projects / OpenSPARC-T2-SAM / blame
| Commit | Line | Data |
|---|---|---|
| 920dae64 AT |
1 | """Parse tree transformation module. |
| 2 | ||
| 3 | Transforms Python source code into an abstract syntax tree (AST) | |
| 4 | defined in the ast module. | |
| 5 | ||
| 6 | The simplest ways to invoke this module are via parse and parseFile. | |
| 7 | parse(buf) -> AST | |
| 8 | parseFile(path) -> AST | |
| 9 | """ | |
| 10 | ||
| 11 | # Original version written by Greg Stein (gstein@lyra.org) | |
| 12 | # and Bill Tutt (rassilon@lima.mudlib.org) | |
| 13 | # February 1997. | |
| 14 | # | |
| 15 | # Modifications and improvements for Python 2.0 by Jeremy Hylton and | |
| 16 | # Mark Hammond | |
| 17 | # | |
| 18 | # Some fixes to try to have correct line number on almost all nodes | |
| 19 | # (except Module, Discard and Stmt) added by Sylvain Thenault | |
| 20 | # | |
| 21 | # Portions of this file are: | |
| 22 | # Copyright (C) 1997-1998 Greg Stein. All Rights Reserved. | |
| 23 | # | |
| 24 | # This module is provided under a BSD-ish license. See | |
| 25 | # http://www.opensource.org/licenses/bsd-license.html | |
| 26 | # and replace OWNER, ORGANIZATION, and YEAR as appropriate. | |
| 27 | ||
| 28 | from compiler.ast import * | |
| 29 | import parser | |
| 30 | import symbol | |
| 31 | import token | |
| 32 | import sys | |
| 33 | ||
| 34 | class WalkerError(StandardError): | |
| 35 | pass | |
| 36 | ||
| 37 | from consts import CO_VARARGS, CO_VARKEYWORDS | |
| 38 | from consts import OP_ASSIGN, OP_DELETE, OP_APPLY | |
| 39 | ||
| 40 | def parseFile(path): | |
| 41 | f = open(path, "U") | |
| 42 | # XXX The parser API tolerates files without a trailing newline, | |
| 43 | # but not strings without a trailing newline. Always add an extra | |
| 44 | # newline to the file contents, since we're going through the string | |
| 45 | # version of the API. | |
| 46 | src = f.read() + "\n" | |
| 47 | f.close() | |
| 48 | return parse(src) | |
| 49 | ||
| 50 | def parse(buf, mode="exec"): | |
| 51 | if mode == "exec" or mode == "single": | |
| 52 | return Transformer().parsesuite(buf) | |
| 53 | elif mode == "eval": | |
| 54 | return Transformer().parseexpr(buf) | |
| 55 | else: | |
| 56 | raise ValueError("compile() arg 3 must be" | |
| 57 | " 'exec' or 'eval' or 'single'") | |
| 58 | ||
| 59 | def asList(nodes): | |
| 60 | l = [] | |
| 61 | for item in nodes: | |
| 62 | if hasattr(item, "asList"): | |
| 63 | l.append(item.asList()) | |
| 64 | else: | |
| 65 | if type(item) is type( (None, None) ): | |
| 66 | l.append(tuple(asList(item))) | |
| 67 | elif type(item) is type( [] ): | |
| 68 | l.append(asList(item)) | |
| 69 | else: | |
| 70 | l.append(item) | |
| 71 | return l | |
| 72 | ||
| 73 | def extractLineNo(ast): | |
| 74 | if not isinstance(ast[1], tuple): | |
| 75 | # get a terminal node | |
| 76 | return ast[2] | |
| 77 | for child in ast[1:]: | |
| 78 | if isinstance(child, tuple): | |
| 79 | lineno = extractLineNo(child) | |
| 80 | if lineno is not None: | |
| 81 | return lineno | |
| 82 | ||
| 83 | def Node(*args): | |
| 84 | kind = args[0] | |
| 85 | if nodes.has_key(kind): | |
| 86 | try: | |
| 87 | return nodes[kind](*args[1:]) | |
| 88 | except TypeError: | |
| 89 | print nodes[kind], len(args), args | |
| 90 | raise | |
| 91 | else: | |
| 92 | raise WalkerError, "Can't find appropriate Node type: %s" % str(args) | |
| 93 | #return apply(ast.Node, args) | |
| 94 | ||
| 95 | class Transformer: | |
| 96 | """Utility object for transforming Python parse trees. | |
| 97 | ||
| 98 | Exposes the following methods: | |
| 99 | tree = transform(ast_tree) | |
| 100 | tree = parsesuite(text) | |
| 101 | tree = parseexpr(text) | |
| 102 | tree = parsefile(fileob | filename) | |
| 103 | """ | |
| 104 | ||
| 105 | def __init__(self): | |
| 106 | self._dispatch = {} | |
| 107 | for value, name in symbol.sym_name.items(): | |
| 108 | if hasattr(self, name): | |
| 109 | self._dispatch[value] = getattr(self, name) | |
| 110 | self._dispatch[token.NEWLINE] = self.com_NEWLINE | |
| 111 | self._atom_dispatch = {token.LPAR: self.atom_lpar, | |
| 112 | token.LSQB: self.atom_lsqb, | |
| 113 | token.LBRACE: self.atom_lbrace, | |
| 114 | token.BACKQUOTE: self.atom_backquote, | |
| 115 | token.NUMBER: self.atom_number, | |
| 116 | token.STRING: self.atom_string, | |
| 117 | token.NAME: self.atom_name, | |
| 118 | } | |
| 119 | self.encoding = None | |
| 120 | ||
| 121 | def transform(self, tree): | |
| 122 | """Transform an AST into a modified parse tree.""" | |
| 123 | if not (isinstance(tree, tuple) or isinstance(tree, list)): | |
| 124 | tree = parser.ast2tuple(tree, line_info=1) | |
| 125 | return self.compile_node(tree) | |
| 126 | ||
| 127 | def parsesuite(self, text): | |
| 128 | """Return a modified parse tree for the given suite text.""" | |
| 129 | return self.transform(parser.suite(text)) | |
| 130 | ||
| 131 | def parseexpr(self, text): | |
| 132 | """Return a modified parse tree for the given expression text.""" | |
| 133 | return self.transform(parser.expr(text)) | |
| 134 | ||
| 135 | def parsefile(self, file): | |
| 136 | """Return a modified parse tree for the contents of the given file.""" | |
| 137 | if type(file) == type(''): | |
| 138 | file = open(file) | |
| 139 | return self.parsesuite(file.read()) | |
| 140 | ||
| 141 | # -------------------------------------------------------------- | |
| 142 | # | |
| 143 | # PRIVATE METHODS | |
| 144 | # | |
| 145 | ||
| 146 | def compile_node(self, node): | |
| 147 | ### emit a line-number node? | |
| 148 | n = node[0] | |
| 149 | ||
| 150 | if n == symbol.encoding_decl: | |
| 151 | self.encoding = node[2] | |
| 152 | node = node[1] | |
| 153 | n = node[0] | |
| 154 | ||
| 155 | if n == symbol.single_input: | |
| 156 | return self.single_input(node[1:]) | |
| 157 | if n == symbol.file_input: | |
| 158 | return self.file_input(node[1:]) | |
| 159 | if n == symbol.eval_input: | |
| 160 | return self.eval_input(node[1:]) | |
| 161 | if n == symbol.lambdef: | |
| 162 | return self.lambdef(node[1:]) | |
| 163 | if n == symbol.funcdef: | |
| 164 | return self.funcdef(node[1:]) | |
| 165 | if n == symbol.classdef: | |
| 166 | return self.classdef(node[1:]) | |
| 167 | ||
| 168 | raise WalkerError, ('unexpected node type', n) | |
| 169 | ||
| 170 | def single_input(self, node): | |
| 171 | ### do we want to do anything about being "interactive" ? | |
| 172 | ||
| 173 | # NEWLINE | simple_stmt | compound_stmt NEWLINE | |
| 174 | n = node[0][0] | |
| 175 | if n != token.NEWLINE: | |
| 176 | return self.com_stmt(node[0]) | |
| 177 | ||
| 178 | return Pass() | |
| 179 | ||
| 180 | def file_input(self, nodelist): | |
| 181 | doc = self.get_docstring(nodelist, symbol.file_input) | |
| 182 | if doc is not None: | |
