test_parse.py 44.1 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
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
55
class TestParserError:
    def test_parser_error_str(self):
56
        pe = ParserError(Node('TAG', 'test').with_pos(0), StringView('Beispiel'), None, True)
57
58
        assert str(pe).find('Beispiel') >= 0 and str(pe).find('TAG') >= 0

59
60
61
62
63
64
65
66
    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)()
67
68
        set_tracer(gr, trace_history)
        st = gr('hard-time')
69
        assert not st.errors
70
        st = gr('hard-')
71
72
        assert st.errors and not any(e.code == 1045 for e in st.errors)

eckhart's avatar
eckhart committed
73

eckhart's avatar
eckhart committed
74
75
76
77
78
79
80
81
82
83
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
84
            l.append(p.pname + p.ptype)
eckhart's avatar
eckhart committed
85
86
87
88
89
90
91
92
93
94
95
        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


96
class TestInfiLoopsAndRecursion:
97
    def test_direct_left_recursion1(self):
eckhart's avatar
eckhart committed
98
        minilang = """
99
100
101
102
            expr = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
103
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
104
        parser = grammar_provider(minilang)()
105
        assert parser
Eckhart Arnold's avatar
Eckhart Arnold committed
106
        syntax_tree = parser(snippet)
107
108
        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
109
        if is_logging():
110
            log_ST(syntax_tree, "test_LeftRecursion_direct.cst")
111
            log_parsing_history(parser, "test_LeftRecursion_direct")
112

113
    def test_direct_left_recursion2(self):
114
115
116
117
118
119
        minilang = """
            expr = ex
            ex   = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
120
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
121
        parser = grammar_provider(minilang)()
122
        assert parser
123
        syntax_tree = parser(snippet)
124
125
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
126

127
    def test_indirect_left_recursion1(self):
128
129
        minilang = """
            Expr    = //~ (Product | Sum | Value)
130
131
            Product = Expr { ('*' | '/') Expr }+
            Sum     = Expr { ('+' | '-') Expr }+
132
            Value   = /[0-9.]+/~ | '(' §Expr ')'
133
            """
134
        parser = grammar_provider(minilang)()
135
        assert parser
136
        snippet = "8 * 4"
137
        syntax_tree = parser(snippet)
138
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
139
140
        snippet = "7 + 8 * 4"
        syntax_tree = parser(snippet)
141
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
142
143
        snippet = "9 + 8 * (4 + 3)"
        syntax_tree = parser(snippet)
144
145
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
146
        if is_logging():
147
            log_ST(syntax_tree, "test_LeftRecursion_indirect.cst")
148
            log_parsing_history(parser, "test_LeftRecursion_indirect")
149

150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
    # # 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
170

171
172
173
    def test_break_inifnite_loop_ZeroOrMore(self):
        forever = ZeroOrMore(RegExp(''))
        result = Grammar(forever)('')  # infinite loops will automatically be broken
174
        assert repr(result) == "Node(':EMPTY', '')", repr(result)
175
176
177
178

    def test_break_inifnite_loop_OneOrMore(self):
        forever = OneOrMore(RegExp(''))
        result = Grammar(forever)('')  # infinite loops will automatically be broken
179
        assert repr(result) == "Node(':EMPTY', '')", repr(result)
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
205
    # 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)
206

eckhart's avatar
eckhart committed
207

Eckhart Arnold's avatar
Eckhart Arnold committed
208
class TestFlowControl:
209
210
    def setup(self):
        self.t1 = """
eckhart's avatar
eckhart committed
211
        All work and no play
212
213
214
215
216
        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
217
    def test_lookbehind(self):
218
        ws = RegExp(r'\s*')
Eckhart Arnold's avatar
Eckhart Arnold committed
219
        end = RegExp("END")
220
        doc_end = Lookbehind(RegExp('\\s*?\\n')) + end
221
        word = RegExp(r'\w+')
Eckhart Arnold's avatar
Eckhart Arnold committed
222
223
224
225
        sequence = OneOrMore(NegativeLookahead(end) + word + ws)
        document = ws + sequence + doc_end + ws

