Make substitutions for most invalid characters rather than stripping them.
[pyTivo/wmcbrine.git] / Cheetah / Parser.py
blob3b8c14504839f0de4c791f7f472a1117568ec164
1 #!/usr/bin/env python
2 # $Id: Parser.py,v 1.135 2007/11/16 18:26:01 tavis_rudd Exp $
3 """Parser classes for Cheetah's Compiler
5 Classes:
6 ParseError( Exception )
7 _LowLevelParser( Cheetah.SourceReader.SourceReader ), basically a lexer
8 _HighLevelParser( _LowLevelParser )
9 Parser === _HighLevelParser (an alias)
11 Meta-Data
12 ================================================================================
13 Author: Tavis Rudd <tavis@damnsimple.com>
14 Version: $Revision: 1.135 $
15 Start Date: 2001/08/01
16 Last Revision Date: $Date: 2007/11/16 18:26:01 $
17 """
18 __author__ = "Tavis Rudd <tavis@damnsimple.com>"
19 __revision__ = "$Revision: 1.135 $"[11:-2]
21 import os
22 import sys
23 import re
24 from re import DOTALL, MULTILINE
25 from types import StringType, ListType, TupleType, ClassType, TypeType
26 import time
27 from tokenize import pseudoprog
28 import inspect
29 import new
30 import traceback
32 from Cheetah.SourceReader import SourceReader
33 from Cheetah import Filters
34 from Cheetah import ErrorCatchers
35 from Cheetah.Unspecified import Unspecified
36 from Cheetah.Macros.I18n import I18n
38 # re tools
39 _regexCache = {}
40 def cachedRegex(pattern):
41 if pattern not in _regexCache:
42 _regexCache[pattern] = re.compile(pattern)
43 return _regexCache[pattern]
45 def escapeRegexChars(txt,
46 escapeRE=re.compile(r'([\$\^\*\+\.\?\{\}\[\]\(\)\|\\])')):
48 """Return a txt with all special regular expressions chars escaped."""
50 return escapeRE.sub(r'\\\1' , txt)
52 def group(*choices): return '(' + '|'.join(choices) + ')'
53 def nongroup(*choices): return '(?:' + '|'.join(choices) + ')'
54 def namedGroup(name, *choices): return '(P:<' + name +'>' + '|'.join(choices) + ')'
55 def any(*choices): return apply(group, choices) + '*'
56 def maybe(*choices): return apply(group, choices) + '?'
58 ##################################################
59 ## CONSTANTS & GLOBALS ##
61 NO_CACHE = 0
62 STATIC_CACHE = 1
63 REFRESH_CACHE = 2
65 SET_LOCAL = 0
66 SET_GLOBAL = 1
67 SET_MODULE = 2
69 ##################################################
70 ## Tokens for the parser ##
72 #generic
73 identchars = "abcdefghijklmnopqrstuvwxyz" \
74 "ABCDEFGHIJKLMNOPQRSTUVWXYZ_"
75 namechars = identchars + "0123456789"
77 #operators
78 powerOp = '**'
79 unaryArithOps = ('+', '-', '~')
80 binaryArithOps = ('+', '-', '/', '//','%')
81 shiftOps = ('>>','<<')
82 bitwiseOps = ('&','|','^')
83 assignOp = '='
84 augAssignOps = ('+=','-=','/=','*=', '**=','^=','%=',
85 '>>=','<<=','&=','|=', )
86 assignmentOps = (assignOp,) + augAssignOps
88 compOps = ('<','>','==','!=','<=','>=', '<>', 'is', 'in',)
89 booleanOps = ('and','or','not')
90 operators = (powerOp,) + unaryArithOps + binaryArithOps \
91 + shiftOps + bitwiseOps + assignmentOps \
92 + compOps + booleanOps
94 delimeters = ('(',')','{','}','[',']',
95 ',','.',':',';','=','`') + augAssignOps
98 keywords = ('and', 'del', 'for', 'is', 'raise',
99 'assert', 'elif', 'from', 'lambda', 'return',
100 'break', 'else', 'global', 'not', 'try',
101 'class', 'except', 'if', 'or', 'while',
102 'continue', 'exec', 'import', 'pass',
103 'def', 'finally', 'in', 'print',
106 single3 = "'''"
107 double3 = '"""'
109 tripleQuotedStringStarts = ("'''", '"""',
110 "r'''", 'r"""', "R'''", 'R"""',
111 "u'''", 'u"""', "U'''", 'U"""',
112 "ur'''", 'ur"""', "Ur'''", 'Ur"""',
113 "uR'''", 'uR"""', "UR'''", 'UR"""')
115 tripleQuotedStringPairs = {"'''": single3, '"""': double3,
116 "r'''": single3, 'r"""': double3,
117 "u'''": single3, 'u"""': double3,
118 "ur'''": single3, 'ur"""': double3,
119 "R'''": single3, 'R"""': double3,
120 "U'''": single3, 'U"""': double3,
121 "uR'''": single3, 'uR"""': double3,
122 "Ur'''": single3, 'Ur"""': double3,
123 "UR'''": single3, 'UR"""': double3,
126 closurePairs= {')':'(',']':'[','}':'{'}
127 closurePairsRev= {'(':')','[':']','{':'}'}
129 ##################################################
130 ## Regex chunks for the parser ##
132 tripleQuotedStringREs = {}
133 def makeTripleQuoteRe(start, end):
134 start = escapeRegexChars(start)
135 end = escapeRegexChars(end)
136 return re.compile(r'(?:' + start + r').*?' + r'(?:' + end + r')', re.DOTALL)
138 for start, end in tripleQuotedStringPairs.items():
139 tripleQuotedStringREs[start] = makeTripleQuoteRe(start, end)
141 WS = r'[ \f\t]*'
142 EOL = r'\r\n|\n|\r'
143 EOLZ = EOL + r'|\Z'
144 escCharLookBehind = nongroup(r'(?<=\A)',r'(?<!\\)')
145 nameCharLookAhead = r'(?=[A-Za-z_])'
146 identRE=re.compile(r'[a-zA-Z_][a-zA-Z_0-9]*')
147 EOLre=re.compile(r'(?:\r\n|\r|\n)')
149 specialVarRE=re.compile(r'([a-zA-z_]+)@') # for matching specialVar comments
150 # e.g. ##author@ Tavis Rudd
152 unicodeDirectiveRE = re.compile(
153 r'(?:^|\r\n|\r|\n)\s*#\s{0,5}unicode[:\s]*([-\w.]*)\s*(?:\r\n|\r|\n)', re.MULTILINE)
154 encodingDirectiveRE = re.compile(
155 r'(?:^|\r\n|\r|\n)\s*#\s{0,5}encoding[:\s]*([-\w.]*)\s*(?:\r\n|\r|\n)', re.MULTILINE)
157 escapedNewlineRE = re.compile(r'(?<!\\)\\n')
159 directiveNamesAndParsers = {
160 # importing and inheritance
161 'import':None,
162 'from':None,
163 'extends': 'eatExtends',
164 'implements': 'eatImplements',
165 'super': 'eatSuper',
167 # output, filtering, and caching
168 'slurp': 'eatSlurp',
169 'raw': 'eatRaw',
170 'include': 'eatInclude',
171 'cache': 'eatCache',
172 'filter': 'eatFilter',
173 'echo': None,
174 'silent': None,
176 'call': 'eatCall',
177 'arg': 'eatCallArg',
179 'capture': 'eatCapture',
181 # declaration, assignment, and deletion
182 'attr': 'eatAttr',
183 'def': 'eatDef',
184 'block': 'eatBlock',
185 '@': 'eatDecorator',
186 'defmacro': 'eatDefMacro',
188 'closure': 'eatClosure',
190 'set': 'eatSet',
191 'del': None,
193 # flow control
194 'if': 'eatIf',
195 'while': None,
196 'for': None,
197 'else': None,
198 'elif': None,
199 'pass': None,
200 'break': None,
201 'continue': None,
202 'stop': None,
203 'return': None,
204 'yield': None,
206 # little wrappers
207 'repeat': None,
208 'unless': None,
210 # error handling
211 'assert': None,
212 'raise': None,
213 'try': None,
214 'except': None,
215 'finally': None,
216 'errorCatcher': 'eatErrorCatcher',
218 # intructions to the parser and compiler
219 'breakpoint': 'eatBreakPoint',
220 'compiler': 'eatCompiler',
221 'compiler-settings': 'eatCompilerSettings',
223 # misc
224 'shBang': 'eatShbang',
225 'encoding': 'eatEncoding',
227 'end': 'eatEndDirective',
230 endDirectiveNamesAndHandlers = {
231 'def': 'handleEndDef', # has short-form
232 'block': None, # has short-form
233 'closure': None, # has short-form
234 'cache': None, # has short-form
235 'call': None, # has short-form
236 'capture': None, # has short-form
237 'filter': None,
238 'errorCatcher':None,
239 'while': None, # has short-form
240 'for': None, # has short-form
241 'if': None, # has short-form
242 'try': None, # has short-form
243 'repeat': None, # has short-form
244 'unless': None, # has short-form
247 ##################################################
248 ## CLASSES ##
250 # @@TR: SyntaxError doesn't call exception.__str__ for some reason!
251 #class ParseError(SyntaxError):
252 class ParseError(ValueError):
253 def __init__(self, stream, msg='Invalid Syntax', extMsg='', lineno=None, col=None):
254 self.stream = stream
255 if stream.pos() >= len(stream):
256 stream.setPos(len(stream) -1)
257 self.msg = msg
258 self.extMsg = extMsg
259 self.lineno = lineno
260 self.col = col
262 def __str__(self):
263 return self.report()
265 def report(self):
266 stream = self.stream
267 if stream.filename():
268 f = " in file %s" % stream.filename()
269 else:
270 f = ''
271 report = ''
272 if self.lineno:
273 lineno = self.lineno
274 row, col, line = (lineno, (self.col or 0),
275 self.stream.splitlines()[lineno-1])
276 else:
277 row, col, line = self.stream.getRowColLine()
279 ## get the surrounding lines
280 lines = stream.splitlines()
281 prevLines = [] # (rowNum, content)
282 for i in range(1,4):
283 if row-1-i <=0:
284 break
285 prevLines.append( (row-i,lines[row-1-i]) )
287 nextLines = [] # (rowNum, content)
288 for i in range(1,4):
289 if not row-1+i < len(lines):
290 break
291 nextLines.append( (row+i,lines[row-1+i]) )
292 nextLines.reverse()
294 ## print the main message
295 report += "\n\n%s\n" %self.msg
296 report += "Line %i, column %i%s\n\n" % (row, col, f)
297 report += 'Line|Cheetah Code\n'
298 report += '----|-------------------------------------------------------------\n'
299 while prevLines:
300 lineInfo = prevLines.pop()
301 report += "%(row)-4d|%(line)s\n"% {'row':lineInfo[0], 'line':lineInfo[1]}
302 report += "%(row)-4d|%(line)s\n"% {'row':row, 'line':line}
303 report += ' '*5 +' '*(col-1) + "^\n"
305 while nextLines:
306 lineInfo = nextLines.pop()
307 report += "%(row)-4d|%(line)s\n"% {'row':lineInfo[0], 'line':lineInfo[1]}
308 ## add the extra msg
309 if self.extMsg:
310 report += self.extMsg + '\n'
312 return report
314 class ForbiddenSyntax(ParseError): pass
315 class ForbiddenExpression(ForbiddenSyntax): pass
316 class ForbiddenDirective(ForbiddenSyntax): pass
318 class CheetahVariable:
319 def __init__(self, nameChunks, useNameMapper=True, cacheToken=None,
320 rawSource=None):
321 self.nameChunks = nameChunks
322 self.useNameMapper = useNameMapper
323 self.cacheToken = cacheToken
324 self.rawSource = rawSource
326 class Placeholder(CheetahVariable): pass
328 class ArgList:
329 """Used by _LowLevelParser.getArgList()"""
331 def __init__(self):
332 self.argNames = []
333 self.defVals = []
334 self.i = 0
336 def addArgName(self, name):
337 self.argNames.append( name )
338 self.defVals.append( None )
340 def next(self):
341 self.i += 1
343 def addToDefVal(self, token):
344 i = self.i
345 if self.defVals[i] == None:
346 self.defVals[i] = ''
347 self.defVals[i] += token
349 def merge(self):
350 defVals = self.defVals
351 for i in range(len(defVals)):
352 if type(defVals[i]) == StringType:
353 defVals[i] = defVals[i].strip()
355 return map(None, [i.strip() for i in self.argNames], defVals)
357 def __str__(self):
358 return str(self.merge())
360 class _LowLevelParser(SourceReader):
361 """This class implements the methods to match or extract ('get*') the basic
362 elements of Cheetah's grammar. It does NOT handle any code generation or
363 state management.