| 183 | i = 1 | |
| 184 | else: | |
| 185 | i = 0 | |
| 186 | stmts = [] | |
| 187 | for node in nodelist[i:]: | |
| 188 | if node[0] != token.ENDMARKER and node[0] != token.NEWLINE: | |
| 189 | self.com_append_stmt(stmts, node) | |
| 190 | return Module(doc, Stmt(stmts)) | |
| 191 | ||
| 192 | def eval_input(self, nodelist): | |
| 193 | # from the built-in function input() | |
| 194 | ### is this sufficient? | |
| 195 | return Expression(self.com_node(nodelist[0])) | |
| 196 | ||
| 197 | def decorator_name(self, nodelist): | |
| 198 | listlen = len(nodelist) | |
| 199 | assert listlen >= 1 and listlen % 2 == 1 | |
| 200 | ||
| 201 | item = self.atom_name(nodelist) | |
| 202 | i = 1 | |
| 203 | while i < listlen: | |
| 204 | assert nodelist[i][0] == token.DOT | |
| 205 | assert nodelist[i + 1][0] == token.NAME | |
| 206 | item = Getattr(item, nodelist[i + 1][1]) | |
| 207 | i += 2 | |
| 208 | ||
| 209 | return item | |
| 210 | ||
| 211 | def decorator(self, nodelist): | |
| 212 | # '@' dotted_name [ '(' [arglist] ')' ] | |
| 213 | assert len(nodelist) in (3, 5, 6) | |
| 214 | assert nodelist[0][0] == token.AT | |
| 215 | assert nodelist[-1][0] == token.NEWLINE | |
| 216 | ||
| 217 | assert nodelist[1][0] == symbol.dotted_name | |
| 218 | funcname = self.decorator_name(nodelist[1][1:]) | |
| 219 | ||
| 220 | if len(nodelist) > 3: | |
| 221 | assert nodelist[2][0] == token.LPAR | |
| 222 | expr = self.com_call_function(funcname, nodelist[3]) | |
| 223 | else: | |
| 224 | expr = funcname | |
| 225 | ||
| 226 | return expr | |
| 227 | ||
| 228 | def decorators(self, nodelist): | |
| 229 | # decorators: decorator ([NEWLINE] decorator)* NEWLINE | |
| 230 | items = [] | |
| 231 | for dec_nodelist in nodelist: | |
| 232 | assert dec_nodelist[0] == symbol.decorator | |
| 233 | items.append(self.decorator(dec_nodelist[1:])) | |
| 234 | return Decorators(items) | |
| 235 | ||
| 236 | def funcdef(self, nodelist): | |
| 237 | # -6 -5 -4 -3 -2 -1 | |
| 238 | # funcdef: [decorators] 'def' NAME parameters ':' suite | |
| 239 | # parameters: '(' [varargslist] ')' | |
| 240 | ||
| 241 | if len(nodelist) == 6: | |
| 242 | assert nodelist[0][0] == symbol.decorators | |
| 243 | decorators = self.decorators(nodelist[0][1:]) | |
| 244 | else: | |
| 245 | assert len(nodelist) == 5 | |
| 246 | decorators = None | |
| 247 | ||
| 248 | lineno = nodelist[-4][2] | |
| 249 | name = nodelist[-4][1] | |
| 250 | args = nodelist[-3][2] | |
| 251 | ||
| 252 | if args[0] == symbol.varargslist: | |
| 253 | names, defaults, flags = self.com_arglist(args[1:]) | |
| 254 | else: | |
| 255 | names = defaults = () | |
| 256 | flags = 0 | |
| 257 | doc = self.get_docstring(nodelist[-1]) | |
| 258 | ||
| 259 | # code for function | |
| 260 | code = self.com_node(nodelist[-1]) | |
| 261 | ||
| 262 | if doc is not None: | |
| 263 | assert isinstance(code, Stmt) | |
| 264 | assert isinstance(code.nodes[0], Discard) | |
| 265 | del code.nodes[0] | |
| 266 | return Function(decorators, name, names, defaults, flags, doc, code, | |
| 267 | lineno=lineno) | |
| 268 | ||
| 269 | def lambdef(self, nodelist): | |
| 270 | # lambdef: 'lambda' [varargslist] ':' test | |
| 271 | if nodelist[2][0] == symbol.varargslist: | |
| 272 | names, defaults, flags = self.com_arglist(nodelist[2][1:]) | |
| 273 | else: | |
| 274 | names = defaults = () | |
| 275 | flags = 0 | |
| 276 | ||
| 277 | # code for lambda | |
| 278 | code = self.com_node(nodelist[-1]) | |
| 279 | ||
| 280 | return Lambda(names, defaults, flags, code, lineno=nodelist[1][2]) | |
| 281 | ||
| 282 | def classdef(self, nodelist): | |
| 283 | # classdef: 'class' NAME ['(' testlist ')'] ':' suite | |
| 284 | ||
| 285 | name = nodelist[1][1] | |
| 286 | doc = self.get_docstring(nodelist[-1]) | |
| 287 | if nodelist[2][0] == token.COLON: | |
| 288 | bases = [] | |
| 289 | else: | |
| 290 | bases = self.com_bases(nodelist[3]) | |
| 291 | ||
| 292 | # code for class | |
| 293 | code = self.com_node(nodelist[-1]) | |
| 294 | ||
| 295 | if doc is not None: | |
| 296 | assert isinstance(code, Stmt) | |
| 297 | assert isinstance(code.nodes[0], Discard) | |
| 298 | del code.nodes[0] | |
| 299 | ||
| 300 | return Class(name, bases, doc, code, lineno=nodelist[1][2]) | |
| 301 | ||
| 302 | def stmt(self, nodelist): | |
| 303 | return self.com_stmt(nodelist[0]) | |
| 304 | ||
| 305 | small_stmt = stmt | |
| 306 | flow_stmt = stmt | |
| 307 | compound_stmt = stmt | |
| 308 | ||
| 309 | def simple_stmt(self, nodelist): | |
| 310 | # small_stmt (';' small_stmt)* [';'] NEWLINE | |
| 311 | stmts = [] | |
| 312 | for i in range(0, len(nodelist), 2): | |
| 313 | self.com_append_stmt(stmts, nodelist[i]) | |
| 314 | return Stmt(stmts) | |
| 315 | ||
| 316 | def parameters(self, nodelist): | |
| 317 | raise WalkerError | |
| 318 | ||
| 319 | def varargslist(self, nodelist): | |
| 320 | raise WalkerError | |
| 321 | ||
| 322 | def fpdef(self, nodelist): | |
| 323 | raise WalkerError | |
| 324 | ||
| 325 | def fplist(self, nodelist): | |
| 326 | raise WalkerError | |
| 327 | ||
| 328 | def dotted_name(self, nodelist): | |
| 329 | raise WalkerError | |
| 330 | ||
| 331 | def comp_op(self, nodelist): | |
| 332 | raise WalkerError | |
| 333 | ||
| 334 | def trailer(self, nodelist): | |
| 335 | raise WalkerError | |
| 336 | ||
| 337 | def sliceop(self, nodelist): | |
| 338 | raise WalkerError | |
| 339 | ||
| 340 | def argument(self, nodelist): | |
| 341 | raise WalkerError | |
| 342 | ||
| 343 | # -------------------------------------------------------------- | |
| 344 | # | |
| 345 | # STATEMENT NODES (invoked by com_node()) | |
| 346 | # | |
| 347 | ||
| 348 | def expr_stmt(self, nodelist): | |
| 349 | # augassign testlist | testlist ('=' testlist)* | |
| 350 | en = nodelist[-1] | |
| 351 | exprNode = self.lookup_node(en)(en[1:]) | |
| 352 | if len(nodelist) == 1: | |
| 353 | return Discard(exprNode, lineno=exprNode.lineno) | |
| 354 | if nodelist[1][0] == token.EQUAL: | |
| 355 | nodesl = [] | |
| 356 | for i in range(0, len(nodelist) - 2, 2): | |
| 357 | nodesl.append(self.com_assign(nodelist[i], OP_ASSIGN)) | |
| 358 | return Assign(nodesl, exprNode, lineno=nodelist[1][2]) | |
| 359 | else: | |
| 360 | lval = self.com_augassign(nodelist[0]) | |
| 361 | op = self.com_augassign_op(nodelist[1]) | |
| 362 | return AugAssign(lval, op[1], exprNode, lineno=op[2]) | |
| 363 | raise WalkerError, "can't get here" | |
| 364 | ||
| 365 | def print_stmt(self, nodelist): | |
| 366 | # print ([ test (',' test)* [','] ] | '>>' test [ (',' test)+ [','] ]) | |
| 367 | items = [] | |
| 368 | if len(nodelist) == 1: | |
| 369 | start = 1 | |
| 370 | dest = None | |
| 371 | elif nodelist[1][0] == token.RIGHTSHIFT: | |
| 372 | assert len(nodelist) == 3 \ | |
| 373 | or nodelist[3][0] == token.COMMA | |
