1""" @package antlr3.tree
2@brief ANTLR3 runtime package, tree module
3
4This module contains all support classes for AST construction and tree parsers.
5
6"""
7
8# begin[licence]
9#
10# [The "BSD licence"]
11# Copyright (c) 2005-2008 Terence Parr
12# All rights reserved.
13#
14# Redistribution and use in source and binary forms, with or without
15# modification, are permitted provided that the following conditions
16# are met:
17# 1. Redistributions of source code must retain the above copyright
18#    notice, this list of conditions and the following disclaimer.
19# 2. Redistributions in binary form must reproduce the above copyright
20#    notice, this list of conditions and the following disclaimer in the
21#    documentation and/or other materials provided with the distribution.
22# 3. The name of the author may not be used to endorse or promote products
23#    derived from this software without specific prior written permission.
24#
25# THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
26# IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
27# OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
28# IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
29# INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
30# NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
31# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
32# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
33# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
34# THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
35#
36# end[licence]
37
38# lot's of docstrings are missing, don't complain for now...
39# pylint: disable-msg=C0111
40
41import re
42
43from antlr3.constants import UP, DOWN, EOF, INVALID_TOKEN_TYPE
44from antlr3.recognizers import BaseRecognizer, RuleReturnScope
45from antlr3.streams import IntStream
46from antlr3.tokens import CommonToken, Token, INVALID_TOKEN
47from antlr3.exceptions import MismatchedTreeNodeException, \
48     MissingTokenException, UnwantedTokenException, MismatchedTokenException, \
49     NoViableAltException
50
51
52############################################################################
53#
54# tree related exceptions
55#
56############################################################################
57
58
59class RewriteCardinalityException(RuntimeError):
60    """
61    @brief Base class for all exceptions thrown during AST rewrite construction.
62
63    This signifies a case where the cardinality of two or more elements
64    in a subrule are different: (ID INT)+ where |ID|!=|INT|
65    """
66
67    def __init__(self, elementDescription):
68        RuntimeError.__init__(self, elementDescription)
69
70        self.elementDescription = elementDescription
71
72
73    def getMessage(self):
74        return self.elementDescription
75
76
77class RewriteEarlyExitException(RewriteCardinalityException):
78    """@brief No elements within a (...)+ in a rewrite rule"""
79
80    def __init__(self, elementDescription=None):
81        RewriteCardinalityException.__init__(self, elementDescription)
82
83
84class RewriteEmptyStreamException(RewriteCardinalityException):
85    """
86    @brief Ref to ID or expr but no tokens in ID stream or subtrees in expr stream
87    """
88
89    pass
90
91
92############################################################################
93#
94# basic Tree and TreeAdaptor interfaces
95#
96############################################################################
97
98class Tree(object):
99    """
100    @brief Abstract baseclass for tree nodes.
101
102    What does a tree look like?  ANTLR has a number of support classes
103    such as CommonTreeNodeStream that work on these kinds of trees.  You
104    don't have to make your trees implement this interface, but if you do,
105    you'll be able to use more support code.
106
107    NOTE: When constructing trees, ANTLR can build any kind of tree; it can
108    even use Token objects as trees if you add a child list to your tokens.
109
110    This is a tree node without any payload; just navigation and factory stuff.
111    """
112
113
114    def getChild(self, i):
115        raise NotImplementedError
116
117
118    def getChildCount(self):
119        raise NotImplementedError
120
121
122    def getParent(self):
123        """Tree tracks parent and child index now > 3.0"""
124
125        raise NotImplementedError
126
127    def setParent(self, t):
128        """Tree tracks parent and child index now > 3.0"""
129
130        raise NotImplementedError
131
132
133    def hasAncestor(self, ttype):
134        """Walk upwards looking for ancestor with this token type."""
135
136        raise NotImplementedError
137
138    def getAncestor(self, ttype):
139        """Walk upwards and get first ancestor with this token type."""
140
141        raise NotImplementedError
142
143    def getAncestors(self):
144        """Return a list of all ancestors of this node.
145
146        The first node of list is the root and the last is the parent of
147        this node.
148        """
149
150        raise NotImplementedError
151
152
153    def getChildIndex(self):
154        """This node is what child index? 0..n-1"""
155
156        raise NotImplementedError
157
158    def setChildIndex(self, index):
159        """This node is what child index? 0..n-1"""
160
161        raise NotImplementedError
162
163
164    def freshenParentAndChildIndexes(self):
165        """Set the parent and child index values for all children"""
166
167        raise NotImplementedError
168
169
170    def addChild(self, t):
171        """
172        Add t as a child to this node.  If t is null, do nothing.  If t
173        is nil, add all children of t to this' children.
174        """
175
176        raise NotImplementedError
177
178
179    def setChild(self, i, t):
180        """Set ith child (0..n-1) to t; t must be non-null and non-nil node"""
181
182        raise NotImplementedError
183
184
185    def deleteChild(self, i):
186        raise NotImplementedError
187
188
189    def replaceChildren(self, startChildIndex, stopChildIndex, t):
190        """
191        Delete children from start to stop and replace with t even if t is
192        a list (nil-root tree).  num of children can increase or decrease.
193        For huge child lists, inserting children can force walking rest of
194        children to set their childindex; could be slow.
195        """
196
197        raise NotImplementedError
198
199
200    def isNil(self):
201        """
202        Indicates the node is a nil node but may still have children, meaning
203        the tree is a flat list.
204        """
205
206        raise NotImplementedError
207
208
209    def getTokenStartIndex(self):
210        """
211        What is the smallest token index (indexing from 0) for this node
212           and its children?
213        """
214
215        raise NotImplementedError
216
217
218    def setTokenStartIndex(self, index):
219        raise NotImplementedError
220
221
222    def getTokenStopIndex(self):
223        """
224        What is the largest token index (indexing from 0) for this node
225        and its children?
226        """
227
228        raise NotImplementedError
229
230
231    def setTokenStopIndex(self, index):
232        raise NotImplementedError
233
234
235    def dupNode(self):
236        raise NotImplementedError
237
238
239    def getType(self):
240        """Return a token type; needed for tree parsing."""
241
242        raise NotImplementedError
243
244
245    def getText(self):
246        raise NotImplementedError
247
248
249    def getLine(self):
250        """
251        In case we don't have a token payload, what is the line for errors?
252        """
253
254        raise NotImplementedError
255
256
257    def getCharPositionInLine(self):
258        raise NotImplementedError
259
260
261    def toStringTree(self):
262        raise NotImplementedError
263
264
265    def toString(self):
266        raise NotImplementedError
267
268
269
270class TreeAdaptor(object):
271    """
272    @brief Abstract baseclass for tree adaptors.
273
274    How to create and navigate trees.  Rather than have a separate factory
275    and adaptor, I've merged them.  Makes sense to encapsulate.
276
277    This takes the place of the tree construction code generated in the
278    generated code in 2.x and the ASTFactory.
279
280    I do not need to know the type of a tree at all so they are all
281    generic Objects.  This may increase the amount of typecasting needed. :(
282    """
283
284    # C o n s t r u c t i o n
285
286    def createWithPayload(self, payload):
287        """
288        Create a tree node from Token object; for CommonTree type trees,
289        then the token just becomes the payload.  This is the most
290        common create call.
291
292        Override if you want another kind of node to be built.
293        """
294
295        raise NotImplementedError
296
297
298    def dupNode(self, treeNode):
299        """Duplicate a single tree node.
300
301        Override if you want another kind of node to be built."""
302
303        raise NotImplementedError
304
305
306    def dupTree(self, tree):
307        """Duplicate tree recursively, using dupNode() for each node"""
308
309        raise NotImplementedError
310
311
312    def nil(self):
313        """
314        Return a nil node (an empty but non-null node) that can hold
315        a list of element as the children.  If you want a flat tree (a list)
316        use "t=adaptor.nil(); t.addChild(x); t.addChild(y);"
317        """
318
319        raise NotImplementedError
320
321
322    def errorNode(self, input, start, stop, exc):
323        """
324        Return a tree node representing an error.  This node records the
325        tokens consumed during error recovery.  The start token indicates the
326        input symbol at which the error was detected.  The stop token indicates
327        the last symbol consumed during recovery.
328
329        You must specify the input stream so that the erroneous text can
330        be packaged up in the error node.  The exception could be useful
331        to some applications; default implementation stores ptr to it in
332        the CommonErrorNode.
333
334        This only makes sense during token parsing, not tree parsing.
335        Tree parsing should happen only when parsing and tree construction
336        succeed.
337        """
338
339        raise NotImplementedError
340
341
342    def isNil(self, tree):
343        """Is tree considered a nil node used to make lists of child nodes?"""
344
345        raise NotImplementedError
346
347
348    def addChild(self, t, child):
349        """
350        Add a child to the tree t.  If child is a flat tree (a list), make all
351        in list children of t.  Warning: if t has no children, but child does
352        and child isNil then you can decide it is ok to move children to t via
353        t.children = child.children; i.e., without copying the array.  Just
354        make sure that this is consistent with have the user will build
355        ASTs. Do nothing if t or child is null.
356        """
357
358        raise NotImplementedError
359
360
361    def becomeRoot(self, newRoot, oldRoot):
362        """
363        If oldRoot is a nil root, just copy or move the children to newRoot.
364        If not a nil root, make oldRoot a child of newRoot.
365
366           old=^(nil a b c), new=r yields ^(r a b c)
367           old=^(a b c), new=r yields ^(r ^(a b c))
368
369        If newRoot is a nil-rooted single child tree, use the single
370        child as the new root node.
371
372           old=^(nil a b c), new=^(nil r) yields ^(r a b c)
373           old=^(a b c), new=^(nil r) yields ^(r ^(a b c))
374
375        If oldRoot was null, it's ok, just return newRoot (even if isNil).
376
377           old=null, new=r yields r
378           old=null, new=^(nil r) yields ^(nil r)
379
380        Return newRoot.  Throw an exception if newRoot is not a
381        simple node or nil root with a single child node--it must be a root
382        node.  If newRoot is ^(nil x) return x as newRoot.