366 _settingsManager = None
368 def setSettingsManager(self, settingsManager):
369 self._settingsManager = settingsManager
371 def setting(self, key, default=Unspecified):
372 if default is Unspecified:
373 return self._settingsManager.setting(key)
374 else:
375 return self._settingsManager.setting(key, default=default)
377 def setSetting(self, key, val):
378 self._settingsManager.setSetting(key, val)
380 def settings(self):
381 return self._settingsManager.settings()
383 def updateSettings(self, settings):
384 self._settingsManager.updateSettings(settings)
386 def _initializeSettings(self):
387 self._settingsManager._initializeSettings()
389 def configureParser(self):
390 """Is called by the Compiler instance after the parser has had a
391 settingsManager assigned with self.setSettingsManager()
393 self._makeCheetahVarREs()
394 self._makeCommentREs()
395 self._makeDirectiveREs()
396 self._makePspREs()
397 self._possibleNonStrConstantChars = (
398 self.setting('commentStartToken')[0] +
399 self.setting('multiLineCommentStartToken')[0] +
400 self.setting('cheetahVarStartToken')[0] +
401 self.setting('directiveStartToken')[0] +
402 self.setting('PSPStartToken')[0])
403 self._nonStrConstMatchers = [
404 self.matchCommentStartToken,
405 self.matchMultiLineCommentStartToken,
406 self.matchVariablePlaceholderStart,
407 self.matchExpressionPlaceholderStart,
408 self.matchDirective,
409 self.matchPSPStartToken,
410 self.matchEOLSlurpToken,
413 ## regex setup ##
415 def _makeCheetahVarREs(self):
417 """Setup the regexs for Cheetah $var parsing."""
419 num = r'[0-9\.]+'
420 interval = (r'(?P<interval>' +
421 num + r's|' +
422 num + r'm|' +
423 num + r'h|' +
424 num + r'd|' +
425 num + r'w|' +
426 num + ')'
429 cacheToken = (r'(?:' +
430 r'(?P<REFRESH_CACHE>\*' + interval + '\*)'+
431 '|' +
432 r'(?P<STATIC_CACHE>\*)' +
433 '|' +
434 r'(?P<NO_CACHE>)' +
435 ')')
436 self.cacheTokenRE = cachedRegex(cacheToken)
438 silentPlaceholderToken = (r'(?:' +
439 r'(?P<SILENT>' +escapeRegexChars('!')+')'+
440 '|' +
441 r'(?P<NOT_SILENT>)' +
442 ')')
443 self.silentPlaceholderTokenRE = cachedRegex(silentPlaceholderToken)
445 self.cheetahVarStartRE = cachedRegex(
446 escCharLookBehind +
447 r'(?P<startToken>'+escapeRegexChars(self.setting('cheetahVarStartToken'))+')'+
448 r'(?P<silenceToken>'+silentPlaceholderToken+')'+
449 r'(?P<cacheToken>'+cacheToken+')'+
450 r'(?P<enclosure>|(?:(?:\{|\(|\[)[ \t\f]*))' + # allow WS after enclosure
451 r'(?=[A-Za-z_])')
452 validCharsLookAhead = r'(?=[A-Za-z_\*!\{\(\[])'
453 self.cheetahVarStartToken = self.setting('cheetahVarStartToken')
454 self.cheetahVarStartTokenRE = cachedRegex(
455 escCharLookBehind +
456 escapeRegexChars(self.setting('cheetahVarStartToken'))
457 +validCharsLookAhead
460 self.cheetahVarInExpressionStartTokenRE = cachedRegex(
461 escapeRegexChars(self.setting('cheetahVarStartToken'))
462 +r'(?=[A-Za-z_])'
465 self.expressionPlaceholderStartRE = cachedRegex(
466 escCharLookBehind +
467 r'(?P<startToken>' + escapeRegexChars(self.setting('cheetahVarStartToken')) + ')' +
468 r'(?P<cacheToken>' + cacheToken + ')' +
469 #r'\[[ \t\f]*'
470 r'(?:\{|\(|\[)[ \t\f]*'
471 + r'(?=[^\)\}\]])'
474 if self.setting('EOLSlurpToken'):
475 self.EOLSlurpRE = cachedRegex(
476 escapeRegexChars(self.setting('EOLSlurpToken'))
477 + r'[ \t\f]*'
478 + r'(?:'+EOL+')'
480 else:
481 self.EOLSlurpRE = None
484 def _makeCommentREs(self):
485 """Construct the regex bits that are used in comment parsing."""
486 startTokenEsc = escapeRegexChars(self.setting('commentStartToken'))
487 self.commentStartTokenRE = cachedRegex(escCharLookBehind + startTokenEsc)
488 del startTokenEsc
490 startTokenEsc = escapeRegexChars(
491 self.setting('multiLineCommentStartToken'))
492 endTokenEsc = escapeRegexChars(
493 self.setting('multiLineCommentEndToken'))
494 self.multiLineCommentTokenStartRE = cachedRegex(escCharLookBehind +
495 startTokenEsc)
496 self.multiLineCommentEndTokenRE = cachedRegex(escCharLookBehind +
497 endTokenEsc)
499 def _makeDirectiveREs(self):
500 """Construct the regexs that are used in directive parsing."""
501 startToken = self.setting('directiveStartToken')
502 endToken = self.setting('directiveEndToken')
503 startTokenEsc = escapeRegexChars(startToken)
504 endTokenEsc = escapeRegexChars(endToken)
505 validSecondCharsLookAhead = r'(?=[A-Za-z_@])'
506 reParts = [escCharLookBehind, startTokenEsc]
507 if self.setting('allowWhitespaceAfterDirectiveStartToken'):
508 reParts.append('[ \t]*')
509 reParts.append(validSecondCharsLookAhead)
510 self.directiveStartTokenRE = cachedRegex(''.join(reParts))
511 self.directiveEndTokenRE = cachedRegex(escCharLookBehind + endTokenEsc)
513 def _makePspREs(self):
514 """Setup the regexs for PSP parsing."""
515 startToken = self.setting('PSPStartToken')
516 startTokenEsc = escapeRegexChars(startToken)
517 self.PSPStartTokenRE = cachedRegex(escCharLookBehind + startTokenEsc)
518 endToken = self.setting('PSPEndToken')
519 endTokenEsc = escapeRegexChars(endToken)
520 self.PSPEndTokenRE = cachedRegex(escCharLookBehind + endTokenEsc)
523 def isLineClearToStartToken(self, pos=None):
524 return self.isLineClearToPos(pos)
526 def matchTopLevelToken(self):
527 """Returns the first match found from the following methods:
528 self.matchCommentStartToken
529 self.matchMultiLineCommentStartToken
530 self.matchVariablePlaceholderStart
531 self.matchExpressionPlaceholderStart
532 self.matchDirective
533 self.matchPSPStartToken
534 self.matchEOLSlurpToken
536 Returns None if no match.
538 match = None
539 if self.peek() in self._possibleNonStrConstantChars:
540 for matcher in self._nonStrConstMatchers:
541 match = matcher()
542 if match:
543 break
544 return match
546 def matchPyToken(self):
547 match = pseudoprog.match(self.src(), self.pos())
549 if match and match.group() in tripleQuotedStringStarts:
550 TQSmatch = tripleQuotedStringREs[match.group()].match(self.src(), self.pos())
551 if TQSmatch:
552 return TQSmatch
553 return match
555 def getPyToken(self):
556 match = self.matchPyToken()
557 if match is None:
558 raise ParseError(self)
559 elif match.group() in tripleQuotedStringStarts:
560 raise ParseError(self, msg='Malformed triple-quoted string')
561 return self.readTo(match.end())
563 def matchEOLSlurpToken(self):
564 if self.EOLSlurpRE:
565 return self.EOLSlurpRE.match(self.src(), self.pos())
567 def getEOLSlurpToken(self):
568 match = self.matchEOLSlurpToken()
569 if not match:
570 raise ParseError(self, msg='Invalid EOL slurp token')
571 return self.readTo(match.end())
573 def matchCommentStartToken(self):
574 return self.commentStartTokenRE.match(self.src(), self.pos())
576 def getCommentStartToken(self):
577 match = self.matchCommentStartToken()
578 if not match:
579 raise ParseError(self, msg='Invalid single-line comment start token')
580 return self.readTo(match.end())
582 def matchMultiLineCommentStartToken(self):
583 return self.multiLineCommentTokenStartRE.match(self.src(), self.pos())
585 def getMultiLineCommentStartToken(self):
586 match = self.matchMultiLineCommentStartToken()
587 if not match:
588 raise ParseError(self, msg='Invalid multi-line comment start token')
589 return self.readTo(match.end())
591 def matchMultiLineCommentEndToken(self):
592 return self.multiLineCommentEndTokenRE.match(self.src(), self.pos())
594 def getMultiLineCommentEndToken(self):
595 match = self.matchMultiLineCommentEndToken()
596 if not match:
597 raise ParseError(self, msg='Invalid multi-line comment end token')
598 return self.readTo(match.end())
600 def getDottedName(self):
601 srcLen = len(self)
602 nameChunks = []
604 if not self.peek() in identchars:
605 raise ParseError(self)
607 while self.pos() < srcLen:
608 c = self.peek()
609 if c in namechars:
610 nameChunk = self.getIdentifier()
611 nameChunks.append(nameChunk)
612 elif c == '.':
613 if self.pos()+1 <srcLen and self.peek(1) in identchars:
614 nameChunks.append(self.getc())
615 else:
616 break
617 else:
618 break
620 return ''.join(nameChunks)
622 def matchIdentifier(self):
623 return identRE.match(self.src(), self.pos())
625 def getIdentifier(self):
626 match = self.matchIdentifier()
627 if not match:
628 raise ParseError(self, msg='Invalid identifier')
629 return self.readTo(match.end())
631 def matchOperator(self):
632 match = self.matchPyToken()
633 if match and match.group() not in operators:
634 match = None
635 return match
637 def getOperator(self):
638 match = self.matchOperator()
639 if not match:
640 raise ParseError(self, msg='Expected operator')
641 return self.readTo( match.end() )
643 def matchAssignmentOperator(self):
644 match = self.matchPyToken()
645 if match and match.group() not in assignmentOps:
646 match = None
647 return match
649 def getAssignmentOperator(self):
650 match = self.matchAssignmentOperator()
651 if not match:
652 raise ParseError(self, msg='Expected assignment operator')
653 return self.readTo( match.end() )
655 def matchDirective(self):
656 """Returns False or the name of the directive matched.
