test_parse.py 39.4 KB
Newer Older
1
2
#!/usr/bin/python3

3
"""test_parse.py - tests of the parsers-module of DHParser
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

Author: Eckhart Arnold <arnold@badw.de>

Copyright 2017 Bavarian Academy of Sciences and Humanities

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.
"""

22
import os
23
import sys
24
from functools import partial
25

26
27
scriptpath = os.path.dirname(__file__) or '.'
sys.path.append(os.path.abspath(os.path.join(scriptpath, '..')))
28

eckhart's avatar
eckhart committed
29
from DHParser.configuration import get_config_value, set_config_value
30
from DHParser.toolkit import compile_python_object, re
31
from DHParser.log import is_logging, log_ST, log_parsing_history
32
from DHParser.error import Error, is_error
33
from DHParser.parse import ParserError, Parser, Grammar, Forward, TKN, ZeroOrMore, RE, \
eckhart's avatar
eckhart committed
34
    RegExp, Lookbehind, NegativeLookahead, OneOrMore, Series, Alternative, AllOf, SomeOf, \
35
    UnknownParserError, MetaParser, EMPTY_NODE
36
from DHParser import compile_source
37
from DHParser.ebnf import get_ebnf_grammar, get_ebnf_transformer, get_ebnf_compiler, DHPARSER_IMPORTS
38
from DHParser.dsl import grammar_provider
39
from DHParser.syntaxtree import Node, parse_sxpr
40
from DHParser.stringview import StringView
41
42


43
44
45
46
47
48

class TestWhitespace:
    # TODO: add test cases here
    def test_whitespace_comment_mangling(self):
        pass

Eckhart Arnold's avatar
Eckhart Arnold committed
49
    def test_non_empty_derivation(self):
50
51
        pass

52
53
class TestParserError:
    def test_parser_error_str(self):
54
        pe = ParserError(Node('TAG', 'test').with_pos(0), StringView('Beispiel'), None, True)
55
56
        assert str(pe).find('Beispiel') >= 0 and str(pe).find('TAG') >= 0

57
58
59
60
61
62
63
64
65
66
67
68
69
    def test_false_lookahead_only_message(self):
        """Error.PARSER_LOOKAHEAD_*_ONLY errors must not be reported if there
        no lookahead parser in the history!"""
        lang = """
        word = letters { letters | `-` letters }
        letters = /[A-Za-z]+/
        """
        gr = grammar_provider(lang)()
        st = gr('hard-time', track_history=True)
        assert not st.errors
        st = gr('hard-', track_history=True)
        assert st.errors and not any(e.code == 1045 for e in st.errors)

eckhart's avatar
eckhart committed
70

eckhart's avatar
eckhart committed
71
72
73
74
75
76
77
78
79
80
class TestParserClass:
    def test_apply(self):
        minilang ="""
            expr = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
        gr = grammar_provider(minilang)()
        l = []
        def visitor(p: Parser):
di68kap's avatar
di68kap committed
81
            l.append(p.pname + p.ptype)
eckhart's avatar
eckhart committed
82
83
84
85
86
87
88
89
90
91
92
        gr.root__.apply(visitor)
        s1 = ", ".join(l)
        l = []
        gr.root__.apply(visitor)
        s2 = ", ".join(l)
        l = []
        gr.root__.apply(visitor)
        s3 = ", ".join(l)
        assert s1 == s2 == s3


93
class TestInfiLoopsAndRecursion:
94
    def test_direct_left_recursion1(self):
eckhart's avatar
eckhart committed
95
        minilang = """
96
97
98
99
            expr = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
100
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
101
        parser = grammar_provider(minilang)()
102
        assert parser
Eckhart Arnold's avatar
Eckhart Arnold committed
103
        syntax_tree = parser(snippet)
104
105
        assert not is_error(syntax_tree.error_flag), str(syntax_tree.errors_sorted)
        assert snippet == syntax_tree.content, str(syntax_tree)
Eckhart Arnold's avatar
Eckhart Arnold committed
106
        if is_logging():
107
            log_ST(syntax_tree, "test_LeftRecursion_direct.cst")
108
            log_parsing_history(parser, "test_LeftRecursion_direct")
109

110
    def test_direct_left_recursion2(self):
111
112
113
114
115
116
        minilang = """
            expr = ex
            ex   = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
117
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
118
        parser = grammar_provider(minilang)()
119
        assert parser
120
        syntax_tree = parser(snippet)
121
122
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
123

124
    def test_indirect_left_recursion1(self):
125
126
        minilang = """
            Expr    = //~ (Product | Sum | Value)
127
128
            Product = Expr { ('*' | '/') Expr }+
            Sum     = Expr { ('+' | '-') Expr }+
129
            Value   = /[0-9.]+/~ | '(' §Expr ')'
130
            """
131
        parser = grammar_provider(minilang)()
132
        assert parser
133
        snippet = "8 * 4"
134
        syntax_tree = parser(snippet)
135
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
136
137
        snippet = "7 + 8 * 4"
        syntax_tree = parser(snippet)
138
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
139
140
        snippet = "9 + 8 * (4 + 3)"
        syntax_tree = parser(snippet)
141
142
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
143
        if is_logging():
144
            log_ST(syntax_tree, "test_LeftRecursion_indirect.cst")
145
            log_parsing_history(parser, "test_LeftRecursion_indirect")
146