383
384        Be advised that it's ok for newRoot to point at oldRoot's
385        children; i.e., you don't have to copy the list.  We are
386        constructing these nodes so we should have this control for
387        efficiency.
388        """
389
390        raise NotImplementedError
391
392
393    def rulePostProcessing(self, root):
394        """
395        Given the root of the subtree created for this rule, post process
396        it to do any simplifications or whatever you want.  A required
397        behavior is to convert ^(nil singleSubtree) to singleSubtree
398        as the setting of start/stop indexes relies on a single non-nil root
399        for non-flat trees.
400
401        Flat trees such as for lists like "idlist : ID+ ;" are left alone
402        unless there is only one ID.  For a list, the start/stop indexes
403        are set in the nil node.
404
405        This method is executed after all rule tree construction and right
406        before setTokenBoundaries().
407        """
408
409        raise NotImplementedError
410
411
412    def getUniqueID(self, node):
413        """For identifying trees.
414
415        How to identify nodes so we can say "add node to a prior node"?
416        Even becomeRoot is an issue.  Use System.identityHashCode(node)
417        usually.
418        """
419
420        raise NotImplementedError
421
422
423    # R e w r i t e  R u l e s
424
425    def createFromToken(self, tokenType, fromToken, text=None):
426        """
427        Create a new node derived from a token, with a new token type and
428        (optionally) new text.
429
430        This is invoked from an imaginary node ref on right side of a
431        rewrite rule as IMAG[$tokenLabel] or IMAG[$tokenLabel "IMAG"].
432
433        This should invoke createToken(Token).
434        """
435
436        raise NotImplementedError
437
438
439    def createFromType(self, tokenType, text):
440        """Create a new node derived from a token, with a new token type.
441
442        This is invoked from an imaginary node ref on right side of a
443        rewrite rule as IMAG["IMAG"].
444
445        This should invoke createToken(int,String).
446        """
447
448        raise NotImplementedError
449
450
451    # C o n t e n t
452
453    def getType(self, t):
454        """For tree parsing, I need to know the token type of a node"""
455
456        raise NotImplementedError
457
458
459    def setType(self, t, type):
460        """Node constructors can set the type of a node"""
461
462        raise NotImplementedError
463
464
465    def getText(self, t):
466        raise NotImplementedError
467
468    def setText(self, t, text):
469        """Node constructors can set the text of a node"""
470
471        raise NotImplementedError
472
473
474    def getToken(self, t):
475        """Return the token object from which this node was created.
476
477        Currently used only for printing an error message.
478        The error display routine in BaseRecognizer needs to
479        display where the input the error occurred. If your
480        tree of limitation does not store information that can
481        lead you to the token, you can create a token filled with
482        the appropriate information and pass that back.  See
483        BaseRecognizer.getErrorMessage().
484        """
485
486        raise NotImplementedError
487
488
489    def setTokenBoundaries(self, t, startToken, stopToken):
490        """
491        Where are the bounds in the input token stream for this node and
492        all children?  Each rule that creates AST nodes will call this
493        method right before returning.  Flat trees (i.e., lists) will
494        still usually have a nil root node just to hold the children list.
495        That node would contain the start/stop indexes then.
496        """
497
498        raise NotImplementedError
499
500
501    def getTokenStartIndex(self, t):
502        """
503        Get the token start index for this subtree; return -1 if no such index
504        """
505
506        raise NotImplementedError
507
508
509    def getTokenStopIndex(self, t):
510        """
511        Get the token stop index for this subtree; return -1 if no such index
512        """
513
514        raise NotImplementedError
515
516
517    # N a v i g a t i o n  /  T r e e  P a r s i n g
518
519    def getChild(self, t, i):
520        """Get a child 0..n-1 node"""
521
522        raise NotImplementedError
523
524
525    def setChild(self, t, i, child):
526        """Set ith child (0..n-1) to t; t must be non-null and non-nil node"""
527
528        raise NotImplementedError
529
530
531    def deleteChild(self, t, i):
532        """Remove ith child and shift children down from right."""
533
534        raise NotImplementedError
535
536
537    def getChildCount(self, t):
538        """How many children?  If 0, then this is a leaf node"""
539
540        raise NotImplementedError
541
542
543    def getParent(self, t):
544        """
545        Who is the parent node of this node; if null, implies node is root.
546        If your node type doesn't handle this, it's ok but the tree rewrites
547        in tree parsers need this functionality.
548        """
549
550        raise NotImplementedError
551
552
553    def setParent(self, t, parent):
554        """
555        Who is the parent node of this node; if null, implies node is root.
556        If your node type doesn't handle this, it's ok but the tree rewrites
557        in tree parsers need this functionality.
558        """
559
560        raise NotImplementedError
561
562
563    def getChildIndex(self, t):
564        """
565        What index is this node in the child list? Range: 0..n-1
566        If your node type doesn't handle this, it's ok but the tree rewrites
567        in tree parsers need this functionality.
568        """
569
570        raise NotImplementedError
571
572
573    def setChildIndex(self, t, index):
574        """
575        What index is this node in the child list? Range: 0..n-1
576        If your node type doesn't handle this, it's ok but the tree rewrites
577        in tree parsers need this functionality.
578        """
579
580        raise NotImplementedError
581
582
583    def replaceChildren(self, parent, startChildIndex, stopChildIndex, t):
584        """
585        Replace from start to stop child index of parent with t, which might
586        be a list.  Number of children may be different
587        after this call.
588
589        If parent is null, don't do anything; must be at root of overall tree.
590        Can't replace whatever points to the parent externally.  Do nothing.
591        """
592
593        raise NotImplementedError
594
595
596    # Misc
597
598    def create(self, *args):
599        """
600        Deprecated, use createWithPayload, createFromToken or createFromType.
601
602        This method only exists to mimic the Java interface of TreeAdaptor.
603
604        """
605
606        if len(args) == 1 and isinstance(args[0], Token):
607            # Object create(Token payload);
608##             warnings.warn(
609##                 "Using create() is deprecated, use createWithPayload()",
610##                 DeprecationWarning,
611##                 stacklevel=2
612##                 )
613            return self.createWithPayload(args[0])
614
615        if (len(args) == 2
616            and isinstance(args[0], (int, long))
617            and isinstance(args[1], Token)
618            ):
619            # Object create(int tokenType, Token fromToken);
620##             warnings.warn(
621##                 "Using create() is deprecated, use createFromToken()",
622##                 DeprecationWarning,
623##                 stacklevel=2
624##                 )
625            return self.createFromToken(args[0], args[1])
626
627        if (len(args) == 3
628            and isinstance(args[0], (int, long))
629            and isinstance(args[1], Token)
630            and isinstance(args[2], basestring)
631            ):
632            # Object create(int tokenType, Token fromToken, String text);
633##             warnings.warn(
634##                 "Using create() is deprecated, use createFromToken()",
635##                 DeprecationWarning,
636##                 stacklevel=2
637##                 )
638            return self.createFromToken(args[0], args[1], args[2])
639
640        if (len(args) == 2
641            and isinstance(args[0], (int, long))
642            and isinstance(args[1], basestring)
643            ):
644            # Object create(int tokenType, String text);
645##             warnings.warn(
646##                 "Using create() is deprecated, use createFromType()",
647##                 DeprecationWarning,
648##                 stacklevel=2
649##                 )
650            return self.createFromType(args[0], args[1])
651
652        raise TypeError(
653            "No create method with this signature found: %s"
654            % (', '.join(type(v).__name__ for v in args))
655            )
656
657
658############################################################################
659#
660# base implementation of Tree and TreeAdaptor
661#
662# Tree
663# \- BaseTree
664#
665# TreeAdaptor
666# \- BaseTreeAdaptor
667#
668############################################################################
669
670
671class BaseTree(Tree):
672    """
673    @brief A generic tree implementation with no payload.
674
675    You must subclass to
676    actually have any user data.  ANTLR v3 uses a list of children approach
677    instead of the child-sibling approach in v2.  A flat tree (a list) is
678    an empty node whose children represent the list.  An empty, but
679    non-null node is called "nil".
680    """
681
682    # BaseTree is abstract, no need to complain about not implemented abstract
683    # methods
684    # pylint: disable-msg=W0223
685
686    def __init__(self, node=None):
687        """
688        Create a new node from an existing node does nothing for BaseTree
689        as there are no fields other than the children list, which cannot
690        be copied as the children are not considered part of this node.
691        """
692
693        Tree.__init__(self)
694        self.children = []
695        self.parent = None
696        self.childIndex = 0
697
698
699    def getChild(self, i):
700        try:
701            return self.children[i]
702        except IndexError:
703            return None
704
705
706    def getChildren(self):
707        """@brief Get the children internal List
708
709        Note that if you directly mess with
710        the list, do so at your own risk.
711        """
712
713        # FIXME: mark as deprecated
714        return self.children
715
716
717    def getFirstChildWithType(self, treeType):
718        for child in self.children:
719            if child.getType() == treeType:
720                return child
721
722        return None
723
724
725    def getChildCount(self):
726        return len(self.children)
727
728
729    def addChild(self, childTree):
730        """Add t as child of this node.
731
732        Warning: if t has no children, but child does
733        and child isNil then this routine moves children to t via
734        t.children = child.children; i.e., without copying the array.
735        """
736
737        # this implementation is much simpler and probably less efficient
738        # than the mumbo-jumbo that Ter did for the Java runtime.