| 374 | dest = self.com_node(nodelist[2]) | |
| 375 | start = 4 | |
| 376 | else: | |
| 377 | dest = None | |
| 378 | start = 1 | |
| 379 | for i in range(start, len(nodelist), 2): | |
| 380 | items.append(self.com_node(nodelist[i])) | |
| 381 | if nodelist[-1][0] == token.COMMA: | |
| 382 | return Print(items, dest, lineno=nodelist[0][2]) | |
| 383 | return Printnl(items, dest, lineno=nodelist[0][2]) | |
| 384 | ||
| 385 | def del_stmt(self, nodelist): | |
| 386 | return self.com_assign(nodelist[1], OP_DELETE) | |
| 387 | ||
| 388 | def pass_stmt(self, nodelist): | |
| 389 | return Pass(lineno=nodelist[0][2]) | |
| 390 | ||
| 391 | def break_stmt(self, nodelist): | |
| 392 | return Break(lineno=nodelist[0][2]) | |
| 393 | ||
| 394 | def continue_stmt(self, nodelist): | |
| 395 | return Continue(lineno=nodelist[0][2]) | |
| 396 | ||
| 397 | def return_stmt(self, nodelist): | |
| 398 | # return: [testlist] | |
| 399 | if len(nodelist) < 2: | |
| 400 | return Return(Const(None), lineno=nodelist[0][2]) | |
| 401 | return Return(self.com_node(nodelist[1]), lineno=nodelist[0][2]) | |
| 402 | ||
| 403 | def yield_stmt(self, nodelist): | |
| 404 | return Yield(self.com_node(nodelist[1]), lineno=nodelist[0][2]) | |
| 405 | ||
| 406 | def raise_stmt(self, nodelist): | |
| 407 | # raise: [test [',' test [',' test]]] | |
| 408 | if len(nodelist) > 5: | |
| 409 | expr3 = self.com_node(nodelist[5]) | |
| 410 | else: | |
| 411 | expr3 = None | |
| 412 | if len(nodelist) > 3: | |
| 413 | expr2 = self.com_node(nodelist[3]) | |
| 414 | else: | |
| 415 | expr2 = None | |
| 416 | if len(nodelist) > 1: | |
| 417 | expr1 = self.com_node(nodelist[1]) | |
| 418 | else: | |
| 419 | expr1 = None | |
| 420 | return Raise(expr1, expr2, expr3, lineno=nodelist[0][2]) | |
| 421 | ||
| 422 | def import_stmt(self, nodelist): | |
| 423 | # import_stmt: import_name | import_from | |
| 424 | assert len(nodelist) == 1 | |
| 425 | return self.com_node(nodelist[0]) | |
| 426 | ||
| 427 | def import_name(self, nodelist): | |
| 428 | # import_name: 'import' dotted_as_names | |
| 429 | return Import(self.com_dotted_as_names(nodelist[1]), | |
| 430 | lineno=nodelist[0][2]) | |
| 431 | ||
| 432 | def import_from(self, nodelist): | |
| 433 | # import_from: 'from' dotted_name 'import' ('*' | | |
| 434 | # '(' import_as_names ')' | import_as_names) | |
| 435 | assert nodelist[0][1] == 'from' | |
| 436 | assert nodelist[1][0] == symbol.dotted_name | |
| 437 | assert nodelist[2][1] == 'import' | |
| 438 | fromname = self.com_dotted_name(nodelist[1]) | |
| 439 | if nodelist[3][0] == token.STAR: | |
| 440 | return From(fromname, [('*', None)], | |
| 441 | lineno=nodelist[0][2]) | |
| 442 | else: | |
| 443 | node = nodelist[3 + (nodelist[3][0] == token.LPAR)] | |
| 444 | return From(fromname, self.com_import_as_names(node), | |
| 445 | lineno=nodelist[0][2]) | |
| 446 | ||
| 447 | def global_stmt(self, nodelist): | |
| 448 | # global: NAME (',' NAME)* | |
| 449 | names = [] | |
| 450 | for i in range(1, len(nodelist), 2): | |
| 451 | names.append(nodelist[i][1]) | |
| 452 | return Global(names, lineno=nodelist[0][2]) | |
| 453 | ||
| 454 | def exec_stmt(self, nodelist): | |
| 455 | # exec_stmt: 'exec' expr ['in' expr [',' expr]] | |
| 456 | expr1 = self.com_node(nodelist[1]) | |
| 457 | if len(nodelist) >= 4: | |
| 458 | expr2 = self.com_node(nodelist[3]) | |
| 459 | if len(nodelist) >= 6: | |
| 460 | expr3 = self.com_node(nodelist[5]) | |
| 461 | else: | |
| 462 | expr3 = None | |
| 463 | else: | |
| 464 | expr2 = expr3 = None | |
| 465 | ||
| 466 | return Exec(expr1, expr2, expr3, lineno=nodelist[0][2]) | |
| 467 | ||
| 468 | def assert_stmt(self, nodelist): | |
| 469 | # 'assert': test, [',' test] | |
| 470 | expr1 = self.com_node(nodelist[1]) | |
| 471 | if (len(nodelist) == 4): | |
| 472 | expr2 = self.com_node(nodelist[3]) | |
| 473 | else: | |
| 474 | expr2 = None | |
| 475 | return Assert(expr1, expr2, lineno=nodelist[0][2]) | |
| 476 | ||
| 477 | def if_stmt(self, nodelist): | |
| 478 | # if: test ':' suite ('elif' test ':' suite)* ['else' ':' suite] | |
| 479 | tests = [] | |
| 480 | for i in range(0, len(nodelist) - 3, 4): | |
| 481 | testNode = self.com_node(nodelist[i + 1]) | |
| 482 | suiteNode = self.com_node(nodelist[i + 3]) | |
| 483 | tests.append((testNode, suiteNode)) | |
| 484 | ||
| 485 | if len(nodelist) % 4 == 3: | |
| 486 | elseNode = self.com_node(nodelist[-1]) | |
| 487 | ## elseNode.lineno = nodelist[-1][1][2] | |
| 488 | else: | |
| 489 | elseNode = None | |
| 490 | return If(tests, elseNode, lineno=nodelist[0][2]) | |
| 491 | ||
| 492 | def while_stmt(self, nodelist): | |
| 493 | # 'while' test ':' suite ['else' ':' suite] | |
| 494 | ||
| 495 | testNode = self.com_node(nodelist[1]) | |
| 496 | bodyNode = self.com_node(nodelist[3]) | |
| 497 | ||
| 498 | if len(nodelist) > 4: | |
| 499 | elseNode = self.com_node(nodelist[6]) | |
| 500 | else: | |
| 501 | elseNode = None | |
| 502 | ||
| 503 | return While(testNode, bodyNode, elseNode, lineno=nodelist[0][2]) | |
| 504 | ||
| 505 | def for_stmt(self, nodelist): | |
| 506 | # 'for' exprlist 'in' exprlist ':' suite ['else' ':' suite] | |
| 507 | ||
| 508 | assignNode = self.com_assign(nodelist[1], OP_ASSIGN) | |
| 509 | listNode = self.com_node(nodelist[3]) | |
| 510 | bodyNode = self.com_node(nodelist[5]) | |
| 511 | ||
| 512 | if len(nodelist) > 8: | |
| 513 | elseNode = self.com_node(nodelist[8]) | |
| 514 | else: | |
| 515 | elseNode = None | |
| 516 | ||
| 517 | return For(assignNode, listNode, bodyNode, elseNode, | |
| 518 | lineno=nodelist[0][2]) | |
| 519 | ||
| 520 | def try_stmt(self, nodelist): | |
| 521 | # 'try' ':' suite (except_clause ':' suite)+ ['else' ':' suite] | |
| 522 | # | 'try' ':' suite 'finally' ':' suite | |
| 523 | if nodelist[3][0] != symbol.except_clause: | |
| 524 | return self.com_try_finally(nodelist) | |
| 525 | ||
| 526 | return self.com_try_except(nodelist) | |
| 527 | ||
| 528 | def suite(self, nodelist): | |
| 529 | # simple_stmt | NEWLINE INDENT NEWLINE* (stmt NEWLINE*)+ DEDENT | |
| 530 | if len(nodelist) == 1: | |
| 531 | return self.com_stmt(nodelist[0]) | |
| 532 | ||
| 533 | stmts = [] | |
| 534 | for node in nodelist: | |
| 535 | if node[0] == symbol.stmt: | |
| 536 | self.com_append_stmt(stmts, node) | |
| 537 | return Stmt(stmts) | |
| 538 | ||
| 539 | # -------------------------------------------------------------- | |
| 540 | # | |
| 541 | # EXPRESSION NODES (invoked by com_node()) | |
| 542 | # | |
| 543 | ||
| 544 | def testlist(self, nodelist): | |
| 545 | # testlist: expr (',' expr)* [','] | |
| 546 | # testlist_safe: test [(',' test)+ [',']] | |
| 547 | # exprlist: expr (',' expr)* [','] | |
| 548 | return self.com_binary(Tuple, nodelist) | |
| 549 | ||