147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
    # # BEWARE: EXPERIMENTAL TEST can be long running
    # def test_indirect_left_recursion2(self):
    #     arithmetic_syntax = """
    #         expression     = addition | subtraction
    #         addition       = (expression | term) "+" (expression | term)
    #         subtraction    = (expression | term) "-" (expression | term)
    #         term           = multiplication | division
    #         multiplication = (term | factor) "*" (term | factor)
    #         division       = (term | factor) "/" (term | factor)
    #         factor         = [SIGN] ( NUMBER | VARIABLE | group ) { VARIABLE | group }
    #         group          = "(" expression ")"
    #         SIGN           = /[+-]/
    #         NUMBER         = /(?:0|(?:[1-9]\d*))(?:\.\d+)?/~
    #         VARIABLE       = /[A-Za-z]/~
    #         """
    #     arithmetic = grammar_provider(arithmetic_syntax)()
    #     arithmetic.left_recursion_depth__ = 2
    #     assert arithmetic
    #     syntax_tree = arithmetic("(a + b) * (a - b)")
    #     assert syntax_tree.errors
167

168
169
170
    def test_break_inifnite_loop_ZeroOrMore(self):
        forever = ZeroOrMore(RegExp(''))
        result = Grammar(forever)('')  # infinite loops will automatically be broken
171
        assert repr(result) == "Node(:EMPTY, )", repr(result)
172
173
174
175

    def test_break_inifnite_loop_OneOrMore(self):
        forever = OneOrMore(RegExp(''))
        result = Grammar(forever)('')  # infinite loops will automatically be broken
176
        assert repr(result) == "Node(:EMPTY, )", str(result)
177

178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
    # def test_infinite_loops(self):
    #     minilang = """forever = { // } \n"""
    #     try:
    #         parser_class = grammar_provider(minilang)
    #     except CompilationError as error:
    #         assert all(e.code == Error.INFINITE_LOOP for e in error.errors)
    #     save = get_config_value('static_analysis')
    #     set_config_value('static_analysis', 'late')
    #     provider = grammar_provider(minilang)
    #     try:
    #         parser = provider()
    #     except GrammarError as error:
    #         assert error.errors[0][2].code == Error.INFINITE_LOOP
    #     set_config_value('static_analysis', 'none')
    #     parser = provider()
    #     snippet = " "
    #     syntax_tree = parser(snippet)
    #     assert any(e.code == Error.INFINITE_LOOP for e in syntax_tree.errors)
    #     res = parser.static_analysis()
    #     assert res and res[0][2].code == Error.INFINITE_LOOP
    #     minilang = """not_forever = { / / } \n"""
    #     parser = grammar_provider(minilang)()
    #     res = parser.static_analysis()
    #     assert not res
    #     set_config_value('static_analysis', save)
203

eckhart's avatar
eckhart committed
204

Eckhart Arnold's avatar
Eckhart Arnold committed
205
class TestFlowControl:
206
207
    def setup(self):
        self.t1 = """
eckhart's avatar
eckhart committed
208
        All work and no play
209
210
211
212
213
        makes Jack a dull boy
        END
        """
        self.t2 = "All word and not play makes Jack a dull boy END\n"

Eckhart Arnold's avatar
Eckhart Arnold committed
214
    def test_lookbehind(self):
215
        ws = RegExp(r'\s*')
Eckhart Arnold's avatar
Eckhart Arnold committed
216
        end = RegExp("END")
217
        doc_end = Lookbehind(RegExp('\\s*?\\n')) + end
218
        word = RegExp(r'\w+')
Eckhart Arnold's avatar
Eckhart Arnold committed
219
220
221
222
        sequence = OneOrMore(NegativeLookahead(end) + word + ws)
        document = ws + sequence + doc_end + ws

        parser = Grammar(document)
223
224
225
226
227
228
229
        cst = parser(self.t1)
        assert not cst.error_flag, cst.as_sxpr()
        cst = parser(self.t2)
        assert cst.error_flag, cst.as_sxpr()

    def test_lookbehind_indirect(self):
        class LookbehindTestGrammar(Grammar):
eckhart's avatar
eckhart committed
230
            parser_initialization__ = ["upon instantiation"]
231
            ws = RegExp(r'\s*')
232
            end = RegExp('END')
233
            SUCC_LB = RegExp('\\s*?\\n')
234
            doc_end = Series(Lookbehind(SUCC_LB), end)
235
            word = RegExp(r'\w+')
236
237
238
239
240
241
            sequence = OneOrMore(Series(NegativeLookahead(end), word, ws))
            document = Series(ws, sequence, doc_end, ws)
            root__ = document

        parser = LookbehindTestGrammar()
        cst = parser(self.t1)
Eckhart Arnold's avatar
Eckhart Arnold committed
242
        assert not cst.error_flag, cst.as_sxpr()
243
        cst = parser(self.t2)
Eckhart Arnold's avatar
Eckhart Arnold committed
244
245
246
        assert cst.error_flag, cst.as_sxpr()


