 | 2014-02-17 23:12:16 -0500 | [diff] [blame] | 287 | |
| 288 | for tok in chain([token], iterable): |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 289 | toknum, tokval = tok[:2] |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 290 | if toknum == ENCODING: |
| 291 | self.encoding = tokval |
| 292 | continue |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 293 | |
Yury Selivanov | 7544508 | 2015-05-11 22:57:16 -0400 | [diff] [blame] | 294 | if toknum in (NAME, NUMBER, ASYNC, AWAIT): |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 295 | tokval += ' ' |
| 296 | |
Christian Heimes | ba4af49 | 2008-03-28 00:55:15 +0000 | [diff] [blame] | 297 | # Insert a space between two consecutive strings |
| 298 | if toknum == STRING: |
| 299 | if prevstring: |
| 300 | tokval = ' ' + tokval |
| 301 | prevstring = True |
| 302 | else: |
| 303 | prevstring = False |
| 304 | |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 305 | if toknum == INDENT: |
| 306 | indents.append(tokval) |
| 307 | continue |
| 308 | elif toknum == DEDENT: |
| 309 | indents.pop() |
| 310 | continue |
| 311 | elif toknum in (NEWLINE, NL): |
| 312 | startline = True |
| 313 | elif startline and indents: |
| 314 | toks_append(indents[-1]) |
| 315 | startline = False |
| 316 | toks_append(tokval) |
Raymond Hettinger | 68c0453 | 2005-06-10 11:05:19 +0000 | [diff] [blame] | 317 | |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 318 | |
Raymond Hettinger | 68c0453 | 2005-06-10 11:05:19 +0000 | [diff] [blame] | 319 | def untokenize(iterable): |
| 320 | """Transform tokens back into Python source code. |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 321 | It returns a bytes object, encoded using the ENCODING |
| 322 | token, which is the first token sequence output by tokenize. |
Raymond Hettinger | 68c0453 | 2005-06-10 11:05:19 +0000 | [diff] [blame] | 323 | |
| 324 | Each element returned by the iterable must be a token sequence |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 325 | with at least two elements, a token number and token value. If |
| 326 | only two tokens are passed, the resulting output is poor. |
Raymond Hettinger | 68c0453 | 2005-06-10 11:05:19 +0000 | [diff] [blame] | 327 | |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 328 | Round-trip invariant for full input: |
| 329 | Untokenized source will match input source exactly |
| 330 | |
Berker Peksag | ff8d087 | 2015-12-30 01:41:58 +0200 | [diff] [blame] | 331 | Round-trip invariant for limited input: |
| 332 | # Output bytes will tokenize back to the input |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 333 | t1 = [tok[:2] for tok in tokenize(f.readline)] |
Raymond Hettinger | 68c0453 | 2005-06-10 11:05:19 +0000 | [diff] [blame] | 334 | newcode = untokenize(t1) |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 335 | readline = BytesIO(newcode).readline |
| 336 | t2 = [tok[:2] for tok in tokenize(readline)] |
Raymond Hettinger | 68c0453 | 2005-06-10 11:05:19 +0000 | [diff] [blame] | 337 | assert t1 == t2 |
| 338 | """ |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 339 | ut = Untokenizer() |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 340 | out = ut.untokenize(iterable) |
| 341 | if ut.encoding is not None: |
| 342 | out = out.encode(ut.encoding) |
| 343 | return out |
Raymond Hettinger | 68c0453 | 2005-06-10 11:05:19 +0000 | [diff] [blame] | 344 | |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 345 | |
Benjamin Peterson | d3afada | 2009-10-09 21:43:09 +0000 | [diff] [blame] | 346 | def _get_normal_name(orig_enc): |
| 347 | """Imitates get_normal_name in tokenizer.c.""" |
| 348 | # Only care about the first 12 characters. |
| 349 | enc = orig_enc[:12].lower().replace("_", "-") |
| 350 | if enc == "utf-8" or enc.startswith("utf-8-"): |
| 351 | return "utf-8" |
| 352 | if enc in ("latin-1", "iso-8859-1", "iso-latin-1") or \ |
| 353 | enc.startswith(("latin-1-", "iso-8859-1-", "iso-latin-1-")): |
| 354 | return "iso-8859-1" |