| 550 | testlist_safe = testlist # XXX | |
| 551 | testlist1 = testlist | |
| 552 | exprlist = testlist | |
| 553 | ||
| 554 | def testlist_gexp(self, nodelist): | |
| 555 | if len(nodelist) == 2 and nodelist[1][0] == symbol.gen_for: | |
| 556 | test = self.com_node(nodelist[0]) | |
| 557 | return self.com_generator_expression(test, nodelist[1]) | |
| 558 | return self.testlist(nodelist) | |
| 559 | ||
| 560 | def test(self, nodelist): | |
| 561 | # and_test ('or' and_test)* | lambdef | |
| 562 | if len(nodelist) == 1 and nodelist[0][0] == symbol.lambdef: | |
| 563 | return self.lambdef(nodelist[0]) | |
| 564 | return self.com_binary(Or, nodelist) | |
| 565 | ||
| 566 | def and_test(self, nodelist): | |
| 567 | # not_test ('and' not_test)* | |
| 568 | return self.com_binary(And, nodelist) | |
| 569 | ||
| 570 | def not_test(self, nodelist): | |
| 571 | # 'not' not_test | comparison | |
| 572 | result = self.com_node(nodelist[-1]) | |
| 573 | if len(nodelist) == 2: | |
| 574 | return Not(result, lineno=nodelist[0][2]) | |
| 575 | return result | |
| 576 | ||
| 577 | def comparison(self, nodelist): | |
| 578 | # comparison: expr (comp_op expr)* | |
| 579 | node = self.com_node(nodelist[0]) | |
| 580 | if len(nodelist) == 1: | |
| 581 | return node | |
| 582 | ||
| 583 | results = [] | |
| 584 | for i in range(2, len(nodelist), 2): | |
| 585 | nl = nodelist[i-1] | |
| 586 | ||
| 587 | # comp_op: '<' | '>' | '=' | '>=' | '<=' | '<>' | '!=' | '==' | |
| 588 | # | 'in' | 'not' 'in' | 'is' | 'is' 'not' | |
| 589 | n = nl[1] | |
| 590 | if n[0] == token.NAME: | |
| 591 | type = n[1] | |
| 592 | if len(nl) == 3: | |
| 593 | if type == 'not': | |
| 594 | type = 'not in' | |
| 595 | else: | |
| 596 | type = 'is not' | |
| 597 | else: | |
| 598 | type = _cmp_types[n[0]] | |
| 599 | ||
| 600 | lineno = nl[1][2] | |
| 601 | results.append((type, self.com_node(nodelist[i]))) | |
| 602 | ||
| 603 | # we need a special "compare" node so that we can distinguish | |
| 604 | # 3 < x < 5 from (3 < x) < 5 | |
| 605 | # the two have very different semantics and results (note that the | |
| 606 | # latter form is always true) | |
| 607 | ||
| 608 | return Compare(node, results, lineno=lineno) | |
| 609 | ||
| 610 | def expr(self, nodelist): | |
| 611 | # xor_expr ('|' xor_expr)* | |
| 612 | return self.com_binary(Bitor, nodelist) | |
| 613 | ||
| 614 | def xor_expr(self, nodelist): | |
| 615 | # xor_expr ('^' xor_expr)* | |
| 616 | return self.com_binary(Bitxor, nodelist) | |
| 617 | ||
| 618 | def and_expr(self, nodelist): | |
| 619 | # xor_expr ('&' xor_expr)* | |
| 620 | return self.com_binary(Bitand, nodelist) | |
| 621 | ||
| 622 | def shift_expr(self, nodelist): | |
| 623 | # shift_expr ('<<'|'>>' shift_expr)* | |
| 624 | node = self.com_node(nodelist[0]) | |
| 625 | for i in range(2, len(nodelist), 2): | |
| 626 | right = self.com_node(nodelist[i]) | |
| 627 | if nodelist[i-1][0] == token.LEFTSHIFT: | |
| 628 | node = LeftShift([node, right], lineno=nodelist[1][2]) | |
| 629 | elif nodelist[i-1][0] == token.RIGHTSHIFT: | |
| 630 | node = RightShift([node, right], lineno=nodelist[1][2]) | |
| 631 | else: | |
| 632 | raise ValueError, "unexpected token: %s" % nodelist[i-1][0] | |
| 633 | return node | |
| 634 | ||
| 635 | def arith_expr(self, nodelist): | |
| 636 | node = self.com_node(nodelist[0]) | |
| 637 | for i in range(2, len(nodelist), 2): | |
| 638 | right = self.com_node(nodelist[i]) | |
| 639 | if nodelist[i-1][0] == token.PLUS: | |
| 640 | node = Add([node, right], lineno=nodelist[1][2]) | |
| 641 | elif nodelist[i-1][0] == token.MINUS: | |
| 642 | node = Sub([node, right], lineno=nodelist[1][2]) | |
| 643 | else: | |
| 644 | raise ValueError, "unexpected token: %s" % nodelist[i-1][0] | |
| 645 | return node | |
| 646 | ||
| 647 | def term(self, nodelist): | |
| 648 | node = self.com_node(nodelist[0]) | |
| 649 | for i in range(2, len(nodelist), 2): | |
| 650 | right = self.com_node(nodelist[i]) | |
| 651 | t = nodelist[i-1][0] | |
| 652 | if t == token.STAR: | |
| 653 | node = Mul([node, right]) | |
| 654 | elif t == token.SLASH: | |
| 655 | node = Div([node, right]) | |
| 656 | elif t == token.PERCENT: | |
| 657 | node = Mod([node, right]) | |
| 658 | elif t == token.DOUBLESLASH: | |
| 659 | node = FloorDiv([node, right]) | |
| 660 | else: | |
| 661 | raise ValueError, "unexpected token: %s" % t | |
| 662 | node.lineno = nodelist[1][2] | |
| 663 | return node | |
| 664 | ||
| 665 | def factor(self, nodelist): | |
| 666 | elt = nodelist[0] | |
| 667 | t = elt[0] | |
| 668 | node = self.lookup_node(nodelist[-1])(nodelist[-1][1:]) | |
| 669 | # need to handle (unary op)constant here... | |
| 670 | if t == token.PLUS: | |
| 671 | return UnaryAdd(node, lineno=elt[2]) | |
| 672 | elif t == token.MINUS: | |
| 673 | return UnarySub(node, lineno=elt[2]) | |
| 674 | elif t == token.TILDE: | |
| 675 | node = Invert(node, lineno=elt[2]) | |
| 676 | return node | |
| 677 | ||
| 678 | def power(self, nodelist): | |
| 679 | # power: atom trailer* ('**' factor)* | |
| 680 | node = self.com_node(nodelist[0]) | |
| 681 | for i in range(1, len(nodelist)): | |
| 682 | elt = nodelist[i] | |
| 683 | if elt[0] == token.DOUBLESTAR: | |
| 684 | return Power([node, self.com_node(nodelist[i+1])], | |
| 685 | lineno=elt[2]) | |
| 686 | ||
| 687 | node = self.com_apply_trailer(node, elt) | |
| 688 | ||
| 689 | return node | |
| 690 | ||
| 691 | def atom(self, nodelist): | |
| 692 | return self._atom_dispatch[nodelist[0][0]](nodelist) | |
| 693 | n.lineno = nodelist[0][2] | |
| 694 | return n | |
| 695 | ||
| 696 | def atom_lpar(self, nodelist): | |
| 697 | if nodelist[1][0] == token.RPAR: | |
| 698 | return Tuple(()) | |
| 699 | return self.com_node(nodelist[1]) | |
| 700 | ||
| 701 | def atom_lsqb(self, nodelist): | |
| 702 | if nodelist[1][0] == token.RSQB: | |
| 703 | return List(()) | |
| 704 | return self.com_list_constructor(nodelist[1]) | |
| 705 | ||
| 706 | def atom_lbrace(self, nodelist): | |
| 707 | if nodelist[1][0] == token.RBRACE: | |
| 708 | return Dict(()) | |
| 709 | return self.com_dictmaker(nodelist[1]) | |
| 710 | ||
| 711 | def atom_backquote(self, nodelist): | |
| 712 | return Backquote(self.com_node(nodelist[1])) | |
| 713 | ||
| 714 | def atom_number(self, nodelist): | |
| 715 | ### need to verify this matches compile.c | |
| 716 | k = eval(nodelist[0][1]) | |
| 717 | return Const(k, lineno=nodelist[0][2]) | |
| 718 | ||
| 719 | def decode_literal(self, lit): | |
| 720 | if self.encoding: | |
| 721 | # this is particularly fragile & a bit of a | |
| 722 | # hack... changes in compile.c:parsestr and | |
| 723 | # tokenizer.c must be reflected here. | |
| 724 | if self.encoding not in ['utf-8', 'iso-8859-1']: | |
| 725 | lit = unicode(lit, 'utf-8').encode(self.encoding) | |
