test_parse.py 39.6 KB
Newer Older
1
#!/usr/bin/env python3
2

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
from DHParser.trace import set_tracer, trace_history, resume_notices_on
42
43


44
45
46
47
48
49

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

Eckhart Arnold's avatar
Eckhart Arnold committed
50
    def test_non_empty_derivation(self):
51
52
        pass

53
54
class TestParserError:
    def test_parser_error_str(self):
55
        pe = ParserError(Node('TAG', 'test').with_pos(0), StringView('Beispiel'), None, True)
56
57
        assert str(pe).find('Beispiel') >= 0 and str(pe).find('TAG') >= 0

58
59
60
61
62
63
64
65
    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)()
66
67
        set_tracer(gr, trace_history)
        st = gr('hard-time')
68
        assert not st.errors
69
        st = gr('hard-')
70
71
        assert st.errors and not any(e.code == 1045 for e in st.errors)

eckhart's avatar
eckhart committed
72

eckhart's avatar
eckhart committed
73
74
75
76
77
78
79
80
81
82
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
83
            l.append(p.pname + p.ptype)
eckhart's avatar
eckhart committed
84
85
86
87
88
89
90
91
92
93
94
        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


95
class TestInfiLoopsAndRecursion:
96
    def test_direct_left_recursion1(self):
eckhart's avatar
eckhart committed
97
        minilang = """
98
99
100
101
            expr = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
102
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
103
        parser = grammar_provider(minilang)()
104
        assert parser
Eckhart Arnold's avatar
Eckhart Arnold committed
105
        syntax_tree = parser(snippet)
106
107
        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
108
        if is_logging():
109
            log_ST(syntax_tree, "test_LeftRecursion_direct.cst")
110
            log_parsing_history(parser, "test_LeftRecursion_direct")
111

112
    def test_direct_left_recursion2(self):
113
114
115
116
117
118
        minilang = """
            expr = ex
            ex   = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
119
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
120
        parser = grammar_provider(minilang)()
121
        assert parser
122
        syntax_tree = parser(snippet)
123
124
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
125

126
    def test_indirect_left_recursion1(self):
127
128
        minilang = """
            Expr    = //~ (Product | Sum | Value)
129
130
            Product = Expr { ('*' | '/') Expr }+
            Sum     = Expr { ('+' | '-') Expr }+
131
            Value   = /[0-9.]+/~ | '(' §Expr ')'
132
            """
133
        parser = grammar_provider(minilang)()
134
        assert parser
135
        snippet = "8 * 4"
136
        syntax_tree = parser(snippet)
137
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
138
139
        snippet = "7 + 8 * 4"
        syntax_tree = parser(snippet)
140
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
141
142
        snippet = "9 + 8 * (4 + 3)"
        syntax_tree = parser(snippet)
143
144
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
145
        if is_logging():
146
            log_ST(syntax_tree, "test_LeftRecursion_indirect.cst")
147
            log_parsing_history(parser, "test_LeftRecursion_indirect")
148

149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
    # # 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
169

170
171
172
    def test_break_inifnite_loop_ZeroOrMore(self):
        forever = ZeroOrMore(RegExp(''))
        result = Grammar(forever)('')  # infinite loops will automatically be broken
173
        assert repr(result) == "Node(':EMPTY', '')", repr(result)
174
175
176
177

    def test_break_inifnite_loop_OneOrMore(self):
        forever = OneOrMore(RegExp(''))
        result = Grammar(forever)('')  # infinite loops will automatically be broken
178
        assert repr(result) == "Node(':EMPTY', '')", repr(result)
179

180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
    # 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)
205

eckhart's avatar
eckhart committed
206

Eckhart Arnold's avatar
Eckhart Arnold committed
207
class TestFlowControl:
208
209
    def setup(self):
        self.t1 = """
eckhart's avatar
eckhart committed
210
        All work and no play
211
212
213
214
215
        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
216
    def test_lookbehind(self):
217
        ws = RegExp(r'\s*')
Eckhart Arnold's avatar
Eckhart Arnold committed
218
        end = RegExp("END")
219
        doc_end = Lookbehind(RegExp('\\s*?\\n')) + end
220
        word = RegExp(r'\w+')
Eckhart Arnold's avatar
Eckhart Arnold committed
221
222
223
224
        sequence = OneOrMore(NegativeLookahead(end) + word + ws)
        document = ws + sequence + doc_end + ws