739
740        if childTree is None:
741            return
742
743        if childTree.isNil():
744            # t is an empty node possibly with children
745
746            if self.children is childTree.children:
747                raise ValueError("attempt to add child list to itself")
748
749            # fix parent pointer and childIndex for new children
750            for idx, child in enumerate(childTree.children):
751                child.parent = self
752                child.childIndex = len(self.children) + idx
753
754            self.children += childTree.children
755
756        else:
757            # child is not nil (don't care about children)
758            self.children.append(childTree)
759            childTree.parent = self
760            childTree.childIndex = len(self.children) - 1
761
762
763    def addChildren(self, children):
764        """Add all elements of kids list as children of this node"""
765
766        self.children += children
767
768
769    def setChild(self, i, t):
770        if t is None:
771            return
772
773        if t.isNil():
774            raise ValueError("Can't set single child to a list")
775
776        self.children[i] = t
777        t.parent = self
778        t.childIndex = i
779
780
781    def deleteChild(self, i):
782        killed = self.children[i]
783
784        del self.children[i]
785
786        # walk rest and decrement their child indexes
787        for idx, child in enumerate(self.children[i:]):
788            child.childIndex = i + idx
789
790        return killed
791
792
793    def replaceChildren(self, startChildIndex, stopChildIndex, newTree):
794        """
795        Delete children from start to stop and replace with t even if t is
796        a list (nil-root tree).  num of children can increase or decrease.
797        For huge child lists, inserting children can force walking rest of
798        children to set their childindex; could be slow.
799        """
800
801        if (startChildIndex >= len(self.children)
802            or stopChildIndex >= len(self.children)
803            ):
804            raise IndexError("indexes invalid")
805
806        replacingHowMany = stopChildIndex - startChildIndex + 1
807
808        # normalize to a list of children to add: newChildren
809        if newTree.isNil():
810            newChildren = newTree.children
811
812        else:
813            newChildren = [newTree]
814
815        replacingWithHowMany = len(newChildren)
816        delta = replacingHowMany - replacingWithHowMany
817
818
819        if delta == 0:
820            # if same number of nodes, do direct replace
821            for idx, child in enumerate(newChildren):
822                self.children[idx + startChildIndex] = child
823                child.parent = self
824                child.childIndex = idx + startChildIndex
825
826        else:
827            # length of children changes...
828
829            # ...delete replaced segment...
830            del self.children[startChildIndex:stopChildIndex+1]
831
832            # ...insert new segment...
833            self.children[startChildIndex:startChildIndex] = newChildren
834
835            # ...and fix indeces
836            self.freshenParentAndChildIndexes(startChildIndex)
837
838
839    def isNil(self):
840        return False
841
842
843    def freshenParentAndChildIndexes(self, offset=0):
844        for idx, child in enumerate(self.children[offset:]):
845            child.childIndex = idx + offset
846            child.parent = self
847
848
849    def sanityCheckParentAndChildIndexes(self, parent=None, i=-1):
850        if parent != self.parent:
851            raise ValueError(
852                "parents don't match; expected %r found %r"
853                % (parent, self.parent)
854                )
855
856        if i != self.childIndex:
857            raise ValueError(
858                "child indexes don't match; expected %d found %d"
859                % (i, self.childIndex)
860                )
861
862        for idx, child in enumerate(self.children):
863            child.sanityCheckParentAndChildIndexes(self, idx)
864
865
866    def getChildIndex(self):
867        """BaseTree doesn't track child indexes."""
868
869        return 0
870
871
872    def setChildIndex(self, index):
873        """BaseTree doesn't track child indexes."""
874
875        pass
876
877
878    def getParent(self):
879        """BaseTree doesn't track parent pointers."""
880
881        return None
882
883    def setParent(self, t):
884        """BaseTree doesn't track parent pointers."""
885
886        pass
887
888
889    def hasAncestor(self, ttype):
890        """Walk upwards looking for ancestor with this token type."""
891        return self.getAncestor(ttype) is not None
892
893    def getAncestor(self, ttype):
894        """Walk upwards and get first ancestor with this token type."""
895        t = self.getParent()
896        while t is not None:
897            if t.getType() == ttype:
898                return t
899            t = t.getParent()
900
901        return None
902
903    def getAncestors(self):
904        """Return a list of all ancestors of this node.
905
906        The first node of list is the root and the last is the parent of
907        this node.
908        """
909        if selfgetParent() is None:
910            return None
911
912        ancestors = []
913        t = self.getParent()
914        while t is not None:
915            ancestors.insert(0, t) # insert at start
916            t = t.getParent()
917
918        return ancestors
919
920
921    def toStringTree(self):
922        """Print out a whole tree not just a node"""
923
924        if len(self.children) == 0:
925            return self.toString()
926
927        buf = []
928        if not self.isNil():
929            buf.append('(')
930            buf.append(self.toString())
931            buf.append(' ')
932
933        for i, child in enumerate(self.children):
934            if i > 0:
935                buf.append(' ')
936            buf.append(child.toStringTree())
937
938        if not self.isNil():
939            buf.append(')')
940
941        return ''.join(buf)
942
943
944    def getLine(self):
945        return 0
946
947
948    def getCharPositionInLine(self):
949        return 0
950
951
952    def toString(self):
953        """Override to say how a node (not a tree) should look as text"""
954
955        raise NotImplementedError
956
957
958
959class BaseTreeAdaptor(TreeAdaptor):
960    """
961    @brief A TreeAdaptor that works with any Tree implementation.
962    """
963
964    # BaseTreeAdaptor is abstract, no need to complain about not implemented
965    # abstract methods
966    # pylint: disable-msg=W0223
967
968    def nil(self):
969        return self.createWithPayload(None)
970
971
972    def errorNode(self, input, start, stop, exc):
973        """
974        create tree node that holds the start and stop tokens associated
975        with an error.
976
977        If you specify your own kind of tree nodes, you will likely have to
978        override this method. CommonTree returns Token.INVALID_TOKEN_TYPE
979        if no token payload but you might have to set token type for diff
980        node type.
981
982        You don't have to subclass CommonErrorNode; you will likely need to
983        subclass your own tree node class to avoid class cast exception.
984        """
985
986        return CommonErrorNode(input, start, stop, exc)
987
988
989    def isNil(self, tree):
990        return tree.isNil()
991
992
993    def dupTree(self, t, parent=None):
994        """
995        This is generic in the sense that it will work with any kind of
996        tree (not just Tree interface).  It invokes the adaptor routines
997        not the tree node routines to do the construction.
998        """
999
1000        if t is None:
1001            return None
1002
1003        newTree = self.dupNode(t)
1004
1005        # ensure new subtree root has parent/child index set
1006
1007        # same index in new tree
1008        self.setChildIndex(newTree, self.getChildIndex(t))
1009
1010        self.setParent(newTree, parent)
1011
1012        for i in range(self.getChildCount(t)):
1013            child = self.getChild(t, i)
1014            newSubTree = self.dupTree(child, t)
1015            self.addChild(newTree, newSubTree)
1016
1017        return newTree
1018
1019
1020    def addChild(self, tree, child):
1021        """
1022        Add a child to the tree t.  If child is a flat tree (a list), make all
1023        in list children of t.  Warning: if t has no children, but child does
1024        and child isNil then you can decide it is ok to move children to t via
1025        t.children = child.children; i.e., without copying the array.  Just
1026        make sure that this is consistent with have the user will build
1027        ASTs.
1028        """
1029
1030        #if isinstance(child, Token):
1031        #    child = self.createWithPayload(child)
1032
1033        if tree is not None and child is not None:
1034            tree.addChild(child)
1035
1036
1037    def becomeRoot(self, newRoot, oldRoot):
1038        """
1039        If oldRoot is a nil root, just copy or move the children to newRoot.
1040        If not a nil root, make oldRoot a child of newRoot.
1041
1042          old=^(nil a b c), new=r yields ^(r a b c)
1043          old=^(a b c), new=r yields ^(r ^(a b c))
1044
1045        If newRoot is a nil-rooted single child tree, use the single
1046        child as the new root node.
1047
1048          old=^(nil a b c), new=^(nil r) yields ^(r a b c)
1049          old=^(a b c), new=^(nil r) yields ^(r ^(a b c))
1050
1051        If oldRoot was null, it's ok, just return newRoot (even if isNil).
1052
1053          old=null, new=r yields r
1054          old=null, new=^(nil r) yields ^(nil r)
1055
1056        Return newRoot.  Throw an exception if newRoot is not a
1057        simple node or nil root with a single child node--it must be a root
1058        node.  If newRoot is ^(nil x) return x as newRoot.
1059
1060        Be advised that it's ok for newRoot to point at oldRoot's
1061        children; i.e., you don't have to copy the list.  We are
1062        constructing these nodes so we should have this control for
1063        efficiency.
1064        """
1065
1066        if isinstance(newRoot, Token):
1067            newRoot = self.create(newRoot)
1068
1069        if oldRoot is None:
1070            return newRoot
1071
1072        if not isinstance(newRoot, CommonTree):
1073            newRoot = self.createWithPayload(newRoot)
1074
1075        # handle ^(nil real-node)
1076        if newRoot.isNil():
1077            nc = newRoot.getChildCount()
1078            if nc == 1:
1079                newRoot = newRoot.getChild(0)
1080
1081            elif nc > 1:
1082                # TODO: make tree run time exceptions hierarchy
1083                raise RuntimeError("more than one node as root")
1084
1085        # add oldRoot to newRoot; addChild takes care of case where oldRoot
1086        # is a flat list (i.e., nil-rooted tree).  All children of oldRoot
1087        # are added to newRoot.