| 355 | return orig_enc |
| 356 | |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 357 | def detect_encoding(readline): |
Raymond Hettinger | d1fa3db | 2002-05-15 02:56:03 +0000 | [diff] [blame] | 358 | """ |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 359 | The detect_encoding() function is used to detect the encoding that should |
Ezio Melotti | 4bcc796 | 2013-11-25 05:14:51 +0200 | [diff] [blame] | 360 | be used to decode a Python source file. It requires one argument, readline, |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 361 | in the same way as the tokenize() generator. |
| 362 | |
| 363 | It will call readline a maximum of twice, and return the encoding used |
Florent Xicluna | 43e4ea1 | 2010-09-03 19:54:02 +0000 | [diff] [blame] | 364 | (as a string) and a list of any lines (left as bytes) it has read in. |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 365 | |
| 366 | It detects the encoding from the presence of a utf-8 bom or an encoding |
Florent Xicluna | 43e4ea1 | 2010-09-03 19:54:02 +0000 | [diff] [blame] | 367 | cookie as specified in pep-0263. If both a bom and a cookie are present, |
| 368 | but disagree, a SyntaxError will be raised. If the encoding cookie is an |
| 369 | invalid charset, raise a SyntaxError. Note that if a utf-8 bom is found, |
Benjamin Peterson | 689a558 | 2010-03-18 22:29:52 +0000 | [diff] [blame] | 370 | 'utf-8-sig' is returned. |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 371 | |
| 372 | If no encoding is specified, then the default of 'utf-8' will be returned. |
| 373 | """ |
Brett Cannon | c33f3f2 | 2012-04-20 13:23:54 -0400 | [diff] [blame] | 374 | try: |
| 375 | filename = readline.__self__.name |
| 376 | except AttributeError: |
| 377 | filename = None |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 378 | bom_found = False |
| 379 | encoding = None |
Benjamin Peterson | 689a558 | 2010-03-18 22:29:52 +0000 | [diff] [blame] | 380 | default = 'utf-8' |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 381 | def read_or_stop(): |
| 382 | try: |
| 383 | return readline() |
| 384 | except StopIteration: |
| 385 | return b'' |
| 386 | |
| 387 | def find_cookie(line): |
| 388 | try: |
Martin v. Löwis | 63674f4 | 2012-04-20 14:36:47 +0200 | [diff] [blame] | 389 | # Decode as UTF-8. Either the line is an encoding declaration, |
| 390 | # in which case it should be pure ASCII, or it must be UTF-8 |
| 391 | # per default encoding. |
| 392 | line_string = line.decode('utf-8') |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 393 | except UnicodeDecodeError: |
Brett Cannon | c33f3f2 | 2012-04-20 13:23:54 -0400 | [diff] [blame] | 394 | msg = "invalid or missing encoding declaration" |
| 395 | if filename is not None: |
| 396 | msg = '{} for {!r}'.format(msg, filename) |
| 397 | raise SyntaxError(msg) |
Benjamin Peterson | 433f32c | 2008-12-12 01:25:05 +0000 | [diff] [blame] | 398 | |
Serhiy Storchaka | dafea85 | 2013-09-16 23:51:56 +0300 | [diff] [blame] | 399 | match = cookie_re.match(line_string) |
| 400 | if not match: |
Benjamin Peterson | 433f32c | 2008-12-12 01:25:05 +0000 | [diff] [blame] | 401 | return None |
Serhiy Storchaka | dafea85 | 2013-09-16 23:51:56 +0300 | [diff] [blame] | 402 | encoding = _get_normal_name(match.group(1)) |
Benjamin Peterson | 433f32c | 2008-12-12 01:25:05 +0000 | [diff] [blame] | 403 | try: |
| 404 | codec = lookup(encoding) |
| 405 | except LookupError: |
| 406 | # This behaviour mimics the Python interpreter |
Brett Cannon | c33f3f2 | 2012-04-20 13:23:54 -0400 | [diff] [blame] | 407 | if filename is None: |
| 408 | msg = "unknown encoding: " + encoding |
| 409 | else: |
| 410 | msg = "unknown encoding for {!r}: {}".format(filename, |
| 411 | encoding) |
| 412 | raise SyntaxError(msg) |
Benjamin Peterson | 433f32c | 2008-12-12 01:25:05 +0000 | [diff] [blame] | 413 | |
Benjamin Peterson | 1613ed8 | 2010-03-18 22:34:15 +0000 | [diff] [blame] | 414 | if bom_found: |
