treewizard.py

Go to the documentation of this file.

""" @package antlr3.tree
@brief ANTLR3 runtime package, treewizard module

A utility module to create ASTs at runtime.
See <http://www.antlr.org/wiki/display/~admin/2007/07/02/Exploring+Concept+of+TreeWizard> for an overview. Note that the API of the Python implementation is slightly different.

"""

# begin[licence]
#
# [The "BSD licence"]
# Copyright (c) 2005-2008 Terence Parr
# All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions
# are met:
# 1. Redistributions of source code must retain the above copyright
#    notice, this list of conditions and the following disclaimer.
# 2. Redistributions in binary form must reproduce the above copyright
#    notice, this list of conditions and the following disclaimer in the
#    documentation and/or other materials provided with the distribution.
# 3. The name of the author may not be used to endorse or promote products
#    derived from this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
# IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
# OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
# IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
# INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
# NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
# THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#
# end[licence]

from constants import INVALID_TOKEN_TYPE
from tokens import CommonToken
from tree import CommonTree, CommonTreeAdaptor


def computeTokenTypes(tokenNames):
    """
    Compute a dict that is an inverted index of
    tokenNames (which maps int token types to names).
    """

    if tokenNames is None:
        return {}

    return dict((name, type) for type, name in enumerate(tokenNames))


## token types for pattern parser
EOF = -1
BEGIN = 1
END = 2
ID = 3
ARG = 4
PERCENT = 5
COLON = 6
DOT = 7

class TreePatternLexer(object):
    def __init__(self, pattern):
        ## The tree pattern to lex like "(A B C)"
        self.pattern = pattern

        ## Index into input string
        self.p = -1

        ## Current char
        self.c = None

        ## How long is the pattern in char?
        self.n = len(pattern)

        ## Set when token type is ID or ARG
        self.sval = None

        self.error = False

        self.consume()


    __idStartChar = frozenset(
        'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_'
        )
    __idChar = __idStartChar | frozenset('0123456789')
    
    def nextToken(self):
        self.sval = ""
        while self.c != EOF:
            if self.c in (' ', '\n', '\r', '\t'):
                self.consume()
                continue

            if self.c in self.__idStartChar:
                self.sval += self.c
                self.consume()
                while self.c in self.__idChar:
                    self.sval += self.c
                    self.consume()

                return ID

            if self.c == '(':
                self.consume()
                return BEGIN

            if self.c == ')':
                self.consume()
                return END

            if self.c == '%':
                self.consume()
                return PERCENT

            if self.c == ':':
                self.consume()
                return COLON

            if self.c == '.':
                self.consume()
                return DOT

            if self.c == '[': # grab [x] as a string, returning x
                self.consume()
                while self.c != ']':
                    if self.c == '\\':
                        self.consume()
                        if self.c != ']':
                            self.sval += '\\'

                        self.sval += self.c

                    else:
                        self.sval += self.c

                    self.consume()

                self.consume()
                return ARG

            self.consume()
            self.error = True
            return EOF

        return EOF


    def consume(self):
        self.p += 1
        if self.p >= self.n:
            self.c = EOF

        else:
            self.c = self.pattern[self.p]


class TreePatternParser(object):
    def __init__(self, tokenizer, wizard, adaptor):
        self.tokenizer = tokenizer
        self.wizard = wizard
        self.adaptor = adaptor
        self.ttype = tokenizer.nextToken() # kickstart


    def pattern(self):
        if self.ttype == BEGIN:
            return self.parseTree()

        elif self.ttype == ID:
            node = self.parseNode()
            if self.ttype == EOF:
                return node

            return None # extra junk on end

        return None


    def parseTree(self):
        if self.ttype != BEGIN:
            return None

        self.ttype = self.tokenizer.nextToken()
        root = self.parseNode()
        if root is None:
            return None

        while self.ttype in (BEGIN, ID, PERCENT, DOT):
            if self.ttype == BEGIN:
                subtree = self.parseTree()
                self.adaptor.addChild(root, subtree)

            else:
                child = self.parseNode()
                if child is None:
                    return None

                self.adaptor.addChild(root, child)

        if self.ttype != END:
            return None

        self.ttype = self.tokenizer.nextToken()
        return root


    def parseNode(self):
        # "%label:" prefix
        label = None
        
        if self.ttype == PERCENT:
            self.ttype = self.tokenizer.nextToken()
            if self.ttype != ID:
                return None

            label = self.tokenizer.sval
            self.ttype = self.tokenizer.nextToken()
            if self.ttype != COLON:
                return None
            
            self.ttype = self.tokenizer.nextToken() # move to ID following colon

        # Wildcard?
        if self.ttype == DOT:
            self.ttype = self.tokenizer.nextToken()
            wildcardPayload = CommonToken(0, ".")
            node = WildcardTreePattern(wildcardPayload)
            if label is not None:
                node.label = label
            return node