        parser = Grammar(document)
225
226
227
228
229
230
231
        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
232
            parser_initialization__ = ["upon instantiation"]
233
            ws = RegExp(r'\s*')
234
            end = RegExp('END')
235
            SUCC_LB = RegExp('\\s*?\\n')
236
            doc_end = Series(Lookbehind(SUCC_LB), end)
237
            word = RegExp(r'\w+')
238
239
240
241
242
243
            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
244
        assert not cst.error_flag, cst.as_sxpr()
245
        cst = parser(self.t2)
Eckhart Arnold's avatar
Eckhart Arnold committed
246
247
248
        assert cst.error_flag, cst.as_sxpr()


249
250
251
252
253
254
255
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
256
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
Eckhart Arnold's avatar
Eckhart Arnold committed
257
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
258
        assert result
259
        assert not messages, str(messages)
260
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
eckhart's avatar
eckhart committed
261
262
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
263
        assert node.tag_name == "regex"
264
265
        assert str(node) == 'abc+def'

266
267
    def test_multilineRegex_wo_Comments(self):
        mlregex = r"""
eckhart's avatar
eckhart committed
268
269
        regex =  /\w+
                  [+]
270
271
                  \w* /
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
272
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
273
274
275
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages, str(messages)
276
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
277
278
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
279
        assert node.tag_name == "regex"
280
281
        assert str(node) == 'abc+def'

282
283
284
285
286
    def text_ignore_case(self):
        mlregex = r"""
        @ ignorecase = True
        regex = /alpha/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
287
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
288
289
290
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages
291
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
292
293
294
295
        node, rest = parser.regex('Alpha')
        assert node
        assert not node.error_flag
        assert rest == ''
296
        assert node.tag_name == "regex"
297
298
299
300
301
302
        assert str(node) == 'Alpha'

        mlregex = r"""
        @ ignorecase = False
        regex = /alpha/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
303
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
304
305
306
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages
307
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
308
309
310
311
        node, rest = parser.regex('Alpha')
        assert node.error_flag


312
313
314
315
316
317
318
319
320
321
322
323
    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
324
        result, messages, _ = compile_source(
eckhart's avatar
eckhart committed
325
326
            tokenlang, None, get_ebnf_grammar(), get_ebnf_transformer(),
            get_ebnf_compiler("TokenTest"))
327
        assert result
eckhart's avatar
eckhart committed
328
        assert not messages, str(messages)
329
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
Eckhart Arnold's avatar
Eckhart Arnold committed
330
        result = parser(testdoc)
331
        # log_parsing_history(parser, "test.log")
332
333
        assert not result.error_flag

334

335
class TestGrammar:
336
    def setup(self):
337
338
339
340
341
342
        grammar = r"""@whitespace = horizontal
        haupt        = textzeile LEERZEILE
        textzeile    = { WORT }+
        WORT         = /[^ \t]+/~
        LEERZEILE    = /\n[ \t]*(?=\n)/~
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
343
344
        self.pyparser, messages, _ = compile_source(grammar, None, get_ebnf_grammar(),
                                    get_ebnf_transformer(), get_ebnf_compiler("PosTest"))
345
        assert self.pyparser
346
        assert not messages
347
348
349
350

    def test_pos_values_initialized(self):
        # checks whether pos values in the parsing result and in the
        # history record have been initialized
351
352
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
        grammar("no_file_name*")
353
        for record in grammar.history__:
354
355
            assert not record.node or record.node.pos >= 0

356
    def test_select_parsing(self):
357
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
358
359
360
361
        grammar("wort", "WORT")
        grammar("eine Zeile", "textzeile")
        grammar("kein Haupt", "haupt")
        grammar("so ist es richtig", "haupt")
362

di68kap's avatar
di68kap committed
363
364
    def test_grammar_subclassing(self):
        class Arithmetic(Grammar):
365
            r'''
di68kap's avatar
di68kap committed
366
367
368
369
370
371
            expression =  term  { ("+" | "-") term }
            term       =  factor  { ("*" | "/") factor }
            factor     =  INTEGER | "("  expression  ")"
            INTEGER    =  /\d+/~
            '''
            expression = Forward()
372
373
374
375
            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
376
377
378
379
380
381
            root__ = expression

