test_parse.py 46.7 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
        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
98
99
100
101
102
103
104
105
106
    def test_symbol(self):
        class MyGrammar(Grammar):
            wrong = Token('wrong')
            word = OneOrMore(wrong) + Whitespace(r'\s*') + OneOrMore(RegExp(r'\w+'))
            root__ = word
        gr = MyGrammar()
        regex = gr['word'].parsers[-1].parser
        result = gr.associated_symbol(regex).symbol
        # assert result == 'word', result

eckhart's avatar
eckhart committed
107

108
class TestInfiLoopsAndRecursion:
109
    def test_direct_left_recursion1(self):
eckhart's avatar
eckhart committed
110
        minilang = """
111
112
113
114
            expr = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
115
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
116
        parser = grammar_provider(minilang)()
117
        assert parser
Eckhart Arnold's avatar
Eckhart Arnold committed
118
        syntax_tree = parser(snippet)
119
120
        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
121
        if is_logging():
122
            log_ST(syntax_tree, "test_LeftRecursion_direct.cst")
123
            log_parsing_history(parser, "test_LeftRecursion_direct")
124

125
    def test_direct_left_recursion2(self):
126
127
128
129
130
131
        minilang = """
            expr = ex
            ex   = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
132
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
133
        parser = grammar_provider(minilang)()
134
        assert parser
135
        syntax_tree = parser(snippet)
136
137
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
138

139
    def test_indirect_left_recursion1(self):
140
141
        minilang = """
            Expr    = //~ (Product | Sum | Value)
142
143
            Product = Expr { ('*' | '/') Expr }+
            Sum     = Expr { ('+' | '-') Expr }+
144
            Value   = /[0-9.]+/~ | '(' §Expr ')'
145
            """
146
        parser = grammar_provider(minilang)()
147
        assert parser
148
        snippet = "8 * 4"
149
        syntax_tree = parser(snippet)
150
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
151
152
        snippet = "7 + 8 * 4"
        syntax_tree = parser(snippet)
153
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
154
155
        snippet = "9 + 8 * (4 + 3)"
        syntax_tree = parser(snippet)
156
157
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
158
        if is_logging():
159
            log_ST(syntax_tree, "test_LeftRecursion_indirect.cst")
160
            log_parsing_history(parser, "test_LeftRecursion_indirect")
161

162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
    # # 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
182

183
184
185
    def test_break_inifnite_loop_ZeroOrMore(self):
        forever = ZeroOrMore(RegExp(''))
        result = Grammar(forever)('')  # infinite loops will automatically be broken
186
        assert repr(result) == "Node(':EMPTY', '')", repr(result)
187
188
189
190

    def test_break_inifnite_loop_OneOrMore(self):
        forever = OneOrMore(RegExp(''))
        result = Grammar(forever)('')  # infinite loops will automatically be broken
191
        assert repr(result) == "Node(':EMPTY', '')", repr(result)
192

193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
    # 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)
218

eckhart's avatar
eckhart committed
219

Eckhart Arnold's avatar
Eckhart Arnold committed
220
class TestFlowControl:
221
222
    def setup(self):
        self.t1 = """
eckhart's avatar
eckhart committed
223
        All work and no play
224
225
226
227
228
        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
229
    def test_lookbehind(self):
230
        ws = RegExp(r'\s*')
Eckhart Arnold's avatar
Eckhart Arnold committed
231
        end = RegExp("END")
232
        doc_end = Lookbehind(RegExp('\\s*?\\n')) + end
233
        word = RegExp(r'\w+')
Eckhart Arnold's avatar
Eckhart Arnold committed
234
235
236
237
        sequence = OneOrMore(NegativeLookahead(end) + word + ws)
        document = ws + sequence + doc_end + ws

        parser = Grammar(document)
238
239
240
241
242
243
244
        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
245
            parser_initialization__ = ["upon instantiation"]
246
            ws = RegExp(r'\s*')
247
            end = RegExp('END')
248
            SUCC_LB = RegExp('\\s*?\\n')
249
            doc_end = Series(Lookbehind(SUCC_LB), end)
250
            word = RegExp(r'\w+')
251
252
253
254
255
256
            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
257
        assert not cst.error_flag, cst.as_sxpr()
258
        cst = parser(self.t2)
Eckhart Arnold's avatar
Eckhart Arnold committed
259
260
261
        assert cst.error_flag, cst.as_sxpr()


