Notice to GitKraken users: A vulnerability has been found in the SSH key generation of GitKraken versions 7.6.0 to 8.0.0 (https://www.gitkraken.com/blog/weak-ssh-key-fix). If you use GitKraken and have generated a SSH key using one of these versions, please remove it both from your local workstation and from your LRZ GitLab profile.

21.10.2021, 9:00 - 11:00: Due to updates GitLab may be unavailable for some minutes between 09:00 and 11:00.

syntaxtree.py 28.3 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
# syntaxtree.py - syntax tree classes for DHParser
#
# Copyright 2016  by Eckhart Arnold (arnold@badw.de)
#                 Bavarian Academy of Sciences an Humanities (badw.de)
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
# implied.  See the License for the specific language governing
# permissions and limitations under the License.
17
18
19


"""
20
21
Module ``syntaxtree`` defines the ``Node``-class for syntax trees as well
as an abstract base class for parser-objects. The latter is defined
eckhart's avatar
eckhart committed
22
here, because node-objects refer to parser-objects. All concrete
23
24
25
parser classes are defined in the ``parse`` module.
"""

26

Eckhart Arnold's avatar
Eckhart Arnold committed
27
import collections.abc
28
import copy
29

30
from DHParser.error import Error, linebreaks, line_col
31
from DHParser.stringview import StringView
eckhart's avatar
eckhart committed
32
from DHParser.toolkit import re, typing
33
from typing import Callable, cast, Iterator, List, Set, Union, Tuple, Optional
di68kap's avatar
di68kap committed
34

35

36
37
38
39
40
41
42
__all__ = ('ParserBase',
           'WHITESPACE_PTYPE',
           'TOKEN_PTYPE',
           'MockParser',
           'ZombieParser',
           'ZOMBIE_PARSER',
           'Node',
43
           'mock_syntax_tree',
44
           'flatten_sxpr')
Eckhart Arnold's avatar
Eckhart Arnold committed
45
46


47
48
49
50
51
52
53
54
55
56
57
58
59
60
#######################################################################
#
# parser base and mock parsers
#
#######################################################################


class ParserBase:
    """
    ParserBase is the base class for all real and mock parser classes.
    It is defined here, because Node objects require a parser object
    for instantiation.
    """
    def __init__(self, name=''):  # , pbases=frozenset()):
61
        self._name = name  # type: str
62
63
64
        self._ptype = ':' + self.__class__.__name__  # type: str

    def __repr__(self):
di68kap's avatar
di68kap committed
65
         return self.name + self.ptype
66
67
68
69

    def __str__(self):
        return self.name + (' = ' if self.name else '') + repr(self)

70
71
72
    def __call__(self, text: StringView) -> Tuple[Optional['Node'], StringView]:
        return None, text

73
74
    @property
    def name(self):
75
76
        """Returns the name of the parser or the empty string '' for unnamed
        parsers."""
77
78
        return self._name

79
80
    @property
    def ptype(self) -> str:
81
82
        """Returns the type of the parser. By default this is the parser's
        class name preceded by a colon, e.g. ':ZeroOrMore'."""
83
84
85
86
        return self._ptype

    @property
    def repr(self) -> str:
di68kap's avatar
di68kap committed
87
        """Returns the parser's name if it has a name and repr()"""
88
89
        return self.name if self.name else repr(self)

90
    def reset(self):
91
        """Resets any parser variables. (Should be overridden.)"""
92
93
        pass

eckhart's avatar
eckhart committed
94
    def grammar(self) -> Optional[object]:
95
96
        """Returns the Grammar object to which the parser belongs. If not
        yet connected to any Grammar object, None is returned."""
97
98
99
        return None

    def apply(self, func: Callable) -> bool:
eckhart's avatar
eckhart committed
100
        """Applies the function `func` to the parser. Returns False, if
101
        - for whatever reason - the functions has not been applied, True
eckhart's avatar
eckhart committed
102
        otherwise."""
103
104
        return False

105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121

WHITESPACE_PTYPE = ':Whitespace'
TOKEN_PTYPE = ':Token'


class MockParser(ParserBase):
    """
    MockParser objects can be used to reconstruct syntax trees from a
    serialized form like S-expressions or XML. Mock objects can mimic
    different parser types by assigning them a ptype on initialization.