1088        newRoot.addChild(oldRoot)
1089        return newRoot
1090
1091
1092    def rulePostProcessing(self, root):
1093        """Transform ^(nil x) to x and nil to null"""
1094
1095        if root is not None and root.isNil():
1096            if root.getChildCount() == 0:
1097                root = None
1098
1099            elif root.getChildCount() == 1:
1100                root = root.getChild(0)
1101                # whoever invokes rule will set parent and child index
1102                root.setParent(None)
1103                root.setChildIndex(-1)
1104
1105        return root
1106
1107
1108    def createFromToken(self, tokenType, fromToken, text=None):
1109        if fromToken is None:
1110            return self.createFromType(tokenType, text)
1111
1112        assert isinstance(tokenType, (int, long)), type(tokenType).__name__
1113        assert isinstance(fromToken, Token), type(fromToken).__name__
1114        assert text is None or isinstance(text, basestring), type(text).__name__
1115
1116        fromToken = self.createToken(fromToken)
1117        fromToken.type = tokenType
1118        if text is not None:
1119            fromToken.text = text
1120        t = self.createWithPayload(fromToken)
1121        return t
1122
1123
1124    def createFromType(self, tokenType, text):
1125        assert isinstance(tokenType, (int, long)), type(tokenType).__name__
1126        assert isinstance(text, basestring) or text is None, type(text).__name__
1127
1128        fromToken = self.createToken(tokenType=tokenType, text=text)
1129        t = self.createWithPayload(fromToken)
1130        return t
1131
1132
1133    def getType(self, t):
1134        return t.getType()
1135
1136
1137    def setType(self, t, type):
1138        raise RuntimeError("don't know enough about Tree node")
1139
1140
1141    def getText(self, t):
1142        return t.getText()
1143
1144
1145    def setText(self, t, text):
1146        raise RuntimeError("don't know enough about Tree node")
1147
1148
1149    def getChild(self, t, i):
1150        return t.getChild(i)
1151
1152
1153    def setChild(self, t, i, child):
1154        t.setChild(i, child)
1155
1156
1157    def deleteChild(self, t, i):
1158        return t.deleteChild(i)
1159
1160
1161    def getChildCount(self, t):
1162        return t.getChildCount()
1163
1164
1165    def getUniqueID(self, node):
1166        return hash(node)
1167
1168
1169    def createToken(self, fromToken=None, tokenType=None, text=None):
1170        """
1171        Tell me how to create a token for use with imaginary token nodes.
1172        For example, there is probably no input symbol associated with imaginary
1173        token DECL, but you need to create it as a payload or whatever for
1174        the DECL node as in ^(DECL type ID).
1175
1176        If you care what the token payload objects' type is, you should
1177        override this method and any other createToken variant.
1178        """
1179
1180        raise NotImplementedError
1181
1182
1183############################################################################
1184#
1185# common tree implementation
1186#
1187# Tree
1188# \- BaseTree
1189#    \- CommonTree
1190#       \- CommonErrorNode
1191#
1192# TreeAdaptor
1193# \- BaseTreeAdaptor
1194#    \- CommonTreeAdaptor
1195#
1196############################################################################
1197
1198
1199class CommonTree(BaseTree):
1200    """@brief A tree node that is wrapper for a Token object.
1201
1202    After 3.0 release
1203    while building tree rewrite stuff, it became clear that computing
1204    parent and child index is very difficult and cumbersome.  Better to
1205    spend the space in every tree node.  If you don't want these extra
1206    fields, it's easy to cut them out in your own BaseTree subclass.
1207
1208    """
1209
1210    def __init__(self, payload):
1211        BaseTree.__init__(self)
1212
1213        # What token indexes bracket all tokens associated with this node
1214        # and below?
1215        self.startIndex = -1
1216        self.stopIndex = -1
1217
1218        # Who is the parent node of this node; if null, implies node is root
1219        self.parent = None
1220
1221        # What index is this node in the child list? Range: 0..n-1
1222        self.childIndex = -1
1223
1224        # A single token is the payload
1225        if payload is None:
1226            self.token = None
1227
1228        elif isinstance(payload, CommonTree):
1229            self.token = payload.token
1230            self.startIndex = payload.startIndex
1231            self.stopIndex = payload.stopIndex
1232
1233        elif payload is None or isinstance(payload, Token):
1234            self.token = payload
1235
1236        else:
1237            raise TypeError(type(payload).__name__)
1238
1239
1240
1241    def getToken(self):
1242        return self.token
1243
1244
1245    def dupNode(self):
1246        return CommonTree(self)
1247
1248
1249    def isNil(self):
1250        return self.token is None
1251
1252
1253    def getType(self):
1254        if self.token is None:
1255            return INVALID_TOKEN_TYPE
1256
1257        return self.token.getType()
1258
1259    type = property(getType)
1260
1261
1262    def getText(self):
1263        if self.token is None:
1264            return None
1265
1266        return self.token.text
1267
1268    text = property(getText)
1269
1270
1271    def getLine(self):
1272        if self.token is None or self.token.getLine() == 0:
1273            if self.getChildCount():
1274                return self.getChild(0).getLine()
1275            else:
1276                return 0
1277
1278        return self.token.getLine()
1279
1280    line = property(getLine)
1281
1282
1283    def getCharPositionInLine(self):
1284        if self.token is None or self.token.getCharPositionInLine() == -1:
1285            if self.getChildCount():
1286                return self.getChild(0).getCharPositionInLine()
1287            else:
1288                return 0
1289
1290        else:
1291            return self.token.getCharPositionInLine()
1292
1293    charPositionInLine = property(getCharPositionInLine)
1294
1295
1296    def getTokenStartIndex(self):
1297        if self.startIndex == -1 and self.token is not None:
1298            return self.token.getTokenIndex()
1299
1300        return self.startIndex
1301
1302    def setTokenStartIndex(self, index):
1303        self.startIndex = index
1304
1305    tokenStartIndex = property(getTokenStartIndex, setTokenStartIndex)
1306
1307
1308    def getTokenStopIndex(self):
1309        if self.stopIndex == -1 and self.token is not None:
1310            return self.token.getTokenIndex()
1311
1312        return self.stopIndex
1313
1314    def setTokenStopIndex(self, index):
1315        self.stopIndex = index
1316
1317    tokenStopIndex = property(getTokenStopIndex, setTokenStopIndex)
1318
1319
1320    def setUnknownTokenBoundaries(self):
1321        """For every node in this subtree, make sure it's start/stop token's
1322        are set.  Walk depth first, visit bottom up.  Only updates nodes
1323        with at least one token index < 0.
1324        """
1325
1326        if self.children is None:
1327            if self.startIndex < 0 or self.stopIndex < 0:
1328                self.startIndex = self.stopIndex = self.token.getTokenIndex()
1329
1330            return
1331
1332        for child in self.children:
1333            child.setUnknownTokenBoundaries()
1334
1335        if self.startIndex >= 0 and self.stopIndex >= 0:
1336            # already set
1337            return
1338
1339        if self.children:
1340            firstChild = self.children[0]
1341            lastChild = self.children[-1]
1342            self.startIndex = firstChild.getTokenStartIndex()
1343            self.stopIndex = lastChild.getTokenStopIndex()
1344
1345
1346    def getChildIndex(self):
1347        #FIXME: mark as deprecated
1348        return self.childIndex
1349
1350
1351    def setChildIndex(self, idx):
1352        #FIXME: mark as deprecated
1353        self.childIndex = idx
1354
1355
1356    def getParent(self):
1357        #FIXME: mark as deprecated
1358        return self.parent
1359
1360
1361    def setParent(self, t):
1362        #FIXME: mark as deprecated
1363        self.parent = t
1364
1365
1366    def toString(self):
1367        if self.isNil():
1368            return "nil"
1369
1370        if self.getType() == INVALID_TOKEN_TYPE:
1371            return "<errornode>"
1372
1373        return self.token.text
1374
1375    __str__ = toString
1376
1377
1378
1379    def toStringTree(self):
1380        if not self.children:
1381            return self.toString()
1382
1383        ret = ''
1384        if not self.isNil():
1385            ret += '(%s ' % (self.toString())
1386
1387        ret += ' '.join([child.toStringTree() for child in self.children])
1388
1389        if not self.isNil():
1390            ret += ')'
1391
1392        return ret
1393
1394
1395INVALID_NODE = CommonTree(INVALID_TOKEN)
1396
1397
1398class CommonErrorNode(CommonTree):
1399    """A node representing erroneous token range in token stream"""
1400
1401    def __init__(self, input, start, stop, exc):
1402        CommonTree.__init__(self, None)
1403
1404        if (stop is None or
1405            (stop.getTokenIndex() < start.getTokenIndex() and
1406             stop.getType() != EOF
1407             )
1408            ):
1409            # sometimes resync does not consume a token (when LT(1) is
1410            # in follow set.  So, stop will be 1 to left to start. adjust.
1411            # Also handle case where start is the first token and no token
1412            # is consumed during recovery; LT(-1) will return null.
1413            stop = start
1414
1415        self.input = input
1416        self.start = start
1417        self.stop = stop
1418        self.trappedException = exc
1419
1420
1421    def isNil(self):
1422        return False
1423
1424
1425    def getType(self):
1426        return INVALID_TOKEN_TYPE
1427
1428
1429    def getText(self):
1430        if isinstance(self.start, Token):
1431            i = self.start.getTokenIndex()
1432            j = self.stop.getTokenIndex()
1433            if self.stop.getType() == EOF:
1434                j = self.input.size()
1435
1436            badText = self.input.toString(i, j)
1437
1438        elif isinstance(self.start, Tree):
1439            badText = self.input.toString(self.start, self.stop)
1440
1441        else:
1442            # people should subclass if they alter the tree type so this
1443            # next one is for sure correct.
1444            badText = "<unknown>"
1445
1446        return badText
1447
1448
1449    def toString(self):
1450        if isinstance(self.trappedException, MissingTokenException):
1451            return ("<missing type: "
1452                    + str(self.trappedException.getMissingType())
1453                    + ">")
1454
1455        elif isinstance(self.trappedException, UnwantedTokenException):
1456            return ("<extraneous: "
1457                    + str(self.trappedException.getUnexpectedToken())
1458                    + ", resync=" + self.getText() + ">")
1459
1460        elif isinstance(self.trappedException, MismatchedTokenException):
1461            return ("<mismatched token: "
1462                    + str(self.trappedException.token)
1463                    + ", resync=" + self.getText() + ">")
1464
1465        elif isinstance(self.trappedException, NoViableAltException):
1466            return ("<unexpected: "
1467                    + str(self.trappedException.token)
1468                    + ", resync=" + self.getText() + ">")
1469
1470        return "<error: "+self.getText()+">"
1471
1472
1473class CommonTreeAdaptor(BaseTreeAdaptor):
1474    """
1475    @brief A TreeAdaptor that works with any Tree implementation.
1476
1477    It provides
1478    really just factory methods; all the work is done by BaseTreeAdaptor.
1479    If you would like to have different tokens created than ClassicToken
1480    objects, you need to override this and then set the parser tree adaptor to
1481    use your subclass.