262
263
264
265
266
267
268
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
269
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
Eckhart Arnold's avatar
Eckhart Arnold committed
270
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
271
        assert result
272
        assert not messages, str(messages)
273
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
eckhart's avatar
eckhart committed
274
275
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
276
        assert node.tag_name == "regex"
277
278
        assert str(node) == 'abc+def'

279
280
    def test_multilineRegex_wo_Comments(self):
        mlregex = r"""
eckhart's avatar
eckhart committed
281
282
        regex =  /\w+
                  [+]
283
284
                  \w* /
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
285
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
286
287
288
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages, str(messages)
289
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
290
291
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
292
        assert node.tag_name == "regex"
293
294
        assert str(node) == 'abc+def'

295
    def test_ignore_case(self):
296
297
298
299
        mlregex = r"""
        @ ignorecase = True
        regex = /alpha/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
300
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
301
302
303
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages
304
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
305
        node, rest = parser.regex(StringView('Alpha'))
306
307
        assert node
        assert rest == ''
308
        assert node.tag_name == "regex"
309
310
311
312
313
314
        assert str(node) == 'Alpha'

        mlregex = r"""
        @ ignorecase = False
        regex = /alpha/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
315
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
316
317
318
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages
319
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
320
321
        node, rest = parser.regex(StringView('Alpha'))
        assert node is None
322

323
324
325
326
327
328
329
330
331
332
333
334
    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
335
        result, messages, _ = compile_source(
eckhart's avatar
eckhart committed
336
337
            tokenlang, None, get_ebnf_grammar(), get_ebnf_transformer(),
            get_ebnf_compiler("TokenTest"))
338
        assert result
eckhart's avatar
eckhart committed
339
        assert not messages, str(messages)
340
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
Eckhart Arnold's avatar
Eckhart Arnold committed
341
        result = parser(testdoc)
342
        # log_parsing_history(parser, "test.log")
343
344
        assert not result.error_flag

345

346
class TestGrammar:
347
    def setup(self):
348
349
350
351
352
353
        grammar = r"""@whitespace = horizontal
        haupt        = textzeile LEERZEILE
        textzeile    = { WORT }+
        WORT         = /[^ \t]+/~
        LEERZEILE    = /\n[ \t]*(?=\n)/~
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
354
355
        self.pyparser, messages, _ = compile_source(grammar, None, get_ebnf_grammar(),
                                    get_ebnf_transformer(), get_ebnf_compiler("PosTest"))
356
        assert self.pyparser
357
        assert not messages
358
359
360
361

    def test_pos_values_initialized(self):
        # checks whether pos values in the parsing result and in the
        # history record have been initialized
362
363
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
        grammar("no_file_name*")
364
        for record in grammar.history__:
365
366
            assert not record.node or record.node.pos >= 0

367
    def test_select_parsing(self):
368
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
369
370
371
372
        grammar("wort", "WORT")
        grammar("eine Zeile", "textzeile")
        grammar("kein Haupt", "haupt")
        grammar("so ist es richtig", "haupt")
373

di68kap's avatar
di68kap committed
374
375
    def test_grammar_subclassing(self):
        class Arithmetic(Grammar):
376
            r'''
di68kap's avatar
di68kap committed
377
378
379
380
381
382
            expression =  term  { ("+" | "-") term }
            term       =  factor  { ("*" | "/") factor }
            factor     =  INTEGER | "("  expression  ")"
            INTEGER    =  /\d+/~
            '''
            expression = Forward()
383
384
385
386
            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
387
388
389
390
391
392
            root__ = expression