    Mock objects should not be used for anything other than
    syntax tree (re-)construction. In all other cases where a parser
    object substitute is needed, chose the singleton ZOMBIE_PARSER.
    """
    def __init__(self, name='', ptype=''):  # , pbases=frozenset()):
        assert not ptype or ptype[0] == ':'
122
        super().__init__(name)
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
        self._ptype = ptype or ':' + self.__class__.__name__


class ZombieParser(MockParser):
    """
    Serves as a substitute for a Parser instance.

    ``ZombieParser`` is the class of the singelton object
    ``ZOMBIE_PARSER``. The  ``ZOMBIE_PARSER`` has a name and can be
    called, but it never matches. It serves as a substitute where only
    these (or one of these properties) is needed, but no real Parser-
    object is instantiated.
    """
    alive = False

    def __init__(self):
        super(ZombieParser, self).__init__("__ZOMBIE__")
        assert not self.__class__.alive, "There can be only one!"
        assert self.__class__ == ZombieParser, "No derivatives, please!"
        self.__class__.alive = True

    def __copy__(self):
        return self

    def __deepcopy__(self, memo):
        return self

    def __call__(self, text):
        """Better call Saul ;-)"""
        return None, text


ZOMBIE_PARSER = ZombieParser()


#######################################################################
#
# syntaxtree nodes
#
#######################################################################


165
ChildrenType = Tuple['Node', ...]
Eckhart Arnold's avatar
Eckhart Arnold committed
166
NoChildren = cast(ChildrenType, ())  # type: ChildrenType
eckhart's avatar
eckhart committed
167
StrictResultType = Union[ChildrenType, StringView, str]
168
ResultType = Union[ChildrenType, 'Node', StringView, str, None]
169
170


Eckhart Arnold's avatar
Eckhart Arnold committed
171
def flatten_sxpr(sxpr: str) -> str:
172
    """Returns S-expression ``sxpr`` as a one-liner without unnecessary
Eckhart Arnold's avatar
Eckhart Arnold committed
173
174
175
    whitespace.

    Example:
Eckhart Arnold's avatar
Eckhart Arnold committed
176
    >>> flatten_sxpr('(a\\n    (b\\n        c\\n    )\\n)\\n')
Eckhart Arnold's avatar
Eckhart Arnold committed
177
178
    '(a (b c))'
    """
179
    return re.sub(r'\s(?=\))', '', re.sub(r'\s+', ' ', sxpr)).strip()
Eckhart Arnold's avatar
Eckhart Arnold committed
180
181


Eckhart Arnold's avatar
Eckhart Arnold committed
182
class Node(collections.abc.Sized):
183
184
185
    """
    Represents a node in the concrete or abstract syntax tree.

186
    Attributes:
187
188
        tag_name (str):  The name of the node, which is either its
            parser's name or, if that is empty, the parser's class name
189

190
191
192
        result (str or tuple):  The result of the parser which
            generated this node, which can be either a string or a
            tuple of child nodes.
193

194
195
        children (tuple):  The tuple of child nodes or an empty tuple
            if there are no child nodes. READ ONLY!
196

eckhart's avatar
eckhart committed
197
198
199
200
        content (str):  Yields the contents of the tree as string. The
            difference to ``str(node)`` is that ``node.content`` does
            not add the error messages to the returned string.

Eckhart Arnold's avatar
Eckhart Arnold committed
201
        parser (Parser):  The parser which generated this node.
202
203
            WARNING: In case you use mock syntax trees for testing or
            parser replacement during the AST-transformation: DO NOT
Eckhart Arnold's avatar
Eckhart Arnold committed
204
205
            rely on this being a real parser object in any phase after
            parsing (i.e. AST-transformation and compiling), for
206
            example by calling ``isinstance(node.parer, ...)``.
207

208
209
        errors (list):  A list of parser- or compiler-errors:
            tuple(position, string) attached to this node
210

211
212
213
        error_flag (int):  0 if no error occurred in either the node
            itself or any of its descendants. Otherwise contains the
            highest warning or error level or all errors that occurred.
214

215
216
217
        len (int):  The full length of the node's string result if the
            node is a leaf node or, otherwise, the concatenated string
            result's of its descendants. The figure always represents
218
            the length before AST-transformation and will never change
219
            through AST-transformation. READ ONLY!
220

221
222
        pos (int):  the position of the node within the parsed text.

