test_parse.py 45.5 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
    Interleave, UnknownParserError, MetaParser, Token, EMPTY_NODE
36
from DHParser import compile_source
37
38
from DHParser.ebnf import get_ebnf_grammar, get_ebnf_transformer, get_ebnf_compiler, \
    compile_ebnf, DHPARSER_IMPORTS
39
from DHParser.dsl import grammar_provider
40
from DHParser.syntaxtree import Node, parse_sxpr
41
from DHParser.stringview import StringView
42
from DHParser.trace import set_tracer, trace_history, resume_notices_on
43
44


45
46
47
48
49
50

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

Eckhart Arnold's avatar
Eckhart Arnold committed
51
    def test_non_empty_derivation(self):
52
53
        pass

54

55
56
class TestParserError:
    def test_parser_error_str(self):
57
        pe = ParserError(Node('TAG', 'test').with_pos(0), StringView('Beispiel'), None, True)
58
59
        assert str(pe).find('Beispiel') >= 0 and str(pe).find('TAG') >= 0

60
61
62
63
64
65
66
67
    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)()
68
69
        set_tracer(gr, trace_history)
        st = gr('hard-time')
70
        assert not st.errors
71
        st = gr('hard-')
72
73
        assert st.errors and not any(e.code == 1045 for e in st.errors)

eckhart's avatar
eckhart committed
74

eckhart's avatar
eckhart committed
75
76
77
78
79
80
81
82
83
84
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
85
            l.append(p.pname + p.ptype)
eckhart's avatar
eckhart committed
86
87
88
89
90
91
92
93
94
95
96
        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


97
class TestInfiLoopsAndRecursion:
98
    def test_direct_left_recursion1(self):
eckhart's avatar
eckhart committed
99
        minilang = """
100
101
102
103
            expr = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
104
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
105
        parser = grammar_provider(minilang)()
106
        assert parser
Eckhart Arnold's avatar
Eckhart Arnold committed
107
        syntax_tree = parser(snippet)
108
109
        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
110
        if is_logging():
111
            log_ST(syntax_tree, "test_LeftRecursion_direct.cst")
112
            log_parsing_history(parser, "test_LeftRecursion_direct")
113

114
    def test_direct_left_recursion2(self):
115
116
117
118
119
120
        minilang = """
            expr = ex
            ex   = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
121
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
122
        parser = grammar_provider(minilang)()
123
        assert parser
124
        syntax_tree = parser(snippet)
125
126
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
127

128
    def test_indirect_left_recursion1(self):
129
130
        minilang = """
            Expr    = //~ (Product | Sum | Value)
131
132
            Product = Expr { ('*' | '/') Expr }+
            Sum     = Expr { ('+' | '-') Expr }+
133
            Value   = /[0-9.]+/~ | '(' §Expr ')'
134
            """
135
        parser = grammar_provider(minilang)()
136
        assert parser
137
        snippet = "8 * 4"
138
        syntax_tree = parser(snippet)
139
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
140
141
        snippet = "7 + 8 * 4"
        syntax_tree = parser(snippet)
142
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
143
144
        snippet = "9 + 8 * (4 + 3)"
        syntax_tree = parser(snippet)
145
146
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
147
        if is_logging():
148
            log_ST(syntax_tree, "test_LeftRecursion_indirect.cst")
149
            log_parsing_history(parser, "test_LeftRecursion_indirect")
150

151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
    # # 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
171

172
173
174
    def test_break_inifnite_loop_ZeroOrMore(self):
        forever = ZeroOrMore(RegExp(''))
        result = Grammar(forever)('')  # infinite loops will automatically be broken
175
        assert repr(result) == "Node(':EMPTY', '')", repr(result)
176
177
178
179

    def test_break_inifnite_loop_OneOrMore(self):
        forever = OneOrMore(RegExp(''))
        result = Grammar(forever)('')  # infinite loops will automatically be broken
180
        assert repr(result) == "Node(':EMPTY', '')", repr(result)
181