1482
1483    To get your parser to build nodes of a different type, override
1484    create(Token), errorNode(), and to be safe, YourTreeClass.dupNode().
1485    dupNode is called to duplicate nodes during rewrite operations.
1486    """
1487
1488    def dupNode(self, treeNode):
1489        """
1490        Duplicate a node.  This is part of the factory;
1491        override if you want another kind of node to be built.
1492
1493        I could use reflection to prevent having to override this
1494        but reflection is slow.
1495        """
1496
1497        if treeNode is None:
1498            return None
1499
1500        return treeNode.dupNode()
1501
1502
1503    def createWithPayload(self, payload):
1504        return CommonTree(payload)
1505
1506
1507    def createToken(self, fromToken=None, tokenType=None, text=None):
1508        """
1509        Tell me how to create a token for use with imaginary token nodes.
1510        For example, there is probably no input symbol associated with imaginary
1511        token DECL, but you need to create it as a payload or whatever for
1512        the DECL node as in ^(DECL type ID).
1513
1514        If you care what the token payload objects' type is, you should
1515        override this method and any other createToken variant.
1516        """
1517
1518        if fromToken is not None:
1519            return CommonToken(oldToken=fromToken)
1520
1521        return CommonToken(type=tokenType, text=text)
1522
1523
1524    def setTokenBoundaries(self, t, startToken, stopToken):
1525        """
1526        Track start/stop token for subtree root created for a rule.
1527        Only works with Tree nodes.  For rules that match nothing,
1528        seems like this will yield start=i and stop=i-1 in a nil node.
1529        Might be useful info so I'll not force to be i..i.
1530        """
1531
1532        if t is None:
1533            return
1534
1535        start = 0
1536        stop = 0
1537
1538        if startToken is not None:
1539            start = startToken.index
1540
1541        if stopToken is not None:
1542            stop = stopToken.index
1543
1544        t.setTokenStartIndex(start)
1545        t.setTokenStopIndex(stop)
1546
1547
1548    def getTokenStartIndex(self, t):
1549        if t is None:
1550            return -1
1551        return t.getTokenStartIndex()
1552
1553
1554    def getTokenStopIndex(self, t):
1555        if t is None:
1556            return -1
1557        return t.getTokenStopIndex()
1558
1559
1560    def getText(self, t):
1561        if t is None:
1562            return None
1563        return t.getText()
1564
1565
1566    def getType(self, t):
1567        if t is None:
1568            return INVALID_TOKEN_TYPE
1569
1570        return t.getType()
1571
1572
1573    def getToken(self, t):
1574        """
1575        What is the Token associated with this node?  If
1576        you are not using CommonTree, then you must
1577        override this in your own adaptor.
1578        """
1579
1580        if isinstance(t, CommonTree):
1581            return t.getToken()
1582
1583        return None # no idea what to do
1584
1585
1586    def getChild(self, t, i):
1587        if t is None:
1588            return None
1589        return t.getChild(i)
1590
1591
1592    def getChildCount(self, t):
1593        if t is None:
1594            return 0
1595        return t.getChildCount()
1596
1597
1598    def getParent(self, t):
1599        return t.getParent()
1600
1601
1602    def setParent(self, t, parent):
1603        t.setParent(parent)
1604
1605
1606    def getChildIndex(self, t):
1607        if t is None:
1608            return 0
1609        return t.getChildIndex()
1610
1611
1612    def setChildIndex(self, t, index):
1613        t.setChildIndex(index)
1614
1615
1616    def replaceChildren(self, parent, startChildIndex, stopChildIndex, t):
1617        if parent is not None:
1618            parent.replaceChildren(startChildIndex, stopChildIndex, t)
1619
1620
1621############################################################################
1622#
1623# streams
1624#
1625# TreeNodeStream
1626# \- BaseTree
1627#    \- CommonTree
1628#
1629# TreeAdaptor
1630# \- BaseTreeAdaptor
1631#    \- CommonTreeAdaptor
1632#
1633############################################################################
1634
1635
1636
1637class TreeNodeStream(IntStream):
1638    """@brief A stream of tree nodes
1639
1640    It accessing nodes from a tree of some kind.
1641    """
1642
1643    # TreeNodeStream is abstract, no need to complain about not implemented
1644    # abstract methods
1645    # pylint: disable-msg=W0223
1646
1647    def get(self, i):
1648        """Get a tree node at an absolute index i; 0..n-1.
1649        If you don't want to buffer up nodes, then this method makes no
1650        sense for you.
1651        """
1652
1653        raise NotImplementedError
1654
1655
1656    def LT(self, k):
1657        """
1658        Get tree node at current input pointer + i ahead where i=1 is next node.
1659        i<0 indicates nodes in the past.  So LT(-1) is previous node, but
1660        implementations are not required to provide results for k < -1.
1661        LT(0) is undefined.  For i>=n, return null.
1662        Return null for LT(0) and any index that results in an absolute address
1663        that is negative.
1664
1665        This is analogus to the LT() method of the TokenStream, but this
1666        returns a tree node instead of a token.  Makes code gen identical
1667        for both parser and tree grammars. :)
1668        """
1669
1670        raise NotImplementedError
1671
1672
1673    def getTreeSource(self):
1674        """
1675        Where is this stream pulling nodes from?  This is not the name, but
1676        the object that provides node objects.
1677        """
1678
1679        raise NotImplementedError
1680
1681
1682    def getTokenStream(self):
1683        """
1684        If the tree associated with this stream was created from a TokenStream,
1685        you can specify it here.  Used to do rule $text attribute in tree
1686        parser.  Optional unless you use tree parser rule text attribute
1687        or output=template and rewrite=true options.
1688        """
1689
1690        raise NotImplementedError
1691
1692
1693    def getTreeAdaptor(self):
1694        """
1695        What adaptor can tell me how to interpret/navigate nodes and
1696        trees.  E.g., get text of a node.
1697        """
1698
1699        raise NotImplementedError
1700
1701
1702    def setUniqueNavigationNodes(self, uniqueNavigationNodes):
1703        """
1704        As we flatten the tree, we use UP, DOWN nodes to represent
1705        the tree structure.  When debugging we need unique nodes
1706        so we have to instantiate new ones.  When doing normal tree
1707        parsing, it's slow and a waste of memory to create unique
1708        navigation nodes.  Default should be false;
1709        """
1710
1711        raise NotImplementedError
1712
1713
1714    def reset(self):
1715        """
1716        Reset the tree node stream in such a way that it acts like
1717        a freshly constructed stream.
1718        """
1719
1720        raise NotImplementedError
1721
1722
1723    def toString(self, start, stop):
1724        """
1725        Return the text of all nodes from start to stop, inclusive.
1726        If the stream does not buffer all the nodes then it can still
1727        walk recursively from start until stop.  You can always return
1728        null or "" too, but users should not access $ruleLabel.text in
1729        an action of course in that case.
1730        """
1731
1732        raise NotImplementedError
1733
1734
1735    # REWRITING TREES (used by tree parser)
1736    def replaceChildren(self, parent, startChildIndex, stopChildIndex, t):
1737        """
1738 	Replace from start to stop child index of parent with t, which might
1739        be a list.  Number of children may be different
1740        after this call.  The stream is notified because it is walking the
1741        tree and might need to know you are monkeying with the underlying
1742        tree.  Also, it might be able to modify the node stream to avoid
1743        restreaming for future phases.
1744
1745        If parent is null, don't do anything; must be at root of overall tree.
1746        Can't replace whatever points to the parent externally.  Do nothing.
1747        """
1748
1749        raise NotImplementedError
1750
1751
1752class CommonTreeNodeStream(TreeNodeStream):
1753    """@brief A buffered stream of tree nodes.
1754
1755    Nodes can be from a tree of ANY kind.
1756
1757    This node stream sucks all nodes out of the tree specified in
1758    the constructor during construction and makes pointers into
1759    the tree using an array of Object pointers. The stream necessarily
1760    includes pointers to DOWN and UP and EOF nodes.
1761
1762    This stream knows how to mark/release for backtracking.
1763
1764    This stream is most suitable for tree interpreters that need to
1765    jump around a lot or for tree parsers requiring speed (at cost of memory).
1766    There is some duplicated functionality here with UnBufferedTreeNodeStream
1767    but just in bookkeeping, not tree walking etc...
1768
1769    @see UnBufferedTreeNodeStream
1770    """
1771
1772    def __init__(self, *args):
1773        TreeNodeStream.__init__(self)
1774
1775        if len(args) == 1:
1776            adaptor = CommonTreeAdaptor()
1777            tree = args[0]
1778
1779            nodes = None
1780            down = None
1781            up = None
1782            eof = None
1783
1784        elif len(args) == 2:
1785            adaptor = args[0]
1786            tree = args[1]
1787
1788            nodes = None
1789            down = None
1790            up = None
1791            eof = None
1792
1793        elif len(args) == 3:
1794            parent = args[0]
1795            start = args[1]
1796            stop = args[2]
1797
1798            adaptor = parent.adaptor
1799            tree = parent.root
1800
1801            nodes = parent.nodes[start:stop]
1802            down = parent.down
1803            up = parent.up
1804            eof = parent.eof
1805
1806        else:
1807            raise TypeError("Invalid arguments")
1808
1809        # all these navigation nodes are shared and hence they
1810        # cannot contain any line/column info
1811        if down is not None:
1812            self.down = down
1813        else:
1814            self.down = adaptor.createFromType(DOWN, "DOWN")
1815
1816        if up is not None:
1817            self.up = up
1818        else:
1819            self.up = adaptor.createFromType(UP, "UP")
1820
1821        if eof is not None:
1822            self.eof = eof
1823        else:
1824            self.eof = adaptor.createFromType(EOF, "EOF")
1825
1826        # The complete mapping from stream index to tree node.
1827        # This buffer includes pointers to DOWN, UP, and EOF nodes.
1828        # It is built upon ctor invocation.  The elements are type
1829        #  Object as we don't what the trees look like.
1830
1831        # Load upon first need of the buffer so we can set token types
1832        # of interest for reverseIndexing.  Slows us down a wee bit to
1833        # do all of the if p==-1 testing everywhere though.