658 startPos = self.pos()
659 if not self.matchDirectiveStartToken():
660 return False
661 self.getDirectiveStartToken()
662 directiveName = self.matchDirectiveName()
663 self.setPos(startPos)
664 return directiveName
666 def matchDirectiveName(self, directiveNameChars=identchars+'0123456789-@'):
667 startPos = self.pos()
668 possibleMatches = self._directiveNamesAndParsers.keys()
669 name = ''
670 match = None
672 while not self.atEnd():
673 c = self.getc()
674 if not c in directiveNameChars:
675 break
676 name += c
677 if name == '@':
678 if not self.atEnd() and self.peek() in identchars:
679 match = '@'
680 break
681 possibleMatches = [dn for dn in possibleMatches if dn.startswith(name)]
682 if not possibleMatches:
683 break
684 elif (name in possibleMatches and (self.atEnd() or self.peek() not in directiveNameChars)):
685 match = name
686 break
688 self.setPos(startPos)
689 return match
691 def matchDirectiveStartToken(self):
692 return self.directiveStartTokenRE.match(self.src(), self.pos())
694 def getDirectiveStartToken(self):
695 match = self.matchDirectiveStartToken()
696 if not match:
697 raise ParseError(self, msg='Invalid directive start token')
698 return self.readTo(match.end())
700 def matchDirectiveEndToken(self):
701 return self.directiveEndTokenRE.match(self.src(), self.pos())
703 def getDirectiveEndToken(self):
704 match = self.matchDirectiveEndToken()
705 if not match:
706 raise ParseError(self, msg='Invalid directive end token')
707 return self.readTo(match.end())
710 def matchColonForSingleLineShortFormDirective(self):
711 if not self.atEnd() and self.peek()==':':
712 restOfLine = self[self.pos()+1:self.findEOL()]
713 restOfLine = restOfLine.strip()
714 if not restOfLine:
715 return False
716 elif self.commentStartTokenRE.match(restOfLine):
717 return False
718 else: # non-whitespace, non-commment chars found
719 return True
720 return False
722 def matchPSPStartToken(self):
723 return self.PSPStartTokenRE.match(self.src(), self.pos())
725 def matchPSPEndToken(self):
726 return self.PSPEndTokenRE.match(self.src(), self.pos())
728 def getPSPStartToken(self):
729 match = self.matchPSPStartToken()
730 if not match:
731 raise ParseError(self, msg='Invalid psp start token')
732 return self.readTo(match.end())
734 def getPSPEndToken(self):
735 match = self.matchPSPEndToken()
736 if not match:
737 raise ParseError(self, msg='Invalid psp end token')
738 return self.readTo(match.end())
740 def matchCheetahVarStart(self):
741 """includes the enclosure and cache token"""
742 return self.cheetahVarStartRE.match(self.src(), self.pos())
744 def matchCheetahVarStartToken(self):
745 """includes the enclosure and cache token"""
746 return self.cheetahVarStartTokenRE.match(self.src(), self.pos())
748 def matchCheetahVarInExpressionStartToken(self):
749 """no enclosures or cache tokens allowed"""
750 return self.cheetahVarInExpressionStartTokenRE.match(self.src(), self.pos())
752 def matchVariablePlaceholderStart(self):
753 """includes the enclosure and cache token"""
754 return self.cheetahVarStartRE.match(self.src(), self.pos())
756 def matchExpressionPlaceholderStart(self):
757 """includes the enclosure and cache token"""
758 return self.expressionPlaceholderStartRE.match(self.src(), self.pos())
760 def getCheetahVarStartToken(self):
761 """just the start token, not the enclosure or cache token"""
762 match = self.matchCheetahVarStartToken()
763 if not match:
764 raise ParseError(self, msg='Expected Cheetah $var start token')
765 return self.readTo( match.end() )
768 def getCacheToken(self):
769 try:
770 token = self.cacheTokenRE.match(self.src(), self.pos())
771 self.setPos( token.end() )
772 return token.group()
773 except:
774 raise ParseError(self, msg='Expected cache token')
776 def getSilentPlaceholderToken(self):
777 try:
778 token = self.silentPlaceholderTokenRE.match(self.src(), self.pos())
779 self.setPos( token.end() )
780 return token.group()
781 except:
782 raise ParseError(self, msg='Expected silent placeholder token')
786 def getTargetVarsList(self):
787 varnames = []
788 while not self.atEnd():
789 if self.peek() in ' \t\f':
790 self.getWhiteSpace()
791 elif self.peek() in '\r\n':
792 break
793 elif self.startswith(','):
794 self.advance()
795 elif self.startswith('in ') or self.startswith('in\t'):
796 break
797 #elif self.matchCheetahVarStart():
798 elif self.matchCheetahVarInExpressionStartToken():
799 self.getCheetahVarStartToken()
800 self.getSilentPlaceholderToken()
801 self.getCacheToken()
802 varnames.append( self.getDottedName() )
803 elif self.matchIdentifier():
804 varnames.append( self.getDottedName() )
805 else:
806 break
807 return varnames
809 def getCheetahVar(self, plain=False, skipStartToken=False):
810 """This is called when parsing inside expressions. Cache tokens are only
811 valid in placeholders so this method discards any cache tokens found.
813 if not skipStartToken:
814 self.getCheetahVarStartToken()
815 self.getSilentPlaceholderToken()
816 self.getCacheToken()
817 return self.getCheetahVarBody(plain=plain)
819 def getCheetahVarBody(self, plain=False):
820 # @@TR: this should be in the compiler
821 return self._compiler.genCheetahVar(self.getCheetahVarNameChunks(), plain=plain)
823 def getCheetahVarNameChunks(self):
826 nameChunks = list of Cheetah $var subcomponents represented as tuples
827 [ (namemapperPart,autoCall,restOfName),
829 where:
830 namemapperPart = the dottedName base
831 autocall = where NameMapper should use autocalling on namemapperPart
832 restOfName = any arglist, index, or slice
834 If restOfName contains a call arglist (e.g. '(1234)') then autocall is
835 False, otherwise it defaults to True.
837 EXAMPLE
838 ------------------------------------------------------------------------
840 if the raw CheetahVar is
841 $a.b.c[1].d().x.y.z
843 nameChunks is the list
844 [ ('a.b.c',True,'[1]'),
845 ('d',False,'()'),
846 ('x.y.z',True,''),
851 chunks = []
852 while self.pos() < len(self):
853 rest = ''
854 autoCall = True
855 if not self.peek() in identchars + '.':
856 break
857 elif self.peek() == '.':
859 if self.pos()+1 < len(self) and self.peek(1) in identchars:
860 self.advance() # discard the period as it isn't needed with NameMapper
861 else:
862 break
864 dottedName = self.getDottedName()
865 if not self.atEnd() and self.peek() in '([':
866 if self.peek() == '(':
867 rest = self.getCallArgString()
868 else:
869 rest = self.getExpression(enclosed=True)
871 period = max(dottedName.rfind('.'), 0)
872 if period:
873 chunks.append( (dottedName[:period], autoCall, '') )
874 dottedName = dottedName[period+1:]
875 if rest and rest[0]=='(':
876 autoCall = False
877 chunks.append( (dottedName, autoCall, rest) )
879 return chunks
882 def getCallArgString(self,
883 enclosures=[], # list of tuples (char, pos), where char is ({ or [
884 useNameMapper=Unspecified):
886 """ Get a method/function call argument string.
888 This method understands *arg, and **kw
891 # @@TR: this settings mangling should be removed
892 if useNameMapper is not Unspecified:
893 useNameMapper_orig = self.setting('useNameMapper')
894 self.setSetting('useNameMapper', useNameMapper)
896 if enclosures:
897 pass
898 else:
899 if not self.peek() == '(':
900 raise ParseError(self, msg="Expected '('")
901 startPos = self.pos()
902 self.getc()
903 enclosures = [('(', startPos),
906 argStringBits = ['(']
907 addBit = argStringBits.append
909 while 1:
910 if self.atEnd():
911 open = enclosures[-1][0]
912 close = closurePairsRev[open]
913 self.setPos(enclosures[-1][1])
914 raise ParseError(
915 self, msg="EOF was reached before a matching '" + close +
916 "' was found for the '" + open + "'")
918 c = self.peek()
919 if c in ")}]": # get the ending enclosure and break
920 if not enclosures:
921 raise ParseError(self)
922 c = self.getc()
923 open = closurePairs[c]
924 if enclosures[-1][0] == open:
925 enclosures.pop()
926 addBit(')')
927 break
928 else:
929 raise ParseError(self)
930 elif c in " \t\f\r\n":
931 addBit(self.getc())
932 elif self.matchCheetahVarInExpressionStartToken():
933 startPos = self.pos()
934 codeFor1stToken = self.getCheetahVar()
935 WS = self.getWhiteSpace()
936 if not self.atEnd() and self.peek() == '=':
937 nextToken = self.getPyToken()
938 if nextToken == '=':
939 endPos = self.pos()
940 self.setPos(startPos)
941 codeFor1stToken = self.getCheetahVar(plain=True)
942 self.setPos(endPos)
944 ## finally
945 addBit( codeFor1stToken + WS + nextToken )
946 else:
947 addBit( codeFor1stToken + WS)
948 elif self.matchCheetahVarStart():
949 # it has syntax that is only valid at the top level
950 self._raiseErrorAboutInvalidCheetahVarSyntaxInExpr()
951 else:
952 beforeTokenPos = self.pos()
953 token = self.getPyToken()
954 if token in ('{','(','['):
955 self.rev()
956 token = self.getExpression(enclosed=True)
957 token = self.transformToken(token, beforeTokenPos)
958 addBit(token)
960 if useNameMapper is not Unspecified:
961 self.setSetting('useNameMapper', useNameMapper_orig) # @@TR: see comment above
963 return ''.join(argStringBits)
965 def getDefArgList(self, exitPos=None, useNameMapper=False):
967 """ Get an argument list. Can be used for method/function definition
968 argument lists or for #directive argument lists. Returns a list of
969 tuples in the form (argName, defVal=None) with one tuple for each arg
970 name.
972 These defVals are always strings, so (argName, defVal=None) is safe even
973 with a case like (arg1, arg2=None, arg3=1234*2), which would be returned as
974 [('arg1', None),
975 ('arg2', 'None'),
976 ('arg3', '1234*2'),
979 This method understands *arg, and **kw
983 if self.peek() == '(':
984 self.advance()
985 else:
986 exitPos = self.findEOL() # it's a directive so break at the EOL
987 argList = ArgList()
988 onDefVal = False
990 # @@TR: this settings mangling should be removed
991 useNameMapper_orig = self.setting('useNameMapper')
992 self.setSetting('useNameMapper', useNameMapper)
994 while 1:
995 if self.atEnd():
996 raise ParseError(
997 self, msg="EOF was reached before a matching ')'"+
998 " was found for the '('")
1000 if self.pos() == exitPos:
1001 break
1003 c = self.peek()
1004 if c == ")" or self.matchDirectiveEndToken():
1005 break
1006 elif c == ":":
1007 break
1008 elif c in " \t\f\r\n":
1009 if onDefVal:
1010 argList.addToDefVal(c)
1011 self.advance()
1012 elif c == '=':
1013 onDefVal = True
1014 self.advance()
1015 elif c == ",":
1016 argList.next()
1017 onDefVal = False
1018 self.advance()
1019 elif self.startswith(self.cheetahVarStartToken) and not onDefVal:
1020 self.advance(len(self.cheetahVarStartToken))
1021 elif self.matchIdentifier() and not onDefVal:
1022 argList.addArgName( self.getIdentifier() )
1023 elif onDefVal:
1024 if self.matchCheetahVarInExpressionStartToken():
1025 token = self.getCheetahVar()
1026 elif self.matchCheetahVarStart():
1027 # it has syntax that is only valid at the top level
1028 self._raiseErrorAboutInvalidCheetahVarSyntaxInExpr()
1029 else:
1030 beforeTokenPos = self.pos()
1031 token = self.getPyToken()
1032 if token in ('{','(','['):
1033 self.rev()
1034 token = self.getExpression(enclosed=True)
1035 token = self.transformToken(token, beforeTokenPos)
1036 argList.addToDefVal(token)
1037 elif c == '*' and not onDefVal:
1038 varName = self.getc()
1039 if self.peek() == '*':
1040 varName += self.getc()
1041 if not self.matchIdentifier():
1042 raise ParseError(self)
1043 varName += self.getIdentifier()
1044 argList.addArgName(varName)
1045 else:
1046 raise ParseError(self)
1049 self.setSetting('useNameMapper', useNameMapper_orig) # @@TR: see comment above
1050 return argList.merge()
1052 def getExpressionParts(self,
1053 enclosed=False,
1054 enclosures=None, # list of tuples (char, pos), where char is ({ or [
1055 pyTokensToBreakAt=None, # only works if not enclosed
1056 useNameMapper=Unspecified,
1059 """ Get a Cheetah expression that includes $CheetahVars and break at
1060 directive end tokens, the end of an enclosure, or at a specified
1061 pyToken.