        parser = Grammar(document)
226
227
228
229
230
231
232
        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
233
            parser_initialization__ = ["upon instantiation"]
234
            ws = RegExp(r'\s*')
235
            end = RegExp('END')
236
            SUCC_LB = RegExp('\\s*?\\n')
237
            doc_end = Series(Lookbehind(SUCC_LB), end)
238
            word = RegExp(r'\w+')
239
240
241
242
243
244
            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
245
        assert not cst.error_flag, cst.as_sxpr()
246
        cst = parser(self.t2)
Eckhart Arnold's avatar
Eckhart Arnold committed
247
248
249
        assert cst.error_flag, cst.as_sxpr()


250
251
252
253
254
255
256
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
257
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
Eckhart Arnold's avatar
Eckhart Arnold committed
258
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
259
        assert result
260
        assert not messages, str(messages)
261
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
eckhart's avatar
eckhart committed
262
263
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
264
        assert node.tag_name == "regex"
265
266
        assert str(node) == 'abc+def'

267
268
    def test_multilineRegex_wo_Comments(self):
        mlregex = r"""
eckhart's avatar
eckhart committed
269
270
        regex =  /\w+
                  [+]
271
272
                  \w* /
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
273
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
274
275
276
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages, str(messages)
277
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
278
279
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
280
        assert node.tag_name == "regex"
281
282
        assert str(node) == 'abc+def'

283
284
285
286
287
    def text_ignore_case(self):
        mlregex = r"""
        @ ignorecase = True
        regex = /alpha/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
288
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
289
290
291
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages
292
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
293
294
295
296
        node, rest = parser.regex('Alpha')
        assert node
        assert not node.error_flag
        assert rest == ''
297
        assert node.tag_name == "regex"
298
299
300
301
302
303
        assert str(node) == 'Alpha'

        mlregex = r"""
        @ ignorecase = False
        regex = /alpha/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
304
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
305
306
307
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages
308
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
309
310
311
312
        node, rest = parser.regex('Alpha')
        assert node.error_flag


313
314
315
316
317
318
319
320
321
322
323
324
    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
325
        result, messages, _ = compile_source(
eckhart's avatar
eckhart committed
326
327
            tokenlang, None, get_ebnf_grammar(), get_ebnf_transformer(),
            get_ebnf_compiler("TokenTest"))
328
        assert result
eckhart's avatar
eckhart committed
329
        assert not messages, str(messages)
330
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
Eckhart Arnold's avatar
Eckhart Arnold committed
331
        result = parser(testdoc)
332
        # log_parsing_history(parser, "test.log")
333
334
        assert not result.error_flag

335

336
class TestGrammar:
337
    def setup(self):
338
339
340
341
342
343
        grammar = r"""@whitespace = horizontal
        haupt        = textzeile LEERZEILE
        textzeile    = { WORT }+
        WORT         = /[^ \t]+/~
        LEERZEILE    = /\n[ \t]*(?=\n)/~
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
344
345
        self.pyparser, messages, _ = compile_source(grammar, None, get_ebnf_grammar(),
                                    get_ebnf_transformer(), get_ebnf_compiler("PosTest"))
346
        assert self.pyparser
347
        assert not messages
348
349
350
351

    def test_pos_values_initialized(self):
        # checks whether pos values in the parsing result and in the
        # history record have been initialized
352
353
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
        grammar("no_file_name*")
354
        for record in grammar.history__:
355
356
            assert not record.node or record.node.pos >= 0