Florent Xicluna | 11f0b41 | 2012-07-07 12:13:35 +0200 | [diff] [blame] | 415 | if encoding != 'utf-8': |
Benjamin Peterson | 1613ed8 | 2010-03-18 22:34:15 +0000 | [diff] [blame] | 416 | # This behaviour mimics the Python interpreter |
Brett Cannon | c33f3f2 | 2012-04-20 13:23:54 -0400 | [diff] [blame] | 417 | if filename is None: |
| 418 | msg = 'encoding problem: utf-8' |
| 419 | else: |
| 420 | msg = 'encoding problem for {!r}: utf-8'.format(filename) |
| 421 | raise SyntaxError(msg) |
Benjamin Peterson | 1613ed8 | 2010-03-18 22:34:15 +0000 | [diff] [blame] | 422 | encoding += '-sig' |
Benjamin Peterson | 433f32c | 2008-12-12 01:25:05 +0000 | [diff] [blame] | 423 | return encoding |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 424 | |
| 425 | first = read_or_stop() |
Benjamin Peterson | 433f32c | 2008-12-12 01:25:05 +0000 | [diff] [blame] | 426 | if first.startswith(BOM_UTF8): |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 427 | bom_found = True |
| 428 | first = first[3:] |
Benjamin Peterson | 689a558 | 2010-03-18 22:29:52 +0000 | [diff] [blame] | 429 | default = 'utf-8-sig' |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 430 | if not first: |
Benjamin Peterson | 689a558 | 2010-03-18 22:29:52 +0000 | [diff] [blame] | 431 | return default, [] |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 432 | |
| 433 | encoding = find_cookie(first) |
| 434 | if encoding: |
| 435 | return encoding, [first] |
Serhiy Storchaka | 768c16c | 2014-01-09 18:36:09 +0200 | [diff] [blame] | 436 | if not blank_re.match(first): |
| 437 | return default, [first] |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 438 | |
| 439 | second = read_or_stop() |
| 440 | if not second: |
Benjamin Peterson | 689a558 | 2010-03-18 22:29:52 +0000 | [diff] [blame] | 441 | return default, [first] |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 442 | |
| 443 | encoding = find_cookie(second) |
| 444 | if encoding: |
| 445 | return encoding, [first, second] |
| 446 | |
Benjamin Peterson | 689a558 | 2010-03-18 22:29:52 +0000 | [diff] [blame] | 447 | return default, [first, second] |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 448 | |
| 449 | |
Victor Stinner | 58c0752 | 2010-11-09 01:08:59 +0000 | [diff] [blame] | 450 | def open(filename): |
| 451 | """Open a file in read only mode using the encoding detected by |
| 452 | detect_encoding(). |
| 453 | """ |
Victor Stinner | 9691750 | 2014-12-05 10:17:10 +0100 | [diff] [blame] | 454 | buffer = _builtin_open(filename, 'rb') |
Victor Stinner | 387729e | 2015-05-26 00:43:58 +0200 | [diff] [blame] | 455 | try: |
| 456 | encoding, lines = detect_encoding(buffer.readline) |
| 457 | buffer.seek(0) |
| 458 | text = TextIOWrapper(buffer, encoding, line_buffering=True) |
| 459 | text.mode = 'r' |
| 460 | return text |
| 461 | except: |
| 462 | buffer.close() |
| 463 | raise |
Victor Stinner | 58c0752 | 2010-11-09 01:08:59 +0000 | [diff] [blame] | 464 | |
| 465 | |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 466 | def tokenize(readline): |
| 467 | """ |
Berker Peksag | ff8d087 | 2015-12-30 01:41:58 +0200 | [diff] [blame] | 468 | The tokenize() generator requires one argument, readline, which |
Raymond Hettinger | d1fa3db | 2002-05-15 02:56:03 +0000 | [diff] [blame] | 469 | must be a callable object which provides the same interface as the |
Florent Xicluna | 43e4ea1 | 2010-09-03 19:54:02 +0000 | [diff] [blame] | 470 | readline() method of built-in file objects. Each call to the function |
Berker Peksag | ff8d087 | 2015-12-30 01:41:58 +0200 | [diff] [blame] | 471 | should return one line of input as bytes. Alternatively, readline |
Raymond Hettinger | 68c0453 | 2005-06-10 11:05:19 +0000 | [diff] [blame] | 472 | can be a callable function terminating with StopIteration: |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 473 | readline = open(myfile, 'rb').__next__ # Example of alternate readline |
Tim Peters | 8ac1495 | 2002-05-23 15:15:30 +0000 | [diff] [blame] | 474 | |
