test_parse.py 46.1 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, \
35
36
    Interleave, UnknownParserError, MetaParser, Token, EMPTY_NODE, Capture, Drop, Whitespace, \
    GrammarError
37
from DHParser import compile_source
38
39
from DHParser.ebnf import get_ebnf_grammar, get_ebnf_transformer, get_ebnf_compiler, \
    compile_ebnf, DHPARSER_IMPORTS
40
from DHParser.dsl import grammar_provider
41
from DHParser.syntaxtree import Node, parse_sxpr
42
from DHParser.stringview import StringView
43
from DHParser.trace import set_tracer, trace_history, resume_notices_on
44
45


46
47
48
49
50
51

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

Eckhart Arnold's avatar
Eckhart Arnold committed
52
    def test_non_empty_derivation(self):
53
54
        pass

55

56
57
class TestParserError:
    def test_parser_error_str(self):
58
        pe = ParserError(Node('TAG', 'test').with_pos(0), StringView('Beispiel'), None, True)
59
60
        assert str(pe).find('Beispiel') >= 0 and str(pe).find('TAG') >= 0

61
62
63
64
65
66
67
68
    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)()
69
70
        set_tracer(gr, trace_history)
        st = gr('hard-time')
71
        assert not st.errors
72
        st = gr('hard-')
73
74
        assert st.errors and not any(e.code == 1045 for e in st.errors)

eckhart's avatar
eckhart committed
75

eckhart's avatar
eckhart committed
76
77
78
79
80
81
82
83
84
85
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
86
            l.append(p.pname + p.ptype)
eckhart's avatar
eckhart committed
87
88
89
90
91
92
93
94
95
96
97
        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


98
class TestInfiLoopsAndRecursion:
99
    def test_direct_left_recursion1(self):
eckhart's avatar
eckhart committed
100
        minilang = """
101
102
103
104
            expr = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
105
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
106
        parser = grammar_provider(minilang)()
107
        assert parser
Eckhart Arnold's avatar
Eckhart Arnold committed
108
        syntax_tree = parser(snippet)
109
110
        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
111
        if is_logging():
112
            log_ST(syntax_tree, "test_LeftRecursion_direct.cst")
113
            log_parsing_history(parser, "test_LeftRecursion_direct")
114

115
    def test_direct_left_recursion2(self):
116
117
118
119
120
121
        minilang = """
            expr = ex
            ex   = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
122
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
123
        parser = grammar_provider(minilang)()
124
        assert parser
125
        syntax_tree = parser(snippet)
126
127
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
128

129
    def test_indirect_left_recursion1(self):
130
131
        minilang = """
            Expr    = //~ (Product | Sum | Value)
132
133
            Product = Expr { ('*' | '/') Expr }+
            Sum     = Expr { ('+' | '-') Expr }+
134
            Value   = /[0-9.]+/~ | '(' §Expr ')'
135
            """
136
        parser = grammar_provider(minilang)()
137
        assert parser
138
        snippet = "8 * 4"
139
        syntax_tree = parser(snippet)
140
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
141
142
        snippet = "7 + 8 * 4"
        syntax_tree = parser(snippet)
143
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
144
145
        snippet = "9 + 8 * (4 + 3)"
        syntax_tree = parser(snippet)
146
147
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
148
        if is_logging():
149
            log_ST(syntax_tree, "test_LeftRecursion_indirect.cst")
150
            log_parsing_history(parser, "test_LeftRecursion_indirect")
151

152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
    # # 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
172

173
174
175
    def test_break_inifnite_loop_ZeroOrMore(self):
        forever = ZeroOrMore(RegExp(''))
        result = Grammar(forever)('')  # infinite loops will automatically be broken
176
        assert repr(result) == "Node(':EMPTY', '')", repr(result)
177
178
179
180

    def test_break_inifnite_loop_OneOrMore(self):
        forever = OneOrMore(RegExp(''))
        result = Grammar(forever)('')  # infinite loops will automatically be broken
181
        assert repr(result) == "Node(':EMPTY', '')", repr(result)
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
207
    # 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)
208