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

393
394
395
396
397
398
399
400
401
402
403
    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

404
    def test_synonym(self):
di68kap's avatar
di68kap committed
405
        lang = r"""
406
407
408
409
410
411
412
413
414
415
416
417
418
            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')

419

420
421
422
423
424
425
426
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
427
        st = parser("ABCD")
428
        assert not st.error_flag
Eckhart Arnold's avatar
Eckhart Arnold committed
429
        st = parser("A_CD")
430
        assert not st.error_flag
Eckhart Arnold's avatar
Eckhart Arnold committed
431
        st = parser("AB_D")
432
433
434
435
436
437
438
439
440
441
442
        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)()
443
444
445
        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
446
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
447
        # transitivity of mandatory-operator
448
        st = parser("ABC_");  assert st.error_flag
449
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
450
451

    def test_series_composition(self):
452
        TA, TB, TC, TD, TE = (TKN(b) for b in "ABCDE")
453
454
455
456
457
        s1 = Series(TA, TB, TC, mandatory=2)
        s2 = Series(TD, TE)

        combined = Alternative(s1 + s2, RegExp('.*'))
        parser = Grammar(combined)
458
459
460
        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
461
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
462
        st = parser("ABC_E");  assert st.error_flag
463
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
464
465
466

        combined = Alternative(s2 + s1, RegExp('.*'))
        parser = Grammar(combined)
467
468
469
470
471
472
        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
473
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
474

eckhart's avatar
eckhart committed
475
476
477
478
479
480
481
482
483
484
    # 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
485

486
487
    def test_ebnf_serialization(self):
        ebnf_grammar = get_ebnf_grammar()
eckhart's avatar
eckhart committed
488
        # TODO: Add test here
489
        ebnf = ebnf_grammar.as_ebnf()
490
        # print(ebnf)
491
492


493
494
495
class TestAllOfSomeOf:
    def test_allOf_order(self):
        """Test that parsers of an AllOf-List can match in arbitrary order."""
eckhart's avatar
eckhart committed
496
        prefixes = Interleave(TKN("A"), TKN("B"))
497
498
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
499
500
501
502

    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
503
        prefixes = Interleave(TKN("A"), TKN("B"))
504
505
506
507
508
        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
509
        prefixes = Interleave(TKN("A"), TKN("B"), TKN("A"))
510
511
512
        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'
513
514
515
516
        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
517
        prefixes = Interleave(TKN("A"), TKN("B"))
518
519
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
eckhart's avatar
eckhart committed
520
521
522
        st = Grammar(prefixes)('B')
        assert st.error_flag
        prefixes = Interleave(TKN("B"), TKN("A"), repetitions=((0, 1), (0, 1)))
523
        assert Grammar(prefixes)('A B').content == 'A B'
eckhart's avatar
eckhart committed
524
525
526
        st = Grammar(prefixes)('B')
        assert not st.error_flag
        assert st.content == 'B'
527
528
529
530

    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
531
        prefixes = Interleave(TKN("A"), TKN("B"), TKN("A"))
532
533
534
        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'
535
536
537
        assert Grammar(prefixes)('A B B').error_flag


538
539
540
541
542
class TestInterleave:
    def test_interleave_most_simple(self):
        letterset = Interleave(Token("A"), Token("B"), Token("C"))
        gr = Grammar(letterset)
        st = gr('ABC')
543
        assert not st.errors, str(st.errors)
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
        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
559
560
561
562
563
564
565
566
567
568
        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


569
570
571
572
class TestErrorRecovery:
    def test_series_skip(self):
        lang = """
        document = series | /.*/
573
        @series_skip = /(?=[A-Z])/
574
575
576
577
578
        series = "A" "B" §"C" "D"
        """
        parser = grammar_provider(lang)()
        st = parser('AB_D')
        assert len(st.errors) == 1  # no additional "stopped before end"-error!
579
        resume_notices_on(parser)
580
581
        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
582
        assert 'Skipping' in str(st.errors_sorted[1])
583