1064 if useNameMapper is not Unspecified:
1065 useNameMapper_orig = self.setting('useNameMapper')
1066 self.setSetting('useNameMapper', useNameMapper)
1068 if enclosures is None:
1069 enclosures = []
1071 srcLen = len(self)
1072 exprBits = []
1073 while 1:
1074 if self.atEnd():
1075 if enclosures:
1076 open = enclosures[-1][0]
1077 close = closurePairsRev[open]
1078 self.setPos(enclosures[-1][1])
1079 raise ParseError(
1080 self, msg="EOF was reached before a matching '" + close +
1081 "' was found for the '" + open + "'")
1082 else:
1083 break
1085 c = self.peek()
1086 if c in "{([":
1087 exprBits.append(c)
1088 enclosures.append( (c, self.pos()) )
1089 self.advance()
1090 elif enclosed and not enclosures:
1091 break
1092 elif c in "])}":
1093 if not enclosures:
1094 raise ParseError(self)
1095 open = closurePairs[c]
1096 if enclosures[-1][0] == open:
1097 enclosures.pop()
1098 exprBits.append(c)
1099 else:
1100 open = enclosures[-1][0]
1101 close = closurePairsRev[open]
1102 row, col = self.getRowCol()
1103 self.setPos(enclosures[-1][1])
1104 raise ParseError(
1105 self, msg= "A '" + c + "' was found at line " + str(row) +
1106 ", col " + str(col) +
1107 " before a matching '" + close +
1108 "' was found\nfor the '" + open + "'")
1109 self.advance()
1111 elif c in " \f\t":
1112 exprBits.append(self.getWhiteSpace())
1113 elif self.matchDirectiveEndToken() and not enclosures:
1114 break
1115 elif c == "\\" and self.pos()+1 < srcLen:
1116 eolMatch = EOLre.match(self.src(), self.pos()+1)
1117 if not eolMatch:
1118 self.advance()
1119 raise ParseError(self, msg='Line ending expected')
1120 self.setPos( eolMatch.end() )
1121 elif c in '\r\n':
1122 if enclosures:
1123 self.advance()
1124 else:
1125 break
1126 elif self.matchCheetahVarInExpressionStartToken():
1127 expr = self.getCheetahVar()
1128 exprBits.append(expr)
1129 elif self.matchCheetahVarStart():
1130 # it has syntax that is only valid at the top level
1131 self._raiseErrorAboutInvalidCheetahVarSyntaxInExpr()
1132 else:
1133 beforeTokenPos = self.pos()
1134 token = self.getPyToken()
1135 if (not enclosures
1136 and pyTokensToBreakAt
1137 and token in pyTokensToBreakAt):
1139 self.setPos(beforeTokenPos)
1140 break
1142 token = self.transformToken(token, beforeTokenPos)
1144 exprBits.append(token)
1145 if identRE.match(token):
1146 if token == 'for':
1147 expr = self.getExpression(useNameMapper=False, pyTokensToBreakAt=['in'])
1148 exprBits.append(expr)
1149 else:
1150 exprBits.append(self.getWhiteSpace())
1151 if not self.atEnd() and self.peek() == '(':
1152 exprBits.append(self.getCallArgString())
1154 if useNameMapper is not Unspecified:
1155 self.setSetting('useNameMapper', useNameMapper_orig) # @@TR: see comment above
1156 return exprBits
1158 def getExpression(self,
1159 enclosed=False,
1160 enclosures=None, # list of tuples (char, pos), where # char is ({ or [
1161 pyTokensToBreakAt=None,
1162 useNameMapper=Unspecified,
1164 """Returns the output of self.getExpressionParts() as a concatenated
1165 string rather than as a list.
1167 return ''.join(self.getExpressionParts(
1168 enclosed=enclosed, enclosures=enclosures,
1169 pyTokensToBreakAt=pyTokensToBreakAt,
1170 useNameMapper=useNameMapper))
1173 def transformToken(self, token, beforeTokenPos):
1174 """Takes a token from the expression being parsed and performs and
1175 special transformations required by Cheetah.
1177 At the moment only Cheetah's c'$placeholder strings' are transformed.
1179 if token=='c' and not self.atEnd() and self.peek() in '\'"':
1180 nextToken = self.getPyToken()
1181 token = nextToken.upper()
1182 theStr = eval(token)
1183 endPos = self.pos()
1184 if not theStr:
1185 return
1187 if token.startswith(single3) or token.startswith(double3):
1188 startPosIdx = 3
1189 else:
1190 startPosIdx = 1
1191 #print 'CHEETAH STRING', nextToken, theStr, startPosIdx
1192 self.setPos(beforeTokenPos+startPosIdx+1)
1193 outputExprs = []
1194 strConst = ''
1195 while self.pos() < (endPos-startPosIdx):
1196 if self.matchCheetahVarStart() or self.matchExpressionPlaceholderStart():
1197 if strConst:
1198 outputExprs.append(repr(strConst))
1199 strConst = ''
1200 placeholderExpr = self.getPlaceholder()
1201 outputExprs.append('str('+placeholderExpr+')')
1202 else:
1203 strConst += self.getc()
1204 self.setPos(endPos)
1205 if strConst:
1206 outputExprs.append(repr(strConst))
1207 #if not self.atEnd() and self.matches('.join('):
1208 # print 'DEBUG***'
1209 token = "''.join(["+','.join(outputExprs)+"])"
1210 return token
1212 def _raiseErrorAboutInvalidCheetahVarSyntaxInExpr(self):
1213 match = self.matchCheetahVarStart()
1214 groupdict = match.groupdict()
1215 if groupdict.get('cacheToken'):
1216 raise ParseError(
1217 self,
1218 msg='Cache tokens are not valid inside expressions. '
1219 'Use them in top-level $placeholders only.')
1220 elif groupdict.get('enclosure'):
1221 raise ParseError(
1222 self,
1223 msg='Long-form placeholders - ${}, $(), $[], etc. are not valid inside expressions. '
1224 'Use them in top-level $placeholders only.')
1225 else:
1226 raise ParseError(
1227 self,
1228 msg='This form of $placeholder syntax is not valid here.')
1231 def getPlaceholder(self, allowCacheTokens=False, plain=False, returnEverything=False):
1232 # filtered
1233 for callback in self.setting('preparsePlaceholderHooks'):
1234 callback(parser=self)
1236 startPos = self.pos()
1237 lineCol = self.getRowCol(startPos)
1238 startToken = self.getCheetahVarStartToken()
1239 silentPlaceholderToken = self.getSilentPlaceholderToken()
1240 if silentPlaceholderToken:
1241 isSilentPlaceholder = True
1242 else:
1243 isSilentPlaceholder = False
1246 if allowCacheTokens:
1247 cacheToken = self.getCacheToken()
1248 cacheTokenParts = self.cacheTokenRE.match(cacheToken).groupdict()
1249 else:
1250 cacheTokenParts = {}
1252 if self.peek() in '({[':
1253 pos = self.pos()
1254 enclosureOpenChar = self.getc()
1255 enclosures = [ (enclosureOpenChar, pos) ]
1256 self.getWhiteSpace()
1257 else:
1258 enclosures = []
1260 filterArgs = None
1261 if self.matchIdentifier():
1262 nameChunks = self.getCheetahVarNameChunks()
1263 expr = self._compiler.genCheetahVar(nameChunks[:], plain=plain)
1264 restOfExpr = None
1265 if enclosures:
1266 WS = self.getWhiteSpace()
1267 expr += WS
1268 if self.setting('allowPlaceholderFilterArgs') and self.peek()==',':
1269 filterArgs = self.getCallArgString(enclosures=enclosures)[1:-1]
1270 else:
1271 if self.peek()==closurePairsRev[enclosureOpenChar]:
1272 self.getc()
1273 else:
1274 restOfExpr = self.getExpression(enclosed=True, enclosures=enclosures)
1275 if restOfExpr[-1] == closurePairsRev[enclosureOpenChar]:
1276 restOfExpr = restOfExpr[:-1]
1277 expr += restOfExpr
1278 rawPlaceholder = self[startPos: self.pos()]
1279 else:
1280 expr = self.getExpression(enclosed=True, enclosures=enclosures)
1281 if expr[-1] == closurePairsRev[enclosureOpenChar]:
1282 expr = expr[:-1]
1283 rawPlaceholder=self[startPos: self.pos()]
1285 expr = self._applyExpressionFilters(expr,'placeholder',
1286 rawExpr=rawPlaceholder,startPos=startPos)
1287 for callback in self.setting('postparsePlaceholderHooks'):
1288 callback(parser=self)
1290 if returnEverything:
1291 return (expr, rawPlaceholder, lineCol, cacheTokenParts,
1292 filterArgs, isSilentPlaceholder)
1293 else:
1294 return expr
1297 class _HighLevelParser(_LowLevelParser):
1298 """This class is a StateMachine for parsing Cheetah source and
1299 sending state dependent code generation commands to
1300 Cheetah.Compiler.Compiler.
1302 def __init__(self, src, filename=None, breakPoint=None, compiler=None):
1303 _LowLevelParser.__init__(self, src, filename=filename, breakPoint=breakPoint)
1304 self.setSettingsManager(compiler)
1305 self._compiler = compiler
1306 self.setupState()
1307 self.configureParser()
1309 def setupState(self):
1310 self._macros = {}
1311 self._macroDetails = {}
1312 self._openDirectivesStack = []
1314 def cleanup(self):
1315 """Cleanup to remove any possible reference cycles
1317 self._macros.clear()
1318 for macroname, macroDetails in self._macroDetails.items():
1319 macroDetails.template.shutdown()
1320 del macroDetails.template
1321 self._macroDetails.clear()
1323 def configureParser(self):
1324 _LowLevelParser.configureParser(self)
1325 self._initDirectives()
1327 def _initDirectives(self):
1328 def normalizeParserVal(val):
1329 if isinstance(val, (str,unicode)):
1330 handler = getattr(self, val)
1331 elif type(val) in (ClassType, TypeType):
1332 handler = val(self)
1333 elif callable(val):
1334 handler = val
1335 elif val is None:
1336 handler = val
1337 else:
1338 raise Exception('Invalid parser/handler value %r for %s'%(val, name))
1339 return handler
1341 normalizeHandlerVal = normalizeParserVal
1343 _directiveNamesAndParsers = directiveNamesAndParsers.copy()
1344 customNamesAndParsers = self.setting('directiveNamesAndParsers',{})
1345 _directiveNamesAndParsers.update(customNamesAndParsers)
1347 _endDirectiveNamesAndHandlers = endDirectiveNamesAndHandlers.copy()
1348 customNamesAndHandlers = self.setting('endDirectiveNamesAndHandlers',{})
1349 _endDirectiveNamesAndHandlers.update(customNamesAndHandlers)
1351 self._directiveNamesAndParsers = {}
1352 for name, val in _directiveNamesAndParsers.items():
1353 if val in (False, 0):
1354 continue
1355 self._directiveNamesAndParsers[name] = normalizeParserVal(val)
1357 self._endDirectiveNamesAndHandlers = {}
1358 for name, val in _endDirectiveNamesAndHandlers.items():
1359 if val in (False, 0):
1360 continue
1361 self._endDirectiveNamesAndHandlers[name] = normalizeHandlerVal(val)
1363 self._closeableDirectives = ['def','block','closure','defmacro',
1364 'call',
1365 'capture',
1366 'cache',
1367 'filter',
1368 'if','unless',
1369 'for','while','repeat',
1370 'try',
1372 for directiveName in self.setting('closeableDirectives',[]):
1373 self._closeableDirectives.append(directiveName)
1377 macroDirectives = self.setting('macroDirectives',{})
1378 macroDirectives['i18n'] = I18n
1381 for macroName, callback in macroDirectives.items():
1382 if type(callback) in (ClassType, TypeType):
1383 callback = callback(parser=self)
1384 assert callback
1385 self._macros[macroName] = callback
1386 self._directiveNamesAndParsers[macroName] = self.eatMacroCall
1388 def _applyExpressionFilters(self, expr, exprType, rawExpr=None, startPos=None):
1389 """Pipes cheetah expressions through a set of optional filter hooks.