182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
    # 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)
207

eckhart's avatar
eckhart committed
208

Eckhart Arnold's avatar
Eckhart Arnold committed
209
class TestFlowControl:
210
211
    def setup(self):
        self.t1 = """
eckhart's avatar
eckhart committed
212
        All work and no play
213
214
215
216
217
        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
218
    def test_lookbehind(self):
219
        ws = RegExp(r'\s*')
Eckhart Arnold's avatar
Eckhart Arnold committed
220
        end = RegExp("END")
221
        doc_end = Lookbehind(RegExp('\\s*?\\n')) + end
222
        word = RegExp(r'\w+')
Eckhart Arnold's avatar
Eckhart Arnold committed
223
224
225
226
        sequence = OneOrMore(NegativeLookahead(end) + word + ws)
        document = ws + sequence + doc_end + ws

        parser = Grammar(document)
227
228
229
230
231
232
233
        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
234
            parser_initialization__ = ["upon instantiation"]
235
            ws = RegExp(r'\s*')
236
            end = RegExp('END')
237
            SUCC_LB = RegExp('\\s*?\\n')
238
            doc_end = Series(Lookbehind(SUCC_LB), end)
239
            word = RegExp(r'\w+')
240
241
242
243
244
245
            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
246
        assert not cst.error_flag, cst.as_sxpr()
247
        cst = parser(self.t2)
Eckhart Arnold's avatar
Eckhart Arnold committed
248
249
250
        assert cst.error_flag, cst.as_sxpr()


251
252
253
254
255
256
257
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
258
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
Eckhart Arnold's avatar
Eckhart Arnold committed
259
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
260
        assert result
261
        assert not messages, str(messages)
262
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
eckhart's avatar
eckhart committed
263
264
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
265
        assert node.tag_name == "regex"
266
267
        assert str(node) == 'abc+def'

268
269
    def test_multilineRegex_wo_Comments(self):
        mlregex = r"""
eckhart's avatar
eckhart committed
270
271
        regex =  /\w+
                  [+]
272
273
                  \w* /
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
274
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
275
276
277
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages, str(messages)
278
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
279
280
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
281
        assert node.tag_name == "regex"
282
283
        assert str(node) == 'abc+def'

284
285
286
287
288
    def text_ignore_case(self):
        mlregex = r"""
        @ ignorecase = True
        regex = /alpha/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
289
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
290
291
292
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages
293
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
294
295
296
297
        node, rest = parser.regex('Alpha')
        assert node
        assert not node.error_flag
        assert rest == ''
298
        assert node.tag_name == "regex"
299
300
301
302
303
304
        assert str(node) == 'Alpha'

        mlregex = r"""
        @ ignorecase = False
        regex = /alpha/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
305
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
306
307
308
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages
309
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
310
311
312
313
        node, rest = parser.regex('Alpha')
        assert node.error_flag


314
315
316
317
318
319
320
321
322
323
324
325
    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
326
        result, messages, _ = compile_source(
eckhart's avatar
eckhart committed
327
328
            tokenlang, None, get_ebnf_grammar(), get_ebnf_transformer(),
            get_ebnf_compiler("TokenTest"))
329
        assert result
eckhart's avatar
eckhart committed
330
        assert not messages, str(messages)
331
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
Eckhart Arnold's avatar
Eckhart Arnold committed
332
        result = parser(testdoc)
333
        # log_parsing_history(parser, "test.log")
334
335
        assert not result.error_flag

336

337
class TestGrammar:
338
    def setup(self):
339
340
341
342
343
344
        grammar = r"""@whitespace = horizontal
        haupt        = textzeile LEERZEILE
        textzeile    = { WORT }+
        WORT         = /[^ \t]+/~
        LEERZEILE    = /\n[ \t]*(?=\n)/~
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
345
346
        self.pyparser, messages, _ = compile_source(grammar, None, get_ebnf_grammar(),
                                    get_ebnf_transformer(), get_ebnf_compiler("PosTest"))
347
        assert self.pyparser
348
        assert not messages
349
350
351
352

    def test_pos_values_initialized(self):
        # checks whether pos values in the parsing result and in the
        # history record have been initialized
353
354
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
        grammar("no_file_name*")
355
        for record in grammar.history__:
356
357
            assert not record.node or record.node.pos >= 0