| 726 | return eval("# coding: %s\n%s" % (self.encoding, lit)) | |
| 727 | else: | |
| 728 | return eval(lit) | |
| 729 | ||
| 730 | def atom_string(self, nodelist): | |
| 731 | k = '' | |
| 732 | for node in nodelist: | |
| 733 | k += self.decode_literal(node[1]) | |
| 734 | return Const(k, lineno=nodelist[0][2]) | |
| 735 | ||
| 736 | def atom_name(self, nodelist): | |
| 737 | return Name(nodelist[0][1], lineno=nodelist[0][2]) | |
| 738 | ||
| 739 | # -------------------------------------------------------------- | |
| 740 | # | |
| 741 | # INTERNAL PARSING UTILITIES | |
| 742 | # | |
| 743 | ||
| 744 | # The use of com_node() introduces a lot of extra stack frames, | |
| 745 | # enough to cause a stack overflow compiling test.test_parser with | |
| 746 | # the standard interpreter recursionlimit. The com_node() is a | |
| 747 | # convenience function that hides the dispatch details, but comes | |
| 748 | # at a very high cost. It is more efficient to dispatch directly | |
| 749 | # in the callers. In these cases, use lookup_node() and call the | |
| 750 | # dispatched node directly. | |
| 751 | ||
| 752 | def lookup_node(self, node): | |
| 753 | return self._dispatch[node[0]] | |
| 754 | ||
| 755 | _callers = {} | |
| 756 | ||
| 757 | def com_node(self, node): | |
| 758 | # Note: compile.c has handling in com_node for del_stmt, pass_stmt, | |
| 759 | # break_stmt, stmt, small_stmt, flow_stmt, simple_stmt, | |
| 760 | # and compound_stmt. | |
| 761 | # We'll just dispatch them. | |
| 762 | return self._dispatch[node[0]](node[1:]) | |
| 763 | ||
| 764 | def com_NEWLINE(self, *args): | |
| 765 | # A ';' at the end of a line can make a NEWLINE token appear | |
| 766 | # here, Render it harmless. (genc discards ('discard', | |
| 767 | # ('const', xxxx)) Nodes) | |
| 768 | return Discard(Const(None)) | |
| 769 | ||
| 770 | def com_arglist(self, nodelist): | |
| 771 | # varargslist: | |
| 772 | # (fpdef ['=' test] ',')* ('*' NAME [',' '**' NAME] | '**' NAME) | |
| 773 | # | fpdef ['=' test] (',' fpdef ['=' test])* [','] | |
| 774 | # fpdef: NAME | '(' fplist ')' | |
| 775 | # fplist: fpdef (',' fpdef)* [','] | |
| 776 | names = [] | |
| 777 | defaults = [] | |
| 778 | flags = 0 | |
| 779 | ||
| 780 | i = 0 | |
| 781 | while i < len(nodelist): | |
| 782 | node = nodelist[i] | |
| 783 | if node[0] == token.STAR or node[0] == token.DOUBLESTAR: | |
| 784 | if node[0] == token.STAR: | |
| 785 | node = nodelist[i+1] | |
| 786 | if node[0] == token.NAME: | |
| 787 | names.append(node[1]) | |
| 788 | flags = flags | CO_VARARGS | |
| 789 | i = i + 3 | |
| 790 | ||
| 791 | if i < len(nodelist): | |
| 792 | # should be DOUBLESTAR | |
| 793 | t = nodelist[i][0] | |
| 794 | if t == token.DOUBLESTAR: | |
| 795 | node = nodelist[i+1] | |
| 796 | else: | |
| 797 | raise ValueError, "unexpected token: %s" % t | |
| 798 | names.append(node[1]) | |
| 799 | flags = flags | CO_VARKEYWORDS | |
| 800 | ||
| 801 | break | |
| 802 | ||
| 803 | # fpdef: NAME | '(' fplist ')' | |
| 804 | names.append(self.com_fpdef(node)) | |
| 805 | ||
| 806 | i = i + 1 | |
| 807 | if i >= len(nodelist): | |
| 808 | break | |
| 809 | ||
| 810 | if nodelist[i][0] == token.EQUAL: | |
| 811 | defaults.append(self.com_node(nodelist[i + 1])) | |
| 812 | i = i + 2 | |
| 813 | elif len(defaults): | |
| 814 | # XXX This should be a syntax error. | |
| 815 | # Treat "(a=1, b)" as "(a=1, b=None)" | |
| 816 | defaults.append(Const(None)) | |
| 817 | ||
| 818 | i = i + 1 | |
| 819 | ||
| 820 | return names, defaults, flags | |
| 821 | ||
| 822 | def com_fpdef(self, node): | |
| 823 | # fpdef: NAME | '(' fplist ')' | |
| 824 | if node[1][0] == token.LPAR: | |
| 825 | return self.com_fplist(node[2]) | |
| 826 | return node[1][1] | |
| 827 | ||
| 828 | def com_fplist(self, node): | |
| 829 | # fplist: fpdef (',' fpdef)* [','] | |
| 830 | if len(node) == 2: | |
| 831 | return self.com_fpdef(node[1]) | |
| 832 | list = [] | |
| 833 | for i in range(1, len(node), 2): | |
| 834 | list.append(self.com_fpdef(node[i])) | |
| 835 | return tuple(list) | |
| 836 | ||
| 837 | def com_dotted_name(self, node): | |
| 838 | # String together the dotted names and return the string | |
| 839 | name = "" | |
| 840 | for n in node: | |
| 841 | if type(n) == type(()) and n[0] == 1: | |
| 842 | name = name + n[1] + '.' | |
| 843 | return name[:-1] | |
| 844 | ||
| 845 | def com_dotted_as_name(self, node): | |
| 846 | assert node[0] == symbol.dotted_as_name | |
| 847 | node = node[1:] | |
| 848 | dot = self.com_dotted_name(node[0][1:]) | |
| 849 | if len(node) == 1: | |
| 850 | return dot, None | |
| 851 | assert node[1][1] == 'as' | |
| 852 | assert node[2][0] == token.NAME | |
| 853 | return dot, node[2][1] | |
| 854 | ||
| 855 | def com_dotted_as_names(self, node): | |
| 856 | assert node[0] == symbol.dotted_as_names | |
| 857 | node = node[1:] | |
| 858 | names = [self.com_dotted_as_name(node[0])] | |
| 859 | for i in range(2, len(node), 2): | |
| 860 | names.append(self.com_dotted_as_name(node[i])) | |
| 861 | return names | |
| 862 | ||
| 863 | def com_import_as_name(self, node): | |
| 864 | assert node[0] == symbol.import_as_name | |
| 865 | node = node[1:] | |
| 866 | assert node[0][0] == token.NAME | |
| 867 | if len(node) == 1: | |
| 868 | return node[0][1], None | |
| 869 | assert node[1][1] == 'as', node | |
| 870 | assert node[2][0] == token.NAME | |
| 871 | return node[0][1], node[2][1] | |
| 872 | ||
| 873 | def com_import_as_names(self, node): | |
| 874 | assert node[0] == symbol.import_as_names | |
| 875 | node = node[1:] | |
| 876 | names = [self.com_import_as_name(node[0])] | |
| 877 | for i in range(2, len(node), 2): | |
| 878 | names.append(self.com_import_as_name(node[i])) | |
| 879 | return names | |
| 880 | ||
| 881 | def com_bases(self, node): | |
| 882 | bases = [] | |
| 883 | for i in range(1, len(node), 2): | |
| 884 | bases.append(self.com_node(node[i])) | |
| 885 | return bases | |
| 886 | ||
| 887 | def com_try_finally(self, nodelist): | |
| 888 | # try_fin_stmt: "try" ":" suite "finally" ":" suite | |
| 889 | return TryFinally(self.com_node(nodelist[2]), | |
| 890 | self.com_node(nodelist[5]), | |
| 891 | lineno=nodelist[0][2]) | |
| 892 | ||
| 893 | def com_try_except(self, nodelist): | |
| 894 | # try_except: 'try' ':' suite (except_clause ':' suite)* ['else' suite] | |
| 895 | #tryexcept: [TryNode, [except_clauses], elseNode)] | |
| 896 | stmt = self.com_node(nodelist[2]) | |
| 897 | clauses = [] | |
| 898 | elseNode = None | |
| 899 | for i in range(3, len(nodelist), 3): | |
| 900 | node = nodelist[i] | |
| 901 | if node[0] == symbol.except_clause: | |
| 902 | # except_clause: 'except' [expr [',' expr]] */ | |
| 903 | if len(node) > 2: | |
| 904 | expr1 = self.com_node(node[2]) | |
| 905 | if len(node) > 4: | |