Eckhart Arnold's avatar
Eckhart Arnold committed
223
            The value of ``pos`` is -1 meaning invalid by default.
224
            Setting this value will set the positions of all child
Eckhart Arnold's avatar
Eckhart Arnold committed
225
            nodes relative to this value.
226
227

            To set the pos values of all nodes in a syntax tree, the
Eckhart Arnold's avatar
Eckhart Arnold committed
228
            pos value of the root node should be set to 0 right
229
230
            after parsing.

Eckhart Arnold's avatar
Eckhart Arnold committed
231
            Other than that, this value should be considered READ ONLY.
232
233
            At any rate, it should only be reassigned during the parsing
            stage and never during or after the AST-transformation.
234
235
236
237
238

        xml_attr (dict): An optional dictionary of XML-attributes. This
            dictionary is created lazily upon first usage. The attributes
            will only be shown in the XML-Representation, not in the
            S-Expression-output.
eckhart's avatar
eckhart committed
239
240
241


        _parent (Node): SLOT RESERVED FOR FUTURE USE!
242
    """
243

di68kap's avatar
di68kap committed
244
    __slots__ = ['_result', 'children', '_errors', '_len', '_pos', 'parser', 'error_flag',
245
                 '_xml_attr', '_content', '_parent']
246

247
248
249
    def __init__(self, parser, result: ResultType, leafhint: bool = False) -> None:
        """
        Initializes the ``Node``-object with the ``Parser``-Instance
250
251
        that generated the node and the parser's result.
        """
252
253
        self.error_flag = 0             # type: int
        self._errors = []               # type: List[Error]
eckhart's avatar
eckhart committed
254
        self._pos = -1                  # type: int
di68kap's avatar
di68kap committed
255
256
        # Assignment to self.result initializes the attributes _result, children and _len
        # The following if-clause is merely an optimization, i.e. a fast-path for leaf-Nodes
257
        if leafhint:
eckhart's avatar
eckhart committed
258
            self._result = result       # type: StrictResultType
259
            self._content = None        # type: str
260
            self.children = NoChildren  # type: ChildrenType
261
            self._len = -1              # type: int  # lazy evaluation
262
263
        else:
            self.result = result
264
265
        self.parser = parser or ZOMBIE_PARSER

Eckhart Arnold's avatar
Eckhart Arnold committed
266

267
    def __str__(self):
268
        s = "".join(str(child) for child in self.children) if self.children else self.content
269
270
271
272
        if self._errors:
            return ' <<< Error on "%s" | %s >>> ' % \
                   (s, '; '.join(e.message for e in self._errors))
        return s
273

Eckhart Arnold's avatar
Eckhart Arnold committed
274

275
276
277
278
279
280
    def __repr__(self):
        mpargs = {'name': self.parser.name, 'ptype': self.parser.ptype}
        parg = "MockParser({name}, {ptype})".format(**mpargs)
        rarg = str(self) if not self.children else \
               "(" + ", ".join(repr(child) for child in self.children) + ")"
        return "Node(%s, %s)" % (parg, rarg)
281

Eckhart Arnold's avatar
Eckhart Arnold committed
282

283
    def __len__(self):
284
        if self._len < 0:
di68kap's avatar
di68kap committed
285
            self._len = sum(len(child) for child in self.children) \
286
                if self.children else len(self._result)
287
288
289
290
291
292
293
294
        return self._len


    def __bool__(self):
        # A node that is not None is always True, even if it's empty
        return True


295
    def __eq__(self, other):
296
297
298
299
        """
        Equality of nodes: Two nodes are considered as equal, if their tag
        name is the same and if their results are equal.
        """
300
        return self.tag_name == other.tag_name and self.result == other.result
301