1834        if nodes is not None:
1835            self.nodes = nodes
1836        else:
1837            self.nodes = []
1838
1839        # Pull nodes from which tree?
1840        self.root = tree
1841
1842        # IF this tree (root) was created from a token stream, track it.
1843        self.tokens = None
1844
1845        # What tree adaptor was used to build these trees
1846        self.adaptor = adaptor
1847
1848        # Reuse same DOWN, UP navigation nodes unless this is true
1849        self.uniqueNavigationNodes = False
1850
1851        # The index into the nodes list of the current node (next node
1852        # to consume).  If -1, nodes array not filled yet.
1853        self.p = -1
1854
1855        # Track the last mark() call result value for use in rewind().
1856        self.lastMarker = None
1857
1858        # Stack of indexes used for push/pop calls
1859        self.calls = []
1860
1861
1862    def __iter__(self):
1863        return TreeIterator(self.root, self.adaptor)
1864
1865
1866    def fillBuffer(self):
1867        """Walk tree with depth-first-search and fill nodes buffer.
1868        Don't do DOWN, UP nodes if its a list (t is isNil).
1869        """
1870
1871        self._fillBuffer(self.root)
1872        self.p = 0 # buffer of nodes intialized now
1873
1874
1875    def _fillBuffer(self, t):
1876        nil = self.adaptor.isNil(t)
1877
1878        if not nil:
1879            self.nodes.append(t) # add this node
1880
1881        # add DOWN node if t has children
1882        n = self.adaptor.getChildCount(t)
1883        if not nil and n > 0:
1884            self.addNavigationNode(DOWN)
1885
1886        # and now add all its children
1887        for c in range(n):
1888            self._fillBuffer(self.adaptor.getChild(t, c))
1889
1890        # add UP node if t has children
1891        if not nil and n > 0:
1892            self.addNavigationNode(UP)
1893
1894
1895    def getNodeIndex(self, node):
1896        """What is the stream index for node? 0..n-1
1897        Return -1 if node not found.
1898        """
1899
1900        if self.p == -1:
1901            self.fillBuffer()
1902
1903        for i, t in enumerate(self.nodes):
1904            if t == node:
1905                return i
1906
1907        return -1
1908
1909
1910    def addNavigationNode(self, ttype):
1911        """
1912        As we flatten the tree, we use UP, DOWN nodes to represent
1913        the tree structure.  When debugging we need unique nodes
1914        so instantiate new ones when uniqueNavigationNodes is true.
1915        """
1916
1917        navNode = None
1918
1919        if ttype == DOWN:
1920            if self.hasUniqueNavigationNodes():
1921                navNode = self.adaptor.createFromType(DOWN, "DOWN")
1922
1923            else:
1924                navNode = self.down
1925
1926        else:
1927            if self.hasUniqueNavigationNodes():
1928                navNode = self.adaptor.createFromType(UP, "UP")
1929
1930            else:
1931                navNode = self.up
1932
1933        self.nodes.append(navNode)
1934
1935
1936    def get(self, i):
1937        if self.p == -1:
1938            self.fillBuffer()
1939
1940        return self.nodes[i]
1941
1942
1943    def LT(self, k):
1944        if self.p == -1:
1945            self.fillBuffer()
1946
1947        if k == 0:
1948            return None
1949
1950        if k < 0:
1951            return self.LB(-k)
1952
1953        if self.p + k - 1 >= len(self.nodes):
1954            return self.eof
1955
1956        return self.nodes[self.p + k - 1]
1957
1958
1959    def getCurrentSymbol(self):
1960        return self.LT(1)
1961
1962
1963    def LB(self, k):
1964        """Look backwards k nodes"""
1965
1966        if k == 0:
1967            return None
1968
1969        if self.p - k < 0:
1970            return None
1971
1972        return self.nodes[self.p - k]
1973
1974
1975    def isEOF(self, obj):
1976        return self.adaptor.getType(obj) == EOF
1977
1978
1979    def getTreeSource(self):
1980        return self.root
1981
1982
1983    def getSourceName(self):
1984        return self.getTokenStream().getSourceName()
1985
1986
1987    def getTokenStream(self):
1988        return self.tokens
1989
1990
1991    def setTokenStream(self, tokens):
1992        self.tokens = tokens
1993
1994
1995    def getTreeAdaptor(self):
1996        return self.adaptor
1997
1998
1999    def hasUniqueNavigationNodes(self):
2000        return self.uniqueNavigationNodes
2001
2002
2003    def setUniqueNavigationNodes(self, uniqueNavigationNodes):
2004        self.uniqueNavigationNodes = uniqueNavigationNodes
2005
2006
2007    def consume(self):
2008        if self.p == -1:
2009            self.fillBuffer()
2010
2011        self.p += 1
2012
2013
2014    def LA(self, i):
2015        return self.adaptor.getType(self.LT(i))
2016
2017
2018    def mark(self):
2019        if self.p == -1:
2020            self.fillBuffer()
2021
2022
2023        self.lastMarker = self.index()
2024        return self.lastMarker
2025
2026
2027    def release(self, marker=None):
2028        # no resources to release
2029
2030        pass
2031
2032
2033    def index(self):
2034        return self.p
2035
2036
2037    def rewind(self, marker=None):
2038        if marker is None:
2039            marker = self.lastMarker
2040
2041        self.seek(marker)
2042
2043
2044    def seek(self, index):
2045        if self.p == -1:
2046            self.fillBuffer()
2047
2048        self.p = index
2049
2050
2051    def push(self, index):
2052        """
2053        Make stream jump to a new location, saving old location.
2054        Switch back with pop().
2055        """
2056
2057        self.calls.append(self.p) # save current index
2058        self.seek(index)
2059
2060
2061    def pop(self):
2062        """
2063        Seek back to previous index saved during last push() call.
2064        Return top of stack (return index).
2065        """
2066
2067        ret = self.calls.pop(-1)
2068        self.seek(ret)
2069        return ret
2070
2071
2072    def reset(self):
2073        self.p = 0
2074        self.lastMarker = 0
2075        self.calls = []
2076
2077
2078    def size(self):
2079        if self.p == -1:
2080            self.fillBuffer()
2081
2082        return len(self.nodes)
2083
2084
2085    # TREE REWRITE INTERFACE
2086
2087    def replaceChildren(self, parent, startChildIndex, stopChildIndex, t):
2088        if parent is not None:
2089            self.adaptor.replaceChildren(
2090                parent, startChildIndex, stopChildIndex, t
2091                )
2092
2093
2094    def __str__(self):
2095        """Used for testing, just return the token type stream"""
2096
2097        if self.p == -1:
2098            self.fillBuffer()
2099
2100        return ' '.join([str(self.adaptor.getType(node))
2101                         for node in self.nodes
2102                         ])
2103
2104
2105    def toString(self, start, stop):
2106        if start is None or stop is None:
2107            return None
2108
2109        if self.p == -1:
2110            self.fillBuffer()
2111
2112        #System.out.println("stop: "+stop);
2113        #if ( start instanceof CommonTree )
2114        #    System.out.print("toString: "+((CommonTree)start).getToken()+", ");
2115        #else
2116        #    System.out.println(start);
2117        #if ( stop instanceof CommonTree )
2118        #    System.out.println(((CommonTree)stop).getToken());
2119        #else
2120        #    System.out.println(stop);
2121
2122        # if we have the token stream, use that to dump text in order
2123        if self.tokens is not None:
2124            beginTokenIndex = self.adaptor.getTokenStartIndex(start)
2125            endTokenIndex = self.adaptor.getTokenStopIndex(stop)
2126
2127            # if it's a tree, use start/stop index from start node
2128            # else use token range from start/stop nodes
2129            if self.adaptor.getType(stop) == UP:
2130                endTokenIndex = self.adaptor.getTokenStopIndex(start)
2131
2132            elif self.adaptor.getType(stop) == EOF:
2133                endTokenIndex = self.size() -2 # don't use EOF
2134
2135            return self.tokens.toString(beginTokenIndex, endTokenIndex)
2136
2137        # walk nodes looking for start
2138        i, t = 0, None
2139        for i, t in enumerate(self.nodes):
2140            if t == start:
2141                break
2142
2143        # now walk until we see stop, filling string buffer with text
2144        buf = []
2145        t = self.nodes[i]
2146        while t != stop:
2147            text = self.adaptor.getText(t)
2148            if text is None:
2149                text = " " + self.adaptor.getType(t)
2150
2151            buf.append(text)
2152            i += 1
2153            t = self.nodes[i]
2154
2155        # include stop node too
2156        text = self.adaptor.getText(stop)
2157        if text is None:
2158            text = " " +self.adaptor.getType(stop)
2159
2160        buf.append(text)
2161
2162        return ''.join(buf)
2163
2164
2165    ## iterator interface
2166    def __iter__(self):
2167        if self.p == -1:
2168            self.fillBuffer()
2169
2170        for node in self.nodes:
2171            yield node
2172
2173
2174#############################################################################
2175#
2176# tree parser
2177#
2178#############################################################################
2179
2180class TreeParser(BaseRecognizer):
2181    """@brief Baseclass for generated tree parsers.
2182
2183    A parser for a stream of tree nodes.  "tree grammars" result in a subclass
2184    of this.  All the error reporting and recovery is shared with Parser via
2185    the BaseRecognizer superclass.