1391 The filters are functions which may modify the expressions or raise
1392 a ForbiddenExpression exception if the expression is not allowed. They
1393 are defined in the compiler setting 'expressionFilterHooks'.
1395 Some intended use cases:
1397 - to implement 'restricted execution' safeguards in cases where you
1398 can't trust the author of the template.
1400 - to enforce style guidelines
1402 filter call signature: (parser, expr, exprType, rawExpr=None, startPos=None)
1403 - parser is the Cheetah parser
1404 - expr is the expression to filter. In some cases the parser will have
1405 already modified it from the original source code form. For example,
1406 placeholders will have been translated into namemapper calls. If you
1407 need to work with the original source, see rawExpr.
1408 - exprType is the name of the directive, 'psp', or 'placeholder'. All
1409 lowercase. @@TR: These will eventually be replaced with a set of
1410 constants.
1411 - rawExpr is the original source string that Cheetah parsed. This
1412 might be None in some cases.
1413 - startPos is the character position in the source string/file
1414 where the parser started parsing the current expression.
1416 @@TR: I realize this use of the term 'expression' is a bit wonky as many
1417 of the 'expressions' are actually statements, but I haven't thought of
1418 a better name yet. Suggestions?
1420 for callback in self.setting('expressionFilterHooks'):
1421 expr = callback(parser=self, expr=expr, exprType=exprType,
1422 rawExpr=rawExpr, startPos=startPos)
1423 return expr
1425 def _filterDisabledDirectives(self, directiveName):
1426 directiveName = directiveName.lower()
1427 if (directiveName in self.setting('disabledDirectives')
1428 or (self.setting('enabledDirectives')
1429 and directiveName not in self.setting('enabledDirectives'))):
1430 for callback in self.setting('disabledDirectiveHooks'):
1431 callback(parser=self, directiveName=directiveName)
1432 raise ForbiddenDirective(self, msg='This %r directive is disabled'%directiveName)
1434 ## main parse loop
1436 def parse(self, breakPoint=None, assertEmptyStack=True):
1437 if breakPoint:
1438 origBP = self.breakPoint()
1439 self.setBreakPoint(breakPoint)
1440 assertEmptyStack = False
1442 while not self.atEnd():
1443 if self.matchCommentStartToken():
1444 self.eatComment()
1445 elif self.matchMultiLineCommentStartToken():
1446 self.eatMultiLineComment()
1447 elif self.matchVariablePlaceholderStart():
1448 self.eatPlaceholder()
1449 elif self.matchExpressionPlaceholderStart():
1450 self.eatPlaceholder()
1451 elif self.matchDirective():
1452 self.eatDirective()
1453 elif self.matchPSPStartToken():
1454 self.eatPSP()
1455 elif self.matchEOLSlurpToken():
1456 self.eatEOLSlurpToken()
1457 else:
1458 self.eatPlainText()
1459 if assertEmptyStack:
1460 self.assertEmptyOpenDirectivesStack()
1461 if breakPoint:
1462 self.setBreakPoint(origBP)
1464 ## non-directive eat methods
1466 def eatPlainText(self):
1467 startPos = self.pos()
1468 match = None
1469 while not self.atEnd():
1470 match = self.matchTopLevelToken()
1471 if match:
1472 break
1473 else:
1474 self.advance()
1475 strConst = self.readTo(self.pos(), start=startPos)
1476 self._compiler.addStrConst(strConst)
1477 return match
1479 def eatComment(self):
1480 isLineClearToStartToken = self.isLineClearToStartToken()
1481 if isLineClearToStartToken:
1482 self._compiler.handleWSBeforeDirective()
1483 self.getCommentStartToken()
1484 comm = self.readToEOL(gobble=isLineClearToStartToken)
1485 self._compiler.addComment(comm)
1487 def eatMultiLineComment(self):
1488 isLineClearToStartToken = self.isLineClearToStartToken()
1489 endOfFirstLine = self.findEOL()
1491 self.getMultiLineCommentStartToken()
1492 endPos = startPos = self.pos()
1493 level = 1
1494 while 1:
1495 endPos = self.pos()
1496 if self.atEnd():
1497 break
1498 if self.matchMultiLineCommentStartToken():
1499 self.getMultiLineCommentStartToken()
1500 level += 1
1501 elif self.matchMultiLineCommentEndToken():
1502 self.getMultiLineCommentEndToken()
1503 level -= 1
1504 if not level:
1505 break
1506 self.advance()
1507 comm = self.readTo(endPos, start=startPos)
1509 if not self.atEnd():
1510 self.getMultiLineCommentEndToken()
1512 if (not self.atEnd()) and self.setting('gobbleWhitespaceAroundMultiLineComments'):
1513 restOfLine = self[self.pos():self.findEOL()]
1514 if not restOfLine.strip(): # WS only to EOL
1515 self.readToEOL(gobble=isLineClearToStartToken)
1517 if isLineClearToStartToken and (self.atEnd() or self.pos() > endOfFirstLine):
1518 self._compiler.handleWSBeforeDirective()
1520 self._compiler.addComment(comm)
1522 def eatPlaceholder(self):
1523 (expr, rawPlaceholder,
1524 lineCol, cacheTokenParts,
1525 filterArgs, isSilentPlaceholder) = self.getPlaceholder(
1526 allowCacheTokens=True, returnEverything=True)
1528 self._compiler.addPlaceholder(
1529 expr,
1530 filterArgs=filterArgs,
1531 rawPlaceholder=rawPlaceholder,
1532 cacheTokenParts=cacheTokenParts,
1533 lineCol=lineCol,
1534 silentMode=isSilentPlaceholder)
1535 return
1537 def eatPSP(self):
1538 # filtered
1539 self._filterDisabledDirectives(directiveName='psp')
1540 self.getPSPStartToken()
1541 endToken = self.setting('PSPEndToken')
1542 startPos = self.pos()
1543 while not self.atEnd():
1544 if self.peek() == endToken[0]:
1545 if self.matchPSPEndToken():
1546 break
1547 self.advance()
1548 pspString = self.readTo(self.pos(), start=startPos).strip()
1549 pspString = self._applyExpressionFilters(pspString, 'psp', startPos=startPos)
1550 self._compiler.addPSP(pspString)
1551 self.getPSPEndToken()
1553 ## generic directive eat methods
1554 _simpleIndentingDirectives = '''
1555 else elif for while repeat unless try except finally'''.split()
1556 _simpleExprDirectives = '''
1557 pass continue stop return yield break
1558 del assert raise
1559 silent echo
1560 import from'''.split()
1561 _directiveHandlerNames = {'import':'addImportStatement',
1562 'from':'addImportStatement', }
1563 def eatDirective(self):
1564 directiveName = self.matchDirective()
1565 self._filterDisabledDirectives(directiveName)
1567 for callback in self.setting('preparseDirectiveHooks'):
1568 callback(parser=self, directiveName=directiveName)
1570 # subclasses can override the default behaviours here by providing an
1571 # eater method in self._directiveNamesAndParsers[directiveName]
1572 directiveParser = self._directiveNamesAndParsers.get(directiveName)
1573 if directiveParser:
1574 directiveParser()
1575 elif directiveName in self._simpleIndentingDirectives:
1576 handlerName = self._directiveHandlerNames.get(directiveName)
1577 if not handlerName:
1578 handlerName = 'add'+directiveName.capitalize()
1579 handler = getattr(self._compiler, handlerName)
1580 self.eatSimpleIndentingDirective(directiveName, callback=handler)
1581 elif directiveName in self._simpleExprDirectives:
1582 handlerName = self._directiveHandlerNames.get(directiveName)
1583 if not handlerName:
1584 handlerName = 'add'+directiveName.capitalize()
1585 handler = getattr(self._compiler, handlerName)
1586 if directiveName in ('silent', 'echo'):
1587 includeDirectiveNameInExpr = False
1588 else:
1589 includeDirectiveNameInExpr = True
1590 expr = self.eatSimpleExprDirective(
1591 directiveName,
1592 includeDirectiveNameInExpr=includeDirectiveNameInExpr)
1593 handler(expr)
1595 for callback in self.setting('postparseDirectiveHooks'):
1596 callback(parser=self, directiveName=directiveName)
1598 def _eatRestOfDirectiveTag(self, isLineClearToStartToken, endOfFirstLinePos):
1599 foundComment = False
1600 if self.matchCommentStartToken():
1601 pos = self.pos()
1602 self.advance()
1603 if not self.matchDirective():
1604 self.setPos(pos)
1605 foundComment = True
1606 self.eatComment() # this won't gobble the EOL
1607 else:
1608 self.setPos(pos)
1610 if not foundComment and self.matchDirectiveEndToken():
1611 self.getDirectiveEndToken()
1612 elif isLineClearToStartToken and (not self.atEnd()) and self.peek() in '\r\n':
1613 # still gobble the EOL if a comment was found.
1614 self.readToEOL(gobble=True)
1616 if isLineClearToStartToken and (self.atEnd() or self.pos() > endOfFirstLinePos):
1617 self._compiler.handleWSBeforeDirective()
1619 def _eatToThisEndDirective(self, directiveName):
1620 finalPos = endRawPos = startPos = self.pos()
1621 directiveChar = self.setting('directiveStartToken')[0]
1622 isLineClearToStartToken = False
1623 while not self.atEnd():
1624 if self.peek() == directiveChar:
1625 if self.matchDirective() == 'end':
1626 endRawPos = self.pos()
1627 self.getDirectiveStartToken()
1628 self.advance(len('end'))
1629 self.getWhiteSpace()
1630 if self.startswith(directiveName):
1631 if self.isLineClearToStartToken(endRawPos):
1632 isLineClearToStartToken = True
1633 endRawPos = self.findBOL(endRawPos)
1634 self.advance(len(directiveName)) # to end of directiveName
1635 self.getWhiteSpace()
1636 finalPos = self.pos()
1637 break
1638 self.advance()
1639 finalPos = endRawPos = self.pos()
1641 textEaten = self.readTo(endRawPos, start=startPos)
1642 self.setPos(finalPos)
1644 endOfFirstLinePos = self.findEOL()
1646 if self.matchDirectiveEndToken():
1647 self.getDirectiveEndToken()
1648 elif isLineClearToStartToken and (not self.atEnd()) and self.peek() in '\r\n':
1649 self.readToEOL(gobble=True)
1651 if isLineClearToStartToken and self.pos() > endOfFirstLinePos:
1652 self._compiler.handleWSBeforeDirective()
1653 return textEaten
1656 def eatSimpleExprDirective(self, directiveName, includeDirectiveNameInExpr=True):
1657 # filtered
1658 isLineClearToStartToken = self.isLineClearToStartToken()
1659 endOfFirstLine = self.findEOL()
1660 self.getDirectiveStartToken()
1661 if not includeDirectiveNameInExpr:
1662 self.advance(len(directiveName))
1663 startPos = self.pos()
1664 expr = self.getExpression().strip()
1665 directiveName = expr.split()[0]
1666 expr = self._applyExpressionFilters(expr, directiveName, startPos=startPos)
1667 if directiveName in self._closeableDirectives:
1668 self.pushToOpenDirectivesStack(directiveName)
1669 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLine)
1670 return expr
1672 def eatSimpleIndentingDirective(self, directiveName, callback,
1673 includeDirectiveNameInExpr=False):
1674 # filtered
1675 isLineClearToStartToken = self.isLineClearToStartToken()
1676 endOfFirstLinePos = self.findEOL()
1677 lineCol = self.getRowCol()
1678 self.getDirectiveStartToken()
1679 if directiveName not in 'else elif for while try except finally'.split():
1680 self.advance(len(directiveName))
1681 startPos = self.pos()
1683 self.getWhiteSpace()
1685 expr = self.getExpression(pyTokensToBreakAt=[':'])
1686 expr = self._applyExpressionFilters(expr, directiveName, startPos=startPos)
1687 if self.matchColonForSingleLineShortFormDirective():
1688 self.advance() # skip over :
1689 if directiveName in 'else elif except finally'.split():
1690 callback(expr, dedent=False, lineCol=lineCol)
1691 else:
1692 callback(expr, lineCol=lineCol)
1694 self.getWhiteSpace(max=1)
1695 self.parse(breakPoint=self.findEOL(gobble=True))
1696 self._compiler.commitStrConst()
1697 self._compiler.dedent()
1698 else:
1699 if self.peek()==':':
1700 self.advance()
1701 self.getWhiteSpace()
1702 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLinePos)
1703 if directiveName in self._closeableDirectives:
1704 self.pushToOpenDirectivesStack(directiveName)
1705 callback(expr, lineCol=lineCol)
1707 def eatEndDirective(self):
1708 isLineClearToStartToken = self.isLineClearToStartToken()
1709 self.getDirectiveStartToken()
1710 self.advance(3) # to end of 'end'
1711 self.getWhiteSpace()
1712 pos = self.pos()
1713 directiveName = False
1714 for key in self._endDirectiveNamesAndHandlers.keys():
1715 if self.find(key, pos) == pos:
1716 directiveName = key
1717 break
1718 if not directiveName:
1719 raise ParseError(self, msg='Invalid end directive')
1721 endOfFirstLinePos = self.findEOL()
1722 self.getExpression() # eat in any extra comment-like crap
1723 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLinePos)
1724 if directiveName in self._closeableDirectives:
1725 self.popFromOpenDirectivesStack(directiveName)
1727 # subclasses can override the default behaviours here by providing an
1728 # end-directive handler in self._endDirectiveNamesAndHandlers[directiveName]
1729 if self._endDirectiveNamesAndHandlers.get(directiveName):
1730 handler = self._endDirectiveNamesAndHandlers[directiveName]
1731 handler()
1732 elif directiveName in 'block capture cache call filter errorCatcher'.split():
1733 if key == 'block':
1734 self._compiler.closeBlock()
1735 elif key == 'capture':
1736 self._compiler.endCaptureRegion()
1737 elif key == 'cache':
1738 self._compiler.endCacheRegion()
1739 elif key == 'call':
1740 self._compiler.endCallRegion()
1741 elif key == 'filter':
1742 self._compiler.closeFilterBlock()
1743 elif key == 'errorCatcher':
1744 self._compiler.turnErrorCatcherOff()
1745 elif directiveName in 'while for if try repeat unless'.split():
1746 self._compiler.commitStrConst()
1747 self._compiler.dedent()
1748 elif directiveName=='closure':
1749 self._compiler.commitStrConst()
1750 self._compiler.dedent()
1751 # @@TR: temporary hack of useSearchList
1752 self.setSetting('useSearchList', self._useSearchList_orig)
1754 ## specific directive eat methods
1756 def eatBreakPoint(self):
1757 """Tells the parser to stop parsing at this point and completely ignore
1758 everything else.