| 906 | expr2 = self.com_assign(node[4], OP_ASSIGN) | |
| 907 | else: | |
| 908 | expr2 = None | |
| 909 | else: | |
| 910 | expr1 = expr2 = None | |
| 911 | clauses.append((expr1, expr2, self.com_node(nodelist[i+2]))) | |
| 912 | ||
| 913 | if node[0] == token.NAME: | |
| 914 | elseNode = self.com_node(nodelist[i+2]) | |
| 915 | return TryExcept(self.com_node(nodelist[2]), clauses, elseNode, | |
| 916 | lineno=nodelist[0][2]) | |
| 917 | ||
| 918 | def com_augassign_op(self, node): | |
| 919 | assert node[0] == symbol.augassign | |
| 920 | return node[1] | |
| 921 | ||
| 922 | def com_augassign(self, node): | |
| 923 | """Return node suitable for lvalue of augmented assignment | |
| 924 | ||
| 925 | Names, slices, and attributes are the only allowable nodes. | |
| 926 | """ | |
| 927 | l = self.com_node(node) | |
| 928 | if l.__class__ in (Name, Slice, Subscript, Getattr): | |
| 929 | return l | |
| 930 | raise SyntaxError, "can't assign to %s" % l.__class__.__name__ | |
| 931 | ||
| 932 | def com_assign(self, node, assigning): | |
| 933 | # return a node suitable for use as an "lvalue" | |
| 934 | # loop to avoid trivial recursion | |
| 935 | while 1: | |
| 936 | t = node[0] | |
| 937 | if t == symbol.exprlist or t == symbol.testlist or t == symbol.testlist_gexp: | |
| 938 | if len(node) > 2: | |
| 939 | return self.com_assign_tuple(node, assigning) | |
| 940 | node = node[1] | |
| 941 | elif t in _assign_types: | |
| 942 | if len(node) > 2: | |
| 943 | raise SyntaxError, "can't assign to operator" | |
| 944 | node = node[1] | |
| 945 | elif t == symbol.power: | |
| 946 | if node[1][0] != symbol.atom: | |
| 947 | raise SyntaxError, "can't assign to operator" | |
| 948 | if len(node) > 2: | |
| 949 | primary = self.com_node(node[1]) | |
| 950 | for i in range(2, len(node)-1): | |
| 951 | ch = node[i] | |
| 952 | if ch[0] == token.DOUBLESTAR: | |
| 953 | raise SyntaxError, "can't assign to operator" | |
| 954 | primary = self.com_apply_trailer(primary, ch) | |
| 955 | return self.com_assign_trailer(primary, node[-1], | |
| 956 | assigning) | |
| 957 | node = node[1] | |
| 958 | elif t == symbol.atom: | |
| 959 | t = node[1][0] | |
| 960 | if t == token.LPAR: | |
| 961 | node = node[2] | |
| 962 | if node[0] == token.RPAR: | |
| 963 | raise SyntaxError, "can't assign to ()" | |
| 964 | elif t == token.LSQB: | |
| 965 | node = node[2] | |
| 966 | if node[0] == token.RSQB: | |
| 967 | raise SyntaxError, "can't assign to []" | |
| 968 | return self.com_assign_list(node, assigning) | |
| 969 | elif t == token.NAME: | |
| 970 | return self.com_assign_name(node[1], assigning) | |
| 971 | else: | |
| 972 | raise SyntaxError, "can't assign to literal" | |
| 973 | else: | |
| 974 | raise SyntaxError, "bad assignment" | |
| 975 | ||
| 976 | def com_assign_tuple(self, node, assigning): | |
| 977 | assigns = [] | |
| 978 | for i in range(1, len(node), 2): | |
| 979 | assigns.append(self.com_assign(node[i], assigning)) | |
| 980 | return AssTuple(assigns, lineno=extractLineNo(node)) | |
| 981 | ||
| 982 | def com_assign_list(self, node, assigning): | |
| 983 | assigns = [] | |
| 984 | for i in range(1, len(node), 2): | |
| 985 | if i + 1 < len(node): | |
| 986 | if node[i + 1][0] == symbol.list_for: | |
| 987 | raise SyntaxError, "can't assign to list comprehension" | |
| 988 | assert node[i + 1][0] == token.COMMA, node[i + 1] | |
| 989 | assigns.append(self.com_assign(node[i], assigning)) | |
| 990 | return AssList(assigns, lineno=extractLineNo(node)) | |
| 991 | ||
| 992 | def com_assign_name(self, node, assigning): | |
| 993 | return AssName(node[1], assigning, lineno=node[2]) | |
| 994 | ||
| 995 | def com_assign_trailer(self, primary, node, assigning): | |
| 996 | t = node[1][0] | |
| 997 | if t == token.DOT: | |
| 998 | return self.com_assign_attr(primary, node[2], assigning) | |
| 999 | if t == token.LSQB: | |
| 1000 | return self.com_subscriptlist(primary, node[2], assigning) | |
| 1001 | if t == token.LPAR: | |
| 1002 | raise SyntaxError, "can't assign to function call" | |
| 1003 | raise SyntaxError, "unknown trailer type: %s" % t | |
| 1004 | ||
| 1005 | def com_assign_attr(self, primary, node, assigning): | |
| 1006 | return AssAttr(primary, node[1], assigning, lineno=node[-1]) | |
| 1007 | ||
| 1008 | def com_binary(self, constructor, nodelist): | |
| 1009 | "Compile 'NODE (OP NODE)*' into (type, [ node1, ..., nodeN ])." | |
| 1010 | l = len(nodelist) | |
| 1011 | if l == 1: | |
| 1012 | n = nodelist[0] | |
| 1013 | return self.lookup_node(n)(n[1:]) | |
| 1014 | items = [] | |
| 1015 | for i in range(0, l, 2): | |
| 1016 | n = nodelist[i] | |
| 1017 | items.append(self.lookup_node(n)(n[1:])) | |
| 1018 | return constructor(items, lineno=extractLineNo(nodelist)) | |
| 1019 | ||
| 1020 | def com_stmt(self, node): | |
| 1021 | result = self.lookup_node(node)(node[1:]) | |
| 1022 | assert result is not None | |
| 1023 | if isinstance(result, Stmt): | |
| 1024 | return result | |
| 1025 | return Stmt([result]) | |
| 1026 | ||
| 1027 | def com_append_stmt(self, stmts, node): | |
| 1028 | result = self.lookup_node(node)(node[1:]) | |
| 1029 | assert result is not None | |
| 1030 | if isinstance(result, Stmt): | |
| 1031 | stmts.extend(result.nodes) | |
| 1032 | else: | |
| 1033 | stmts.append(result) | |
| 1034 | ||
| 1035 | if hasattr(symbol, 'list_for'): | |
| 1036 | def com_list_constructor(self, nodelist): | |
| 1037 | # listmaker: test ( list_for | (',' test)* [','] ) | |
| 1038 | values = [] | |
| 1039 | for i in range(1, len(nodelist)): | |
| 1040 | if nodelist[i][0] == symbol.list_for: | |
| 1041 | assert len(nodelist[i:]) == 1 | |
| 1042 | return self.com_list_comprehension(values[0], | |
| 1043 | nodelist[i]) | |
| 1044 | elif nodelist[i][0] == token.COMMA: | |
| 1045 | continue | |
| 1046 | values.append(self.com_node(nodelist[i])) | |
| 1047 | return List(values, lineno=values[0].lineno) | |
| 1048 | ||
| 1049 | def com_list_comprehension(self, expr, node): | |
| 1050 | # list_iter: list_for | list_if | |
| 1051 | # list_for: 'for' exprlist 'in' testlist [list_iter] | |
| 1052 | # list_if: 'if' test [list_iter] | |
| 1053 | ||
| 1054 | # XXX should raise SyntaxError for assignment | |
| 1055 | ||
| 1056 | lineno = node[1][2] | |
| 1057 | fors = [] | |
| 1058 | while node: | |
| 1059 | t = node[1][1] | |
| 1060 | if t == 'for': | |
| 1061 | assignNode = self.com_assign(node[2], OP_ASSIGN) | |
| 1062 | listNode = self.com_node(node[4]) | |
| 1063 | newfor = ListCompFor(assignNode, listNode, []) | |
| 1064 | newfor.lineno = node[1][2] | |
| 1065 | fors.append(newfor) | |
| 1066 | if len(node) == 5: | |
| 1067 | node = None | |
| 1068 | else: | |
| 1069 | node = self.com_list_iter(node[5]) | |
| 1070 | elif t == 'if': | |