358
    def test_select_parsing(self):
359
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
360
361
362
363
        grammar("wort", "WORT")
        grammar("eine Zeile", "textzeile")
        grammar("kein Haupt", "haupt")
        grammar("so ist es richtig", "haupt")
364

di68kap's avatar
di68kap committed
365
366
    def test_grammar_subclassing(self):
        class Arithmetic(Grammar):
367
            r'''
di68kap's avatar
di68kap committed
368
369
370
371
372
373
            expression =  term  { ("+" | "-") term }
            term       =  factor  { ("*" | "/") factor }
            factor     =  INTEGER | "("  expression  ")"
            INTEGER    =  /\d+/~
            '''
            expression = Forward()
374
375
376
377
            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
378
379
380
381
382
383
            root__ = expression

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

384
385
386
387
388
389
390
391
392
393
394
    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

395
    def test_synonym(self):
di68kap's avatar
di68kap committed
396
        lang = r"""
397
398
399
400
401
402
403
404
405
406
407
408
409
            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')

410

411
412
413
414
415
416
417
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
418
        st = parser("ABCD")
419
        assert not st.error_flag
Eckhart Arnold's avatar
Eckhart Arnold committed
420
        st = parser("A_CD")
421
        assert not st.error_flag
Eckhart Arnold's avatar
Eckhart Arnold committed
422
        st = parser("AB_D")
423
424
425
426
427
428
429
430
431
432
433
        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)()
434
435
436
        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
437
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
438
        # transitivity of mandatory-operator
439
        st = parser("ABC_");  assert st.error_flag
440
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
441
442

    def test_series_composition(self):
443
        TA, TB, TC, TD, TE = (TKN(b) for b in "ABCDE")
444
445
446
447
448
        s1 = Series(TA, TB, TC, mandatory=2)
        s2 = Series(TD, TE)

        combined = Alternative(s1 + s2, RegExp('.*'))
        parser = Grammar(combined)
449
450
451
        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
452
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
453
        st = parser("ABC_E");  assert st.error_flag
454
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
455
456
457

        combined = Alternative(s2 + s1, RegExp('.*'))
        parser = Grammar(combined)
458
459
460
461
462
463
        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
464
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
465

eckhart's avatar
eckhart committed
466
467
468
469
470
471
472
473
474
475
    # 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
476

477
478
    def test_ebnf_serialization(self):
        ebnf_grammar = get_ebnf_grammar()
eckhart's avatar
eckhart committed
479
        # TODO: Add test here
480
        ebnf = ebnf_grammar.as_ebnf()
481
        # print(ebnf)
482
483


484
485
486
class TestAllOfSomeOf:
    def test_allOf_order(self):
        """Test that parsers of an AllOf-List can match in arbitrary order."""
487
        prefixes = AllOf(TKN("A"), TKN("B"))
488
489
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
490
491
492
493

    def test_allOf_completeness(self):
        """Test that an error is raised if not  all parsers of an AllOf-List
        match."""
494
        prefixes = AllOf(TKN("A"), TKN("B"))
495
496
497
498
499
        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."""
500
        prefixes = AllOf(TKN("A"), TKN("B"), TKN("A"))
501
502
503
        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'
504
505
506
507
        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."""
508
        prefixes = SomeOf(TKN("A"), TKN("B"))
509
510
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
511
        prefixes = SomeOf(TKN("B"), TKN("A"))
512
513
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B').content == 'B'
514
515
516
517

    def test_someOf_redundance(self):
        """Test that one and the same parser may be listed several times
        and must be matched several times accordingly."""
518
        prefixes = SomeOf(TKN("A"), TKN("B"), TKN("A"))
519
520
521
        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'
522
523
524
        assert Grammar(prefixes)('A B B').error_flag