eckhart's avatar
eckhart committed
209

Eckhart Arnold's avatar
Eckhart Arnold committed
210
class TestFlowControl:
211
212
    def setup(self):
        self.t1 = """
eckhart's avatar
eckhart committed
213
        All work and no play
214
215
216
217
218
        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
219
    def test_lookbehind(self):
220
        ws = RegExp(r'\s*')
Eckhart Arnold's avatar
Eckhart Arnold committed
221
        end = RegExp("END")
222
        doc_end = Lookbehind(RegExp('\\s*?\\n')) + end
223
        word = RegExp(r'\w+')
Eckhart Arnold's avatar
Eckhart Arnold committed
224
225
226
227
        sequence = OneOrMore(NegativeLookahead(end) + word + ws)
        document = ws + sequence + doc_end + ws

        parser = Grammar(document)
228
229
230
231
232
233
234
        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
235
            parser_initialization__ = ["upon instantiation"]
236
            ws = RegExp(r'\s*')
237
            end = RegExp('END')
238
            SUCC_LB = RegExp('\\s*?\\n')
239
            doc_end = Series(Lookbehind(SUCC_LB), end)
240
            word = RegExp(r'\w+')
241
242
243
244
245
246
            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
247
        assert not cst.error_flag, cst.as_sxpr()
248
        cst = parser(self.t2)
Eckhart Arnold's avatar
Eckhart Arnold committed
249
250
251
        assert cst.error_flag, cst.as_sxpr()


252
253
254
255
256
257
258
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
259
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
Eckhart Arnold's avatar
Eckhart Arnold committed
260
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
261
        assert result
262
        assert not messages, str(messages)
263
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
eckhart's avatar
eckhart committed
264
265
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
266
        assert node.tag_name == "regex"
267
268
        assert str(node) == 'abc+def'

269
270
    def test_multilineRegex_wo_Comments(self):
        mlregex = r"""
eckhart's avatar
eckhart committed
271
272
        regex =  /\w+
                  [+]
273
274
                  \w* /
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
275
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
276
277
278
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages, str(messages)
279
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
280
281
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
282
        assert node.tag_name == "regex"
283
284
        assert str(node) == 'abc+def'

285
286
287
288
289
    def text_ignore_case(self):
        mlregex = r"""
        @ ignorecase = True
        regex = /alpha/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
290
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
291
292
293
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages
294
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
295
296
297
298
        node, rest = parser.regex('Alpha')
        assert node
        assert not node.error_flag
        assert rest == ''
299
        assert node.tag_name == "regex"
300
301
302
303
304
305
        assert str(node) == 'Alpha'

        mlregex = r"""
        @ ignorecase = False
        regex = /alpha/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
306
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
307
308
309
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages
310
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
311
312
313
314
        node, rest = parser.regex('Alpha')
        assert node.error_flag


315
316
317
318
319
320
321
322
323
324
325
326
    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
327
        result, messages, _ = compile_source(
eckhart's avatar
eckhart committed
328
329
            tokenlang, None, get_ebnf_grammar(), get_ebnf_transformer(),
            get_ebnf_compiler("TokenTest"))
330
        assert result
eckhart's avatar
eckhart committed
331
        assert not messages, str(messages)
332
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
Eckhart Arnold's avatar
Eckhart Arnold committed
333
        result = parser(testdoc)
334
        # log_parsing_history(parser, "test.log")
335
336
        assert not result.error_flag

337

338
class TestGrammar:
339
    def setup(self):
340
341
342
343
344
345
        grammar = r"""@whitespace = horizontal
        haupt        = textzeile LEERZEILE
        textzeile    = { WORT }+
        WORT         = /[^ \t]+/~
        LEERZEILE    = /\n[ \t]*(?=\n)/~
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
346
347
        self.pyparser, messages, _ = compile_source(grammar, None, get_ebnf_grammar(),
                                    get_ebnf_transformer(), get_ebnf_compiler("PosTest"))
348
        assert self.pyparser
349
        assert not messages
350
351
352
353

    def test_pos_values_initialized(self):
        # checks whether pos values in the parsing result and in the
        # history record have been initialized
354
355
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
        grammar("no_file_name*")
356
        for record in grammar.history__:
357
358
            assert not record.node or record.node.pos >= 0