584
    def test_Interleave_skip(self):
585
586
587
        lang = """
        document = allof | /.*/
        @allof_skip = /[A-Z]/
588
        allof = "A" ° §"B" ° "C" ° "D"
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
        """
        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"
605
606
        st = parser('A_CB')
        assert st['allof'].content == "A_CB"
607
        st = parser('BC_A')
608
        assert 'allof' not in st
609
610


611
class TestPopRetrieve:
612
    mini_language = r"""
613
614
        document       = { text | codeblock }
        codeblock      = delimiter { text | (!:delimiter delimiter_sign) } ::delimiter
615
        delimiter      = delimiter_sign  # never use delimiter between capture and pop except for retrival!
616
        delimiter_sign = /`+/
eckhart's avatar
eckhart committed
617
        text           = /[^`]+/
618
        """
619
    mini_lang2 = r"""
620
        @braces_filter=matching_bracket
621
622
623
624
625
626
627
        document       = { text | codeblock }
        codeblock      = braces { text | opening_braces | (!:braces closing_braces) } ::braces
        braces         = opening_braces
        opening_braces = /\{+/
        closing_braces = /\}+/
        text           = /[^{}]+/
        """
628
    mini_lang3 = r"""
629
630
631
632
        document       = { text | env }
        env            = (specialtag | opentag) text [closespecial | closetag]
        opentag        = "<" name ">"
        specialtag     = "<" /ABC/ !name ">"
eckhart's avatar
eckhart committed
633
        closetag       = close_slash | close_star
634
635
636
637
638
639
        close_slash    = "<" ::name "/>"
        close_star     = "<" ::name "*>"
        closespecial   = "<" /ABC/~ ">"
        name           = /\w+/~
        text           = /[^<>]+/
        """
640
641
642
643
644
645
646
647
    mini_lang4 = r"""
        document       = { text | env }
        env            = opentag document closetag
        opentag        = "<" name ">"
        closetag       = "</" :?name ">"
        name           = /\w+/~
        text           = /[^<>]+/        
    """
648
649

    def setup(self):
650
651
652
        self.minilang_parser = grammar_provider(self.mini_language)()
        self.minilang_parser2 = grammar_provider(self.mini_lang2)()
        self.minilang_parser3 = grammar_provider(self.mini_lang3)()
653
        self.minilang_parser4 = grammar_provider(self.mini_lang4)()
654
655

    @staticmethod
656
    def has_tag_name(node, name):
657
        return node.tag_name == name # and not isinstance(node.parser, Retrieve)
658

659
660
    def test_capture_assertions(self):
        try:
661
662
663
            _ = Grammar(Capture(Drop(Whitespace(r'\s*'))))
            assert False, "GrammarError expected!"
        except GrammarError as ge:
664
665
            pass
        try:
666
            _ = Grammar(Capture(Series(Token(' '), Drop(Whitespace(r'\s*')))))
667
            assert False, "ValueError expected!"
668
        except GrammarError:
669
            pass
670
671
672
        cp = Capture(RegExp(r'\w+'))
        cp.pname = "capture"
        _ = Grammar(cp)
673

674
675
676
677
    def test_compile_mini_language(self):
        assert self.minilang_parser
        assert self.minilang_parser2
        assert self.minilang_parser3
678
        assert self.minilang_parser4
679
680
681
682

    def test_stackhandling(self):
        ambigous_opening = "<ABCnormal> normal tag <ABCnormal*>"
        syntax_tree = self.minilang_parser3(ambigous_opening)
683
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
684
685
686

        ambigous_opening = "<ABCnormal> normal tag <ABCnormal/>"
        syntax_tree = self.minilang_parser3(ambigous_opening)
687
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
688
689
690

        forgot_closing_tag = "<em> where is the closing tag?"
        syntax_tree = self.minilang_parser3(forgot_closing_tag)
691
        assert syntax_tree.error_flag, str(syntax_tree.errors_sorted)
692
693
694

        proper = "<em> has closing tag <em/>"
        syntax_tree = self.minilang_parser3(proper)
695
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
696
697
698

        proper = "<em> has closing tag <em*>"
        syntax_tree = self.minilang_parser3(proper)
699
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
700