247
248
249
250
251
252
253
class TestRegex:
    def test_multilineRegex(self):
        mlregex = r"""
        regex =  /\w+    # one or more alphabetical characters including the underscore
                  [+]    # followed by a plus sign
                  \w*    # possibly followed by more alpha chracters/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
254
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
Eckhart Arnold's avatar
Eckhart Arnold committed
255
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
256
        assert result
257
        assert not messages, str(messages)
258
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
eckhart's avatar
eckhart committed
259
260
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
261
        assert node.tag_name == "regex"
262
263
        assert str(node) == 'abc+def'

264
265
    def test_multilineRegex_wo_Comments(self):
        mlregex = r"""
eckhart's avatar
eckhart committed
266
267
        regex =  /\w+
                  [+]
268
269
                  \w* /
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
270
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
271
272
273
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages, str(messages)
274
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
275
276
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
277
        assert node.tag_name == "regex"
278
279
        assert str(node) == 'abc+def'

280
281
282
283
284
    def text_ignore_case(self):
        mlregex = r"""
        @ ignorecase = True
        regex = /alpha/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
285
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
286
287
288
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages
289
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
290
291
292
293
        node, rest = parser.regex('Alpha')
        assert node
        assert not node.error_flag
        assert rest == ''
294
        assert node.tag_name == "regex"
295
296
297
298
299
300
        assert str(node) == 'Alpha'

        mlregex = r"""
        @ ignorecase = False
        regex = /alpha/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
301
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
302
303
304
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages
305
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
306
307
308
309
        node, rest = parser.regex('Alpha')
        assert node.error_flag


310
311
312
313
314
315
316
317
318
319
320
321
    def test_token(self):
        tokenlang = r"""
            @whitespace = linefeed
            lang        = "" begin_token {/\w+/ ""} end_token
            begin_token = "\begin{document}"
            end_token   = "\end{document}"
            """
        testdoc = r"""
            \begin{document}
            test
            \end{document}
            """
Eckhart Arnold's avatar
Eckhart Arnold committed
322
        result, messages, _ = compile_source(
eckhart's avatar
eckhart committed
323
324
            tokenlang, None, get_ebnf_grammar(), get_ebnf_transformer(),
            get_ebnf_compiler("TokenTest"))
325
        assert result
eckhart's avatar
eckhart committed
326
        assert not messages, str(messages)
327
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
Eckhart Arnold's avatar
Eckhart Arnold committed
328
        result = parser(testdoc)
329
        # log_parsing_history(parser, "test.log")
330
331
        assert not result.error_flag

332

333
class TestGrammar:
334
    def setup(self):
335
336
337
338
339
340
        grammar = r"""@whitespace = horizontal
        haupt        = textzeile LEERZEILE
        textzeile    = { WORT }+
        WORT         = /[^ \t]+/~
        LEERZEILE    = /\n[ \t]*(?=\n)/~
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
341
342
        self.pyparser, messages, _ = compile_source(grammar, None, get_ebnf_grammar(),
                                    get_ebnf_transformer(), get_ebnf_compiler("PosTest"))
343
        assert self.pyparser
344
        assert not messages
345
346
347
348

    def test_pos_values_initialized(self):
        # checks whether pos values in the parsing result and in the
        # history record have been initialized
349
350
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
        grammar("no_file_name*")
351
        for record in grammar.history__:
352
353
            assert not record.node or record.node.pos >= 0

354
    def test_select_parsing(self):
355
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
356
357
358
359
        grammar("wort", "WORT")
        grammar("eine Zeile", "textzeile")
        grammar("kein Haupt", "haupt")
        grammar("so ist es richtig", "haupt")
360

di68kap's avatar
di68kap committed
361
362
    def test_grammar_subclassing(self):
        class Arithmetic(Grammar):
363
            r'''
di68kap's avatar
di68kap committed
364
365
366
367
368
369
            expression =  term  { ("+" | "-") term }
            term       =  factor  { ("*" | "/") factor }
            factor     =  INTEGER | "("  expression  ")"
            INTEGER    =  /\d+/~
            '''
            expression = Forward()
370
371
372
373
            INTEGER = RE('\\d+')
            factor = INTEGER | TKN("(") + expression + TKN(")")
            term = factor + ZeroOrMore((TKN("*") | TKN("/")) + factor)
            expression.set(term + ZeroOrMore((TKN("+") | TKN("-")) + term))
di68kap's avatar
di68kap committed
374
375
376
377
378
379
            root__ = expression

        grammar = Arithmetic()
        CST = grammar('3+4')
        assert not CST.error_flag, CST.as_sxpr()

380
381
382
383
384
385
386
387
388
389
390
    def test_incomplete_matching(self):
        """Tests whether the flag `complete_match` works as expected when
        calling a grammar object in order to parse a document."""
        gr = grammar_provider('word = ~/\\w+/\n')()
        st = gr('eins')
        assert not st.errors
        st = gr('eins zwei')
        assert st.errors[0].code == Error.PARSER_STOPPED_BEFORE_END
        st = gr('eins zwei', complete_match=False)
        assert not st.errors