2186    """
2187
2188    def __init__(self, input, state=None):
2189        BaseRecognizer.__init__(self, state)
2190
2191        self.input = None
2192        self.setTreeNodeStream(input)
2193
2194
2195    def reset(self):
2196        BaseRecognizer.reset(self) # reset all recognizer state variables
2197        if self.input is not None:
2198            self.input.seek(0) # rewind the input
2199
2200
2201    def setTreeNodeStream(self, input):
2202        """Set the input stream"""
2203
2204        self.input = input
2205
2206
2207    def getTreeNodeStream(self):
2208        return self.input
2209
2210
2211    def getSourceName(self):
2212        return self.input.getSourceName()
2213
2214
2215    def getCurrentInputSymbol(self, input):
2216        return input.LT(1)
2217
2218
2219    def getMissingSymbol(self, input, e, expectedTokenType, follow):
2220        tokenText = "<missing " + self.tokenNames[expectedTokenType] + ">"
2221        adaptor = input.adaptor
2222        return adaptor.createToken(
2223            CommonToken(type=expectedTokenType, text=tokenText))
2224
2225
2226    # precompiled regex used by inContext
2227    dotdot = ".*[^.]\\.\\.[^.].*"
2228    doubleEtc = ".*\\.\\.\\.\\s+\\.\\.\\..*"
2229    dotdotPattern = re.compile(dotdot)
2230    doubleEtcPattern = re.compile(doubleEtc)
2231
2232    def inContext(self, context, adaptor=None, tokenName=None, t=None):
2233        """Check if current node in input has a context.
2234
2235        Context means sequence of nodes towards root of tree.  For example,
2236        you might say context is "MULT" which means my parent must be MULT.
2237        "CLASS VARDEF" says current node must be child of a VARDEF and whose
2238        parent is a CLASS node.  You can use "..." to mean zero-or-more nodes.
2239        "METHOD ... VARDEF" means my parent is VARDEF and somewhere above
2240        that is a METHOD node.  The first node in the context is not
2241        necessarily the root.  The context matcher stops matching and returns
2242        true when it runs out of context.  There is no way to force the first
2243        node to be the root.
2244        """
2245
2246        return _inContext(
2247            self.input.getTreeAdaptor(), self.getTokenNames(),
2248            self.input.LT(1), context)
2249
2250    @classmethod
2251    def _inContext(cls, adaptor, tokenNames, t, context):
2252        """The worker for inContext.
2253
2254        It's static and full of parameters for testing purposes.
2255        """
2256
2257        if cls.dotdotPattern.match(context):
2258            # don't allow "..", must be "..."
2259            raise ValueError("invalid syntax: ..")
2260
2261        if cls.doubleEtcPattern.match(context):
2262            # don't allow double "..."
2263            raise ValueError("invalid syntax: ... ...")
2264
2265        # ensure spaces around ...
2266        context = context.replace("...", " ... ")
2267        context = context.strip()
2268        nodes = context.split()
2269
2270        ni = len(nodes) - 1
2271        t = adaptor.getParent(t)
2272        while ni >= 0 and t is not None:
2273            if nodes[ni] == "...":
2274                # walk upwards until we see nodes[ni-1] then continue walking
2275                if ni == 0:
2276                    # ... at start is no-op
2277                    return True
2278                goal = nodes[ni-1]
2279                ancestor = cls._getAncestor(adaptor, tokenNames, t, goal)
2280                if ancestor is None:
2281                    return False
2282                t = ancestor
2283                ni -= 1
2284
2285            name = tokenNames[adaptor.getType(t)]
2286            if name != nodes[ni]:
2287                return False
2288
2289            # advance to parent and to previous element in context node list
2290            ni -= 1
2291            t = adaptor.getParent(t)
2292
2293        # at root but more nodes to match
2294        if t is None and ni >= 0:
2295            return False
2296
2297        return True
2298
2299    @staticmethod
2300    def _getAncestor(adaptor, tokenNames, t, goal):
2301        """Helper for static inContext."""
2302        while t is not None:
2303            name = tokenNames[adaptor.getType(t)]
2304            if name == goal:
2305                return t
2306            t = adaptor.getParent(t)
2307
2308        return None
2309
2310
2311    def matchAny(self, ignore): # ignore stream, copy of this.input
2312        """
2313        Match '.' in tree parser has special meaning.  Skip node or
2314        entire tree if node has children.  If children, scan until
2315        corresponding UP node.
2316        """
2317
2318        self._state.errorRecovery = False
2319
2320        look = self.input.LT(1)
2321        if self.input.getTreeAdaptor().getChildCount(look) == 0:
2322            self.input.consume() # not subtree, consume 1 node and return
2323            return
2324
2325        # current node is a subtree, skip to corresponding UP.
2326        # must count nesting level to get right UP
2327        level = 0
2328        tokenType = self.input.getTreeAdaptor().getType(look)
2329        while tokenType != EOF and not (tokenType == UP and level==0):
2330            self.input.consume()
2331            look = self.input.LT(1)
2332            tokenType = self.input.getTreeAdaptor().getType(look)
2333            if tokenType == DOWN:
2334                level += 1
2335
2336            elif tokenType == UP:
2337                level -= 1
2338
2339        self.input.consume() # consume UP
2340
2341
2342    def mismatch(self, input, ttype, follow):
2343        """
2344        We have DOWN/UP nodes in the stream that have no line info; override.
2345        plus we want to alter the exception type. Don't try to recover
2346        from tree parser errors inline...
2347        """
2348
2349        raise MismatchedTreeNodeException(ttype, input)
2350
2351
2352    def getErrorHeader(self, e):
2353        """
2354        Prefix error message with the grammar name because message is
2355        always intended for the programmer because the parser built
2356        the input tree not the user.
2357        """
2358
2359        return (self.getGrammarFileName() +
2360                ": node from %sline %s:%s"
2361                % (['', "after "][e.approximateLineInfo],
2362                   e.line,
2363                   e.charPositionInLine
2364                   )
2365                )
2366
2367    def getErrorMessage(self, e, tokenNames):
2368        """
2369        Tree parsers parse nodes they usually have a token object as
2370        payload. Set the exception token and do the default behavior.
2371        """
2372
2373        if isinstance(self, TreeParser):
2374            adaptor = e.input.getTreeAdaptor()
2375            e.token = adaptor.getToken(e.node)
2376            if e.token is not None: # could be an UP/DOWN node
2377                e.token = CommonToken(
2378                    type=adaptor.getType(e.node),
2379                    text=adaptor.getText(e.node)
2380                    )
2381
2382        return BaseRecognizer.getErrorMessage(self, e, tokenNames)
2383
2384
2385    def traceIn(self, ruleName, ruleIndex):
2386        BaseRecognizer.traceIn(self, ruleName, ruleIndex, self.input.LT(1))
2387
2388
2389    def traceOut(self, ruleName, ruleIndex):
2390        BaseRecognizer.traceOut(self, ruleName, ruleIndex, self.input.LT(1))
2391
2392
2393#############################################################################
2394#
2395# tree visitor
2396#
2397#############################################################################
2398
2399class TreeVisitor(object):
2400    """Do a depth first walk of a tree, applying pre() and post() actions
2401    we go.
2402    """
2403
2404    def __init__(self, adaptor=None):
2405        if adaptor is not None:
2406            self.adaptor = adaptor
2407        else:
2408            self.adaptor = CommonTreeAdaptor()
2409
2410    def visit(self, t, pre_action=None, post_action=None):
2411        """Visit every node in tree t and trigger an action for each node
2412        before/after having visited all of its children.  Bottom up walk.
2413        Execute both actions even if t has no children.  Ignore return
2414        results from transforming children since they will have altered
2415        the child list of this node (their parent).  Return result of
2416        applying post action to this node.
2417
2418        The Python version differs from the Java version by taking two
2419        callables 'pre_action' and 'post_action' instead of a class instance
2420        that wraps those methods. Those callables must accept a TreeNode as
2421        their single argument and return the (potentially transformed or
2422        replaced) TreeNode.
2423        """
2424
2425        isNil = self.adaptor.isNil(t)
2426        if pre_action is not None and not isNil:
2427            # if rewritten, walk children of new t
2428            t = pre_action(t)
2429
2430        idx = 0
2431        while idx < self.adaptor.getChildCount(t):
2432            child = self.adaptor.getChild(t, idx)
2433            self.visit(child, pre_action, post_action)
2434            idx += 1
2435
2436        if post_action is not None and not isNil:
2437            t = post_action(t)
2438
2439        return t
2440
2441#############################################################################
2442#
2443# tree iterator
2444#
2445#############################################################################
2446
2447class TreeIterator(object):
2448    """
2449    Return a node stream from a doubly-linked tree whose nodes
2450    know what child index they are.
2451
2452    Emit navigation nodes (DOWN, UP, and EOF) to let show tree structure.
2453    """
2454
2455    def __init__(self, tree, adaptor=None):
2456        if adaptor is None:
2457            adaptor = CommonTreeAdaptor()
2458
2459        self.root = tree
2460        self.adaptor = adaptor
2461
2462        self.first_time = True
2463        self.tree = tree
2464
2465        # If we emit UP/DOWN nodes, we need to spit out multiple nodes per
2466        # next() call.
2467        self.nodes = []
2468
2469        # navigation nodes to return during walk and at end
2470        self.down = adaptor.createFromType(DOWN, "DOWN")
2471        self.up = adaptor.createFromType(UP, "UP")
2472        self.eof = adaptor.createFromType(EOF, "EOF")
2473
2474
2475    def reset(self):
2476        self.first_time = True
2477        self.tree = self.root
2478        self.nodes = []
2479
2480
2481    def __iter__(self):
2482        return self
2483
2484
2485    def has_next(self):
2486        if self.first_time:
2487            return self.root is not None
2488
2489        if len(self.nodes) > 0:
2490            return True
2491
2492        if self.tree is None:
2493            return False
2494
2495        if self.adaptor.getChildCount(self.tree) > 0:
2496            return True
2497
2498        # back at root?
2499        return self.adaptor.getParent(self.tree) is not None
2500
2501
2502    def next(self):
2503        if not self.has_next():
2504            raise StopIteration
2505
2506        if self.first_time:
2507            # initial condition
2508            self.first_time = False
2509            if self.adaptor.getChildCount(self.tree) == 0:
2510                # single node tree (special)
2511                self.nodes.append(self.eof)
2512                return self.tree
2513
2514            return self.tree
2515
2516        # if any queued up, use those first
2517        if len(self.nodes) > 0:
2518            return self.nodes.pop(0)
2519
2520        # no nodes left?