Raymond Hettinger | d1fa3db | 2002-05-15 02:56:03 +0000 | [diff] [blame] | 475 | The generator produces 5-tuples with these members: the token type; the |
| 476 | token string; a 2-tuple (srow, scol) of ints specifying the row and |
| 477 | column where the token begins in the source; a 2-tuple (erow, ecol) of |
| 478 | ints specifying the row and column where the token ends in the source; |
Florent Xicluna | 43e4ea1 | 2010-09-03 19:54:02 +0000 | [diff] [blame] | 479 | and the line on which the token was found. The line passed is the |
Tim Peters | 8ac1495 | 2002-05-23 15:15:30 +0000 | [diff] [blame] | 480 | logical line; continuation lines are included. |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 481 | |
| 482 | The first token sequence will always be an ENCODING token |
| 483 | which tells you which encoding was used to decode the bytes stream. |
Raymond Hettinger | d1fa3db | 2002-05-15 02:56:03 +0000 | [diff] [blame] | 484 | """ |
Benjamin Peterson | 21db77e | 2009-11-14 16:27:26 +0000 | [diff] [blame] | 485 | # This import is here to avoid problems when the itertools module is not |
| 486 | # built yet and tokenize is imported. |
Benjamin Peterson | 81dd8b9 | 2009-11-14 18:09:17 +0000 | [diff] [blame] | 487 | from itertools import chain, repeat |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 488 | encoding, consumed = detect_encoding(readline) |
Benjamin Peterson | 81dd8b9 | 2009-11-14 18:09:17 +0000 | [diff] [blame] | 489 | rl_gen = iter(readline, b"") |
| 490 | empty = repeat(b"") |
| 491 | return _tokenize(chain(consumed, rl_gen, empty).__next__, encoding) |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 492 | |
| 493 | |
| 494 | def _tokenize(readline, encoding): |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 495 | lnum = parenlev = continued = 0 |
Benjamin Peterson | 33856de | 2010-08-30 14:41:20 +0000 | [diff] [blame] | 496 | numchars = '0123456789' |
Guido van Rossum | de65527 | 1997-04-09 17:15:54 +0000 | [diff] [blame] | 497 | contstr, needcont = '', 0 |
Guido van Rossum | a90c78b | 1998-04-03 16:05:38 +0000 | [diff] [blame] | 498 | contline = None |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 499 | indents = [0] |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 500 | |
Yury Selivanov | 96ec934 | 2015-07-23 15:01:58 +0300 | [diff] [blame] | 501 | # 'stashed' and 'async_*' are used for async/await parsing |
Yury Selivanov | 7544508 | 2015-05-11 22:57:16 -0400 | [diff] [blame] | 502 | stashed = None |
Yury Selivanov | 96ec934 | 2015-07-23 15:01:58 +0300 | [diff] [blame] | 503 | async_def = False |
| 504 | async_def_indent = 0 |
| 505 | async_def_nl = False |
Yury Selivanov | 7544508 | 2015-05-11 22:57:16 -0400 | [diff] [blame] | 506 | |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 507 | if encoding is not None: |
Benjamin Peterson | 689a558 | 2010-03-18 22:29:52 +0000 | [diff] [blame] | 508 | if encoding == "utf-8-sig": |
| 509 | # BOM will already have been stripped. |
| 510 | encoding = "utf-8" |
Raymond Hettinger | a48db39 | 2009-04-29 00:34:27 +0000 | [diff] [blame] | 511 | yield TokenInfo(ENCODING, encoding, (0, 0), (0, 0), '') |
Benjamin Peterson | 0fe1438 | 2008-06-05 23:07:42 +0000 | [diff] [blame] | 512 | while True: # loop over lines in stream |
Raymond Hettinger | 68c0453 | 2005-06-10 11:05:19 +0000 | [diff] [blame] | 513 | try: |
| 514 | line = readline() |
| 515 | except StopIteration: |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 516 | line = b'' |
| 517 | |
| 518 | if encoding is not None: |
| 519 | line = line.decode(encoding) |
Benjamin Peterson | a0dfa82 | 2009-11-13 02:25:08 +0000 | [diff] [blame] | 520 | lnum += 1 |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 521 | pos, max = 0, len(line) |
| 522 | |
| 523 | if contstr: # continued string |
Guido van Rossum | de65527 | 1997-04-09 17:15:54 +0000 | [diff] [blame] | 524 | if not line: |
Collin Winter | ce36ad8 | 2007-08-30 01:19:48 +0000 | [diff] [blame] | 525 | raise TokenError("EOF in multi-line string", strstart) |