525
526
527
528
529
class TestInterleave:
    def test_interleave_most_simple(self):
        letterset = Interleave(Token("A"), Token("B"), Token("C"))
        gr = Grammar(letterset)
        st = gr('ABC')
530
        assert not st.errors, str(st.errors)
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
        assert st.content == "ABC"
        st = gr('BCA')
        assert not st.errors
        assert st.content == "BCA"
        st = gr('BCBA')
        assert st.errors
        st = gr('AB')
        assert st.errors

    def test_interleave(self):
        letterset = Interleave(Token("A"), Token("B"), Token("C"),
                               repetitions=[(1, 1000), (0, 1), (1, 1)])
        gr = Grammar(letterset)
        st = gr('AABC')
        assert not st.errors
546
547
548
549
550
551
552
553
554
555
        st = gr('BACAAA')
        assert not st.errors
        st = gr('ABCC')
        assert st.errors
        st = gr('AAACAAA')
        assert not st.errors
        st = gr('AAABAAA')
        assert st.errors


556
557
558
559
class TestErrorRecovery:
    def test_series_skip(self):
        lang = """
        document = series | /.*/
560
        @series_skip = /(?=[A-Z])/
561
562
563
564
565
        series = "A" "B" §"C" "D"
        """
        parser = grammar_provider(lang)()
        st = parser('AB_D')
        assert len(st.errors) == 1  # no additional "stopped before end"-error!
566
        resume_notices_on(parser)
567
568
        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
569
        assert 'Skipping' in str(st.errors_sorted[1])
570

571

572
    def test_Interleave_skip(self):
573
574
575
        lang = """
        document = allof | /.*/
        @allof_skip = /[A-Z]/
576
        allof = "A" ° §"B" ° "C" ° "D"
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
        """
        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"
593
594
        st = parser('A_CB')
        assert st['allof'].content == "A_CB"
595
        st = parser('BC_A')
596
        assert 'allof' not in st
597
598


599
class TestPopRetrieve:
600
    mini_language = r"""
601
602
        document       = { text | codeblock }
        codeblock      = delimiter { text | (!:delimiter delimiter_sign) } ::delimiter
603
        delimiter      = delimiter_sign  # never use delimiter between capture and pop except for retrival!
604
        delimiter_sign = /`+/
eckhart's avatar
eckhart committed
605
        text           = /[^`]+/
606
        """
607
    mini_lang2 = r"""
608
        @braces_filter=matching_bracket
609
610
611
612
613
614
615
        document       = { text | codeblock }
        codeblock      = braces { text | opening_braces | (!:braces closing_braces) } ::braces
        braces         = opening_braces
        opening_braces = /\{+/
        closing_braces = /\}+/
        text           = /[^{}]+/
        """
616
    mini_lang3 = r"""
617
618
619
620
        document       = { text | env }
        env            = (specialtag | opentag) text [closespecial | closetag]
        opentag        = "<" name ">"
        specialtag     = "<" /ABC/ !name ">"
eckhart's avatar
eckhart committed
621
        closetag       = close_slash | close_star
622
623
624
625
626
627
        close_slash    = "<" ::name "/>"
        close_star     = "<" ::name "*>"
        closespecial   = "<" /ABC/~ ">"
        name           = /\w+/~
        text           = /[^<>]+/
        """
628
629
630
631
632
633
634
635
    mini_lang4 = r"""
        document       = { text | env }
        env            = opentag document closetag
        opentag        = "<" name ">"
        closetag       = "</" :?name ">"
        name           = /\w+/~
        text           = /[^<>]+/        
    """
636
637

    def setup(self):
638
639
640
        self.minilang_parser = grammar_provider(self.mini_language)()
        self.minilang_parser2 = grammar_provider(self.mini_lang2)()
        self.minilang_parser3 = grammar_provider(self.mini_lang3)()
641
        self.minilang_parser4 = grammar_provider(self.mini_lang4)()
642
643