        # "ID" or "ID[arg]"
        if self.ttype != ID:
            return None

        tokenName = self.tokenizer.sval
        self.ttype = self.tokenizer.nextToken()
        
        if tokenName == "nil":
            return self.adaptor.nil()

        text = tokenName
        # check for arg
        arg = None
        if self.ttype == ARG:
            arg = self.tokenizer.sval
            text = arg
            self.ttype = self.tokenizer.nextToken()

        # create node
        treeNodeType = self.wizard.getTokenType(tokenName)
        if treeNodeType == INVALID_TOKEN_TYPE:
            return None

        node = self.adaptor.createFromType(treeNodeType, text)
        if label is not None and isinstance(node, TreePattern):
            node.label = label

        if arg is not None and isinstance(node, TreePattern):
            node.hasTextArg = True

        return node


class TreePattern(CommonTree):
    """
    When using %label:TOKENNAME in a tree for parse(), we must
    track the label.
    """

    def __init__(self, payload):
        CommonTree.__init__(self, payload)

        self.label = None
        self.hasTextArg = None
        

    def toString(self):
        if self.label is not None:
            return '%' + self.label + ':' + CommonTree.toString(self)
        
        else:
            return CommonTree.toString(self)


class WildcardTreePattern(TreePattern):
    pass


class TreePatternTreeAdaptor(CommonTreeAdaptor):
    """This adaptor creates TreePattern objects for use during scan()"""

    def createWithPayload(self, payload):
        return TreePattern(payload)


class TreeWizard(object):
    """
    Build and navigate trees with this object.  Must know about the names
    of tokens so you have to pass in a map or array of token names (from which
    this class can build the map).  I.e., Token DECL means nothing unless the
    class can translate it to a token type.

    In order to create nodes and navigate, this class needs a TreeAdaptor.

    This class can build a token type -> node index for repeated use or for
    iterating over the various nodes with a particular type.

    This class works in conjunction with the TreeAdaptor rather than moving
    all this functionality into the adaptor.  An adaptor helps build and
    navigate trees using methods.  This class helps you do it with string
    patterns like "(A B C)".  You can create a tree from that pattern or
    match subtrees against it.
    """

    def __init__(self, adaptor=None, tokenNames=None, typeMap=None):
        self.adaptor = adaptor
        if typeMap is None:
            self.tokenNameToTypeMap = computeTokenTypes(tokenNames)

        else:
            if tokenNames is not None:
                raise ValueError("Can't have both tokenNames and typeMap")

            self.tokenNameToTypeMap = typeMap


    def getTokenType(self, tokenName):
        """Using the map of token names to token types, return the type."""

        try:
            return self.tokenNameToTypeMap[tokenName]
        except KeyError:
            return INVALID_TOKEN_TYPE


    def create(self, pattern):
        """
        Create a tree or node from the indicated tree pattern that closely
        follows ANTLR tree grammar tree element syntax:
        
        (root child1 ... child2).
        
        You can also just pass in a node: ID
         
        Any node can have a text argument: ID[foo]
        (notice there are no quotes around foo--it's clear it's a string).
        
        nil is a special name meaning "give me a nil node".  Useful for
        making lists: (nil A B C) is a list of A B C.
        """
        
        tokenizer = TreePatternLexer(pattern)
        parser = TreePatternParser(tokenizer, self, self.adaptor)
        return parser.pattern()


    def index(self, tree):
        """Walk the entire tree and make a node name to nodes mapping.
        
        For now, use recursion but later nonrecursive version may be
        more efficient.  Returns a dict int -> list where the list is
        of your AST node type.  The int is the token type of the node.
        """

        m = {}
        self._index(tree, m)
        return m


    def _index(self, t, m):
        """Do the work for index"""

        if t is None:
            return

        ttype = self.adaptor.getType(t)
        elements = m.get(ttype)
        if elements is None:
            m[ttype] = elements = []

        elements.append(t)
        for i in range(self.adaptor.getChildCount(t)):
            child = self.adaptor.getChild(t, i)
            self._index(child, m)


    def find(self, tree, what):
        """Return a list of matching token.

        what may either be an integer specifzing the token type to find or
        a string with a pattern that must be matched.
        
        """
        
        if isinstance(what, (int, long)):
            return self._findTokenType(tree, what)

        elif isinstance(what, basestring):
            return self._findPattern(tree, what)

        else:
            raise TypeError("'what' must be string or integer")


    def _findTokenType(self, t, ttype):
        """Return a List of tree nodes with token type ttype"""

        nodes = []

        def visitor(tree, parent, childIndex, labels):
            nodes.append(tree)

        self.visit(t, ttype, visitor)

        return nodes


    def _findPattern(self, t, pattern):
        """Return a List of subtrees matching pattern."""
        
        subtrees = []
        
        # Create a TreePattern from the pattern
        tokenizer = TreePatternLexer(pattern)
        parser = TreePatternParser(tokenizer, self, TreePatternTreeAdaptor())
        tpattern = parser.pattern()
        
        # don't allow invalid patterns
        if (tpattern is None or tpattern.isNil()
            or isinstance(tpattern, WildcardTreePattern)):
            return None

        rootTokenType = tpattern.getType()

        def visitor(tree, parent, childIndex, label):
            if self._parse(tree, tpattern, None):
                subtrees.append(tree)
                
        self.visit(t, rootTokenType, visitor)

        return subtrees


    def visit(self, tree, what, visitor):
        """Visit every node in tree matching what, invoking the visitor.