Eckhart Arnold's avatar
Eckhart Arnold committed
302

303
    def __hash__(self):
304
        return hash(self.tag_name)
305

Eckhart Arnold's avatar
Eckhart Arnold committed
306

307
308
309
310
311
312
    def __deepcopy__(self, memodict={}):
        result = copy.deepcopy(self.result)
        other = Node(self.parser, result)
        other._pos = self._pos
        return other

Eckhart Arnold's avatar
Eckhart Arnold committed
313

314
315
316
    def __getitem__(self, index_or_tagname: Union[int, str]) -> Union['Node', Iterator['Node']]:
        """
        Returns the child node with the given index if ``index_or_tagname`` is
317
        an integer or the first child node with the given tag name. Examples::
318

319
            >>> tree = mock_syntax_tree('(a (b "X") (X (c "d")) (e (X "F")))')
320
321
            >>> flatten_sxpr(tree[0].as_sxpr())
            '(b "X")'
322
323
            >>> flatten_sxpr(tree["X"].as_sxpr())
            '(X (c "d"))'
324
325
326
327

        Args:
            index_or_tagname(str): Either an index of a child node or a
                tag name.
328
        Returns:
329
330
            Node: All nodes which have a given tag name.
        """
331
332
333
        if self.children:
            if isinstance(index_or_tagname, int):
                return self.children[index_or_tagname]
334
            else:
335
336
337
338
339
                for child in self.children:
                    if child.tag_name == index_or_tagname:
                        return child
                raise KeyError(index_or_tagname)
        raise ValueError('Leave nodes have no children that can be indexed!')
340
341
342
343


    def __contains__(self, tag_name: str) -> bool:
        """
344
        Returns true if a child with the given tag name exists.
345
        Args:
346
347
            tag_name (str): tag_name which will be searched among to immediate
                descendants of this node.
348
349
350
351
        Returns:
            bool:  True, if at least one descendant node with the given tag
                name exists, False otherwise
        """
352
353
354
355
356
        # assert isinstance(tag_name, str)
        if self.children:
            for child in self.children:
                if child.tag_name == tag_name:
                    return True
357
            return False
358
359
360
361
362
363
364
        raise ValueError('Leave node cannot contain other nodes')
        # generator = self.select_tags(tag_name, False)
        # try:
        #     generator.__next__()
        #     return True
        # except StopIteration:
        #     return False
365
366


367
    @property   # this needs to be a (dynamic) property, in case sef.parser gets updated
368
    def tag_name(self) -> str:
369
370
371
372
373
374
        """
        Returns the tage name of Node, i.e. the name for XML or
        S-expression representation. By default the tag name is the
        name of the node's parser or, if the node's parser is unnamed, the
        node's parser's `ptype`.
        """
375
        return self.parser.name or self.parser.ptype
376

Eckhart Arnold's avatar
Eckhart Arnold committed
377

378
    @property
Eckhart Arnold's avatar
Eckhart Arnold committed
379
    def result(self) -> StrictResultType:
380
381
382
383
384
        """
        Returns the result from the parser that created the node.
        Error messages are not included in the result. Use `self.content()`
        if the result plus any error messages is needed.
        """
385
386
        return self._result

387

388
    @result.setter
Eckhart Arnold's avatar
Eckhart Arnold committed
389
    def result(self, result: ResultType):
Eckhart Arnold's avatar
Eckhart Arnold committed
390
        # # made obsolete by static type checking with mypy
391
392
393
        # assert ((isinstance(result, tuple) and all(isinstance(child, Node) for child in result))
        #         or isinstance(result, Node)
        #         or isinstance(result, str)), str(result)
Eckhart Arnold's avatar
Eckhart Arnold committed
394
395
        # Possible optimization: Do not allow single nodes as argument:
        # assert not isinstance(result, Node)
396
397
        self._len = -1        # lazy evaluation
        self._content = None
398
399
400
401
402
403
404
405
        if isinstance(result, Node):
            self.children = (result,)
            self._result = self.children
            self.error_flag = result.error_flag
        else:
            if isinstance(result, tuple):
                self.children = result
                self._result = result or ''
406
407
408
                if result:
                    if self.error_flag == 0:
                        self.error_flag = max(child.error_flag for child in self.children)
409
410
            else:
                self.children = NoChildren
411
                self._result = result
412