391
    def test_synonym(self):
di68kap's avatar
di68kap committed
392
        lang = r"""
393
394
395
396
397
398
399
400
401
402
403
404
405
            doc  = { word | number }
            word = /\w+/ S
            number = [VZ] /\d+/ S 
            S    = ~        # let S by a synonym for anonymous whitespace
            VZ   = "-"
        """
        gr = grammar_provider(lang)()
        st = gr('eins 1 zwei2drei 3')
        # set_config_value('compiled_EBNF_log', 'grammar.log')
        gr = grammar_provider("@drop = whitespace, token" + lang)()
        st = gr('eins 1 zwei2drei 3')
        st = gr('-3')

406

407
408
409
410
411
412
413
class TestSeries:
    def test_non_mandatory(self):
        lang = """
        document = series | /.*/
        series = "A" "B" "C" "D"
        """
        parser = grammar_provider(lang)()
Eckhart Arnold's avatar
Eckhart Arnold committed
414
        st = parser("ABCD")
415
        assert not st.error_flag
Eckhart Arnold's avatar
Eckhart Arnold committed
416
        st = parser("A_CD")
417
        assert not st.error_flag
Eckhart Arnold's avatar
Eckhart Arnold committed
418
        st = parser("AB_D")
419
420
421
422
423
424
425
426
427
428
429
        assert not st.error_flag

    def test_mandatory(self):
        """Test for the §-operator. The Series-parser should raise an
        error for any non-match that occurs after the mandatory-operator.
        """
        lang = """
        document = series | /.*/
        series = "A" "B" §"C" "D"
        """
        parser = grammar_provider(lang)()
430
431
432
        st = parser("ABCD");  assert not st.error_flag
        st = parser("A_CD");  assert not st.error_flag
        st = parser("AB_D");  assert st.error_flag
433
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
434
        # transitivity of mandatory-operator
435
        st = parser("ABC_");  assert st.error_flag
436
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
437
438

    def test_series_composition(self):
439
        TA, TB, TC, TD, TE = (TKN(b) for b in "ABCDE")
440
441
442
443
444
        s1 = Series(TA, TB, TC, mandatory=2)
        s2 = Series(TD, TE)

        combined = Alternative(s1 + s2, RegExp('.*'))
        parser = Grammar(combined)
445
446
447
        st = parser("ABCDE");  assert not st.error_flag
        st = parser("A_CDE");  assert not st.error_flag
        st = parser("AB_DE");  assert st.error_flag
448
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
449
        st = parser("ABC_E");  assert st.error_flag
450
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
451
452
453

        combined = Alternative(s2 + s1, RegExp('.*'))
        parser = Grammar(combined)
454
455
456
457
458
459
        st = parser("DEABC");  assert not st.error_flag
        st = parser("_EABC");  assert not st.error_flag
        st = parser("D_ABC");  assert not st.error_flag
        st = parser("DE_BC");  assert not st.error_flag
        st = parser("DEA_C");  assert not st.error_flag
        st = parser("DEAB_");  assert st.error_flag
460
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
461

eckhart's avatar
eckhart committed
462
463
464
465
466
467
468
469
470
471
    # def test_boundary_cases(self):
    #     lang = """
    #     document = series | §!single | /.*/
    #     series = "A" "B" §"C" "D"
    #     single = "E"
    #     """
    #     parser_class = grammar_provider(lang)
    #     parser = parser_class()
    #     print(parser.python_src__)
    #     print(parser_class.python_src__)
eckhart's avatar
eckhart committed
472

473

474
475
476
class TestAllOfSomeOf:
    def test_allOf_order(self):
        """Test that parsers of an AllOf-List can match in arbitrary order."""
477
        prefixes = AllOf(TKN("A"), TKN("B"))
478
479
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
480
        # aternative Form
481
        prefixes = AllOf(Series(TKN("B"), TKN("A")))
482
        assert Grammar(prefixes)('A B').content == 'A B'
483
484
485
486

    def test_allOf_completeness(self):
        """Test that an error is raised if not  all parsers of an AllOf-List
        match."""
487
        prefixes = AllOf(TKN("A"), TKN("B"))
488
489
490
491
492
        assert Grammar(prefixes)('B').error_flag

    def test_allOf_redundance(self):
        """Test that one and the same parser may be listed several times
        and must be matched several times accordingly."""
493
        prefixes = AllOf(TKN("A"), TKN("B"), TKN("A"))
494
495
496
        assert Grammar(prefixes)('A A B').content == 'A A B'
        assert Grammar(prefixes)('A B A').content == 'A B A'
        assert Grammar(prefixes)('B A A').content == 'B A A'
497
498
499
500
        assert Grammar(prefixes)('A B B').error_flag

    def test_someOf_order(self):
        """Test that parsers of an AllOf-List can match in arbitrary order."""
501
        prefixes = SomeOf(TKN("A"), TKN("B"))
502
503
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
504
        # aternative Form
505
        prefixes = SomeOf(Alternative(TKN("B"), TKN("A")))
506
507
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B').content == 'B'
508
509
510
511

    def test_someOf_redundance(self):
        """Test that one and the same parser may be listed several times
        and must be matched several times accordingly."""
512
        prefixes = SomeOf(TKN("A"), TKN("B"), TKN("A"))
513
514
515
        assert Grammar(prefixes)('A A B').content == 'A A B'
        assert Grammar(prefixes)('A B A').content == 'A B A'
        assert Grammar(prefixes)('B A A').content == 'B A A'
516
517
518
        assert Grammar(prefixes)('A B B').error_flag