| 1071 | test = self.com_node(node[2]) | |
| 1072 | newif = ListCompIf(test, lineno=node[1][2]) | |
| 1073 | newfor.ifs.append(newif) | |
| 1074 | if len(node) == 3: | |
| 1075 | node = None | |
| 1076 | else: | |
| 1077 | node = self.com_list_iter(node[3]) | |
| 1078 | else: | |
| 1079 | raise SyntaxError, \ | |
| 1080 | ("unexpected list comprehension element: %s %d" | |
| 1081 | % (node, lineno)) | |
| 1082 | return ListComp(expr, fors, lineno=lineno) | |
| 1083 | ||
| 1084 | def com_list_iter(self, node): | |
| 1085 | assert node[0] == symbol.list_iter | |
| 1086 | return node[1] | |
| 1087 | else: | |
| 1088 | def com_list_constructor(self, nodelist): | |
| 1089 | values = [] | |
| 1090 | for i in range(1, len(nodelist), 2): | |
| 1091 | values.append(self.com_node(nodelist[i])) | |
| 1092 | return List(values) | |
| 1093 | ||
| 1094 | if hasattr(symbol, 'gen_for'): | |
| 1095 | def com_generator_expression(self, expr, node): | |
| 1096 | # gen_iter: gen_for | gen_if | |
| 1097 | # gen_for: 'for' exprlist 'in' test [gen_iter] | |
| 1098 | # gen_if: 'if' test [gen_iter] | |
| 1099 | ||
| 1100 | lineno = node[1][2] | |
| 1101 | fors = [] | |
| 1102 | while node: | |
| 1103 | t = node[1][1] | |
| 1104 | if t == 'for': | |
| 1105 | assignNode = self.com_assign(node[2], OP_ASSIGN) | |
| 1106 | genNode = self.com_node(node[4]) | |
| 1107 | newfor = GenExprFor(assignNode, genNode, [], | |
| 1108 | lineno=node[1][2]) | |
| 1109 | fors.append(newfor) | |
| 1110 | if (len(node)) == 5: | |
| 1111 | node = None | |
| 1112 | else: | |
| 1113 | node = self.com_gen_iter(node[5]) | |
| 1114 | elif t == 'if': | |
| 1115 | test = self.com_node(node[2]) | |
| 1116 | newif = GenExprIf(test, lineno=node[1][2]) | |
| 1117 | newfor.ifs.append(newif) | |
| 1118 | if len(node) == 3: | |
| 1119 | node = None | |
| 1120 | else: | |
| 1121 | node = self.com_gen_iter(node[3]) | |
| 1122 | else: | |
| 1123 | raise SyntaxError, \ | |
| 1124 | ("unexpected generator expression element: %s %d" | |
| 1125 | % (node, lineno)) | |
| 1126 | fors[0].is_outmost = True | |
| 1127 | return GenExpr(GenExprInner(expr, fors), lineno=lineno) | |
| 1128 | ||
| 1129 | def com_gen_iter(self, node): | |
| 1130 | assert node[0] == symbol.gen_iter | |
| 1131 | return node[1] | |
| 1132 | ||
| 1133 | def com_dictmaker(self, nodelist): | |
| 1134 | # dictmaker: test ':' test (',' test ':' value)* [','] | |
| 1135 | items = [] | |
| 1136 | for i in range(1, len(nodelist), 4): | |
| 1137 | items.append((self.com_node(nodelist[i]), | |
| 1138 | self.com_node(nodelist[i+2]))) | |
| 1139 | return Dict(items) | |
| 1140 | ||
| 1141 | def com_apply_trailer(self, primaryNode, nodelist): | |
| 1142 | t = nodelist[1][0] | |
| 1143 | if t == token.LPAR: | |
| 1144 | return self.com_call_function(primaryNode, nodelist[2]) | |
| 1145 | if t == token.DOT: | |
| 1146 | return self.com_select_member(primaryNode, nodelist[2]) | |
| 1147 | if t == token.LSQB: | |
| 1148 | return self.com_subscriptlist(primaryNode, nodelist[2], OP_APPLY) | |
| 1149 | ||
| 1150 | raise SyntaxError, 'unknown node type: %s' % t | |
| 1151 | ||
| 1152 | def com_select_member(self, primaryNode, nodelist): | |
| 1153 | if nodelist[0] != token.NAME: | |
| 1154 | raise SyntaxError, "member must be a name" | |
| 1155 | return Getattr(primaryNode, nodelist[1], lineno=nodelist[2]) | |
| 1156 | ||
| 1157 | def com_call_function(self, primaryNode, nodelist): | |
| 1158 | if nodelist[0] == token.RPAR: | |
| 1159 | return CallFunc(primaryNode, [], lineno=extractLineNo(nodelist)) | |
| 1160 | args = [] | |
| 1161 | kw = 0 | |
| 1162 | len_nodelist = len(nodelist) | |
| 1163 | for i in range(1, len_nodelist, 2): | |
| 1164 | node = nodelist[i] | |
| 1165 | if node[0] == token.STAR or node[0] == token.DOUBLESTAR: | |
| 1166 | break | |
| 1167 | kw, result = self.com_argument(node, kw) | |
| 1168 | ||
| 1169 | if len_nodelist != 2 and isinstance(result, GenExpr) \ | |
| 1170 | and len(node) == 3 and node[2][0] == symbol.gen_for: | |
| 1171 | # allow f(x for x in y), but reject f(x for x in y, 1) | |
| 1172 | # should use f((x for x in y), 1) instead of f(x for x in y, 1) | |
| 1173 | raise SyntaxError, 'generator expression needs parenthesis' | |
| 1174 | ||
| 1175 | args.append(result) | |
| 1176 | else: | |
| 1177 | # No broken by star arg, so skip the last one we processed. | |
| 1178 | i = i + 1 | |
| 1179 | if i < len_nodelist and nodelist[i][0] == token.COMMA: | |
| 1180 | # need to accept an application that looks like "f(a, b,)" | |
| 1181 | i = i + 1 | |
| 1182 | star_node = dstar_node = None | |
| 1183 | while i < len_nodelist: | |
| 1184 | tok = nodelist[i] | |
| 1185 | ch = nodelist[i+1] | |
| 1186 | i = i + 3 | |
| 1187 | if tok[0]==token.STAR: | |
| 1188 | if star_node is not None: | |
| 1189 | raise SyntaxError, 'already have the varargs indentifier' | |
| 1190 | star_node = self.com_node(ch) | |
| 1191 | elif tok[0]==token.DOUBLESTAR: | |
| 1192 | if dstar_node is not None: | |
| 1193 | raise SyntaxError, 'already have the kwargs indentifier' | |
| 1194 | dstar_node = self.com_node(ch) | |
| 1195 | else: | |
| 1196 | raise SyntaxError, 'unknown node type: %s' % tok | |
| 1197 | return CallFunc(primaryNode, args, star_node, dstar_node, | |
| 1198 | lineno=extractLineNo(nodelist)) | |
| 1199 | ||
| 1200 | def com_argument(self, nodelist, kw): | |
| 1201 | if len(nodelist) == 3 and nodelist[2][0] == symbol.gen_for: | |
| 1202 | test = self.com_node(nodelist[1]) | |
| 1203 | return 0, self.com_generator_expression(test, nodelist[2]) | |
| 1204 | if len(nodelist) == 2: | |
| 1205 | if kw: | |
| 1206 | raise SyntaxError, "non-keyword arg after keyword arg" | |
| 1207 | return 0, self.com_node(nodelist[1]) | |
| 1208 | result = self.com_node(nodelist[3]) | |
| 1209 | n = nodelist[1] | |
| 1210 | while len(n) == 2 and n[0] != token.NAME: | |
| 1211 | n = n[1] | |
| 1212 | if n[0] != token.NAME: | |
| 1213 | raise SyntaxError, "keyword can't be an expression (%s)"%n[0] | |
| 1214 | node = Keyword(n[1], result, lineno=n[2]) | |
| 1215 | return 1, node | |
| 1216 | ||
| 1217 | def com_subscriptlist(self, primary, nodelist, assigning): | |
| 1218 | # slicing: simple_slicing | extended_slicing | |
| 1219 | # simple_slicing: primary "[" short_slice "]" | |
| 1220 | # extended_slicing: primary "[" slice_list "]" | |
| 1221 | # slice_list: slice_item ("," slice_item)* [","] | |
| 1222 | ||
| 1223 | # backwards compat slice for '[i:j]' | |
| 1224 | if len(nodelist) == 2: | |
| 1225 | sub = nodelist[1] | |
| 1226 | if (sub[1][0] == token.COLON or \ | |
| 1227 | (len(sub) > 2 and sub[2][0] == token.COLON)) and \ | |
| 1228 | sub[-1][0] != symbol.sliceop: | |
| 1229 | return self.com_slice(primary, sub, assigning) | |
| 1230 | ||
| 1231 | subscripts = [] | |