413
414

    @property
415
    def content(self) -> Union[StringView, str]:
416
        """
417
        Returns content as string, omitting error messages.
418
        """
419
420
421
422
        if self._content is None:
            if self.children:
                self._content = "".join(child.content for child in self.children)
            else:
423
                # self._content = self._result
424
425
426
                self._content = str(self._result)
                self._result = self._content  # self._result might be more efficient as a string!?
        return self._content
427
428
429
430
431
432
433
434
435
436
437


    @property
    def structure(self) -> str:
        """
        Return structure (and content) as S-expression on a single line
        without any line breaks.
        """
        return flatten_sxpr(self.as_sxpr(showerrors=False))


438
    @property
439
    def pos(self) -> int:
eckhart's avatar
eckhart committed
440
441
        """Returns the position of the Node's content in the source text."""
        if self._pos < 0:
442
            raise AssertionError("Position value not initialized!")
443
444
        return self._pos

eckhart's avatar
eckhart committed
445

eckhart's avatar
eckhart committed
446
    def init_pos(self, pos: int, overwrite: bool = False) -> 'Node':
eckhart's avatar
eckhart committed
447
448
449
450
451
452
453
454
455
        """
        (Re-)initialize position value. Usually, the parser guard
        (`parsers.add_parser_guard()`) takes care of assigning the
        position in the document to newly created nodes. However,
        where Nodes are created outside the reach of the parser
        guard, their document-position must be assigned manually.
        This function recursively reassigns the position values
        of the child nodes, too.
        """
eckhart's avatar
eckhart committed
456
457
458
459
460
461
        if overwrite or self._pos < 0:
            self._pos = pos
            for err in self._errors:
                err.pos = pos
        else:
            assert self._pos == pos, str("%i != %i" % (self._pos, pos))
462
        # recursively adjust pos-values of all children
eckhart's avatar
eckhart committed
463
        offset = self.pos
464
        for child in self.children:
eckhart's avatar
eckhart committed
465
            child.init_pos(offset, overwrite)
eckhart's avatar
eckhart committed
466
467
468
            offset = child.pos + len(child)
        return self

469
470

    @property
471
    def errors(self) -> List[Error]:
eckhart's avatar
eckhart committed
472
        """
473
        Returns the errors that occurred at this Node,
eckhart's avatar
eckhart committed
474
475
        not including any errors from child nodes.
        """
476
477
        return self._errors.copy()

478

eckhart's avatar
eckhart committed
479
480
    def add_error(self,
                  message: str,
481
                  code: int = Error.ERROR) -> 'Node':
eckhart's avatar
eckhart committed
482
483
484
        """
        Adds an error to this Node.
        Parameters:
eckhart's avatar
eckhart committed
485
486
            message(str): A string with the error message.abs
            code(int):    An error code to identify the kind of error
eckhart's avatar
eckhart committed
487
        """
488
489
        self._errors.append(Error(message, code))
        self.error_flag = max(self.error_flag, self._errors[-1].code)
Eckhart Arnold's avatar
Eckhart Arnold committed
490
491
        return self

492

493
    def collect_errors(self, clear_errors=False) -> List[Error]:
Eckhart Arnold's avatar
Eckhart Arnold committed
494
        """
495
        Recursively adds line- and column-numbers to all error objects.
Eckhart Arnold's avatar
Eckhart Arnold committed
496
497
498
499
        Returns all errors of this node or any child node in the form
        of a set of tuples (position, error_message), where position
        is always relative to this node.
        """
500
        errors = self.errors
501
502
        for err in errors:
            err.pos = self.pos
503
504
        if self.children:
            for child in self.children:
505
                errors.extend(child.collect_errors(clear_errors))
506
507
508
        if clear_errors:
            self._errors = []
            self.error_flag = 0
509
        else:
510
511
512
513
514
515
516
            if self._errors:
                self.error_flag = max(err.code for err in self.errors)
            if self.children:
                max_child_error = max(child.error_flag for child in self.children)
                self.error_flag = max(self.error_flag, max_child_error)
        return errors

517

Eckhart Arnold's avatar
Eckhart Arnold committed
518

519
    def _tree_repr(self, tab, open_fn, close_fn, data_fn=lambda i: i, density=0) -> str:
520
521
522
523
524
525
526
527
        """
        Generates a tree representation of this node and its children
        in string from.