519
520
521
522
class TestErrorRecovery:
    def test_series_skip(self):
        lang = """
        document = series | /.*/
523
        @series_skip = /(?=[A-Z])/
524
525
526
527
528
        series = "A" "B" §"C" "D"
        """
        parser = grammar_provider(lang)()
        st = parser('AB_D')
        assert len(st.errors) == 1  # no additional "stopped before end"-error!
529
530
531
532
        parser.resume_notices__ = True
        st = parser('AB_D')
        assert len(st.errors) == 2 and any(err.code == Error.RESUME_NOTICE for err in st.errors)

533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558

    def test_AllOf_skip(self):
        lang = """
        document = allof | /.*/
        @allof_skip = /[A-Z]/
        allof = < "A" §"B" "C" "D" >
        """
        parser = grammar_provider(lang)()
        st = parser('CADB')
        assert 'allof' in st and st['allof'].content == "CADB"
        st = parser('_BCD')
        assert st.equals(parse_sxpr('(document "_BCD")'))
        st = parser('_ABC')
        assert st.equals(parse_sxpr('(document "_ABC")'))
        st = parser('A_CD')
        assert st['allof'].content == "A_CD"
        st = parser('AB_D')
        assert st['allof'].content == "AB_D"
        st = parser('A__D')
        assert st['allof'].content == "A__D"
        st = parser('CA_D')
        assert st['allof'].content == "CA_D"
        st = parser('BC_A')
        assert st['allof'].content == "BC_A"


559
class TestPopRetrieve:
560
    mini_language = r"""
561
562
        document       = { text | codeblock }
        codeblock      = delimiter { text | (!:delimiter delimiter_sign) } ::delimiter
563
        delimiter      = delimiter_sign  # never use delimiter between capture and pop except for retrival!
564
        delimiter_sign = /`+/
eckhart's avatar
eckhart committed
565
        text           = /[^`]+/
566
        """
567
    mini_lang2 = r"""
568
569
570
571
572
573
574
575
        @braces_filter=counterpart
        document       = { text | codeblock }
        codeblock      = braces { text | opening_braces | (!:braces closing_braces) } ::braces
        braces         = opening_braces
        opening_braces = /\{+/
        closing_braces = /\}+/
        text           = /[^{}]+/
        """
576
    mini_lang3 = r"""
577
578
579
580
        document       = { text | env }
        env            = (specialtag | opentag) text [closespecial | closetag]
        opentag        = "<" name ">"
        specialtag     = "<" /ABC/ !name ">"
eckhart's avatar
eckhart committed
581
        closetag       = close_slash | close_star
582
583
584
585
586
587
588
589
        close_slash    = "<" ::name "/>"
        close_star     = "<" ::name "*>"
        closespecial   = "<" /ABC/~ ">"
        name           = /\w+/~
        text           = /[^<>]+/
        """

    def setup(self):
590
591
592
        self.minilang_parser = grammar_provider(self.mini_language)()
        self.minilang_parser2 = grammar_provider(self.mini_lang2)()
        self.minilang_parser3 = grammar_provider(self.mini_lang3)()
593
594
595

    @staticmethod
    def opening_delimiter(node, name):
596
        return node.tag_name == name # and not isinstance(node.parser, Retrieve)
597
598
599

    @staticmethod
    def closing_delimiter(node):
600
601
        return node.tag_name in {':Pop', ':Retrieve'}
        # return isinstance(node.parser, Retrieve)
602
603
604
605
606
607
608
609
610

    def test_compile_mini_language(self):
        assert self.minilang_parser
        assert self.minilang_parser2
        assert self.minilang_parser3

    def test_stackhandling(self):
        ambigous_opening = "<ABCnormal> normal tag <ABCnormal*>"
        syntax_tree = self.minilang_parser3(ambigous_opening)
611
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
612
613
614

        ambigous_opening = "<ABCnormal> normal tag <ABCnormal/>"
        syntax_tree = self.minilang_parser3(ambigous_opening)
615
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
616
617
618

        forgot_closing_tag = "<em> where is the closing tag?"
        syntax_tree = self.minilang_parser3(forgot_closing_tag)
619
        assert syntax_tree.error_flag, str(syntax_tree.errors_sorted)
620
621
622

        proper = "<em> has closing tag <em/>"
        syntax_tree = self.minilang_parser3(proper)
623
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
624
625
626

        proper = "<em> has closing tag <em*>"
        syntax_tree = self.minilang_parser3(proper)
627
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
628