Guido van Rossum | 3b63177 | 1997-10-27 20:44:15 +0000 | [diff] [blame] | 526 | endmatch = endprog.match(line) |
| 527 | if endmatch: |
| 528 | pos = end = endmatch.end(0) |
Raymond Hettinger | a48db39 | 2009-04-29 00:34:27 +0000 | [diff] [blame] | 529 | yield TokenInfo(STRING, contstr + line[:end], |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 530 | strstart, (lnum, end), contline + line) |
Guido van Rossum | de65527 | 1997-04-09 17:15:54 +0000 | [diff] [blame] | 531 | contstr, needcont = '', 0 |
Guido van Rossum | a90c78b | 1998-04-03 16:05:38 +0000 | [diff] [blame] | 532 | contline = None |
Guido van Rossum | de65527 | 1997-04-09 17:15:54 +0000 | [diff] [blame] | 533 | elif needcont and line[-2:] != '\\\n' and line[-3:] != '\\\r\n': |
Raymond Hettinger | a48db39 | 2009-04-29 00:34:27 +0000 | [diff] [blame] | 534 | yield TokenInfo(ERRORTOKEN, contstr + line, |
Guido van Rossum | a90c78b | 1998-04-03 16:05:38 +0000 | [diff] [blame] | 535 | strstart, (lnum, len(line)), contline) |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 536 | contstr = '' |
Guido van Rossum | a90c78b | 1998-04-03 16:05:38 +0000 | [diff] [blame] | 537 | contline = None |
Guido van Rossum | de65527 | 1997-04-09 17:15:54 +0000 | [diff] [blame] | 538 | continue |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 539 | else: |
| 540 | contstr = contstr + line |
Guido van Rossum | a90c78b | 1998-04-03 16:05:38 +0000 | [diff] [blame] | 541 | contline = contline + line |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 542 | continue |
| 543 | |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 544 | elif parenlev == 0 and not continued: # new statement |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 545 | if not line: break |
| 546 | column = 0 |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 547 | while pos < max: # measure leading whitespace |
Benjamin Peterson | a0dfa82 | 2009-11-13 02:25:08 +0000 | [diff] [blame] | 548 | if line[pos] == ' ': |
| 549 | column += 1 |
| 550 | elif line[pos] == '\t': |
| 551 | column = (column//tabsize + 1)*tabsize |
| 552 | elif line[pos] == '\f': |
| 553 | column = 0 |
| 554 | else: |
| 555 | break |
| 556 | pos += 1 |
| 557 | if pos == max: |
| 558 | break |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 559 | |
| 560 | if line[pos] in '#\r\n': # skip comments or blank lines |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 561 | if line[pos] == '#': |
| 562 | comment_token = line[pos:].rstrip('\r\n') |
| 563 | nl_pos = pos + len(comment_token) |
Raymond Hettinger | a48db39 | 2009-04-29 00:34:27 +0000 | [diff] [blame] | 564 | yield TokenInfo(COMMENT, comment_token, |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 565 | (lnum, pos), (lnum, pos + len(comment_token)), line) |
Raymond Hettinger | a48db39 | 2009-04-29 00:34:27 +0000 | [diff] [blame] | 566 | yield TokenInfo(NL, line[nl_pos:], |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 567 | (lnum, nl_pos), (lnum, len(line)), line) |
| 568 | else: |
Raymond Hettinger | a48db39 | 2009-04-29 00:34:27 +0000 | [diff] [blame] | 569 | yield TokenInfo((NL, COMMENT)[line[pos] == '#'], line[pos:], |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 570 | (lnum, pos), (lnum, len(line)), line) |
| 571 | continue |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 572 | |
| 573 | if column > indents[-1]: # count indents or dedents |
| 574 | indents.append(column) |
Raymond Hettinger | a48db39 | 2009-04-29 00:34:27 +0000 | [diff] [blame] | 575 | yield TokenInfo(INDENT, line[:pos], (lnum, 0), (lnum, pos), line) |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 576 | while column < indents[-1]: |
Raymond Hettinger | da99d1c | 2005-06-21 07:43:58 +0000 | [diff] [blame] | 577 | if column not in indents: |
| 578 | raise IndentationError( |
Thomas Wouters | 00ee7ba | 2006-08-21 19:07:27 +0000 | [diff] [blame] | 579 | "unindent does not match any outer indentation level", |