        The kind ot tree-representation that is determined by several
        function parameters. This could be an XML-representation or a
        lisp-like S-expression.

528
        Args:
529
            tab (str):  The indentation string, e.g. '\t' or '    '
eckhart's avatar
eckhart committed
530
            open_fn:   (Node->str) A function that returns an opening
531
                string (e.g. an XML-tag_name) for a given node
eckhart's avatar
eckhart committed
532
            close_fn:  (Node->str) A function that returns a closeF
533
                string (e.g. an XML-tag_name) for a given node.
eckhart's avatar
eckhart committed
534
            data_fn:   (str->str) A function that filters the data string
535
536
537
538
539
540
                before printing, e.g. to add quotation marks

        Returns (str):
            A string that contains a (serialized) tree representation
            of the node and its children.
        """
eckhart's avatar
eckhart committed
541
542
        head = open_fn(self)
        tail = close_fn(self)
543
544

        if not self.result:
545
            return head.rstrip() + tail.lstrip()
546

eckhart's avatar
eckhart committed
547
        tail = tail.lstrip(None if density & 2 else '')
548
549
550

        if self.children:
            content = []
551
            for child in self.children:
eckhart's avatar
eckhart committed
552
                subtree = child._tree_repr(tab, open_fn, close_fn, data_fn, density).split('\n')
553
                content.append('\n'.join((tab + s) for s in subtree))
eckhart's avatar
eckhart committed
554
            return head + '\n'.join(content) + tail
555

556
        res = cast(str, self.result)  # safe, because if there are no children, result is a string
557
558
559
        if density & 1 and res.find('\n') < 0:  # and head[0] == "<":
            # except for XML, add a gap between opening statement and content
            gap = ' ' if head.rstrip()[-1] != '>' else ''
eckhart's avatar
eckhart committed
560
            return head.rstrip() + gap + data_fn(self.result) + tail.lstrip()
561
        else:
eckhart's avatar
eckhart committed
562
            return head + '\n'.join([tab + data_fn(s) for s in res.split('\n')]) + tail
563

Eckhart Arnold's avatar
Eckhart Arnold committed
564

565
    def as_sxpr(self, src: str = None, compact: bool = False, showerrors: bool = True) -> str:
566
567
568
        """
        Returns content as S-expression, i.e. in lisp-like form.

569
        Args:
570
571
572
            src:  The source text or `None`. In case the source text is
                given the position of the element in the text will be
                reported as line and column.
Eckhart Arnold's avatar
Eckhart Arnold committed
573
            compact:  If True a compact representation is returned where
574
                brackets are omitted and only the indentation indicates the
Eckhart Arnold's avatar
Eckhart Arnold committed
575
                tree structure.
576
577
        """

eckhart's avatar
eckhart committed
578
        left_bracket, right_bracket, density = ('', '', 1) if compact else ('(', '\n)', 0)
579

580
        def opening(node) -> str:
eckhart's avatar
eckhart committed
581
582
            """Returns the opening string for the representation of `node`."""
            txt = left_bracket + node.tag_name
583
584
            # s += " '(pos %i)" % node.pos
            if src:
eckhart's avatar
eckhart committed
585
                txt += " '(pos %i " % node.pos  # + " %i %i)" % line_col(src, node.pos)
di68kap's avatar
di68kap committed
586
587
            # if node.error_flag:   # just for debugging error collecting
            #     txt += " HAS ERRORS"
588
            if showerrors and node.errors:
eckhart's avatar
eckhart committed
589
590
591
                txt += " '(err '(%s))" % ' '.join(str(err).replace('"', r'\"')
                                                  for err in node.errors)
            return txt + '\n'
592

eckhart's avatar
eckhart committed
593
594
595
        def closing(node) -> str:
            """Returns the closing string for the representation of `node`."""
            return right_bracket
596