1760 This is a debugging tool.
1762 self.setBreakPoint(self.pos())
1764 def eatShbang(self):
1765 # filtered
1766 self.getDirectiveStartToken()
1767 self.advance(len('shBang'))
1768 self.getWhiteSpace()
1769 startPos = self.pos()
1770 shBang = self.readToEOL()
1771 shBang = self._applyExpressionFilters(shBang, 'shbang', startPos=startPos)
1772 self._compiler.setShBang(shBang.strip())
1774 def eatEncoding(self):
1775 # filtered
1776 self.getDirectiveStartToken()
1777 self.advance(len('encoding'))
1778 self.getWhiteSpace()
1779 startPos = self.pos()
1780 encoding = self.readToEOL()
1781 encoding = self._applyExpressionFilters(encoding, 'encoding', startPos=startPos)
1782 self._compiler.setModuleEncoding(encoding.strip())
1784 def eatCompiler(self):
1785 # filtered
1786 isLineClearToStartToken = self.isLineClearToStartToken()
1787 endOfFirstLine = self.findEOL()
1788 startPos = self.pos()
1789 self.getDirectiveStartToken()
1790 self.advance(len('compiler')) # to end of 'compiler'
1791 self.getWhiteSpace()
1793 startPos = self.pos()
1794 settingName = self.getIdentifier()
1796 if settingName.lower() == 'reset':
1797 self.getExpression() # gobble whitespace & junk
1798 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLine)
1799 self._initializeSettings()
1800 self.configureParser()
1801 return
1803 self.getWhiteSpace()
1804 if self.peek() == '=':
1805 self.advance()
1806 else:
1807 raise ParseError(self)
1808 valueExpr = self.getExpression()
1809 endPos = self.pos()
1811 # @@TR: it's unlikely that anyone apply filters would have left this
1812 # directive enabled:
1813 # @@TR: fix up filtering, regardless
1814 self._applyExpressionFilters('%s=%r'%(settingName, valueExpr),
1815 'compiler', startPos=startPos)
1817 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLine)
1818 try:
1819 self._compiler.setCompilerSetting(settingName, valueExpr)
1820 except:
1821 out = sys.stderr
1822 print >> out, 'An error occurred while processing the following #compiler directive.'
1823 print >> out, '-'*80
1824 print >> out, self[startPos:endPos]
1825 print >> out, '-'*80
1826 print >> out, 'Please check the syntax of these settings.'
1827 print >> out, 'A full Python exception traceback follows.'
1828 raise
1831 def eatCompilerSettings(self):
1832 # filtered
1833 isLineClearToStartToken = self.isLineClearToStartToken()
1834 endOfFirstLine = self.findEOL()
1835 self.getDirectiveStartToken()
1836 self.advance(len('compiler-settings')) # to end of 'settings'
1838 keywords = self.getTargetVarsList()
1839 self.getExpression() # gobble any garbage
1841 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLine)
1843 if 'reset' in keywords:
1844 self._compiler._initializeSettings()
1845 self.configureParser()
1846 # @@TR: this implies a single-line #compiler-settings directive, and
1847 # thus we should parse forward for an end directive.
1848 # Subject to change in the future
1849 return
1850 startPos = self.pos()
1851 settingsStr = self._eatToThisEndDirective('compiler-settings')
1852 settingsStr = self._applyExpressionFilters(settingsStr, 'compilerSettings',
1853 startPos=startPos)
1854 try:
1855 self._compiler.setCompilerSettings(keywords=keywords, settingsStr=settingsStr)
1856 except:
1857 out = sys.stderr
1858 print >> out, 'An error occurred while processing the following compiler settings.'
1859 print >> out, '-'*80
1860 print >> out, settingsStr.strip()
1861 print >> out, '-'*80
1862 print >> out, 'Please check the syntax of these settings.'
1863 print >> out, 'A full Python exception traceback follows.'
1864 raise
1866 def eatAttr(self):
1867 # filtered
1868 isLineClearToStartToken = self.isLineClearToStartToken()
1869 endOfFirstLinePos = self.findEOL()
1870 startPos = self.pos()
1871 self.getDirectiveStartToken()
1872 self.advance(len('attr'))
1873 self.getWhiteSpace()
1874 startPos = self.pos()
1875 if self.matchCheetahVarStart():
1876 self.getCheetahVarStartToken()
1877 attribName = self.getIdentifier()
1878 self.getWhiteSpace()
1879 self.getAssignmentOperator()
1880 expr = self.getExpression()
1881 expr = self._applyExpressionFilters(expr, 'attr', startPos=startPos)
1882 self._compiler.addAttribute(attribName, expr)
1883 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLinePos)
1885 def eatDecorator(self):
1886 isLineClearToStartToken = self.isLineClearToStartToken()
1887 endOfFirstLinePos = self.findEOL()
1888 startPos = self.pos()
1889 self.getDirectiveStartToken()
1890 #self.advance() # eat @
1891 startPos = self.pos()
1892 decoratorExpr = self.getExpression()
1893 decoratorExpr = self._applyExpressionFilters(decoratorExpr, 'decorator', startPos=startPos)
1894 self._compiler.addDecorator(decoratorExpr)
1895 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLinePos)
1896 self.getWhiteSpace()
1898 directiveName = self.matchDirective()
1899 if not directiveName or directiveName not in ('def', 'block', 'closure'):
1900 raise ParseError(self, msg='Expected #def, #block or #closure')
1901 self.eatDirective()
1903 def eatDef(self):
1904 # filtered
1905 self._eatDefOrBlock('def')
1907 def eatBlock(self):
1908 # filtered
1909 startPos = self.pos()
1910 methodName, rawSignature = self._eatDefOrBlock('block')
1911 self._compiler._blockMetaData[methodName] = {
1912 'raw':rawSignature,
1913 'lineCol':self.getRowCol(startPos),
1916 def eatClosure(self):
1917 # filtered
1918 self._eatDefOrBlock('closure')
1920 def _eatDefOrBlock(self, directiveName):
1921 # filtered
1922 assert directiveName in ('def','block','closure')
1923 isLineClearToStartToken = self.isLineClearToStartToken()
1924 endOfFirstLinePos = self.findEOL()
1925 startPos = self.pos()
1926 self.getDirectiveStartToken()
1927 self.advance(len(directiveName))
1928 self.getWhiteSpace()
1929 if self.matchCheetahVarStart():
1930 self.getCheetahVarStartToken()
1931 methodName = self.getIdentifier()
1932 self.getWhiteSpace()
1933 if self.peek() == '(':
1934 argsList = self.getDefArgList()
1935 self.advance() # past the closing ')'
1936 if argsList and argsList[0][0] == 'self':
1937 del argsList[0]
1938 else:
1939 argsList=[]
1941 def includeBlockMarkers():
1942 if self.setting('includeBlockMarkers'):
1943 startMarker = self.setting('blockMarkerStart')
1944 self._compiler.addStrConst(startMarker[0] + methodName + startMarker[1])
1946 # @@TR: fix up filtering
1947 self._applyExpressionFilters(self[startPos:self.pos()], 'def', startPos=startPos)
1949 if self.matchColonForSingleLineShortFormDirective():
1950 isNestedDef = (self.setting('allowNestedDefScopes')
1951 and [name for name in self._openDirectivesStack if name=='def'])
1952 self.getc()
1953 rawSignature = self[startPos:endOfFirstLinePos]
1954 self._eatSingleLineDef(directiveName=directiveName,
1955 methodName=methodName,
1956 argsList=argsList,
1957 startPos=startPos,
1958 endPos=endOfFirstLinePos)
1959 if directiveName == 'def' and not isNestedDef:
1960 #@@TR: must come before _eatRestOfDirectiveTag ... for some reason
1961 self._compiler.closeDef()
1962 elif directiveName == 'block':
1963 includeBlockMarkers()
1964 self._compiler.closeBlock()
1965 elif directiveName == 'closure' or isNestedDef:
1966 self._compiler.dedent()
1968 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLinePos)
1969 else:
1970 if self.peek()==':':
1971 self.getc()
1972 self.pushToOpenDirectivesStack(directiveName)
1973 rawSignature = self[startPos:self.pos()]
1974 self._eatMultiLineDef(directiveName=directiveName,
1975 methodName=methodName,
1976 argsList=argsList,
1977 startPos=startPos,
1978 isLineClearToStartToken=isLineClearToStartToken)
1979 if directiveName == 'block':
1980 includeBlockMarkers()
1982 return methodName, rawSignature
1984 def _eatMultiLineDef(self, directiveName, methodName, argsList, startPos,
1985 isLineClearToStartToken=False):
1986 # filtered in calling method
1987 self.getExpression() # slurp up any garbage left at the end
1988 signature = self[startPos:self.pos()]
1989 endOfFirstLinePos = self.findEOL()
1990 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLinePos)
1991 parserComment = ('## CHEETAH: generated from ' + signature +
1992 ' at line %s, col %s' % self.getRowCol(startPos)
1993 + '.')