    @staticmethod
644
    def has_tag_name(node, name):
645
        return node.tag_name == name # and not isinstance(node.parser, Retrieve)
646
647
648
649
650

    def test_compile_mini_language(self):
        assert self.minilang_parser
        assert self.minilang_parser2
        assert self.minilang_parser3
651
        assert self.minilang_parser4
652
653
654
655

    def test_stackhandling(self):
        ambigous_opening = "<ABCnormal> normal tag <ABCnormal*>"
        syntax_tree = self.minilang_parser3(ambigous_opening)
656
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
657
658
659

        ambigous_opening = "<ABCnormal> normal tag <ABCnormal/>"
        syntax_tree = self.minilang_parser3(ambigous_opening)
660
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
661
662
663

        forgot_closing_tag = "<em> where is the closing tag?"
        syntax_tree = self.minilang_parser3(forgot_closing_tag)
664
        assert syntax_tree.error_flag, str(syntax_tree.errors_sorted)
665
666
667

        proper = "<em> has closing tag <em/>"
        syntax_tree = self.minilang_parser3(proper)
668
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
669
670
671

        proper = "<em> has closing tag <em*>"
        syntax_tree = self.minilang_parser3(proper)
672
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
673

674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
    def test_optional_match(self):
        test1 = '<info>Hey, you</info>'
        st = self.minilang_parser4(test1)
        assert not st.error_flag
        test12 = '<info>Hey, <emph>you</emph></info>'
        st = self.minilang_parser4(test1)
        assert not st.error_flag
        test2 = '<info>Hey, you</>'
        st = self.minilang_parser4(test2)
        assert not st.error_flag
        test3 = '<info>Hey, <emph>you</></>'
        st = self.minilang_parser4(test3)
        assert not st.error_flag
        test4 = '<info>Hey, <emph>you</></info>'
        st = self.minilang_parser4(test4)
        assert not st.error_flag

    def test_rollback_behaviour_of_optional_match(self):
        test1 = '<info>Hey, you</info*>'
        st = self.minilang_parser4(test1)
        assert not self.minilang_parser4.variables__['name']
        assert st.error_flag
        test2 = '<info>Hey, you</*>'
        st = self.minilang_parser4(test2)
        assert not self.minilang_parser4.variables__['name']
        assert st.error_flag

701
    def test_cache_neutrality(self):
702
703
        """Test that packrat-caching does not interfere with the variable-
        changing parsers: Capture and Retrieve."""
704
        lang = r"""
705
706
707
708
709
710
711
712
713
            text = opening closing
            opening = (unmarked_package | marked_package)
            closing = ::variable
            unmarked_package = package "."
            marked_package = package "*" "."
            package = "(" variable ")"
            variable = /\w+/~
            """
        case = "(secret)*. secret"
714
        gr = grammar_provider(lang)()
715
        st = gr(case)
716
        assert not st.error_flag, str(st.errors_sorted)
717
718
719
720

    def test_single_line(self):
        teststr = "Anfang ```code block `` <- keine Ende-Zeichen ! ``` Ende"
        syntax_tree = self.minilang_parser(teststr)
721
722
        assert not syntax_tree.errors_sorted, \
            ''.join(str(error) for error in syntax_tree.errors_sorted)
723
724
725
        matchf = partial(self.has_tag_name, name="delimiter")
        delim = str(next(syntax_tree.select_if(matchf)))
        pop = str(next(syntax_tree.select_if(matchf, reverse=True)))
726
727
        assert delim == pop
        if is_logging():
728
            log_ST(syntax_tree, "test_PopRetrieve_single_line.cst")
729
730
731
732
733
734
735
736
737
738
739