eckhart's avatar
eckhart committed
597
598
599
600
601
        def pretty(strg):
            """Encloses `strg` with the right kind of quotation marks."""
            return '"%s"' % strg if strg.find('"') < 0 \
                else "'%s'" % strg if strg.find("'") < 0 \
                else '"%s"' % strg.replace('"', r'\"')
602

eckhart's avatar
eckhart committed
603
        return self._tree_repr('    ', opening, closing, pretty, density=density)
Eckhart Arnold's avatar
Eckhart Arnold committed
604

eckhart's avatar
eckhart committed
605

eckhart's avatar
eckhart committed
606
607
608
609
610
611
612
613
    @property
    def xml_attr(self):
        """Returns a dictionary of XML-Attributes attached to the Node."""
        if not hasattr(self, '_xml_attr'):
            self._xml_attr = dict()
        return self._xml_attr


614
    def as_xml(self, src: str = None, showerrors: bool = True) -> str:
615
616
617
        """
        Returns content as XML-tree.

618
        Args:
619
620
621
622
623
            src:  The source text or `None`. In case the source text is
                given the position will also be reported as line and
                column.
        """

624
        def opening(node) -> str:
eckhart's avatar
eckhart committed
625
            """Returns the opening string for the representation of `node`."""            
eckhart's avatar
eckhart committed
626
            txt = ['<', node.tag_name]
627
            # s += ' pos="%i"' % node.pos
di68kap's avatar
di68kap committed
628
629
            if hasattr(node, '_xml_attr'):
                txt.extend(' %s="%s"' % (k, v) for k, v in node.xml_attr.items())
630
            if src:
eckhart's avatar
eckhart committed
631
                txt.append(' line="%i" col="%i"' % line_col(line_breaks, node.pos))
632
            if showerrors and node.errors:
eckhart's avatar
eckhart committed
633
634
635
                txt.append(' err="%s"' % ''.join(str(err).replace('"', r'\"')
                                                 for err in node.errors))
            return "".join(txt + [">\n"])
636
637

        def closing(node):
eckhart's avatar
eckhart committed
638
            """Returns the closing string for the representation of `node`."""            
639
            return '\n</' + node.tag_name + '>'
640

641
        line_breaks = linebreaks(src) if src else []
642
        return self._tree_repr('    ', opening, closing, density=1)
643

Eckhart Arnold's avatar
Eckhart Arnold committed
644

645
    def select(self, match_function: Callable, include_root: bool=True) -> Iterator['Node']:
eckhart's avatar
eckhart committed
646
        """
647
        Finds nodes in the tree that fulfill a given criterion.
eckhart's avatar
eckhart committed
648

649
        `select` is a generator that yields all nodes for which the
di68kap's avatar
di68kap committed
650
        given `match_function` evaluates to True. The tree is
651
        traversed pre-order, depth last.
eckhart's avatar
eckhart committed
652

653
654
655
        Args:
            match_function (function): A function  that takes as Node
                object as argument and returns True or False
656
657
            include_root (bool): If False, only descendant nodes will be
                checked for a match.
658
        Yields:
659
            Node: All nodes of the tree for which
660
661
            ``match_function(node)`` returns True
        """
662
        if include_root and match_function(self):
663
664
665
            yield self
        else:
            for child in self.children:
666
                for node in child.select(match_function, True):
eckhart's avatar
eckhart committed
667
                    yield node
668

Eckhart Arnold's avatar
Eckhart Arnold committed
669

670
671
    def select_tags(self, tag_names: Union[str, Set[str]],
                    include_root: bool=True) -> Iterator['Node']:
672
        """
673
674
675
676
677
678
679
680
681
682
683
684
685
        Returns an iterator that runs through all descendants that have the
        given tag name.
        Example::

            >>> tree = mock_syntax_tree('(a (b "X") (X (c "d")) (e (X "F")))')
            >>> list(flatten_sxpr(item.as_sxpr()) for item in tree.select_tags("X", False))
            ['(X (c "d"))', '(X "F")']
            >>> list(flatten_sxpr(item.as_sxpr()) for item in tree.select_tags({"X", "b"}, False))
            ['(b "X")', '(X (c "d"))', '(X "F")']
            >>> any(tree.select_tags('a', False))
            False
            >>> list(flatten_sxpr(item.as_sxpr()) for item in tree.select_tags('a', True))
            ['(a (b "X") (X (c "d")) (e (X "F")))']
686
687