357
    def test_select_parsing(self):
358
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
359
360
361
362
        grammar("wort", "WORT")
        grammar("eine Zeile", "textzeile")
        grammar("kein Haupt", "haupt")
        grammar("so ist es richtig", "haupt")
363

di68kap's avatar
di68kap committed
364
365
    def test_grammar_subclassing(self):
        class Arithmetic(Grammar):
366
            r'''
di68kap's avatar
di68kap committed
367
368
369
370
371
372
            expression =  term  { ("+" | "-") term }
            term       =  factor  { ("*" | "/") factor }
            factor     =  INTEGER | "("  expression  ")"
            INTEGER    =  /\d+/~
            '''
            expression = Forward()
373
374
375
376
            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
377
378
379
380
381
382
            root__ = expression

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

383
384
385
386
387
388
389
390
391
392
393
    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

394
    def test_synonym(self):
di68kap's avatar
di68kap committed
395
        lang = r"""
396
397
398
399
400
401
402
403
404
405
406
407
408
            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')

409

410
411
412
413
414
415
416
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
417
        st = parser("ABCD")
418
        assert not st.error_flag
Eckhart Arnold's avatar
Eckhart Arnold committed
419
        st = parser("A_CD")
420
        assert not st.error_flag
Eckhart Arnold's avatar
Eckhart Arnold committed
421
        st = parser("AB_D")
422
423
424
425
426
427
428
429
430
431
432
        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)()
433
434
435
        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
436
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
437
        # transitivity of mandatory-operator
438
        st = parser("ABC_");  assert st.error_flag
439
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
440
441

    def test_series_composition(self):
442
        TA, TB, TC, TD, TE = (TKN(b) for b in "ABCDE")
443
444
445
446
447
        s1 = Series(TA, TB, TC, mandatory=2)
        s2 = Series(TD, TE)

        combined = Alternative(s1 + s2, RegExp('.*'))
        parser = Grammar(combined)
448
449
450
        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
451
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
452
        st = parser("ABC_E");  assert st.error_flag
453
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
454
455
456

        combined = Alternative(s2 + s1, RegExp('.*'))
        parser = Grammar(combined)
457
458
459
460
461
462
        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
463
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
464

eckhart's avatar
eckhart committed
465
466
467
468
469
470
471
472
473
474
    # 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
475

476
477
    def test_ebnf_serialization(self):
        ebnf_grammar = get_ebnf_grammar()
eckhart's avatar
eckhart committed
478
        # TODO: Add test here
479
        ebnf = ebnf_grammar.as_ebnf()
480
        # print(ebnf)
481
482


483
484
485
class TestAllOfSomeOf:
    def test_allOf_order(self):
        """Test that parsers of an AllOf-List can match in arbitrary order."""
486
        prefixes = AllOf(TKN("A"), TKN("B"))
487
488
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
489
490
491
492

    def test_allOf_completeness(self):
        """Test that an error is raised if not  all parsers of an AllOf-List
        match."""
493
        prefixes = AllOf(TKN("A"), TKN("B"))
494
495
496
497
498
        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."""
499
        prefixes = AllOf(TKN("A"), TKN("B"), TKN("A"))
500
501
502
        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'
503
504
505
506
        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."""
507
        prefixes = SomeOf(TKN("A"), TKN("B"))
508
509
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
510
        prefixes = SomeOf(TKN("B"), TKN("A"))
511
512
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B').content == 'B'
513
514
515
516

    def test_someOf_redundance(self):
        """Test that one and the same parser may be listed several times
        and must be matched several times accordingly."""
517
        prefixes = SomeOf(TKN("A"), TKN("B"), TKN("A"))
518
519
520
        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'
521
522
523
        assert Grammar(prefixes)('A B B').error_flag


524
525
526
527
528
class TestInterleave:
    def test_interleave_most_simple(self):
        letterset = Interleave(Token("A"), Token("B"), Token("C"))
        gr = Grammar(letterset)
        st = gr('ABC')
529
        assert not st.errors, str(st.errors)
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
        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
545
546
547
548
549
550
551
552
553
554
        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


555
556
557
558
class TestErrorRecovery:
    def test_series_skip(self):
        lang = """
        document = series | /.*/
559
        @series_skip = /(?=[A-Z])/
560
561
562
563
564
        series = "A" "B" §"C" "D"
        """
        parser = grammar_provider(lang)()
        st = parser('AB_D')
        assert len(st.errors) == 1  # no additional "stopped before end"-error!
565
        resume_notices_on(parser)
566
567
        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
568
        assert 'Skipping' in str(st.errors_sorted[1])
569