    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)
740
        assert not syntax_tree.errors_sorted
741
742
743
        matchf = partial(self.has_tag_name, name="delimiter")
        delim = str(next(syntax_tree.select_if(matchf)))
        pop = str(next(syntax_tree.select_if(matchf, reverse=True)))
744
745
        assert delim == pop
        if is_logging():
746
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
747
748
749
750

    def test_single_line_complement(self):
        teststr = "Anfang {{{code block }} <- keine Ende-Zeichen ! }}} Ende"
        syntax_tree = self.minilang_parser2(teststr)
751
        assert not syntax_tree.errors_sorted
752
753
754
755
756
        matchf = partial(self.has_tag_name, name="braces")
        delim = str(next(syntax_tree.select_if(matchf)))
        pop = str(next(syntax_tree.select_if(matchf, reverse=True)))
        assert len(delim) == len(pop)
        assert delim != pop
757
        if is_logging():
758
            log_ST(syntax_tree, "test_PopRetrieve_single_line.cst")
759
760
761
762
763
764
765
766
767
768
769

    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)
770
        assert not syntax_tree.errors_sorted
771
772
773
        matchf = partial(self.has_tag_name, name="braces")
        delim = str(next(syntax_tree.select_if(matchf)))
        pop = str(next(syntax_tree.select_if(matchf, reverse=True)))
774
775
        assert len(delim) == len(pop) and delim != pop
        if is_logging():
776
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
777

778
    def test_autoretrieve(self):
779
780
        lang = r"""
            document   = { definition } § EOF
781
            definition = symbol :defsign value
782
            symbol     = /\w+/~                      
783
784
            defsign    = "=" | ":="
            value      = /\d+/~
785
            EOF        = !/./ [:?defsign]   # eat up captured defsigns
786
        """
787
788
        # code, _, _ = compile_ebnf(lang)
        # print(code)
789
        parser = grammar_provider(lang)()
790
791
792
        st = parser("X := 1")
        assert not st.error_flag
        st1 = st
793
794
        st = parser("")
        assert not st.error_flag
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818

        lines = [line for line in lang.split('\n') if line.strip()]
        eof_line = lines.pop()
        lines.insert(1, eof_line)
        lang = '\n'.join(lines)
        parser = grammar_provider(lang)()
        st = parser("X := 1")
        assert not st.errors
        assert st.equals(st1)

        del lines[1]
        lines.insert(2, eof_line)
        lang = '\n'.join(lines)
        parser = grammar_provider(lang)()
        st = parser("X := 1")
        assert not st.errors
        assert st.equals(st1)

        # and, finally...
        lang_variant = r"""
            document   = { definition } § EOF
            symbol     = /\w+/~                      
            defsign    = "=" | ":="
            value      = /\d+/~
819
            EOF        = !/./ :?defsign   # eat up captured defsign, only if it has been retrieved
820
821
822
823
824
825
            definition = symbol :defsign value
        """
        parser = grammar_provider(lang_variant)()
        st = parser("X := 1")
        assert not st.errors
        assert st.equals(st1)
826
827
        st = parser('')
        assert "EOF expected" in str(st.errors)
di68kap's avatar
di68kap committed
828

829

830
class TestWhitespaceHandling:
Eckhart Arnold's avatar
Eckhart Arnold committed
831
    minilang = """
832
833
834
835
836
837
838
839
840
        doc = A B
        A = "A"
        B = "B"
        Rdoc = ar br
        ar = /A/
        br = /B/
        """

    def setup(self):
841
        self.gr = grammar_provider(self.minilang)()
842
843
844
845
846
847
848
849
850
851
852
853

    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
854
855
856
857
858


class TestErrorReporting:
    grammar = """
        root      = series alpha | anything
eckhart's avatar
eckhart committed
859
        series    = subseries &alpha
di68kap's avatar
di68kap committed
860
861
862
863
864
865
866
867
868
869
870
871
872
873
        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)
874
        assert not cst.error_flag, str(cst.errors_sorted)
di68kap's avatar
di68kap committed
875
876
877
878
        cst = self.parser(testcode2)
        assert not cst.error_flag
        cst = self.parser(testcode3)
        assert cst.error_flag
879
880