        Args:
688
689
690
691
            tag_name(set): A tag name or set of tag names that is being
                searched for
            include_root (bool): If False, only descendant nodes will be
                checked for a match.
692
693
694
        Yields:
            Node: All nodes which have a given tag name.
        """
695
696
697
        if isinstance(tag_names, str):
            tag_names = frozenset(tag_names)
        return self.select(lambda node: node.tag_name in tag_names, include_root)
698
699


700
    def tree_size(self) -> int:
eckhart's avatar
eckhart committed
701
702
703
        """
        Recursively counts the number of nodes in the tree including the root node.
        """
704
705
706
        return sum(child.tree_size() for child in self.children) + 1


707
def mock_syntax_tree(sxpr):
708
    """
di68kap's avatar
di68kap committed
709
710
    Generates a tree of nodes from an S-expression. The main purpose of this is
    to generate test data.
711
712
713
714
715
716
717

    Example:
    >>> mock_syntax_tree("(a (b c))").as_sxpr()
    '(a\\n    (b\\n        "c"\\n    )\\n)'
    """

    def next_block(s):
eckhart's avatar
eckhart committed
718
719
720
        """Generator that yields all characters until the next closing bracket
        that does not match an opening bracket matched earlier within the same
        package."""
721
        s = s.strip()
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
        try:
            while s[0] != ')':
                if s[0] != '(':
                    raise ValueError('"(" expected, not ' + s[:10])
                # assert s[0] == '(', s
                level = 1
                k = 1
                while level > 0:
                    if s[k] == '(':
                        level += 1
                    elif s[k] == ')':
                        level -= 1
                    k += 1
                yield s[:k]
                s = s[k:].strip()
        except IndexError:
            errmsg = ('Malformed S-expression. Unprocessed part: "%s"' % s) if s \
                else 'Malformed S-expression. Closing bracket(s) ")" missing.'
            raise AssertionError(errmsg)
741

742
    sxpr = sxpr.strip()
eckhart's avatar
eckhart committed
743
744
    if sxpr[0] != '(':
        raise ValueError('"(" expected, not ' + sxpr[:10])
745
746
    # assert sxpr[0] == '(', sxpr
    sxpr = sxpr[1:].strip()
eckhart's avatar
eckhart committed
747
    match = re.match(r'[\w:]+', sxpr)
748
749
750
    if match is None:
        raise AssertionError('Malformed S-expression Node-tagname or identifier expected, '
                             'not "%s"' % sxpr[:40].replace('\n', ''))
eckhart's avatar
eckhart committed
751
752
    name, class_name = (sxpr[:match.end()].split(':') + [''])[:2]
    sxpr = sxpr[match.end():].strip()
753
754
    if sxpr[0] == '(':
        result = tuple(mock_syntax_tree(block) for block in next_block(sxpr))
755
756
757
758
        pos = 0
        for node in result:
            node._pos = pos
            pos += len(node)
759
760
    else:
        lines = []
761
        while sxpr and sxpr[0] != ')':
eckhart's avatar
eckhart committed
762
763
764
765
766
767
            for qtmark in ['"""', "'''", '"', "'"]:
                match = re.match(qtmark + r'.*?' + qtmark, sxpr, re.DOTALL)
                if match:
                    i = len(qtmark)
                    lines.append(sxpr[i:match.end() - i])
                    sxpr = sxpr[match.end():].strip()
768
769
                    break
            else:
eckhart's avatar
eckhart committed
770
771
772
                match = re.match(r'(?:(?!\)).)*', sxpr, re.DOTALL)
                lines.append(sxpr[:match.end()])
                sxpr = sxpr[match.end():]
773
        result = "\n".join(lines)
774
775
776
    node = Node(MockParser(name, ':' + class_name), result)
    node._pos = 0
    return node
777

778
779
780
781
# if __name__ == "__main__":
#     st = mock_syntax_tree("(alpha (beta (gamma i\nj\nk) (delta y)) (epsilon z))")
#     print(st.as_sxpr())
#     print(st.as_xml())