359
    def test_select_parsing(self):
360
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
361
362
363
364
        grammar("wort", "WORT")
        grammar("eine Zeile", "textzeile")
        grammar("kein Haupt", "haupt")
        grammar("so ist es richtig", "haupt")
365

di68kap's avatar
di68kap committed
366
367
    def test_grammar_subclassing(self):
        class Arithmetic(Grammar):
368
            r'''
di68kap's avatar
di68kap committed
369
370
371
372
373
374
            expression =  term  { ("+" | "-") term }
            term       =  factor  { ("*" | "/") factor }
            factor     =  INTEGER | "("  expression  ")"
            INTEGER    =  /\d+/~
            '''
            expression = Forward()
375
376
377
378
            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
379
380
381
382
383
384
            root__ = expression

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

385
386
387
388
389
390
391
392
393
394
395
    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

396
    def test_synonym(self):
di68kap's avatar
di68kap committed
397
        lang = r"""
398
399
400
401
402
403
404
405
406
407
408
409
410
            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')

411

412
413
414
415
416
417
418
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
419
        st = parser("ABCD")
420
        assert not st.error_flag
Eckhart Arnold's avatar
Eckhart Arnold committed
421
        st = parser("A_CD")
422
        assert not st.error_flag
Eckhart Arnold's avatar
Eckhart Arnold committed
423
        st = parser("AB_D")
424
425
426
427
428
429
430
431
432
433
434
        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)()
435
436
437
        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
438
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
439
        # transitivity of mandatory-operator
440
        st = parser("ABC_");  assert st.error_flag
441
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
442
443

    def test_series_composition(self):
444
        TA, TB, TC, TD, TE = (TKN(b) for b in "ABCDE")
445
446
447
448
449
        s1 = Series(TA, TB, TC, mandatory=2)
        s2 = Series(TD, TE)

        combined = Alternative(s1 + s2, RegExp('.*'))
        parser = Grammar(combined)
450
451
452
        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
453
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
454
        st = parser("ABC_E");  assert st.error_flag
455
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
456
457
458

        combined = Alternative(s2 + s1, RegExp('.*'))
        parser = Grammar(combined)
459
460
461
462
463
464
        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
465
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
466

eckhart's avatar
eckhart committed
467
468
469
470
471
472
473
474
475
476
    # 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
477

478
479
    def test_ebnf_serialization(self):
        ebnf_grammar = get_ebnf_grammar()
eckhart's avatar
eckhart committed
480
        # TODO: Add test here
481
        ebnf = ebnf_grammar.as_ebnf()
482
        # print(ebnf)
483
484


485
486
487
class TestAllOfSomeOf:
    def test_allOf_order(self):
        """Test that parsers of an AllOf-List can match in arbitrary order."""
eckhart's avatar
eckhart committed
488
        prefixes = Interleave(TKN("A"), TKN("B"))
489
490
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
491
492
493
494

    def test_allOf_completeness(self):
        """Test that an error is raised if not  all parsers of an AllOf-List
        match."""
eckhart's avatar
eckhart committed
495
        prefixes = Interleave(TKN("A"), TKN("B"))
496
497
498
499
500
        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."""
eckhart's avatar
eckhart committed
501
        prefixes = Interleave(TKN("A"), TKN("B"), TKN("A"))
502
503
504
        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'
505
506
507
508
        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."""
eckhart's avatar
eckhart committed
509
        prefixes = Interleave(TKN("A"), TKN("B"))
510
511
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
eckhart's avatar
eckhart committed
512
513
514
        st = Grammar(prefixes)('B')
        assert st.error_flag
        prefixes = Interleave(TKN("B"), TKN("A"), repetitions=((0, 1), (0, 1)))
515
        assert Grammar(prefixes)('A B').content == 'A B'
eckhart's avatar
eckhart committed
516
517
518
        st = Grammar(prefixes)('B')
        assert not st.error_flag
        assert st.content == 'B'
519
520
521
522

    def test_someOf_redundance(self):
        """Test that one and the same parser may be listed several times
        and must be matched several times accordingly."""
eckhart's avatar
eckhart committed
523
        prefixes = Interleave(TKN("A"), TKN("B"), TKN("A"))
524
525
526
        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'
527
528
529
        assert Grammar(prefixes)('A B B').error_flag