2521        if self.tree is None:
2522            return self.eof
2523
2524        # next node will be child 0 if any children
2525        if self.adaptor.getChildCount(self.tree) > 0:
2526            self.tree = self.adaptor.getChild(self.tree, 0)
2527            # real node is next after DOWN
2528            self.nodes.append(self.tree)
2529            return self.down
2530
2531        # if no children, look for next sibling of tree or ancestor
2532        parent = self.adaptor.getParent(self.tree)
2533        # while we're out of siblings, keep popping back up towards root
2534        while (parent is not None
2535               and self.adaptor.getChildIndex(self.tree)+1 >= self.adaptor.getChildCount(parent)):
2536            # we're moving back up
2537            self.nodes.append(self.up)
2538            self.tree = parent
2539            parent = self.adaptor.getParent(self.tree)
2540
2541        # no nodes left?
2542        if parent is None:
2543            self.tree = None # back at root? nothing left then
2544            self.nodes.append(self.eof) # add to queue, might have UP nodes in there
2545            return self.nodes.pop(0)
2546
2547        # must have found a node with an unvisited sibling
2548        # move to it and return it
2549        nextSiblingIndex = self.adaptor.getChildIndex(self.tree) + 1
2550        self.tree = self.adaptor.getChild(parent, nextSiblingIndex)
2551        self.nodes.append(self.tree) # add to queue, might have UP nodes in there
2552        return self.nodes.pop(0)
2553
2554
2555
2556#############################################################################
2557#
2558# streams for rule rewriting
2559#
2560#############################################################################
2561
2562class RewriteRuleElementStream(object):
2563    """@brief Internal helper class.
2564
2565    A generic list of elements tracked in an alternative to be used in
2566    a -> rewrite rule.  We need to subclass to fill in the next() method,
2567    which returns either an AST node wrapped around a token payload or
2568    an existing subtree.
2569
2570    Once you start next()ing, do not try to add more elements.  It will
2571    break the cursor tracking I believe.
2572
2573    @see org.antlr.runtime.tree.RewriteRuleSubtreeStream
2574    @see org.antlr.runtime.tree.RewriteRuleTokenStream
2575
2576    TODO: add mechanism to detect/puke on modification after reading from
2577    stream
2578    """
2579
2580    def __init__(self, adaptor, elementDescription, elements=None):
2581        # Cursor 0..n-1.  If singleElement!=null, cursor is 0 until you next(),
2582        # which bumps it to 1 meaning no more elements.
2583        self.cursor = 0
2584
2585        # Track single elements w/o creating a list.  Upon 2nd add, alloc list
2586        self.singleElement = None
2587
2588        # The list of tokens or subtrees we are tracking
2589        self.elements = None
2590
2591        # Once a node / subtree has been used in a stream, it must be dup'd
2592        # from then on.  Streams are reset after subrules so that the streams
2593        # can be reused in future subrules.  So, reset must set a dirty bit.
2594        # If dirty, then next() always returns a dup.
2595        self.dirty = False
2596
2597        # The element or stream description; usually has name of the token or
2598        # rule reference that this list tracks.  Can include rulename too, but
2599        # the exception would track that info.
2600        self.elementDescription = elementDescription
2601
2602        self.adaptor = adaptor
2603
2604        if isinstance(elements, (list, tuple)):
2605            # Create a stream, but feed off an existing list
2606            self.singleElement = None
2607            self.elements = elements
2608
2609        else:
2610            # Create a stream with one element
2611            self.add(elements)
2612
2613
2614    def reset(self):
2615        """
2616        Reset the condition of this stream so that it appears we have
2617        not consumed any of its elements.  Elements themselves are untouched.
2618        Once we reset the stream, any future use will need duplicates.  Set
2619        the dirty bit.
2620        """
2621
2622        self.cursor = 0
2623        self.dirty = True
2624
2625
2626    def add(self, el):
2627        if el is None:
2628            return
2629
2630        if self.elements is not None: # if in list, just add
2631            self.elements.append(el)
2632            return
2633
2634        if self.singleElement is None: # no elements yet, track w/o list
2635            self.singleElement = el
2636            return
2637
2638        # adding 2nd element, move to list
2639        self.elements = []
2640        self.elements.append(self.singleElement)
2641        self.singleElement = None
2642        self.elements.append(el)
2643
2644
2645    def nextTree(self):
2646        """
2647        Return the next element in the stream.  If out of elements, throw
2648        an exception unless size()==1.  If size is 1, then return elements[0].
2649
2650        Return a duplicate node/subtree if stream is out of elements and
2651        size==1. If we've already used the element, dup (dirty bit set).
2652        """
2653
2654        if (self.dirty
2655            or (self.cursor >= len(self) and len(self) == 1)
2656            ):
2657            # if out of elements and size is 1, dup
2658            el = self._next()
2659            return self.dup(el)
2660
2661        # test size above then fetch
2662        el = self._next()
2663        return el
2664
2665
2666    def _next(self):
2667        """
2668        do the work of getting the next element, making sure that it's
2669        a tree node or subtree.  Deal with the optimization of single-
2670        element list versus list of size > 1.  Throw an exception
2671        if the stream is empty or we're out of elements and size>1.
2672        protected so you can override in a subclass if necessary.
2673        """
2674
2675        if len(self) == 0:
2676            raise RewriteEmptyStreamException(self.elementDescription)
2677
2678        if self.cursor >= len(self): # out of elements?
2679            if len(self) == 1: # if size is 1, it's ok; return and we'll dup
2680                return self.toTree(self.singleElement)
2681
2682            # out of elements and size was not 1, so we can't dup
2683            raise RewriteCardinalityException(self.elementDescription)
2684
2685        # we have elements
2686        if self.singleElement is not None:
2687            self.cursor += 1 # move cursor even for single element list
2688            return self.toTree(self.singleElement)
2689
2690        # must have more than one in list, pull from elements
2691        o = self.toTree(self.elements[self.cursor])
2692        self.cursor += 1
2693        return o
2694
2695
2696    def dup(self, el):
2697        """
2698        When constructing trees, sometimes we need to dup a token or AST
2699        subtree.  Dup'ing a token means just creating another AST node
2700        around it.  For trees, you must call the adaptor.dupTree() unless
2701        the element is for a tree root; then it must be a node dup.
2702        """
2703
2704        raise NotImplementedError
2705
2706
2707    def toTree(self, el):
2708        """
2709        Ensure stream emits trees; tokens must be converted to AST nodes.
2710        AST nodes can be passed through unmolested.
2711        """
2712
2713        return el
2714
2715
2716    def hasNext(self):
2717        return ( (self.singleElement is not None and self.cursor < 1)
2718                 or (self.elements is not None
2719                     and self.cursor < len(self.elements)
2720                     )
2721                 )
2722
2723
2724    def size(self):
2725        if self.singleElement is not None:
2726            return 1
2727
2728        if self.elements is not None:
2729            return len(self.elements)
2730
2731        return 0
2732
2733    __len__ = size
2734
2735
2736    def getDescription(self):
2737        """Deprecated. Directly access elementDescription attribute"""
2738
2739        return self.elementDescription
2740
2741
2742class RewriteRuleTokenStream(RewriteRuleElementStream):
2743    """@brief Internal helper class."""
2744
2745    def toTree(self, el):
2746        # Don't convert to a tree unless they explicitly call nextTree.
2747        # This way we can do hetero tree nodes in rewrite.
2748        return el
2749
2750
2751    def nextNode(self):
2752        t = self._next()
2753        return self.adaptor.createWithPayload(t)
2754
2755
2756    def nextToken(self):
2757        return self._next()
2758
2759
2760    def dup(self, el):
2761        raise TypeError("dup can't be called for a token stream.")
2762
2763
2764class RewriteRuleSubtreeStream(RewriteRuleElementStream):
2765    """@brief Internal helper class."""
2766
2767    def nextNode(self):
2768        """
2769        Treat next element as a single node even if it's a subtree.
2770        This is used instead of next() when the result has to be a
2771        tree root node.  Also prevents us from duplicating recently-added
2772        children; e.g., ^(type ID)+ adds ID to type and then 2nd iteration
2773        must dup the type node, but ID has been added.
2774
2775        Referencing a rule result twice is ok; dup entire tree as
2776        we can't be adding trees as root; e.g., expr expr.
2777
2778        Hideous code duplication here with super.next().  Can't think of
2779        a proper way to refactor.  This needs to always call dup node
2780        and super.next() doesn't know which to call: dup node or dup tree.
2781        """
2782
2783        if (self.dirty
2784            or (self.cursor >= len(self) and len(self) == 1)
2785            ):
2786            # if out of elements and size is 1, dup (at most a single node
2787            # since this is for making root nodes).
2788            el = self._next()
2789            return self.adaptor.dupNode(el)
2790
2791        # test size above then fetch
2792        el = self._next()
2793        while self.adaptor.isNil(el) and self.adaptor.getChildCount(el) == 1:
2794            el = self.adaptor.getChild(el, 0)
2795
2796        # dup just the root (want node here)
2797        return self.adaptor.dupNode(el)
2798
2799
2800    def dup(self, el):
2801        return self.adaptor.dupTree(el)
2802
2803
2804
2805class RewriteRuleNodeStream(RewriteRuleElementStream):
2806    """
2807    Queues up nodes matched on left side of -> in a tree parser. This is
2808    the analog of RewriteRuleTokenStream for normal parsers.
2809    """
2810
2811    def nextNode(self):
2812        return self._next()
2813
2814
2815    def toTree(self, el):
2816        return self.adaptor.dupNode(el)
2817
2818
2819    def dup(self, el):
2820        # we dup every node, so don't have to worry about calling dup; short-
2821        #circuited next() so it doesn't call.
2822        raise TypeError("dup can't be called for a node stream.")
2823
2824
2825class TreeRuleReturnScope(RuleReturnScope):
2826    """
2827    This is identical to the ParserRuleReturnScope except that
2828    the start property is a tree nodes not Token object
2829    when you are parsing trees.  To be generic the tree node types
2830    have to be Object.
2831    """
2832
2833    def __init__(self):
2834        self.start = None
2835        self.tree = None
2836
2837
2838    def getStart(self):
2839        return self.start
2840
2841
2842    def getTree(self):
2843        return self.tree
2844