629
    def test_cache_neutrality(self):
630
631
        """Test that packrat-caching does not interfere with the variable-
        changing parsers: Capture and Retrieve."""
632
        lang = r"""
633
634
635
636
637
638
639
640
641
            text = opening closing
            opening = (unmarked_package | marked_package)
            closing = ::variable
            unmarked_package = package "."
            marked_package = package "*" "."
            package = "(" variable ")"
            variable = /\w+/~
            """
        case = "(secret)*. secret"
642
        gr = grammar_provider(lang)()
643
        st = gr(case)
644
        assert not st.error_flag, str(st.errors_sorted)
645
646
647
648

    def test_single_line(self):
        teststr = "Anfang ```code block `` <- keine Ende-Zeichen ! ``` Ende"
        syntax_tree = self.minilang_parser(teststr)
649
        assert not syntax_tree.errors_sorted
650
651
        delim = str(next(syntax_tree.select_if(partial(self.opening_delimiter, name="delimiter"))))
        pop = str(next(syntax_tree.select_if(self.closing_delimiter)))
652
653
        assert delim == pop
        if is_logging():
654
            log_ST(syntax_tree, "test_PopRetrieve_single_line.cst")
655
656
657
658
659
660
661
662
663
664
665

    def test_multi_line(self):
        teststr = """
            Anfang ```code block `` <- keine Ende-Zeichen ! ``` Ebde

            Absatz ohne ``` codeblock, aber
            das stellt sich erst am Ende herause...

            Mehrzeliger ```code block
            """
        syntax_tree = self.minilang_parser(teststr)
666
        assert not syntax_tree.errors_sorted
667
668
        delim = str(next(syntax_tree.select_if(partial(self.opening_delimiter, name="delimiter"))))
        pop = str(next(syntax_tree.select_if(self.closing_delimiter)))
669
670
        assert delim == pop
        if is_logging():
671
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
672
673
674
675

    def test_single_line_complement(self):
        teststr = "Anfang {{{code block }} <- keine Ende-Zeichen ! }}} Ende"
        syntax_tree = self.minilang_parser2(teststr)
676
        assert not syntax_tree.errors_sorted
677
678
        delim = str(next(syntax_tree.select_if(partial(self.opening_delimiter, name="braces"))))
        pop = str(next(syntax_tree.select_if(self.closing_delimiter)))
679
680
        assert len(delim) == len(pop) and delim != pop
        if is_logging():
681
            log_ST(syntax_tree, "test_PopRetrieve_single_line.cst")
682
683
684
685
686
687
688
689
690
691
692

    def test_multi_line_complement(self):
        teststr = """
            Anfang {{{code block {{ <- keine Ende-Zeichen ! }}} Ende

            Absatz ohne {{{ codeblock, aber
            das stellt sich erst am Ende heraus...

            Mehrzeliger }}}code block
            """
        syntax_tree = self.minilang_parser2(teststr)
693
        assert not syntax_tree.errors_sorted
694
695
        delim = str(next(syntax_tree.select_if(partial(self.opening_delimiter, name="braces"))))
        pop = str(next(syntax_tree.select_if(self.closing_delimiter)))
696
697
        assert len(delim) == len(pop) and delim != pop
        if is_logging():
698
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
699
700


701
class TestWhitespaceHandling:
Eckhart Arnold's avatar
Eckhart Arnold committed
702
    minilang = """
703
704
705
706
707
708
709
710
711
        doc = A B
        A = "A"
        B = "B"
        Rdoc = ar br
        ar = /A/
        br = /B/
        """

    def setup(self):
712
        self.gr = grammar_provider(self.minilang)()
713
714
715
716
717
718
719
720
721
722
723
724

    def test_token_whitespace(self):
        st = self.gr("AB", 'doc')
        assert not st.error_flag
        st = self.gr("A B", 'doc')
        assert not st.error_flag

    def test_regexp_whitespace(self):
        st = self.gr("AB", 'Rdoc')
        assert not st.error_flag
        st = self.gr("A B", 'Rdoc')
        assert st.error_flag
di68kap's avatar
di68kap committed
725
726
727
728
729


class TestErrorReporting:
    grammar = """
        root      = series alpha | anything
eckhart's avatar
eckhart committed
730
        series    = subseries &alpha
di68kap's avatar
di68kap committed
731
732
733
734
735
736
737
738
739
740
741
742
743
744
        subseries = alpha §beta
        alpha     = /[a-z]+/
        beta      = /[A-Z]+/
        anything  = /.*/
        """

    def setup(self):
        self.parser = grammar_provider(self.grammar)()

    def test_error_propagation(self):
        testcode1 = "halloB"
        testcode2 = "XYZ"
        testcode3 = "hallo "
        cst = self.parser(testcode1)
745
        assert not cst.error_flag, str(cst.errors_sorted)
di68kap's avatar
di68kap committed
746
747
748
749
        cst = self.parser(testcode2)
        assert not cst.error_flag
        cst = self.parser(testcode3)
        assert cst.error_flag
750
751