570

571
    def test_Interleave_skip(self):
572
573
574
        lang = """
        document = allof | /.*/
        @allof_skip = /[A-Z]/
575
        allof = "A" ° §"B" ° "C" ° "D"
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
        """
        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"
592
593
        st = parser('A_CB')
        assert st['allof'].content == "A_CB"
594
        st = parser('BC_A')
595
        assert 'allof' not in st
596
597


598
class TestPopRetrieve:
599
    mini_language = r"""
600
601
        document       = { text | codeblock }
        codeblock      = delimiter { text | (!:delimiter delimiter_sign) } ::delimiter
602
        delimiter      = delimiter_sign  # never use delimiter between capture and pop except for retrival!
603
        delimiter_sign = /`+/
eckhart's avatar
eckhart committed
604
        text           = /[^`]+/
605
        """
606
    mini_lang2 = r"""
607
        @braces_filter=matching_bracket
608
609
610
611
612
613
614
        document       = { text | codeblock }
        codeblock      = braces { text | opening_braces | (!:braces closing_braces) } ::braces
        braces         = opening_braces
        opening_braces = /\{+/
        closing_braces = /\}+/
        text           = /[^{}]+/
        """
615
    mini_lang3 = r"""
616
617
618
619
        document       = { text | env }
        env            = (specialtag | opentag) text [closespecial | closetag]
        opentag        = "<" name ">"
        specialtag     = "<" /ABC/ !name ">"
eckhart's avatar
eckhart committed
620
        closetag       = close_slash | close_star
621
622
623
624
625
626
        close_slash    = "<" ::name "/>"
        close_star     = "<" ::name "*>"
        closespecial   = "<" /ABC/~ ">"
        name           = /\w+/~
        text           = /[^<>]+/
        """
627
628
629
630
631
632
633
634
    mini_lang4 = r"""
        document       = { text | env }
        env            = opentag document closetag
        opentag        = "<" name ">"
        closetag       = "</" :?name ">"
        name           = /\w+/~
        text           = /[^<>]+/        
    """
635
636

    def setup(self):
637
638
639
        self.minilang_parser = grammar_provider(self.mini_language)()
        self.minilang_parser2 = grammar_provider(self.mini_lang2)()
        self.minilang_parser3 = grammar_provider(self.mini_lang3)()
640
        self.minilang_parser4 = grammar_provider(self.mini_lang4)()
641
642

    @staticmethod
643
    def has_tag_name(node, name):
644
        return node.tag_name == name # and not isinstance(node.parser, Retrieve)
645
646
647
648
649

    def test_compile_mini_language(self):
        assert self.minilang_parser
        assert self.minilang_parser2
        assert self.minilang_parser3
650
        assert self.minilang_parser4
651
652
653
654

    def test_stackhandling(self):
        ambigous_opening = "<ABCnormal> normal tag <ABCnormal*>"
        syntax_tree = self.minilang_parser3(ambigous_opening)
655
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
656
657
658

        ambigous_opening = "<ABCnormal> normal tag <ABCnormal/>"
        syntax_tree = self.minilang_parser3(ambigous_opening)
659
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
660
661
662

        forgot_closing_tag = "<em> where is the closing tag?"
        syntax_tree = self.minilang_parser3(forgot_closing_tag)
663
        assert syntax_tree.error_flag, str(syntax_tree.errors_sorted)
664
665
666

        proper = "<em> has closing tag <em/>"
        syntax_tree = self.minilang_parser3(proper)
667
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
668
669
670

        proper = "<em> has closing tag <em*>"
        syntax_tree = self.minilang_parser3(proper)
671
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
672

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
    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