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

382
383
384
385
386
387
388
389
390
391
392
    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

393
    def test_synonym(self):
di68kap's avatar
di68kap committed
394
        lang = r"""
395
396
397
398
399
400
401
402
403
404
405
406
407
            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')

408

409
410
411
412
413
414
415
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
416
        st = parser("ABCD")
417
        assert not st.error_flag
Eckhart Arnold's avatar
Eckhart Arnold committed
418
        st = parser("A_CD")
419
        assert not st.error_flag
Eckhart Arnold's avatar
Eckhart Arnold committed
420
        st = parser("AB_D")
421
422
423
424
425
426
427
428
429
430
431
        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)()
432
433
434
        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
435
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
436
        # transitivity of mandatory-operator
437
        st = parser("ABC_");  assert st.error_flag
438
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
439
440

    def test_series_composition(self):
441
        TA, TB, TC, TD, TE = (TKN(b) for b in "ABCDE")
442
443
444
445
446
        s1 = Series(TA, TB, TC, mandatory=2)
        s2 = Series(TD, TE)

        combined = Alternative(s1 + s2, RegExp('.*'))
        parser = Grammar(combined)
447
448
449
        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
450
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
451
        st = parser("ABC_E");  assert st.error_flag
452
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
453
454
455

        combined = Alternative(s2 + s1, RegExp('.*'))
        parser = Grammar(combined)
456
457
458
459
460
461
        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
462
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
463

eckhart's avatar
eckhart committed
464
465
466
467
468
469
470
471
472
473
    # 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
474

475

476
477
478
class TestAllOfSomeOf:
    def test_allOf_order(self):
        """Test that parsers of an AllOf-List can match in arbitrary order."""
479
        prefixes = AllOf(TKN("A"), TKN("B"))
480
481
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
482
        # aternative Form
483
        prefixes = AllOf(Series(TKN("B"), TKN("A")))
484
        assert Grammar(prefixes)('A B').content == 'A B'
485
486
487
488

    def test_allOf_completeness(self):
        """Test that an error is raised if not  all parsers of an AllOf-List
        match."""
489
        prefixes = AllOf(TKN("A"), TKN("B"))
490
491
492
493
494
        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."""
495
        prefixes = AllOf(TKN("A"), TKN("B"), TKN("A"))
496
497
498
        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'
499
500
501
502
        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."""
503
        prefixes = SomeOf(TKN("A"), TKN("B"))
504
505
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
506
        # aternative Form
507
        prefixes = SomeOf(Alternative(TKN("B"), TKN("A")))
508
509
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B').content == 'B'
510
511
512
513

    def test_someOf_redundance(self):
        """Test that one and the same parser may be listed several times
        and must be matched several times accordingly."""
514
        prefixes = SomeOf(TKN("A"), TKN("B"), TKN("A"))
515
516
517
        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'
518
519
520
        assert Grammar(prefixes)('A B B').error_flag


521
522
523
524
class TestErrorRecovery:
    def test_series_skip(self):
        lang = """
        document = series | /.*/
525
        @series_skip = /(?=[A-Z])/
526
527
528
529
530
        series = "A" "B" §"C" "D"
        """
        parser = grammar_provider(lang)()
        st = parser('AB_D')
        assert len(st.errors) == 1  # no additional "stopped before end"-error!
531
        resume_notices_on(parser)
532
533
        st = parser('AB_D')
        assert len(st.errors) == 2 and any(err.code == Error.RESUME_NOTICE for err in st.errors)
eckhart's avatar
eckhart committed
534
        assert 'Skipping' in str(st.errors_sorted[1])
535

536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561

    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"