| 580 | ("<tokenize>", lnum, pos, line)) |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 581 | indents = indents[:-1] |
Yury Selivanov | 7544508 | 2015-05-11 22:57:16 -0400 | [diff] [blame] | 582 | |
Yury Selivanov | 96ec934 | 2015-07-23 15:01:58 +0300 | [diff] [blame] | 583 | if async_def and async_def_indent >= indents[-1]: |
| 584 | async_def = False |
| 585 | async_def_nl = False |
| 586 | async_def_indent = 0 |
Yury Selivanov | 7544508 | 2015-05-11 22:57:16 -0400 | [diff] [blame] | 587 | |
Raymond Hettinger | a48db39 | 2009-04-29 00:34:27 +0000 | [diff] [blame] | 588 | yield TokenInfo(DEDENT, '', (lnum, pos), (lnum, pos), line) |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 589 | |
Yury Selivanov | 96ec934 | 2015-07-23 15:01:58 +0300 | [diff] [blame] | 590 | if async_def and async_def_nl and async_def_indent >= indents[-1]: |
| 591 | async_def = False |
| 592 | async_def_nl = False |
| 593 | async_def_indent = 0 |
| 594 | |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 595 | else: # continued statement |
Guido van Rossum | de65527 | 1997-04-09 17:15:54 +0000 | [diff] [blame] | 596 | if not line: |
Collin Winter | ce36ad8 | 2007-08-30 01:19:48 +0000 | [diff] [blame] | 597 | raise TokenError("EOF in multi-line statement", (lnum, 0)) |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 598 | continued = 0 |
| 599 | |
| 600 | while pos < max: |
Antoine Pitrou | 10a99b0 | 2011-10-11 15:45:56 +0200 | [diff] [blame] | 601 | pseudomatch = _compile(PseudoToken).match(line, pos) |
Guido van Rossum | 3b63177 | 1997-10-27 20:44:15 +0000 | [diff] [blame] | 602 | if pseudomatch: # scan for tokens |
| 603 | start, end = pseudomatch.span(1) |
Guido van Rossum | de65527 | 1997-04-09 17:15:54 +0000 | [diff] [blame] | 604 | spos, epos, pos = (lnum, start), (lnum, end), end |
Ezio Melotti | 2cc3b4b | 2012-11-03 17:38:43 +0200 | [diff] [blame] | 605 | if start == end: |
| 606 | continue |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 607 | token, initial = line[start:end], line[start] |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 608 | |
Georg Brandl | dde0028 | 2007-03-18 19:01:53 +0000 | [diff] [blame] | 609 | if (initial in numchars or # ordinary number |
| 610 | (initial == '.' and token != '.' and token != '...')): |
Raymond Hettinger | a48db39 | 2009-04-29 00:34:27 +0000 | [diff] [blame] | 611 | yield TokenInfo(NUMBER, token, spos, epos, line) |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 612 | elif initial in '\r\n': |
Yury Selivanov | 7544508 | 2015-05-11 22:57:16 -0400 | [diff] [blame] | 613 | if stashed: |
| 614 | yield stashed |
| 615 | stashed = None |
Yury Selivanov | 96ec934 | 2015-07-23 15:01:58 +0300 | [diff] [blame] | 616 | if parenlev > 0: |
| 617 | yield TokenInfo(NL, token, spos, epos, line) |
| 618 | else: |
| 619 | yield TokenInfo(NEWLINE, token, spos, epos, line) |
| 620 | if async_def: |
| 621 | async_def_nl = True |
| 622 | |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 623 | elif initial == '#': |
Thomas Wouters | 89f507f | 2006-12-13 04:49:30 +0000 | [diff] [blame] | 624 | assert not token.endswith("\n") |
Yury Selivanov | 7544508 | 2015-05-11 22:57:16 -0400 | [diff] [blame] | 625 | if stashed: |
| 626 | yield stashed |
| 627 | stashed = None |
Raymond Hettinger | a48db39 | 2009-04-29 00:34:27 +0000 | [diff] [blame] | 628 | yield TokenInfo(COMMENT, token, spos, epos, line) |
Guido van Rossum | 9d6897a | 2002-08-24 06:54:19 +0000 | [diff] [blame] | 629 | elif token in triple_quoted: |
Antoine Pitrou | 10a99b0 | 2011-10-11 15:45:56 +0200 | [diff] [blame] | 630 | endprog = _compile(endpats[token]) |
Guido van Rossum | 3b63177 | 1997-10-27 20:44:15 +0000 | [diff] [blame] | 631 | endmatch = endprog.match(line, pos) |
| 632 | if endmatch: # all on one line |
| 633 | pos = endmatch.end(0) |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 634 | token = line[start:pos] |