530
531
532
533
534
class TestInterleave:
    def test_interleave_most_simple(self):
        letterset = Interleave(Token("A"), Token("B"), Token("C"))
        gr = Grammar(letterset)
        st = gr('ABC')
535
        assert not st.errors, str(st.errors)
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
        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
551
552
553
554
555
556
557
558
559
560
        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


561
562
563
564
class TestErrorRecovery:
    def test_series_skip(self):
        lang = """
        document = series | /.*/
565
        @series_skip = /(?=[A-Z])/
566
567
568
569
570
        series = "A" "B" §"C" "D"
        """
        parser = grammar_provider(lang)()
        st = parser('AB_D')
        assert len(st.errors) == 1  # no additional "stopped before end"-error!
571
        resume_notices_on(parser)
572
573
        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
574
        assert 'Skipping' in str(st.errors_sorted[1])
575

576

577
    def test_Interleave_skip(self):
578
579
580
        lang = """
        document = allof | /.*/
        @allof_skip = /[A-Z]/
581
        allof = "A" ° §"B" ° "C" ° "D"
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
        """
        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"
598
599
        st = parser('A_CB')
        assert st['allof'].content == "A_CB"
600
        st = parser('BC_A')
601
        assert 'allof' not in st
602
603


604
class TestPopRetrieve:
605
    mini_language = r"""
606
607
        document       = { text | codeblock }
        codeblock      = delimiter { text | (!:delimiter delimiter_sign) } ::delimiter
608
        delimiter      = delimiter_sign  # never use delimiter between capture and pop except for retrival!
609
        delimiter_sign = /`+/
eckhart's avatar
eckhart committed
610
        text           = /[^`]+/
611
        """
612
    mini_lang2 = r"""
613
        @braces_filter=matching_bracket
614
615
616
617
618
619
620
        document       = { text | codeblock }
        codeblock      = braces { text | opening_braces | (!:braces closing_braces) } ::braces
        braces         = opening_braces
        opening_braces = /\{+/
        closing_braces = /\}+/
        text           = /[^{}]+/
        """
621
    mini_lang3 = r"""
622
623
624
625
        document       = { text | env }
        env            = (specialtag | opentag) text [closespecial | closetag]
        opentag        = "<" name ">"
        specialtag     = "<" /ABC/ !name ">"
eckhart's avatar
eckhart committed
626
        closetag       = close_slash | close_star
627
628
629
630
631
632
        close_slash    = "<" ::name "/>"
        close_star     = "<" ::name "*>"
        closespecial   = "<" /ABC/~ ">"
        name           = /\w+/~
        text           = /[^<>]+/
        """
633
634
635
636
637
638
639
640
    mini_lang4 = r"""
        document       = { text | env }
        env            = opentag document closetag
        opentag        = "<" name ">"
        closetag       = "</" :?name ">"
        name           = /\w+/~
        text           = /[^<>]+/        
    """
641
642

    def setup(self):
643
644
645
        self.minilang_parser = grammar_provider(self.mini_language)()
        self.minilang_parser2 = grammar_provider(self.mini_lang2)()
        self.minilang_parser3 = grammar_provider(self.mini_lang3)()
646
        self.minilang_parser4 = grammar_provider(self.mini_lang4)()
647
648

    @staticmethod
649
    def has_tag_name(node, name):
650
        return node.tag_name == name # and not isinstance(node.parser, Retrieve)
651

652
653
654
655
656
657
658
659
660
661
662
663
664
    def test_capture_assertions(self):
        try:
            _ = Capture(Drop(Whitespace(r'\s*')))
            assert False, "ValueError expected!"
        except ValueError:
            pass
        try:
            _ = Capture(Series(Token(' '), Drop(Whitespace(r'\s*'))))
            assert False, "ValueError expected!"
        except ValueError:
            pass
        _ = Capture(RegExp(r'\w+'))