562
class TestPopRetrieve:
563
    mini_language = r"""
564
565
        document       = { text | codeblock }
        codeblock      = delimiter { text | (!:delimiter delimiter_sign) } ::delimiter
566
        delimiter      = delimiter_sign  # never use delimiter between capture and pop except for retrival!
567
        delimiter_sign = /`+/
eckhart's avatar
eckhart committed
568
        text           = /[^`]+/
569
        """
570
    mini_lang2 = r"""
571
572
573
574
575
576
577
578
        @braces_filter=counterpart
        document       = { text | codeblock }
        codeblock      = braces { text | opening_braces | (!:braces closing_braces) } ::braces
        braces         = opening_braces
        opening_braces = /\{+/
        closing_braces = /\}+/
        text           = /[^{}]+/
        """
579
    mini_lang3 = r"""
580
581
582
583
        document       = { text | env }
        env            = (specialtag | opentag) text [closespecial | closetag]
        opentag        = "<" name ">"
        specialtag     = "<" /ABC/ !name ">"
eckhart's avatar
eckhart committed
584
        closetag       = close_slash | close_star
585
586
587
588
589
590
591
592
        close_slash    = "<" ::name "/>"
        close_star     = "<" ::name "*>"
        closespecial   = "<" /ABC/~ ">"
        name           = /\w+/~
        text           = /[^<>]+/
        """

    def setup(self):
593
594
595
        self.minilang_parser = grammar_provider(self.mini_language)()
        self.minilang_parser2 = grammar_provider(self.mini_lang2)()
        self.minilang_parser3 = grammar_provider(self.mini_lang3)()
596
597
598

    @staticmethod
    def opening_delimiter(node, name):
599
        return node.tag_name == name # and not isinstance(node.parser, Retrieve)
600
601
602

    @staticmethod
    def closing_delimiter(node):
603
604
        return node.tag_name in {':Pop', ':Retrieve'}
        # return isinstance(node.parser, Retrieve)
605
606
607
608
609
610
611
612
613

    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)
614
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
615
616
617

        ambigous_opening = "<ABCnormal> normal tag <ABCnormal/>"
        syntax_tree = self.minilang_parser3(ambigous_opening)
618
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
619
620
621

        forgot_closing_tag = "<em> where is the closing tag?"
        syntax_tree = self.minilang_parser3(forgot_closing_tag)
622
        assert syntax_tree.error_flag, str(syntax_tree.errors_sorted)
623
624
625

        proper = "<em> has closing tag <em/>"
        syntax_tree = self.minilang_parser3(proper)
626
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
627
628
629

        proper = "<em> has closing tag <em*>"
        syntax_tree = self.minilang_parser3(proper)
630
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
631

632
    def test_cache_neutrality(self):
633
634
        """Test that packrat-caching does not interfere with the variable-
        changing parsers: Capture and Retrieve."""
635
        lang = r"""
636
637
638
639
640
641
642
643
644
            text = opening closing
            opening = (unmarked_package | marked_package)
            closing = ::variable
            unmarked_package = package "."
            marked_package = package "*" "."
            package = "(" variable ")"
            variable = /\w+/~
            """
        case = "(secret)*. secret"
645
        gr = grammar_provider(lang)()
646
        st = gr(case)
647
        assert not st.error_flag, str(st.errors_sorted)
648
649
650
651

    def test_single_line(self):
        teststr = "Anfang ```code block `` <- keine Ende-Zeichen ! ``` Ende"
        syntax_tree = self.minilang_parser(teststr)
652
        assert not syntax_tree.errors_sorted
653
654
        delim = str(next(syntax_tree.select_if(partial(self.opening_delimiter, name="delimiter"))))
        pop = str(next(syntax_tree.select_if(self.closing_delimiter)))
655
656
        assert delim == pop
        if is_logging():
657
            log_ST(syntax_tree, "test_PopRetrieve_single_line.cst")
658
659
660
661
662
663
664
665
666
667
668

    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)
669
        assert not syntax_tree.errors_sorted
670
671
        delim = str(next(syntax_tree.select_if(partial(self.opening_delimiter, name="delimiter"))))
        pop = str(next(syntax_tree.select_if(self.closing_delimiter)))
672
673
        assert delim == pop
        if is_logging():
674
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
675
676
677
678

    def test_single_line_complement(self):
        teststr = "Anfang {{{code block }} <- keine Ende-Zeichen ! }}} Ende"
        syntax_tree = self.minilang_parser2(teststr)
679
        assert not syntax_tree.errors_sorted
680
681
        delim = str(next(syntax_tree.select_if(partial(self.opening_delimiter, name="braces"))))
        pop = str(next(syntax_tree.select_if(self.closing_delimiter)))
682
683
        assert len(delim) == len(pop) and delim != pop
        if is_logging():
684
            log_ST(syntax_tree, "test_PopRetrieve_single_line.cst")
685
686
687
688
689
690
691
692
693
694
695