Raymond Hettinger | a48db39 | 2009-04-29 00:34:27 +0000 | [diff] [blame] | 635 | yield TokenInfo(STRING, token, spos, (lnum, pos), line) |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 636 | else: |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 637 | strstart = (lnum, start) # multiple lines |
| 638 | contstr = line[start:] |
Guido van Rossum | a90c78b | 1998-04-03 16:05:38 +0000 | [diff] [blame] | 639 | contline = line |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 640 | break |
Guido van Rossum | 9d6897a | 2002-08-24 06:54:19 +0000 | [diff] [blame] | 641 | elif initial in single_quoted or \ |
| 642 | token[:2] in single_quoted or \ |
| 643 | token[:3] in single_quoted: |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 644 | if token[-1] == '\n': # continued string |
Guido van Rossum | 1aec323 | 1997-04-08 14:24:39 +0000 | [diff] [blame] | 645 | strstart = (lnum, start) |
Antoine Pitrou | 10a99b0 | 2011-10-11 15:45:56 +0200 | [diff] [blame] | 646 | endprog = _compile(endpats[initial] or |
| 647 | endpats[token[1]] or |
| 648 | endpats[token[2]]) |
Guido van Rossum | de65527 | 1997-04-09 17:15:54 +0000 | [diff] [blame] | 649 | contstr, needcont = line[start:], 1 |
Guido van Rossum | a90c78b | 1998-04-03 16:05:38 +0000 | [diff] [blame] | 650 | contline = line |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 651 | break |
| 652 | else: # ordinary string |
Raymond Hettinger | a48db39 | 2009-04-29 00:34:27 +0000 | [diff] [blame] | 653 | yield TokenInfo(STRING, token, spos, epos, line) |
Benjamin Peterson | 33856de | 2010-08-30 14:41:20 +0000 | [diff] [blame] | 654 | elif initial.isidentifier(): # ordinary name |
Yury Selivanov | 7544508 | 2015-05-11 22:57:16 -0400 | [diff] [blame] | 655 | if token in ('async', 'await'): |
Yury Selivanov | 96ec934 | 2015-07-23 15:01:58 +0300 | [diff] [blame] | 656 | if async_def: |
Yury Selivanov | 7544508 | 2015-05-11 22:57:16 -0400 | [diff] [blame] | 657 | yield TokenInfo( |
| 658 | ASYNC if token == 'async' else AWAIT, |
| 659 | token, spos, epos, line) |
| 660 | continue |
| 661 | |
| 662 | tok = TokenInfo(NAME, token, spos, epos, line) |
| 663 | if token == 'async' and not stashed: |
| 664 | stashed = tok |
| 665 | continue |
| 666 | |
| 667 | if token == 'def': |
| 668 | if (stashed |
| 669 | and stashed.type == NAME |
| 670 | and stashed.string == 'async'): |
| 671 | |
Yury Selivanov | 96ec934 | 2015-07-23 15:01:58 +0300 | [diff] [blame] | 672 | async_def = True |
| 673 | async_def_indent = indents[-1] |
Yury Selivanov | 7544508 | 2015-05-11 22:57:16 -0400 | [diff] [blame] | 674 | |
| 675 | yield TokenInfo(ASYNC, stashed.string, |
| 676 | stashed.start, stashed.end, |
| 677 | stashed.line) |
| 678 | stashed = None |
Yury Selivanov | 7544508 | 2015-05-11 22:57:16 -0400 | [diff] [blame] | 679 | |
| 680 | if stashed: |
| 681 | yield stashed |
| 682 | stashed = None |
| 683 | |
| 684 | yield tok |
Guido van Rossum | 3b63177 | 1997-10-27 20:44:15 +0000 | [diff] [blame] | 685 | elif initial == '\\': # continued stmt |
| 686 | continued = 1 |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 687 | else: |
Benjamin Peterson | a0dfa82 | 2009-11-13 02:25:08 +0000 | [diff] [blame] | 688 | if initial in '([{': |
| 689 | parenlev += 1 |
| 690 | elif initial in ')]}': |
| 691 | parenlev -= 1 |
Yury Selivanov | 7544508 | 2015-05-11 22:57:16 -0400 | [diff] [blame] | 692 | if stashed: |
| 693 | yield stashed |
| 694 | stashed = None |
Raymond Hettinger | a48db39 | 2009-04-29 00:34:27 +0000 | [diff] [blame] | 695 | yield TokenInfo(OP, token, spos, epos, line) |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 696 | else: |
Raymond Hettinger | a48db39 | 2009-04-29 00:34:27 +0000 | [diff] [blame] | 697 | yield TokenInfo(ERRORTOKEN, line[pos], |
Guido van Rossum | de65527 | 1997-04-09 17:15:54 +0000 | [diff] [blame] | 698 | (lnum, pos), (lnum, pos+1), line) |
Benjamin Peterson | a0dfa82 | 2009-11-13 02:25:08 +0000 | [diff] [blame] | 699 | pos += 1 |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 700 | |
Yury Selivanov | 7544508 | 2015-05-11 22:57:16 -0400 | [diff] [blame] | 701 | if stashed: |