700
    def test_cache_neutrality(self):
701
702
        """Test that packrat-caching does not interfere with the variable-
        changing parsers: Capture and Retrieve."""
703
        lang = r"""
704
705
706
707
708
709
710
711
712
            text = opening closing
            opening = (unmarked_package | marked_package)
            closing = ::variable
            unmarked_package = package "."
            marked_package = package "*" "."
            package = "(" variable ")"
            variable = /\w+/~
            """
        case = "(secret)*. secret"
713
        gr = grammar_provider(lang)()
714
        st = gr(case)
715
        assert not st.error_flag, str(st.errors_sorted)
716
717
718
719

    def test_single_line(self):
        teststr = "Anfang ```code block `` <- keine Ende-Zeichen ! ``` Ende"
        syntax_tree = self.minilang_parser(teststr)
720
721
        assert not syntax_tree.errors_sorted, \
            ''.join(str(error) for error in syntax_tree.errors_sorted)
722
723
724
        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)))
725
726
        assert delim == pop
        if is_logging():
727
            log_ST(syntax_tree, "test_PopRetrieve_single_line.cst")
728
729
730
731
732
733
734
735
736
737
738

    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)
739
        assert not syntax_tree.errors_sorted
740
741
742
        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)))
743
744
        assert delim == pop
        if is_logging():
745
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
746
747
748
749

    def test_single_line_complement(self):
        teststr = "Anfang {{{code block }} <- keine Ende-Zeichen ! }}} Ende"
        syntax_tree = self.minilang_parser2(teststr)
750
        assert not syntax_tree.errors_sorted
751
752
753
754
755
        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
756
        if is_logging():
757
            log_ST(syntax_tree, "test_PopRetrieve_single_line.cst")
758
759
760
761
762
763
764
765
766
767
768

    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)
769
        assert not syntax_tree.errors_sorted
770
771
772
        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)))
773
774
        assert len(delim) == len(pop) and delim != pop
        if is_logging():
775
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
776

777
778
779
780
781
782
783
784
785
786
787
788
    def test_autoretrieve(self):
        lang = """
            document   = { definition }
            definition = symbol :defsign value
            symbol     = /\w+/~
            defsign    = "=" | ":="
            value      = /\d+/~
        """
        parser = grammar_provider(lang)()
        st = parser("X := 1")
        print(st.as_sxpr())

789

790
class TestWhitespaceHandling:
Eckhart Arnold's avatar
Eckhart Arnold committed
791
    minilang = """
792
793
794
795
796
797
798
799
800
        doc = A B
        A = "A"
        B = "B"
        Rdoc = ar br
        ar = /A/
        br = /B/
        """

    def setup(self):
801
        self.gr = grammar_provider(self.minilang)()
802
803
804
805
806
807
808
809
810
811
812
813

    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
814
815
816
817
818


class TestErrorReporting:
    grammar = """
        root      = series alpha | anything
eckhart's avatar
eckhart committed
819
        series    = subseries &alpha
di68kap's avatar
di68kap committed
820
821
822
823
824
825
826
827
828
829
830
831
832
833
        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)
834
        assert not cst.error_flag, str(cst.errors_sorted)
di68kap's avatar
di68kap committed
835
836
837
838
        cst = self.parser(testcode2)
        assert not cst.error_flag
        cst = self.parser(testcode3)
        assert cst.error_flag
839
840


di68kap's avatar
di68kap committed
841
842
class TestBorderlineCases:
    def test_not_matching(self):
843
        minilang = """parser = /X/\n"""
di68kap's avatar
di68kap committed
844
845
846
847
        gr = grammar_provider(minilang)()
        cst = gr('X', 'parser')
        assert not cst.error_flag
        cst = gr(' ', 'parser')
848
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
849
        cst = gr('', 'parser')
850
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
851
852
853
854
855
856
857

    def test_matching(self):
        minilang = """parser = /.?/"""
        gr = grammar_provider(minilang)()
        cst = gr(' ', 'parser')
        assert not cst.error_flag
        cst = gr('  ', 'parser')
858
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_STOPPED_BEFORE_END
di68kap's avatar
di68kap committed
859
860
861
862
        cst = gr('', 'parser')
        assert not cst.error_flag