di68kap's avatar
di68kap committed
881
882
class TestBorderlineCases:
    def test_not_matching(self):
883
        minilang = """parser = /X/\n"""
di68kap's avatar
di68kap committed
884
885
886
887
        gr = grammar_provider(minilang)()
        cst = gr('X', 'parser')
        assert not cst.error_flag
        cst = gr(' ', 'parser')
888
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
889
        cst = gr('', 'parser')
890
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
891
892
893
894
895
896
897

    def test_matching(self):
        minilang = """parser = /.?/"""
        gr = grammar_provider(minilang)()
        cst = gr(' ', 'parser')
        assert not cst.error_flag
        cst = gr('  ', 'parser')
898
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_STOPPED_BEFORE_END
di68kap's avatar
di68kap committed
899
900
901
902
        cst = gr('', 'parser')
        assert not cst.error_flag


Eckhart Arnold's avatar
Eckhart Arnold committed
903
class TestReentryAfterError:
904
    def setup(self):
Eckhart Arnold's avatar
Eckhart Arnold committed
905
906
907
908
909
        lang = """
        document = alpha [beta] gamma "."
          alpha = "ALPHA" abc
            abc = §"a" "b" "c"
          beta = "BETA" (bac | bca)
910
911
            bac = "b" "a" §"c"
            bca = "b" "c" §"a"
Eckhart Arnold's avatar
Eckhart Arnold committed
912
913
914
915
          gamma = "GAMMA" §(cab | cba)
            cab = "c" "a" §"b"
            cba = "c" "b" §"a"
        """
916
        self.gr = grammar_provider(lang)()
917

918
919
    def test_no_resume_rules(self):
        gr = self.gr;  gr.resume_rules = dict()
920
921
        content = 'ALPHA acb BETA bac GAMMA cab .'
        cst = gr(content)
922
923
924
925
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')

eckhart's avatar
eckhart committed
926
927
928
929
930
931
932
933
    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')

934
935
    def test_simple_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
936
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)')]
937
        content = 'ALPHA acb BETA bac GAMMA cab .'
938
939
940
941
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
942
        # because of resuming, there should be only one error message
Eckhart Arnold's avatar
Eckhart Arnold committed
943
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 1
944

945
946
    def test_failing_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
947
        gr.resume_rules__['alpha'] = [re.compile(r'(?=XXX)')]
948
        content = 'ALPHA acb BETA bac GAMMA cab .'
949
950
951
952
953
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        # assert cst.pick('alpha').content.startswith('ALPHA')

eckhart's avatar
eckhart committed
954
    def test_several_reentry_points(self):
955
        gr = self.gr;  gr.resume_rules = dict()
956
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
957
        content = 'ALPHA acb BETA bac GAMMA cab .'
958
959
960
961
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
962
        # because of resuming, there should be only one error message
Eckhart Arnold's avatar
Eckhart Arnold committed
963
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 1
964

965
966
    def test_several_reentry_points_second_point_matching(self):
        gr = self.gr;  gr.resume_rules = dict()
967
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
968
969
        content = 'ALPHA acb GAMMA cab .'
        cst = gr(content)
Eckhart Arnold's avatar
Eckhart Arnold committed
970
        assert cst.error_flag
971
        assert cst.content == content
972
        assert cst.pick('alpha').content.startswith('ALPHA')
973
        # because of resuming, there should be only one error message
974
        assert len(cst.errors_sorted) == 1
975
976
        resume_notices_on(gr)
        cst = gr(content)
977
        assert len(cst.errors) == 2 and any(err.code == Error.RESUME_NOTICE for err in cst.errors)
978

979
980
    def test_several_resume_rules_innermost_rule_matching(self):
        gr = self.gr;  gr.resume_rules = dict()
981
982
983
        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)')]
984
985
986
987
988
        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')
989
        # because of resuming, there should be only one error message
Eckhart Arnold's avatar
Eckhart Arnold committed
990
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 1
991
992
993
994
995
996
        # 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')
997
        # there should be only two error messages
Eckhart Arnold's avatar
Eckhart Arnold committed
998
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 2
999

1000
    def test_skip_comment_on_resume(self):