di68kap's avatar
di68kap committed
752
753
class TestBorderlineCases:
    def test_not_matching(self):
754
        minilang = """parser = /X/\n"""
di68kap's avatar
di68kap committed
755
756
757
758
        gr = grammar_provider(minilang)()
        cst = gr('X', 'parser')
        assert not cst.error_flag
        cst = gr(' ', 'parser')
759
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
760
        cst = gr('', 'parser')
761
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
762
763
764
765
766
767
768

    def test_matching(self):
        minilang = """parser = /.?/"""
        gr = grammar_provider(minilang)()
        cst = gr(' ', 'parser')
        assert not cst.error_flag
        cst = gr('  ', 'parser')
769
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_STOPPED_BEFORE_END
di68kap's avatar
di68kap committed
770
771
772
773
        cst = gr('', 'parser')
        assert not cst.error_flag


Eckhart Arnold's avatar
Eckhart Arnold committed
774
class TestReentryAfterError:
775
    def setup(self):
Eckhart Arnold's avatar
Eckhart Arnold committed
776
777
778
779
780
        lang = """
        document = alpha [beta] gamma "."
          alpha = "ALPHA" abc
            abc = §"a" "b" "c"
          beta = "BETA" (bac | bca)
781
782
            bac = "b" "a" §"c"
            bca = "b" "c" §"a"
Eckhart Arnold's avatar
Eckhart Arnold committed
783
784
785
786
          gamma = "GAMMA" §(cab | cba)
            cab = "c" "a" §"b"
            cba = "c" "b" §"a"
        """
787
        self.gr = grammar_provider(lang)()
788

789
790
    def test_no_resume_rules(self):
        gr = self.gr;  gr.resume_rules = dict()
791
792
        content = 'ALPHA acb BETA bac GAMMA cab .'
        cst = gr(content)
793
794
795
796
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')

eckhart's avatar
eckhart committed
797
798
799
800
801
802
803
804
    def test_no_resume_rules_partial_parsing(self):
        gr = self.gr;  gr.resume_rules = dict()
        content = 'ALPHA acb'
        cst = gr(content, 'alpha')
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')

805
806
    def test_simple_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
807
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)')]
808
        content = 'ALPHA acb BETA bac GAMMA cab .'
809
810
811
812
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
813
        # because of resuming, there should be only one error message
814
        assert len(cst.errors_sorted) == 1
815

816
817
    def test_failing_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
818
        gr.resume_rules__['alpha'] = [re.compile(r'(?=XXX)')]
819
        content = 'ALPHA acb BETA bac GAMMA cab .'
820
821
822
823
824
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        # assert cst.pick('alpha').content.startswith('ALPHA')

825
826
    def test_severl_reentry_points(self):
        gr = self.gr;  gr.resume_rules = dict()
827
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
828
        content = 'ALPHA acb BETA bac GAMMA cab .'
829
830
831
832
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
833
        # because of resuming, there should be only one error message
834
        assert len(cst.errors_sorted) == 1
835

836
837
    def test_several_reentry_points_second_point_matching(self):
        gr = self.gr;  gr.resume_rules = dict()
838
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
839
840
        content = 'ALPHA acb GAMMA cab .'
        cst = gr(content)
Eckhart Arnold's avatar
Eckhart Arnold committed
841
        assert cst.error_flag
842
        assert cst.content == content
843
        assert cst.pick('alpha').content.startswith('ALPHA')
844
        # because of resuming, there should be only one error message
845
        assert len(cst.errors_sorted) == 1
846
847
848
        gr.resume_notices__ = True
        cst = gr(content, track_history=True)
        assert len(cst.errors) == 2 and any(err.code == Error.RESUME_NOTICE for err in cst.errors)
849

850
851
    def test_several_resume_rules_innermost_rule_matching(self):
        gr = self.gr;  gr.resume_rules = dict()
852
853
854
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
        gr.resume_rules__['beta'] = [re.compile(r'(?=GAMMA)')]
        gr.resume_rules__['bac'] = [re.compile(r'(?=GAMMA)')]
855
856
857
858
859
        content = 'ALPHA abc BETA bad GAMMA cab .'
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
860
        # because of resuming, there should be only one error message
861
        assert len(cst.errors_sorted) == 1
862
863
864
865
866
867
        # multiple failures
        content = 'ALPHA acb BETA bad GAMMA cab .'
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
868
        # there should be only two error messages
869
        assert len(cst.errors_sorted) == 2
870

871
    def test_skip_comment_on_resume(self):
872
873
        lang = r"""
            @ comment =  /(?:\/\/.*)|(?:\/\*(?:.|\n)*?\*\/)/  # Kommentare im C++-Stil
874
            document = block_A block_B
875
            @ block_A_resume = /(?=x)/
876
877
            block_A = "a" §"b" "c"
            block_B = "x" "y" "z"
878
        """
879
880
881
882
883
884
885
886
887
888
889
890
        def mini_suite(grammar):
            tree = grammar('abc/*x*/xyz')
            assert not tree.errors
            tree = grammar('abDxyz')
            mandatory_cont = (Error.MANDATORY_CONTINUATION, Error.MANDATORY_CONTINUATION_AT_EOF)
            assert len(tree.errors) == 1 and tree.errors[0].code in mandatory_cont
            tree = grammar('abD/*x*/xyz')
            assert len(tree.errors) == 1 and tree.errors[0].code in mandatory_cont
            tree = grammar('aD /*x*/ c /* a */ /*x*/xyz')
            assert len(tree.errors) == 1 and tree.errors[0].code in mandatory_cont