    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)
696
        assert not syntax_tree.errors_sorted
697
698
        delim = str(next(syntax_tree.select_if(partial(self.opening_delimiter, name="braces"))))
        pop = str(next(syntax_tree.select_if(self.closing_delimiter)))
699
700
        assert len(delim) == len(pop) and delim != pop
        if is_logging():
701
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
702
703


704
class TestWhitespaceHandling:
Eckhart Arnold's avatar
Eckhart Arnold committed
705
    minilang = """
706
707
708
709
710
711
712
713
714
        doc = A B
        A = "A"
        B = "B"
        Rdoc = ar br
        ar = /A/
        br = /B/
        """

    def setup(self):
715
        self.gr = grammar_provider(self.minilang)()
716
717
718
719
720
721
722
723
724
725
726
727

    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
728
729
730
731
732


class TestErrorReporting:
    grammar = """
        root      = series alpha | anything
eckhart's avatar
eckhart committed
733
        series    = subseries &alpha
di68kap's avatar
di68kap committed
734
735
736
737
738
739
740
741
742
743
744
745
746
747
        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)
748
        assert not cst.error_flag, str(cst.errors_sorted)
di68kap's avatar
di68kap committed
749
750
751
752
        cst = self.parser(testcode2)
        assert not cst.error_flag
        cst = self.parser(testcode3)
        assert cst.error_flag
753
754


di68kap's avatar
di68kap committed
755
756
class TestBorderlineCases:
    def test_not_matching(self):
757
        minilang = """parser = /X/\n"""
di68kap's avatar
di68kap committed
758
759
760
761
        gr = grammar_provider(minilang)()
        cst = gr('X', 'parser')
        assert not cst.error_flag
        cst = gr(' ', 'parser')
762
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
763
        cst = gr('', 'parser')
764
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
765
766
767
768
769
770
771

    def test_matching(self):
        minilang = """parser = /.?/"""
        gr = grammar_provider(minilang)()
        cst = gr(' ', 'parser')
        assert not cst.error_flag
        cst = gr('  ', 'parser')
772
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_STOPPED_BEFORE_END
di68kap's avatar
di68kap committed
773
774
775
776
        cst = gr('', 'parser')
        assert not cst.error_flag


Eckhart Arnold's avatar
Eckhart Arnold committed
777
class TestReentryAfterError:
778
    def setup(self):
Eckhart Arnold's avatar
Eckhart Arnold committed
779
780
781
782
783
        lang = """
        document = alpha [beta] gamma "."
          alpha = "ALPHA" abc
            abc = §"a" "b" "c"
          beta = "BETA" (bac | bca)
784
785
            bac = "b" "a" §"c"
            bca = "b" "c" §"a"
Eckhart Arnold's avatar
Eckhart Arnold committed
786
787
788
789
          gamma = "GAMMA" §(cab | cba)
            cab = "c" "a" §"b"
            cba = "c" "b" §"a"
        """
790
        self.gr = grammar_provider(lang)()
791

792
793
    def test_no_resume_rules(self):
        gr = self.gr;  gr.resume_rules = dict()
794
795
        content = 'ALPHA acb BETA bac GAMMA cab .'
        cst = gr(content)
796
797
798
799
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')

eckhart's avatar
eckhart committed
800
801
802
803
804
805
806
807
    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')

808
809
    def test_simple_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
810
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)')]
811
        content = 'ALPHA acb BETA bac GAMMA cab .'
812
813
814
815
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
816
        # because of resuming, there should be only one error message
Eckhart Arnold's avatar
Eckhart Arnold committed
817
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 1
818

819
820
    def test_failing_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
821
        gr.resume_rules__['alpha'] = [re.compile(r'(?=XXX)')]
822
        content = 'ALPHA acb BETA bac GAMMA cab .'
823
824
825
826
827
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        # assert cst.pick('alpha').content.startswith('ALPHA')

eckhart's avatar
eckhart committed
828
    def test_several_reentry_points(self):
829
        gr = self.gr;  gr.resume_rules = dict()
830
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
831
        content = 'ALPHA acb BETA bac GAMMA cab .'
832
833
834
835
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
836
        # because of resuming, there should be only one error message
Eckhart Arnold's avatar
Eckhart Arnold committed
837
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 1
838