1995 isNestedDef = (self.setting('allowNestedDefScopes')
1996 and len([name for name in self._openDirectivesStack if name=='def'])>1)
1997 if directiveName=='block' or (directiveName=='def' and not isNestedDef):
1998 self._compiler.startMethodDef(methodName, argsList, parserComment)
1999 else: #closure
2000 self._useSearchList_orig = self.setting('useSearchList')
2001 self.setSetting('useSearchList', False)
2002 self._compiler.addClosure(methodName, argsList, parserComment)
2004 return methodName
2006 def _eatSingleLineDef(self, directiveName, methodName, argsList, startPos, endPos):
2007 # filtered in calling method
2008 fullSignature = self[startPos:endPos]
2009 parserComment = ('## Generated from ' + fullSignature +
2010 ' at line %s, col %s' % self.getRowCol(startPos)
2011 + '.')
2012 isNestedDef = (self.setting('allowNestedDefScopes')
2013 and [name for name in self._openDirectivesStack if name=='def'])
2014 if directiveName=='block' or (directiveName=='def' and not isNestedDef):
2015 self._compiler.startMethodDef(methodName, argsList, parserComment)
2016 else: #closure
2017 # @@TR: temporary hack of useSearchList
2018 useSearchList_orig = self.setting('useSearchList')
2019 self.setSetting('useSearchList', False)
2020 self._compiler.addClosure(methodName, argsList, parserComment)
2022 self.getWhiteSpace(max=1)
2023 self.parse(breakPoint=endPos)
2024 if directiveName=='closure' or isNestedDef: # @@TR: temporary hack of useSearchList
2025 self.setSetting('useSearchList', useSearchList_orig)
2027 def eatExtends(self):
2028 # filtered
2029 isLineClearToStartToken = self.isLineClearToStartToken()
2030 endOfFirstLine = self.findEOL()
2031 self.getDirectiveStartToken()
2032 self.advance(len('extends'))
2033 self.getWhiteSpace()
2034 startPos = self.pos()
2035 if self.setting('allowExpressionsInExtendsDirective'):
2036 baseName = self.getExpression()
2037 else:
2038 baseName = self.getDottedName()
2040 baseName = self._applyExpressionFilters(baseName, 'extends', startPos=startPos)
2041 self._compiler.setBaseClass(baseName) # in compiler
2042 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLine)
2044 def eatImplements(self):
2045 # filtered
2046 isLineClearToStartToken = self.isLineClearToStartToken()
2047 endOfFirstLine = self.findEOL()
2048 self.getDirectiveStartToken()
2049 self.advance(len('implements'))
2050 self.getWhiteSpace()
2051 startPos = self.pos()
2052 methodName = self.getIdentifier()
2053 if not self.atEnd() and self.peek() == '(':
2054 argsList = self.getDefArgList()
2055 self.advance() # past the closing ')'
2056 if argsList and argsList[0][0] == 'self':
2057 del argsList[0]
2058 else:
2059 argsList=[]
2061 # @@TR: need to split up filtering of the methodname and the args
2062 #methodName = self._applyExpressionFilters(methodName, 'implements', startPos=startPos)
2063 self._applyExpressionFilters(self[startPos:self.pos()], 'implements', startPos=startPos)
2065 self._compiler.setMainMethodName(methodName)
2066 self._compiler.setMainMethodArgs(argsList)
2068 self.getExpression() # throw away and unwanted crap that got added in
2069 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLine)
2071 def eatSuper(self):
2072 # filtered
2073 isLineClearToStartToken = self.isLineClearToStartToken()
2074 endOfFirstLine = self.findEOL()
2075 self.getDirectiveStartToken()
2076 self.advance(len('super'))
2077 self.getWhiteSpace()
2078 startPos = self.pos()
2079 if not self.atEnd() and self.peek() == '(':
2080 argsList = self.getDefArgList()
2081 self.advance() # past the closing ')'
2082 if argsList and argsList[0][0] == 'self':
2083 del argsList[0]
2084 else:
2085 argsList=[]
2087 self._applyExpressionFilters(self[startPos:self.pos()], 'super', startPos=startPos)
2089 #parserComment = ('## CHEETAH: generated from ' + signature +
2090 # ' at line %s, col %s' % self.getRowCol(startPos)
2091 # + '.')
2093 self.getExpression() # throw away and unwanted crap that got added in
2094 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLine)
2095 self._compiler.addSuper(argsList)
2097 def eatSet(self):
2098 # filtered
2099 isLineClearToStartToken = self.isLineClearToStartToken()
2100 endOfFirstLine = self.findEOL()
2101 self.getDirectiveStartToken()
2102 self.advance(3)
2103 self.getWhiteSpace()
2104 style = SET_LOCAL
2105 if self.startswith('local'):
2106 self.getIdentifier()
2107 self.getWhiteSpace()
2108 elif self.startswith('global'):
2109 self.getIdentifier()
2110 self.getWhiteSpace()
2111 style = SET_GLOBAL
2112 elif self.startswith('module'):
2113 self.getIdentifier()
2114 self.getWhiteSpace()
2115 style = SET_MODULE
2117 startsWithDollar = self.matchCheetahVarStart()
2118 startPos = self.pos()
2119 LVALUE = self.getExpression(pyTokensToBreakAt=assignmentOps, useNameMapper=False).strip()
2120 OP = self.getAssignmentOperator()
2121 RVALUE = self.getExpression()
2122 expr = LVALUE + ' ' + OP + ' ' + RVALUE.strip()
2124 expr = self._applyExpressionFilters(expr, 'set', startPos=startPos)
2125 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLine)
2127 class Components: pass # used for 'set global'
2128 exprComponents = Components()
2129 exprComponents.LVALUE = LVALUE
2130 exprComponents.OP = OP
2131 exprComponents.RVALUE = RVALUE
2132 self._compiler.addSet(expr, exprComponents, style)
2134 def eatSlurp(self):
2135 if self.isLineClearToStartToken():
2136 self._compiler.handleWSBeforeDirective()
2137 self._compiler.commitStrConst()
2138 self.readToEOL(gobble=True)
2140 def eatEOLSlurpToken(self):
2141 if self.isLineClearToStartToken():
2142 self._compiler.handleWSBeforeDirective()
2143 self._compiler.commitStrConst()
2144 self.readToEOL(gobble=True)
2146 def eatRaw(self):
2147 isLineClearToStartToken = self.isLineClearToStartToken()
2148 endOfFirstLinePos = self.findEOL()
2149 self.getDirectiveStartToken()
2150 self.advance(len('raw'))
2151 self.getWhiteSpace()
2152 if self.matchColonForSingleLineShortFormDirective():
2153 self.advance() # skip over :
2154 self.getWhiteSpace(max=1)
2155 rawBlock = self.readToEOL(gobble=False)
2156 else:
2157 if self.peek()==':':
2158 self.advance()
2159 self.getWhiteSpace()
2160 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLinePos)
2161 rawBlock = self._eatToThisEndDirective('raw')
2162 self._compiler.addRawText(rawBlock)
2164 def eatInclude(self):
2165 # filtered
2166 isLineClearToStartToken = self.isLineClearToStartToken()
2167 endOfFirstLinePos = self.findEOL()
2168 self.getDirectiveStartToken()
2169 self.advance(len('include'))
2171 self.getWhiteSpace()
2172 includeFrom = 'file'
2173 isRaw = False
2174 if self.startswith('raw'):
2175 self.advance(3)
2176 isRaw=True
2178 self.getWhiteSpace()
2179 if self.startswith('source'):
2180 self.advance(len('source'))
2181 includeFrom = 'str'
2182 self.getWhiteSpace()
2183 if not self.peek() == '=':
2184 raise ParseError(self)
2185 self.advance()
2186 startPos = self.pos()
2187 sourceExpr = self.getExpression()
2188 sourceExpr = self._applyExpressionFilters(sourceExpr, 'include', startPos=startPos)
2189 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLinePos)
2190 self._compiler.addInclude(sourceExpr, includeFrom, isRaw)
2193 def eatDefMacro(self):
2194 # @@TR: not filtered yet
2195 isLineClearToStartToken = self.isLineClearToStartToken()
2196 endOfFirstLinePos = self.findEOL()
2197 self.getDirectiveStartToken()
2198 self.advance(len('defmacro'))
2200 self.getWhiteSpace()
2201 if self.matchCheetahVarStart():
2202 self.getCheetahVarStartToken()
2203 macroName = self.getIdentifier()
2204 self.getWhiteSpace()
2205 if self.peek() == '(':
2206 argsList = self.getDefArgList(useNameMapper=False)
2207 self.advance() # past the closing ')'
2208 if argsList and argsList[0][0] == 'self':
2209 del argsList[0]
2210 else:
2211 argsList=[]
2213 assert not self._directiveNamesAndParsers.has_key(macroName)
2214 argsList.insert(0, ('src',None))
2215 argsList.append(('parser','None'))
2216 argsList.append(('macros','None'))
2217 argsList.append(('compilerSettings','None'))
2218 argsList.append(('isShortForm','None'))
2219 argsList.append(('EOLCharsInShortForm','None'))
2220 argsList.append(('startPos','None'))
2221 argsList.append(('endPos','None'))
2223 if self.matchColonForSingleLineShortFormDirective():
2224 self.advance() # skip over :
2225 self.getWhiteSpace(max=1)
2226 macroSrc = self.readToEOL(gobble=False)
2227 self.readToEOL(gobble=True)
2228 else:
2229 if self.peek()==':':
2230 self.advance()
2231 self.getWhiteSpace()
2232 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLinePos)
2233 macroSrc = self._eatToThisEndDirective('defmacro')
2235 #print argsList
2236 normalizedMacroSrc = ''.join(
2237 ['%def callMacro('+','.join([defv and '%s=%s'%(n,defv) or n
2238 for n,defv in argsList])
2239 +')\n',
2240 macroSrc,
2241 '%end def'])
2244 from Cheetah.Template import Template
2245 templateAPIClass = self.setting('templateAPIClassForDefMacro', default=Template)
2246 compilerSettings = self.setting('compilerSettingsForDefMacro', default={})
2247 searchListForMacros = self.setting('searchListForDefMacro', default=[])
2248 searchListForMacros = list(searchListForMacros) # copy to avoid mutation bugs
2249 searchListForMacros.append({'macros':self._macros,
2250 'parser':self,
2251 'compilerSettings':self.settings(),
2254 templateAPIClass._updateSettingsWithPreprocessTokens(
2255 compilerSettings, placeholderToken='@', directiveToken='%')
2256 macroTemplateClass = templateAPIClass.compile(source=normalizedMacroSrc,
2257 compilerSettings=compilerSettings)
2258 #print normalizedMacroSrc
2259 #t = macroTemplateClass()
2260 #print t.callMacro('src')
2261 #print t.generatedClassCode()
2263 class MacroDetails: pass
2264 macroDetails = MacroDetails()
2265 macroDetails.macroSrc = macroSrc
2266 macroDetails.argsList = argsList
2267 macroDetails.template = macroTemplateClass(searchList=searchListForMacros)
2269 self._macroDetails[macroName] = macroDetails
2270 self._macros[macroName] = macroDetails.template.callMacro
2271 self._directiveNamesAndParsers[macroName] = self.eatMacroCall
2273 def eatMacroCall(self):
2274 isLineClearToStartToken = self.isLineClearToStartToken()
2275 endOfFirstLinePos = self.findEOL()
2276 startPos = self.pos()
2277 self.getDirectiveStartToken()
2278 macroName = self.getIdentifier()
2279 macro = self._macros[macroName]
2280 if hasattr(macro, 'parse'):
2281 return macro.parse(parser=self, startPos=startPos)
2283 if hasattr(macro, 'parseArgs'):
2284 args = macro.parseArgs(parser=self, startPos=startPos)
2285 else:
2286 self.getWhiteSpace()
2287 args = self.getExpression(useNameMapper=False,
2288 pyTokensToBreakAt=[':']).strip()
2290 if self.matchColonForSingleLineShortFormDirective():
2291 isShortForm = True
2292 self.advance() # skip over :
2293 self.getWhiteSpace(max=1)