| 702 | yield stashed |
| 703 | stashed = None |
| 704 | |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 705 | for indent in indents[1:]: # pop remaining indent levels |
Raymond Hettinger | a48db39 | 2009-04-29 00:34:27 +0000 | [diff] [blame] | 706 | yield TokenInfo(DEDENT, '', (lnum, 0), (lnum, 0), '') |
| 707 | yield TokenInfo(ENDMARKER, '', (lnum, 0), (lnum, 0), '') |
Guido van Rossum | fc6f533 | 1997-03-07 00:21:12 +0000 | [diff] [blame] | 708 | |
Trent Nelson | 428de65 | 2008-03-18 22:41:35 +0000 | [diff] [blame] | 709 | |
| 710 | # An undocumented, backwards compatible, API for all the places in the standard |
| 711 | # library that expect to be able to use tokenize with strings |
| 712 | def generate_tokens(readline): |
| 713 | return _tokenize(readline, None) |
Raymond Hettinger | 6c60d09 | 2010-09-09 04:32:39 +0000 | [diff] [blame] | 714 | |
Meador Inge | 14c0f03 | 2011-10-07 08:53:38 -0500 | [diff] [blame] | 715 | def main(): |
| 716 | import argparse |
| 717 | |
| 718 | # Helper error handling routines |
| 719 | def perror(message): |
| 720 | print(message, file=sys.stderr) |
| 721 | |
| 722 | def error(message, filename=None, location=None): |
| 723 | if location: |
| 724 | args = (filename,) + location + (message,) |
| 725 | perror("%s:%d:%d: error: %s" % args) |
| 726 | elif filename: |
| 727 | perror("%s: error: %s" % (filename, message)) |
| 728 | else: |
| 729 | perror("error: %s" % message) |
| 730 | sys.exit(1) |
| 731 | |
| 732 | # Parse the arguments and options |
| 733 | parser = argparse.ArgumentParser(prog='python -m tokenize') |
| 734 | parser.add_argument(dest='filename', nargs='?', |
| 735 | metavar='filename.py', |
| 736 | help='the file to tokenize; defaults to stdin') |
Meador Inge | 00c7f85 | 2012-01-19 00:44:45 -0600 | [diff] [blame] | 737 | parser.add_argument('-e', '--exact', dest='exact', action='store_true', |
| 738 | help='display token names using the exact type') |
Meador Inge | 14c0f03 | 2011-10-07 08:53:38 -0500 | [diff] [blame] | 739 | args = parser.parse_args() |
| 740 | |
| 741 | try: |
| 742 | # Tokenize the input |
| 743 | if args.filename: |
| 744 | filename = args.filename |
Victor Stinner | 9691750 | 2014-12-05 10:17:10 +0100 | [diff] [blame] | 745 | with _builtin_open(filename, 'rb') as f: |
Meador Inge | 14c0f03 | 2011-10-07 08:53:38 -0500 | [diff] [blame] | 746 | tokens = list(tokenize(f.readline)) |
| 747 | else: |
| 748 | filename = "<stdin>" |
| 749 | tokens = _tokenize(sys.stdin.readline, None) |
| 750 | |
| 751 | # Output the tokenization |
| 752 | for token in tokens: |
Meador Inge | 00c7f85 | 2012-01-19 00:44:45 -0600 | [diff] [blame] | 753 | token_type = token.type |
| 754 | if args.exact: |
| 755 | token_type = token.exact_type |
Meador Inge | 14c0f03 | 2011-10-07 08:53:38 -0500 | [diff] [blame] | 756 | token_range = "%d,%d-%d,%d:" % (token.start + token.end) |
| 757 | print("%-20s%-15s%-15r" % |
Meador Inge | 00c7f85 | 2012-01-19 00:44:45 -0600 | [diff] [blame] | 758 | (token_range, tok_name[token_type], token.string)) |
Meador Inge | 14c0f03 | 2011-10-07 08:53:38 -0500 | [diff] [blame] | 759 | except IndentationError as err: |
| 760 | line, column = err.args[1][1:3] |
| 761 | error(err.args[0], filename, (line, column)) |
| 762 | except TokenError as err: |
| 763 | line, column = err.args[1] |
| 764 | error(err.args[0], filename, (line, column)) |
| 765 | except SyntaxError as err: |
| 766 | error(err, filename) |
Andrew Svetlov | f7a17b4 | 2012-12-25 16:47:37 +0200 | [diff] [blame] | 767 | except OSError as err: |
Meador Inge | 14c0f03 | 2011-10-07 08:53:38 -0500 | [diff] [blame] | 768 | error(err) |
| 769 | except KeyboardInterrupt: |
| 770 | print("interrupted\n") |
| 771 | except Exception as err: |
| 772 | perror("unexpected error: %s" % err) |
| 773 | raise |
| 774 | |
Raymond Hettinger | 6c60d09 | 2010-09-09 04:32:39 +0000 | [diff] [blame] | 775 | if __name__ == "__main__": |
Meador Inge | 14c0f03 | 2011-10-07 08:53:38 -0500 | [diff] [blame] | 776 | main() |