839
840
    def test_several_reentry_points_second_point_matching(self):
        gr = self.gr;  gr.resume_rules = dict()
841
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
842
843
        content = 'ALPHA acb GAMMA cab .'
        cst = gr(content)
Eckhart Arnold's avatar
Eckhart Arnold committed
844
        assert cst.error_flag
845
        assert cst.content == content
846
        assert cst.pick('alpha').content.startswith('ALPHA')
847
        # because of resuming, there should be only one error message
848
        assert len(cst.errors_sorted) == 1
849
850
        resume_notices_on(gr)
        cst = gr(content)
851
        assert len(cst.errors) == 2 and any(err.code == Error.RESUME_NOTICE for err in cst.errors)
852

853
854
    def test_several_resume_rules_innermost_rule_matching(self):
        gr = self.gr;  gr.resume_rules = dict()
855
856
857
        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)')]
858
859
860
861
862
        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')
863
        # because of resuming, there should be only one error message
Eckhart Arnold's avatar
Eckhart Arnold committed
864
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 1
865
866
867
868
869
870
        # 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')
871
        # there should be only two error messages
Eckhart Arnold's avatar
Eckhart Arnold committed
872
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 2
873

874
    def test_skip_comment_on_resume(self):
875
876
        lang = r"""
            @ comment =  /(?:\/\/.*)|(?:\/\*(?:.|\n)*?\*\/)/  # Kommentare im C++-Stil
877
            document = block_A block_B
878
            @ block_A_resume = /(?=x)/
879
880
            block_A = "a" §"b" "c"
            block_B = "x" "y" "z"
881
        """
882
883
884
885
886
887
888
889
890
891
892
893
        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
894
        grammar = grammar_provider(lang)()
895
        mini_suite(grammar)
896

897

898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
    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)


921
class TestConfiguredErrorMessages:
922
    def test_configured_error_message(self):
923
924
925
        lang = """
            document = series | /.*/
            @series_error = "a badly configured error message {5}"
926
            series = /X/ | head §"C" "D"
927
928
929
930
            head = "A" "B"
            """
        parser = grammar_provider(lang)()
        st = parser("AB_D");  assert st.error_flag
931
932
        assert st.errors_sorted[0].code == Error.MALFORMED_ERROR_STRING
        assert st.errors_sorted[1].code == Error.MANDATORY_CONTINUATION
933
934


935
936
937
938
939
940
941
942
943
944
class TestUnknownParserError:
    def test_unknown_parser_error(self):
        gr = Grammar()
        try:
            gr("", "NonExistantParser")
            assert False, "UnknownParserError expected!"
        except UnknownParserError:
            pass


945
946
class TestEarlyTokenWhitespaceDrop:
    def setup(self):
947
        self.lang = r"""
948
949
950
            @ drop = token, whitespace
            expression = term  { ("+" | "-") term}
            term       = factor  { ("*"|"/") factor}
eckhart's avatar
eckhart committed
951
            factor     = number | variable | "("  expression  ")"
952
                       | constant | fixed
953
            variable   = /[a-z]/~
954
955
            number     = /\d+/~
            constant   = "A" | "B"
eckhart's avatar
eckhart committed
956
            fixed      = "X"
957
958
959
            """
        self.gr = grammar_provider(self.lang)()

960
    def test_drop(self):
961
962
963
        cst = self.gr('4 + 3 * 5')
        assert not cst.pick(':Token')
        assert not cst.pick(':Whitespace')
964
965
        cst = self.gr('A + B')
        try:
966
            _ = next(cst.select_if(lambda node: node.content == 'A'))
967
968
969
970
            assert False, "Tokens in compound expressions should be dropped!"
        except StopIteration:
            pass
        cst = self.gr('X * y')
971
        assert next(cst.select_if(lambda node: node.content == 'X'))
972

973

Eckhart Arnold's avatar
Eckhart Arnold committed
974
class TestMetaParser:
975
976
977
978
    def setup(self):
        self.mp = MetaParser()
        self.mp.grammar = Grammar()  # override placeholder warning
        self.mp.pname = "named"
979
        self.mp.anonymous = False
980
981
982
        self.mp.tag_name = self.mp.pname

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