Eckhart Arnold's avatar
Eckhart Arnold committed
863
class TestReentryAfterError:
864
    def setup(self):
Eckhart Arnold's avatar
Eckhart Arnold committed
865
866
867
868
869
        lang = """
        document = alpha [beta] gamma "."
          alpha = "ALPHA" abc
            abc = §"a" "b" "c"
          beta = "BETA" (bac | bca)
870
871
            bac = "b" "a" §"c"
            bca = "b" "c" §"a"
Eckhart Arnold's avatar
Eckhart Arnold committed
872
873
874
875
          gamma = "GAMMA" §(cab | cba)
            cab = "c" "a" §"b"
            cba = "c" "b" §"a"
        """
876
        self.gr = grammar_provider(lang)()
877

878
879
    def test_no_resume_rules(self):
        gr = self.gr;  gr.resume_rules = dict()
880
881
        content = 'ALPHA acb BETA bac GAMMA cab .'
        cst = gr(content)
882
883
884
885
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')

eckhart's avatar
eckhart committed
886
887
888
889
890
891
892
893
    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')

894
895
    def test_simple_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
896
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)')]
897
        content = 'ALPHA acb BETA bac GAMMA cab .'
898
899
900
901
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
902
        # because of resuming, there should be only one error message
Eckhart Arnold's avatar
Eckhart Arnold committed
903
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 1
904

905
906
    def test_failing_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
907
        gr.resume_rules__['alpha'] = [re.compile(r'(?=XXX)')]
908
        content = 'ALPHA acb BETA bac GAMMA cab .'
909
910
911
912
913
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        # assert cst.pick('alpha').content.startswith('ALPHA')

eckhart's avatar
eckhart committed
914
    def test_several_reentry_points(self):
915
        gr = self.gr;  gr.resume_rules = dict()
916
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
917
        content = 'ALPHA acb BETA bac GAMMA cab .'
918
919
920
921
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
922
        # because of resuming, there should be only one error message
Eckhart Arnold's avatar
Eckhart Arnold committed
923
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 1
924

925
926
    def test_several_reentry_points_second_point_matching(self):
        gr = self.gr;  gr.resume_rules = dict()
927
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
928
929
        content = 'ALPHA acb GAMMA cab .'
        cst = gr(content)
Eckhart Arnold's avatar
Eckhart Arnold committed
930
        assert cst.error_flag
931
        assert cst.content == content
932
        assert cst.pick('alpha').content.startswith('ALPHA')
933
        # because of resuming, there should be only one error message
934
        assert len(cst.errors_sorted) == 1
935
936
        resume_notices_on(gr)
        cst = gr(content)
937
        assert len(cst.errors) == 2 and any(err.code == Error.RESUME_NOTICE for err in cst.errors)
938

939
940
    def test_several_resume_rules_innermost_rule_matching(self):
        gr = self.gr;  gr.resume_rules = dict()
941
942
943
        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)')]
944
945
946
947
948
        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')
949
        # because of resuming, there should be only one error message
Eckhart Arnold's avatar
Eckhart Arnold committed
950
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 1
951
952
953
954
955
956
        # 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')
957
        # there should be only two error messages
Eckhart Arnold's avatar
Eckhart Arnold committed
958
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 2
959

960
    def test_skip_comment_on_resume(self):
961
962
        lang = r"""
            @ comment =  /(?:\/\/.*)|(?:\/\*(?:.|\n)*?\*\/)/  # Kommentare im C++-Stil
963
            document = block_A block_B
964
            @ block_A_resume = /(?=x)/
965
966
            block_A = "a" §"b" "c"
            block_B = "x" "y" "z"
967
        """
968
969
970
971
972
973
974
975
976
977
978
979
        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
980
        grammar = grammar_provider(lang)()
981
        mini_suite(grammar)
982

983

984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
    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",