        If what is a string, it is parsed as a pattern and only matching
        subtrees will be visited.
        The implementation uses the root node of the pattern in combination
        with visit(t, ttype, visitor) so nil-rooted patterns are not allowed.
        Patterns with wildcard roots are also not allowed.

        If what is an integer, it is used as a token type and visit will match
        all nodes of that type (this is faster than the pattern match).
        The labels arg of the visitor action method is never set (it's None)
        since using a token type rather than a pattern doesn't let us set a
        label.
        """

        if isinstance(what, (int, long)):
            self._visitType(tree, None, 0, what, visitor)

        elif isinstance(what, basestring):
            self._visitPattern(tree, what, visitor)

        else:
            raise TypeError("'what' must be string or integer")
        
              
    def _visitType(self, t, parent, childIndex, ttype, visitor):
        """Do the recursive work for visit"""
        
        if t is None:
            return

        if self.adaptor.getType(t) == ttype:
            visitor(t, parent, childIndex, None)

        for i in range(self.adaptor.getChildCount(t)):
            child = self.adaptor.getChild(t, i)
            self._visitType(child, t, i, ttype, visitor)


    def _visitPattern(self, tree, pattern, visitor):
        """
        For all subtrees that match the pattern, execute the visit action.
        """

        # Create a TreePattern from the pattern
        tokenizer = TreePatternLexer(pattern)
        parser = TreePatternParser(tokenizer, self, TreePatternTreeAdaptor())
        tpattern = parser.pattern()
        
        # don't allow invalid patterns
        if (tpattern is None or tpattern.isNil()
            or isinstance(tpattern, WildcardTreePattern)):
            return

        rootTokenType = tpattern.getType()

        def rootvisitor(tree, parent, childIndex, labels):
            labels = {}
            if self._parse(tree, tpattern, labels):
                visitor(tree, parent, childIndex, labels)
                
        self.visit(tree, rootTokenType, rootvisitor)
        

    def parse(self, t, pattern, labels=None):
        """
        Given a pattern like (ASSIGN %lhs:ID %rhs:.) with optional labels
        on the various nodes and '.' (dot) as the node/subtree wildcard,
        return true if the pattern matches and fill the labels Map with
        the labels pointing at the appropriate nodes.  Return false if
        the pattern is malformed or the tree does not match.

        If a node specifies a text arg in pattern, then that must match
        for that node in t.
        """

        tokenizer = TreePatternLexer(pattern)
        parser = TreePatternParser(tokenizer, self, TreePatternTreeAdaptor())
        tpattern = parser.pattern()

        return self._parse(t, tpattern, labels)


    def _parse(self, t1, tpattern, labels):
        """
        Do the work for parse. Check to see if the tpattern fits the
        structure and token types in t1.  Check text if the pattern has
        text arguments on nodes.  Fill labels map with pointers to nodes
        in tree matched against nodes in pattern with labels.
        """
        
        # make sure both are non-null
        if t1 is None or tpattern is None:
            return False

        # check roots (wildcard matches anything)
        if not isinstance(tpattern, WildcardTreePattern):
            if self.adaptor.getType(t1) != tpattern.getType():
                return False

            # if pattern has text, check node text
            if (tpattern.hasTextArg
                and self.adaptor.getText(t1) != tpattern.getText()):
                return False

        if tpattern.label is not None and labels is not None:
            # map label in pattern to node in t1
            labels[tpattern.label] = t1

        # check children
        n1 = self.adaptor.getChildCount(t1)
        n2 = tpattern.getChildCount()
        if n1 != n2:
            return False

        for i in range(n1):
            child1 = self.adaptor.getChild(t1, i)
            child2 = tpattern.getChild(i)
            if not self._parse(child1, child2, labels):
                return False

        return True


    def equals(self, t1, t2, adaptor=None):
        """
        Compare t1 and t2; return true if token types/text, structure match
        exactly.
        The trees are examined in their entirety so that (A B) does not match
        (A B C) nor (A (B C)). 
        """

        if adaptor is None:
            adaptor = self.adaptor

        return self._equals(t1, t2, adaptor)


    def _equals(self, t1, t2, adaptor):
        # make sure both are non-null
        if t1 is None or t2 is None:
            return False

        # check roots
        if adaptor.getType(t1) != adaptor.getType(t2):
            return False

        if adaptor.getText(t1) != adaptor.getText(t2):
            return False
        
        # check children
        n1 = adaptor.getChildCount(t1)
        n2 = adaptor.getChildCount(t2)
        if n1 != n2:
            return False

        for i in range(n1):
            child1 = adaptor.getChild(t1, i)
            child2 = adaptor.getChild(t2, i)
            if not self._equals(child1, child2, adaptor):
                return False

        return True