665
666
667
668
    def test_compile_mini_language(self):
        assert self.minilang_parser
        assert self.minilang_parser2
        assert self.minilang_parser3
669
        assert self.minilang_parser4
670
671
672
673

    def test_stackhandling(self):
        ambigous_opening = "<ABCnormal> normal tag <ABCnormal*>"
        syntax_tree = self.minilang_parser3(ambigous_opening)
674
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
675
676
677

        ambigous_opening = "<ABCnormal> normal tag <ABCnormal/>"
        syntax_tree = self.minilang_parser3(ambigous_opening)
678
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
679
680
681

        forgot_closing_tag = "<em> where is the closing tag?"
        syntax_tree = self.minilang_parser3(forgot_closing_tag)
682
        assert syntax_tree.error_flag, str(syntax_tree.errors_sorted)
683
684
685

        proper = "<em> has closing tag <em/>"
        syntax_tree = self.minilang_parser3(proper)
686
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
687
688
689

        proper = "<em> has closing tag <em*>"
        syntax_tree = self.minilang_parser3(proper)
690
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
691

692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
    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

719
    def test_cache_neutrality(self):
720
721
        """Test that packrat-caching does not interfere with the variable-
        changing parsers: Capture and Retrieve."""
722
        lang = r"""
723
724
725
726
727
728
729
730
731
            text = opening closing
            opening = (unmarked_package | marked_package)
            closing = ::variable
            unmarked_package = package "."
            marked_package = package "*" "."
            package = "(" variable ")"
            variable = /\w+/~
            """
        case = "(secret)*. secret"
732
        gr = grammar_provider(lang)()
733
        st = gr(case)
734
        assert not st.error_flag, str(st.errors_sorted)
735
736
737
738

    def test_single_line(self):
        teststr = "Anfang ```code block `` <- keine Ende-Zeichen ! ``` Ende"
        syntax_tree = self.minilang_parser(teststr)
739
740
        assert not syntax_tree.errors_sorted, \
            ''.join(str(error) for error in 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_single_line.cst")
747
748
749
750
751
752
753
754
755
756
757

    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)
758
        assert not syntax_tree.errors_sorted
759
760
761
        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)))
762
763
        assert delim == pop
        if is_logging():
764
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
765
766
767
768

    def test_single_line_complement(self):
        teststr = "Anfang {{{code block }} <- keine Ende-Zeichen ! }}} Ende"
        syntax_tree = self.minilang_parser2(teststr)
769
        assert not syntax_tree.errors_sorted
770
771
772
773
774
        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
775
        if is_logging():
776
            log_ST(syntax_tree, "test_PopRetrieve_single_line.cst")
777
778
779
780
781
782
783
784
785
786
787

    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)
788
        assert not syntax_tree.errors_sorted
789
790
791
        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)))
792
793
        assert len(delim) == len(pop) and delim != pop
        if is_logging():
794
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
795

796
    def test_autoretrieve(self):
797
798
        lang = r"""
            document   = { definition } § EOF
799
            definition = symbol :defsign value
800
            symbol     = /\w+/~                      
801
802
            defsign    = "=" | ":="
            value      = /\d+/~
803
            EOF        = !/./ [:?defsign]   # eat up captured defsigns
804
        """
805
806
        # code, _, _ = compile_ebnf(lang)
        # print(code)
807
        parser = grammar_provider(lang)()
808
809
810
        st = parser("X := 1")
        assert not st.error_flag
        st1 = st
811
812
        st = parser("")
        assert not st.error_flag
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836

        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+/~
837
            EOF        = !/./ :?defsign   # eat up captured defsign, only if it has been retrieved
838
839
840
841
842
843
            definition = symbol :defsign value
        """
        parser = grammar_provider(lang_variant)()
        st = parser("X := 1")
        assert not st.errors
        assert st.equals(st1)
844
845
        st = parser('')
        assert "EOF expected" in str(st.errors)
di68kap's avatar
di68kap committed
846

847

848
class TestWhitespaceHandling:
Eckhart Arnold's avatar
Eckhart Arnold committed
849
    minilang = """
850
851
852
853
854
855
856
857
858
        doc = A B
        A = "A"
        B = "B"
        Rdoc = ar br
        ar = /A/
        br = /B/
        """

    def setup(self):
859
        self.gr = grammar_provider(self.minilang)()
860
861
862
863
864
865
866
867
868
869
870
871