2294 srcBlock = self.readToEOL(gobble=False)
2295 EOLCharsInShortForm = self.readToEOL(gobble=True)
2296 #self.readToEOL(gobble=False)
2297 else:
2298 isShortForm = False
2299 if self.peek()==':':
2300 self.advance()
2301 self.getWhiteSpace()
2302 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLinePos)
2303 srcBlock = self._eatToThisEndDirective(macroName)
2306 if hasattr(macro, 'convertArgStrToDict'):
2307 kwArgs = macro.convertArgStrToDict(args, parser=self, startPos=startPos)
2308 else:
2309 def getArgs(*pargs, **kws):
2310 return pargs, kws
2311 exec 'positionalArgs, kwArgs = getArgs(%(args)s)'%locals()
2313 assert not kwArgs.has_key('src')
2314 kwArgs['src'] = srcBlock
2316 if type(macro)==new.instancemethod:
2317 co = macro.im_func.func_code
2318 elif (hasattr(macro, '__call__')
2319 and hasattr(macro.__call__, 'im_func')):
2320 co = macro.__call__.im_func.func_code
2321 else:
2322 co = macro.func_code
2323 availableKwArgs = inspect.getargs(co)[0]
2325 if 'parser' in availableKwArgs:
2326 kwArgs['parser'] = self
2327 if 'macros' in availableKwArgs:
2328 kwArgs['macros'] = self._macros
2329 if 'compilerSettings' in availableKwArgs:
2330 kwArgs['compilerSettings'] = self.settings()
2331 if 'isShortForm' in availableKwArgs:
2332 kwArgs['isShortForm'] = isShortForm
2333 if isShortForm and 'EOLCharsInShortForm' in availableKwArgs:
2334 kwArgs['EOLCharsInShortForm'] = EOLCharsInShortForm
2336 if 'startPos' in availableKwArgs:
2337 kwArgs['startPos'] = startPos
2338 if 'endPos' in availableKwArgs:
2339 kwArgs['endPos'] = self.pos()
2341 srcFromMacroOutput = macro(**kwArgs)
2343 origParseSrc = self._src
2344 origBreakPoint = self.breakPoint()
2345 origPos = self.pos()
2346 # add a comment to the output about the macro src that is being parsed
2347 # or add a comment prefix to all the comments added by the compiler
2348 self._src = srcFromMacroOutput
2349 self.setPos(0)
2350 self.setBreakPoint(len(srcFromMacroOutput))
2352 self.parse(assertEmptyStack=False)
2354 self._src = origParseSrc
2355 self.setBreakPoint(origBreakPoint)
2356 self.setPos(origPos)
2359 #self._compiler.addRawText('end')
2361 def eatCache(self):
2362 isLineClearToStartToken = self.isLineClearToStartToken()
2363 endOfFirstLinePos = self.findEOL()
2364 lineCol = self.getRowCol()
2365 self.getDirectiveStartToken()
2366 self.advance(len('cache'))
2368 startPos = self.pos()
2369 argList = self.getDefArgList(useNameMapper=True)
2370 argList = self._applyExpressionFilters(argList, 'cache', startPos=startPos)
2372 def startCache():
2373 cacheInfo = self._compiler.genCacheInfoFromArgList(argList)
2374 self._compiler.startCacheRegion(cacheInfo, lineCol)
2376 if self.matchColonForSingleLineShortFormDirective():
2377 self.advance() # skip over :
2378 self.getWhiteSpace(max=1)
2379 startCache()
2380 self.parse(breakPoint=self.findEOL(gobble=True))
2381 self._compiler.endCacheRegion()
2382 else:
2383 if self.peek()==':':
2384 self.advance()
2385 self.getWhiteSpace()
2386 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLinePos)
2387 self.pushToOpenDirectivesStack('cache')
2388 startCache()
2390 def eatCall(self):
2391 # @@TR: need to enable single line version of this
2392 isLineClearToStartToken = self.isLineClearToStartToken()
2393 endOfFirstLinePos = self.findEOL()
2394 lineCol = self.getRowCol()
2395 self.getDirectiveStartToken()
2396 self.advance(len('call'))
2397 startPos = self.pos()
2399 useAutocallingOrig = self.setting('useAutocalling')
2400 self.setSetting('useAutocalling', False)
2401 self.getWhiteSpace()
2402 if self.matchCheetahVarStart():
2403 functionName = self.getCheetahVar()
2404 else:
2405 functionName = self.getCheetahVar(plain=True, skipStartToken=True)
2406 self.setSetting('useAutocalling', useAutocallingOrig)
2407 # @@TR: fix up filtering
2408 self._applyExpressionFilters(self[startPos:self.pos()], 'call', startPos=startPos)
2410 self.getWhiteSpace()
2411 args = self.getExpression(pyTokensToBreakAt=[':']).strip()
2412 if self.matchColonForSingleLineShortFormDirective():
2413 self.advance() # skip over :
2414 self._compiler.startCallRegion(functionName, args, lineCol)
2415 self.getWhiteSpace(max=1)
2416 self.parse(breakPoint=self.findEOL(gobble=False))
2417 self._compiler.endCallRegion()
2418 else:
2419 if self.peek()==':':
2420 self.advance()
2421 self.getWhiteSpace()
2422 self.pushToOpenDirectivesStack("call")
2423 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLinePos)
2424 self._compiler.startCallRegion(functionName, args, lineCol)
2426 def eatCallArg(self):
2427 isLineClearToStartToken = self.isLineClearToStartToken()
2428 endOfFirstLinePos = self.findEOL()
2429 lineCol = self.getRowCol()
2430 self.getDirectiveStartToken()
2432 self.advance(len('arg'))
2433 startPos = self.pos()
2434 self.getWhiteSpace()
2435 argName = self.getIdentifier()
2436 self.getWhiteSpace()
2437 argName = self._applyExpressionFilters(argName, 'arg', startPos=startPos)
2438 self._compiler.setCallArg(argName, lineCol)
2439 if self.peek() == ':':
2440 self.getc()
2441 else:
2442 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLinePos)
2444 def eatFilter(self):
2445 isLineClearToStartToken = self.isLineClearToStartToken()
2446 endOfFirstLinePos = self.findEOL()
2448 self.getDirectiveStartToken()
2449 self.advance(len('filter'))
2450 self.getWhiteSpace()
2451 startPos = self.pos()
2452 if self.matchCheetahVarStart():
2453 isKlass = True
2454 theFilter = self.getExpression(pyTokensToBreakAt=[':'])
2455 else:
2456 isKlass = False
2457 theFilter = self.getIdentifier()
2458 self.getWhiteSpace()
2459 theFilter = self._applyExpressionFilters(theFilter, 'filter', startPos=startPos)
2461 if self.matchColonForSingleLineShortFormDirective():
2462 self.advance() # skip over :
2463 self.getWhiteSpace(max=1)
2464 self._compiler.setFilter(theFilter, isKlass)
2465 self.parse(breakPoint=self.findEOL(gobble=False))
2466 self._compiler.closeFilterBlock()
2467 else:
2468 if self.peek()==':':
2469 self.advance()
2470 self.getWhiteSpace()
2471 self.pushToOpenDirectivesStack("filter")
2472 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLinePos)
2473 self._compiler.setFilter(theFilter, isKlass)
2475 def eatErrorCatcher(self):
2476 isLineClearToStartToken = self.isLineClearToStartToken()
2477 endOfFirstLinePos = self.findEOL()
2478 self.getDirectiveStartToken()
2479 self.advance(len('errorCatcher'))
2480 self.getWhiteSpace()
2481 startPos = self.pos()
2482 errorCatcherName = self.getIdentifier()
2483 errorCatcherName = self._applyExpressionFilters(
2484 errorCatcherName, 'errorcatcher', startPos=startPos)
2485 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLinePos)
2486 self._compiler.setErrorCatcher(errorCatcherName)
2488 def eatCapture(self):
2489 # @@TR: this could be refactored to use the code in eatSimpleIndentingDirective
2490 # filtered
2491 isLineClearToStartToken = self.isLineClearToStartToken()
2492 endOfFirstLinePos = self.findEOL()
2493 lineCol = self.getRowCol()
2495 self.getDirectiveStartToken()
2496 self.advance(len('capture'))
2497 startPos = self.pos()
2498 self.getWhiteSpace()
2500 expr = self.getExpression(pyTokensToBreakAt=[':'])
2501 expr = self._applyExpressionFilters(expr, 'capture', startPos=startPos)
2502 if self.matchColonForSingleLineShortFormDirective():
2503 self.advance() # skip over :
2504 self._compiler.startCaptureRegion(assignTo=expr, lineCol=lineCol)
2505 self.getWhiteSpace(max=1)
2506 self.parse(breakPoint=self.findEOL(gobble=False))
2507 self._compiler.endCaptureRegion()
2508 else:
2509 if self.peek()==':':
2510 self.advance()
2511 self.getWhiteSpace()
2512 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLinePos)
2513 self.pushToOpenDirectivesStack("capture")
2514 self._compiler.startCaptureRegion(assignTo=expr, lineCol=lineCol)
2517 def eatIf(self):
2518 # filtered
2519 isLineClearToStartToken = self.isLineClearToStartToken()
2520 endOfFirstLine = self.findEOL()
2521 lineCol = self.getRowCol()
2522 self.getDirectiveStartToken()
2523 startPos = self.pos()
2525 expressionParts = self.getExpressionParts(pyTokensToBreakAt=[':'])
2526 expr = ''.join(expressionParts).strip()
2527 expr = self._applyExpressionFilters(expr, 'if', startPos=startPos)
2529 isTernaryExpr = ('then' in expressionParts and 'else' in expressionParts)
2530 if isTernaryExpr:
2531 conditionExpr = []
2532 trueExpr = []
2533 falseExpr = []
2534 currentExpr = conditionExpr
2535 for part in expressionParts:
2536 if part.strip()=='then':
2537 currentExpr = trueExpr
2538 elif part.strip()=='else':
2539 currentExpr = falseExpr
2540 else:
2541 currentExpr.append(part)
2543 conditionExpr = ''.join(conditionExpr)
2544 trueExpr = ''.join(trueExpr)
2545 falseExpr = ''.join(falseExpr)
2546 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLine)
2547 self._compiler.addTernaryExpr(conditionExpr, trueExpr, falseExpr, lineCol=lineCol)
2548 elif self.matchColonForSingleLineShortFormDirective():
2549 self.advance() # skip over :
2550 self._compiler.addIf(expr, lineCol=lineCol)
2551 self.getWhiteSpace(max=1)
2552 self.parse(breakPoint=self.findEOL(gobble=True))
2553 self._compiler.commitStrConst()
2554 self._compiler.dedent()
2555 else:
2556 if self.peek()==':':
2557 self.advance()
2558 self.getWhiteSpace()
2559 self._eatRestOfDirectiveTag(isLineClearToStartToken, endOfFirstLine)
2560 self.pushToOpenDirectivesStack('if')
2561 self._compiler.addIf(expr, lineCol=lineCol)
2563 ## end directive handlers
2564 def handleEndDef(self):
2565 isNestedDef = (self.setting('allowNestedDefScopes')
2566 and [name for name in self._openDirectivesStack if name=='def'])
2567 if not isNestedDef:
2568 self._compiler.closeDef()
2569 else:
2570 # @@TR: temporary hack of useSearchList
2571 self.setSetting('useSearchList', self._useSearchList_orig)
2572 self._compiler.commitStrConst()
2573 self._compiler.dedent()
2576 def pushToOpenDirectivesStack(self, directiveName):
2577 assert directiveName in self._closeableDirectives
2578 self._openDirectivesStack.append(directiveName)
2580 def popFromOpenDirectivesStack(self, directiveName):
2581 if not self._openDirectivesStack:
2582 raise ParseError(self, msg="#end found, but nothing to end")
2584 if self._openDirectivesStack[-1] == directiveName:
2585 del self._openDirectivesStack[-1]
2586 else:
2587 raise ParseError(self, msg="#end %s found, expected #end %s" %(
2588 directiveName, self._openDirectivesStack[-1]))
2590 def assertEmptyOpenDirectivesStack(self):
2591 if self._openDirectivesStack:
2592 errorMsg = (
2593 "Some #directives are missing their corresponding #end ___ tag: %s" %(
2594 ', '.join(self._openDirectivesStack)))
2595 raise ParseError(self, msg=errorMsg)
2597 ##################################################
2598 ## Make an alias to export
2599 Parser = _HighLevelParser