| 1232 | for i in range(1, len(nodelist), 2): | |
| 1233 | subscripts.append(self.com_subscript(nodelist[i])) | |
| 1234 | return Subscript(primary, assigning, subscripts, | |
| 1235 | lineno=extractLineNo(nodelist)) | |
| 1236 | ||
| 1237 | def com_subscript(self, node): | |
| 1238 | # slice_item: expression | proper_slice | ellipsis | |
| 1239 | ch = node[1] | |
| 1240 | t = ch[0] | |
| 1241 | if t == token.DOT and node[2][0] == token.DOT: | |
| 1242 | return Ellipsis() | |
| 1243 | if t == token.COLON or len(node) > 2: | |
| 1244 | return self.com_sliceobj(node) | |
| 1245 | return self.com_node(ch) | |
| 1246 | ||
| 1247 | def com_sliceobj(self, node): | |
| 1248 | # proper_slice: short_slice | long_slice | |
| 1249 | # short_slice: [lower_bound] ":" [upper_bound] | |
| 1250 | # long_slice: short_slice ":" [stride] | |
| 1251 | # lower_bound: expression | |
| 1252 | # upper_bound: expression | |
| 1253 | # stride: expression | |
| 1254 | # | |
| 1255 | # Note: a stride may be further slicing... | |
| 1256 | ||
| 1257 | items = [] | |
| 1258 | ||
| 1259 | if node[1][0] == token.COLON: | |
| 1260 | items.append(Const(None)) | |
| 1261 | i = 2 | |
| 1262 | else: | |
| 1263 | items.append(self.com_node(node[1])) | |
| 1264 | # i == 2 is a COLON | |
| 1265 | i = 3 | |
| 1266 | ||
| 1267 | if i < len(node) and node[i][0] == symbol.test: | |
| 1268 | items.append(self.com_node(node[i])) | |
| 1269 | i = i + 1 | |
| 1270 | else: | |
| 1271 | items.append(Const(None)) | |
| 1272 | ||
| 1273 | # a short_slice has been built. look for long_slice now by looking | |
| 1274 | # for strides... | |
| 1275 | for j in range(i, len(node)): | |
| 1276 | ch = node[j] | |
| 1277 | if len(ch) == 2: | |
| 1278 | items.append(Const(None)) | |
| 1279 | else: | |
| 1280 | items.append(self.com_node(ch[2])) | |
| 1281 | return Sliceobj(items, lineno=extractLineNo(node)) | |
| 1282 | ||
| 1283 | def com_slice(self, primary, node, assigning): | |
| 1284 | # short_slice: [lower_bound] ":" [upper_bound] | |
| 1285 | lower = upper = None | |
| 1286 | if len(node) == 3: | |
| 1287 | if node[1][0] == token.COLON: | |
| 1288 | upper = self.com_node(node[2]) | |
| 1289 | else: | |
| 1290 | lower = self.com_node(node[1]) | |
| 1291 | elif len(node) == 4: | |
| 1292 | lower = self.com_node(node[1]) | |
| 1293 | upper = self.com_node(node[3]) | |
| 1294 | return Slice(primary, assigning, lower, upper, | |
| 1295 | lineno=extractLineNo(node)) | |
| 1296 | ||
| 1297 | def get_docstring(self, node, n=None): | |
| 1298 | if n is None: | |
| 1299 | n = node[0] | |
| 1300 | node = node[1:] | |
| 1301 | if n == symbol.suite: | |
| 1302 | if len(node) == 1: | |
| 1303 | return self.get_docstring(node[0]) | |
| 1304 | for sub in node: | |
| 1305 | if sub[0] == symbol.stmt: | |
| 1306 | return self.get_docstring(sub) | |
| 1307 | return None | |
| 1308 | if n == symbol.file_input: | |
| 1309 | for sub in node: | |
| 1310 | if sub[0] == symbol.stmt: | |
| 1311 | return self.get_docstring(sub) | |
| 1312 | return None | |
| 1313 | if n == symbol.atom: | |
| 1314 | if node[0][0] == token.STRING: | |
| 1315 | s = '' | |
| 1316 | for t in node: | |
| 1317 | s = s + eval(t[1]) | |
| 1318 | return s | |
| 1319 | return None | |
| 1320 | if n == symbol.stmt or n == symbol.simple_stmt \ | |
| 1321 | or n == symbol.small_stmt: | |
| 1322 | return self.get_docstring(node[0]) | |
| 1323 | if n in _doc_nodes and len(node) == 1: | |
| 1324 | return self.get_docstring(node[0]) | |
| 1325 | return None | |
| 1326 | ||
| 1327 | ||
| 1328 | _doc_nodes = [ | |
| 1329 | symbol.expr_stmt, | |
| 1330 | symbol.testlist, | |
| 1331 | symbol.testlist_safe, | |
| 1332 | symbol.test, | |
| 1333 | symbol.and_test, | |
| 1334 | symbol.not_test, | |
| 1335 | symbol.comparison, | |
| 1336 | symbol.expr, | |
| 1337 | symbol.xor_expr, | |
| 1338 | symbol.and_expr, | |
| 1339 | symbol.shift_expr, | |
| 1340 | symbol.arith_expr, | |
| 1341 | symbol.term, | |
| 1342 | symbol.factor, | |
| 1343 | symbol.power, | |
| 1344 | ] | |
| 1345 | ||
| 1346 | # comp_op: '<' | '>' | '=' | '>=' | '<=' | '<>' | '!=' | '==' | |
| 1347 | # | 'in' | 'not' 'in' | 'is' | 'is' 'not' | |
| 1348 | _cmp_types = { | |
| 1349 | token.LESS : '<', | |
| 1350 | token.GREATER : '>', | |
| 1351 | token.EQEQUAL : '==', | |
| 1352 | token.EQUAL : '==', | |
| 1353 | token.LESSEQUAL : '<=', | |
| 1354 | token.GREATEREQUAL : '>=', | |
| 1355 | token.NOTEQUAL : '!=', | |
| 1356 | } | |
| 1357 | ||
| 1358 | _legal_node_types = [ | |
| 1359 | symbol.funcdef, | |
| 1360 | symbol.classdef, | |
| 1361 | symbol.stmt, | |
| 1362 | symbol.small_stmt, | |
| 1363 | symbol.flow_stmt, | |
| 1364 | symbol.simple_stmt, | |
| 1365 | symbol.compound_stmt, | |
| 1366 | symbol.expr_stmt, | |
| 1367 | symbol.print_stmt, | |
| 1368 | symbol.del_stmt, | |
| 1369 | symbol.pass_stmt, | |
| 1370 | symbol.break_stmt, | |
| 1371 | symbol.continue_stmt, | |
| 1372 | symbol.return_stmt, | |
| 1373 | symbol.raise_stmt, | |
| 1374 | symbol.import_stmt, | |
| 1375 | symbol.global_stmt, | |
| 1376 | symbol.exec_stmt, | |
| 1377 | symbol.assert_stmt, | |
| 1378 | symbol.if_stmt, | |
| 1379 | symbol.while_stmt, | |
| 1380 | symbol.for_stmt, | |
| 1381 | symbol.try_stmt, | |
| 1382 | symbol.suite, | |
| 1383 | symbol.testlist, | |
| 1384 | symbol.testlist_safe, | |
| 1385 | symbol.test, | |
| 1386 | symbol.and_test, | |
| 1387 | symbol.not_test, | |
| 1388 | symbol.comparison, | |
| 1389 | symbol.exprlist, | |
| 1390 | symbol.expr, | |
| 1391 | symbol.xor_expr, | |
| 1392 | symbol.and_expr, | |
| 1393 | symbol.shift_expr, | |
| 1394 | symbol.arith_expr, | |
| 1395 | symbol.term, | |
| 1396 | symbol.factor, | |
| 1397 | symbol.power, | |
| 1398 | symbol.atom, | |
| 1399 | ] | |
| 1400 | ||
| 1401 | if hasattr(symbol, 'yield_stmt'): | |
| 1402 | _legal_node_types.append(symbol.yield_stmt) | |
| 1403 | ||
| 1404 | _assign_types = [ | |
| 1405 | symbol.test, | |
| 1406 | symbol.and_test, | |
| 1407 | symbol.not_test, | |
| 1408 | symbol.comparison, | |
| 1409 | symbol.expr, | |
| 1410 | symbol.xor_expr, | |
| 1411 | symbol.and_expr, | |
| 1412 | symbol.shift_expr, | |
| 1413 | symbol.arith_expr, | |
| 1414 | symbol.term, | |
| 1415 | symbol.factor, | |
| 1416 | ] | |
| 1417 | ||
| 1418 | import types | |
| 1419 | _names = {} | |
| 1420 | for k, v in symbol.sym_name.items(): | |
| 1421 | _names[k] = v | |
| 1422 | for k, v in token.tok_name.items(): | |
| 1423 | _names[k] = v | |
| 1424 | ||
| 1425 | def debug_tree(tree): | |
| 1426 | l = [] | |
| 1427 | for elt in tree: | |
| 1428 | if type(elt) == types.IntType: | |
| 1429 | l.append(_names.get(elt, elt)) | |
| 1430 | elif type(elt) == types.StringType: | |
| 1431 | l.append(elt) | |
| 1432 | else: | |
| 1433 | l.append(debug_tree(elt)) | |
| 1434 | return l |