    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
872
873
874
875
876


class TestErrorReporting:
    grammar = """
        root      = series alpha | anything
eckhart's avatar
eckhart committed
877
        series    = subseries &alpha
di68kap's avatar
di68kap committed
878
879
880
881
882
883
884
885
886
887
888
889
890
891
        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)
892
        assert not cst.error_flag, str(cst.errors_sorted)
di68kap's avatar
di68kap committed
893
894
895
896
        cst = self.parser(testcode2)
        assert not cst.error_flag
        cst = self.parser(testcode3)
        assert cst.error_flag
897
898


di68kap's avatar
di68kap committed
899
900
class TestBorderlineCases:
    def test_not_matching(self):
901
        minilang = """parser = /X/\n"""
di68kap's avatar
di68kap committed
902
903
904
905
        gr = grammar_provider(minilang)()
        cst = gr('X', 'parser')
        assert not cst.error_flag
        cst = gr(' ', 'parser')
906
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
907
        cst = gr('', 'parser')
908
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
909
910
911
912
913
914
915

    def test_matching(self):
        minilang = """parser = /.?/"""
        gr = grammar_provider(minilang)()
        cst = gr(' ', 'parser')
        assert not cst.error_flag
        cst = gr('  ', 'parser')
916
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_STOPPED_BEFORE_END
di68kap's avatar
di68kap committed
917
918
919
920
        cst = gr('', 'parser')
        assert not cst.error_flag


Eckhart Arnold's avatar
Eckhart Arnold committed
921
class TestReentryAfterError:
922
    def setup(self):
Eckhart Arnold's avatar
Eckhart Arnold committed
923
924
925
926
927
        lang = """
        document = alpha [beta] gamma "."
          alpha = "ALPHA" abc
            abc = §"a" "b" "c"
          beta = "BETA" (bac | bca)
928
929
            bac = "b" "a" §"c"
            bca = "b" "c" §"a"
Eckhart Arnold's avatar
Eckhart Arnold committed
930
931
932
933
          gamma = "GAMMA" §(cab | cba)
            cab = "c" "a" §"b"
            cba = "c" "b" §"a"
        """
934
        self.gr = grammar_provider(lang)()
935

936
937
    def test_no_resume_rules(self):
        gr = self.gr;  gr.resume_rules = dict()
938
939
        content = 'ALPHA acb BETA bac GAMMA cab .'
        cst = gr(content)
940
941
942
943
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')

eckhart's avatar
eckhart committed
944
945
946
947
948
949
950
951
    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')

952
953
    def test_simple_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
954
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)')]
955
        content = 'ALPHA acb BETA bac GAMMA cab .'
956
957
958
959
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
960
        # because of resuming, there should be only one error message
Eckhart Arnold's avatar
Eckhart Arnold committed
961
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 1
962

963
964
    def test_failing_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
965
        gr.resume_rules__['alpha'] = [re.compile(r'(?=XXX)')]
966
        content = 'ALPHA acb BETA bac GAMMA cab .'
967
968
969
970
971
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        # assert cst.pick('alpha').content.startswith('ALPHA')

eckhart's avatar
eckhart committed
972
    def test_several_reentry_points(self):
973
        gr = self.gr;  gr.resume_rules = dict()
974
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
975
        content = 'ALPHA acb BETA bac GAMMA cab .'
976
977
978
979
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
980
        # because of resuming, there should be only one error message
Eckhart Arnold's avatar
Eckhart Arnold committed
981
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 1
982

983
984
    def test_several_reentry_points_second_point_matching(self):
        gr = self.gr;  gr.resume_rules = dict()
985
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
986
987
        content = 'ALPHA acb GAMMA cab .'
        cst = gr(content)
Eckhart Arnold's avatar
Eckhart Arnold committed
988
        assert cst.error_flag
989
        assert cst.content == content
990
        assert cst.pick('alpha').content.startswith('ALPHA')
991
        # because of resuming, there should be only one error message
992
        assert len(cst.errors_sorted) == 1
993
994
        resume_notices_on(gr)
        cst = gr(content)
995
        assert len(cst.errors) == 2 and any(err.code == Error.RESUME_NOTICE for err in cst.errors)
996

997
998
    def test_several_resume_rules_innermost_rule_matching(self):
        gr = self.gr;  gr.resume_rules = dict()
999
1000
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
        gr.resume_rules__['beta'] = [re.compile(r'(?=GAMMA)')]