        # test regex-defined resume rule
891
        grammar = grammar_provider(lang)()
892
        mini_suite(grammar)
893

894

895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
    def test_unambiguous_error_location(self):
        lang = r"""
            @ drop        = whitespace, token  # drop tokens and whitespace early
           
            @object_resume = /(?<=\})/
           
            json       = ~ value EOF
            value      = object | string 
            object     = "{" [member { "," §member }] "}"
            member     = string §":" value
            string     = `"` CHARACTERS `"` ~

            CHARACTERS = { /[^"\\]+/ }                  
            EOF      =  !/./        # no more characters ahead, end of file reached
            """
        test_case = """{
                "missing member": "abcdef",
            }"""
        gr = grammar_provider(lang)()
        cst = gr(test_case)
        assert any(err.code == Error.MANDATORY_CONTINUATION for err in cst.errors)


918
class TestConfiguredErrorMessages:
919
    def test_configured_error_message(self):
920
921
922
        lang = """
            document = series | /.*/
            @series_error = "a badly configured error message {5}"
923
            series = /X/ | head §"C" "D"
924
925
926
927
            head = "A" "B"
            """
        parser = grammar_provider(lang)()
        st = parser("AB_D");  assert st.error_flag
928
929
        assert st.errors_sorted[0].code == Error.MALFORMED_ERROR_STRING
        assert st.errors_sorted[1].code == Error.MANDATORY_CONTINUATION
930
931


932
933
934
935
936
937
938
939
940
941
class TestUnknownParserError:
    def test_unknown_parser_error(self):
        gr = Grammar()
        try:
            gr("", "NonExistantParser")
            assert False, "UnknownParserError expected!"
        except UnknownParserError:
            pass


942
943
class TestEarlyTokenWhitespaceDrop:
    def setup(self):
944
        self.lang = r"""
945
946
947
            @ drop = token, whitespace
            expression = term  { ("+" | "-") term}
            term       = factor  { ("*"|"/") factor}
eckhart's avatar
eckhart committed
948
            factor     = number | variable | "("  expression  ")"
949
                       | constant | fixed
950
            variable   = /[a-z]/~
951
952
            number     = /\d+/~
            constant   = "A" | "B"
eckhart's avatar
eckhart committed
953
            fixed      = "X"
954
955
956
            """
        self.gr = grammar_provider(self.lang)()

957
    def test_drop(self):
958
959
960
        cst = self.gr('4 + 3 * 5')
        assert not cst.pick(':Token')
        assert not cst.pick(':Whitespace')
961
962
        cst = self.gr('A + B')
        try:
963
            _ = next(cst.select_if(lambda node: node.content == 'A'))
964
965
966
967
            assert False, "Tokens in compound expressions should be dropped!"
        except StopIteration:
            pass
        cst = self.gr('X * y')
968
        assert next(cst.select_if(lambda node: node.content == 'X'))
969

970

Eckhart Arnold's avatar
Eckhart Arnold committed
971
class TestMetaParser:
972
973
974
975
    def setup(self):
        self.mp = MetaParser()
        self.mp.grammar = Grammar()  # override placeholder warning
        self.mp.pname = "named"
976
        self.mp.anonymous = False
977
978
979
        self.mp.tag_name = self.mp.pname

    def test_return_value(self):
980
981
        save = get_config_value('flatten_tree')
        set_config_value('flatten_tree', True)
982
        nd = self.mp._return_value(Node('tagged', 'non-empty'))
Eckhart Arnold's avatar
Eckhart Arnold committed
983
984
985
986
        assert nd.tag_name == 'named', nd.as_sxpr()
        assert len(nd.children) == 1
        assert nd.children[0].tag_name == 'tagged'
        assert nd.children[0].result == "non-empty"
987
        nd = self.mp._return_value(Node('tagged', ''))
Eckhart Arnold's avatar
Eckhart Arnold committed
988
989
990
991
        assert nd.tag_name == 'named', nd.as_sxpr()
        assert len(nd.children) == 1
        assert nd.children[0].tag_name == 'tagged'
        assert not nd.children[0].result
992
        nd = self.mp._return_value(Node(':anonymous', 'content'))
Eckhart Arnold's avatar
Eckhart Arnold committed
993
994
995
        assert nd.tag_name == 'named', nd.as_sxpr()
        assert not nd.children
        assert nd.result == 'content'
996
        nd = self.mp._return_value(Node(':anonymous', ''))
Eckhart Arnold's avatar
Eckhart Arnold committed
997
998
999
        assert nd.tag_name == 'named', nd.as_sxpr()
        assert not nd.children
        assert not nd.content
1000
        nd = self.mp._return_value(EMPTY_NODE)