701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
    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

728
    def test_cache_neutrality(self):
729
730
        """Test that packrat-caching does not interfere with the variable-
        changing parsers: Capture and Retrieve."""
731
        lang = r"""
732
733
734
735
736
737
738
739
740
            text = opening closing
            opening = (unmarked_package | marked_package)
            closing = ::variable
            unmarked_package = package "."
            marked_package = package "*" "."
            package = "(" variable ")"
            variable = /\w+/~
            """
        case = "(secret)*. secret"
741
        gr = grammar_provider(lang)()
742
        st = gr(case)
743
        assert not st.error_flag, str(st.errors_sorted)
744
745
746
747

    def test_single_line(self):
        teststr = "Anfang ```code block `` <- keine Ende-Zeichen ! ``` Ende"
        syntax_tree = self.minilang_parser(teststr)
748
749
        assert not syntax_tree.errors_sorted, \
            ''.join(str(error) for error in syntax_tree.errors_sorted)
750
751
752
        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)))
753
754
        assert delim == pop
        if is_logging():
755
            log_ST(syntax_tree, "test_PopRetrieve_single_line.cst")
756
757
758
759
760
761
762
763
764
765
766

    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)
767
        assert not syntax_tree.errors_sorted
768
769
770
        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)))
771
772
        assert delim == pop
        if is_logging():
773
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
774
775
776
777

    def test_single_line_complement(self):
        teststr = "Anfang {{{code block }} <- keine Ende-Zeichen ! }}} Ende"
        syntax_tree = self.minilang_parser2(teststr)
778
        assert not syntax_tree.errors_sorted
779
780
781
782
783
        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
784
        if is_logging():
785
            log_ST(syntax_tree, "test_PopRetrieve_single_line.cst")
786
787
788
789
790
791
792
793
794
795
796

    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)
797
        assert not syntax_tree.errors_sorted
798
799
800
        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)))
801
802
        assert len(delim) == len(pop) and delim != pop
        if is_logging():
803
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
804

805
    def test_autoretrieve(self):
806
807
        lang = r"""
            document   = { definition } § EOF
808
            definition = symbol :defsign value
809
            symbol     = /\w+/~                      
810
811
            defsign    = "=" | ":="
            value      = /\d+/~
812
            EOF        = !/./ [:?defsign]   # eat up captured defsigns
813
        """
814
815
        # code, _, _ = compile_ebnf(lang)
        # print(code)
816
        parser = grammar_provider(lang)()
817
818
819
        st = parser("X := 1")
        assert not st.error_flag
        st1 = st
820
821
        st = parser("")
        assert not st.error_flag
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845

        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+/~
846
            EOF        = !/./ :?defsign   # eat up captured defsign, only if it has been retrieved
847
848
849
850
851
852
            definition = symbol :defsign value
        """
        parser = grammar_provider(lang_variant)()
        st = parser("X := 1")
        assert not st.errors
        assert st.equals(st1)
853
854
        st = parser('')
        assert "EOF expected" in str(st.errors)
di68kap's avatar
di68kap committed
855

856

857
class TestWhitespaceHandling:
Eckhart Arnold's avatar
Eckhart Arnold committed
858
    minilang = """
859
860
861
862
863
864
865
866
867
        doc = A B
        A = "A"
        B = "B"
        Rdoc = ar br
        ar = /A/
        br = /B/
        """

    def setup(self):
868
        self.gr = grammar_provider(self.minilang)()
869
870
871
872
873
874
875
876
877
878
879
880

    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
881
882
883
884
885


class TestErrorReporting:
    grammar = """
        root      = series alpha | anything
eckhart's avatar
eckhart committed
886
        series    = subseries &alpha
di68kap's avatar
di68kap committed
887
888
889
890
891
892
893
894
895
896
897
898
899
900
        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)
901
        assert not cst.error_flag, str(cst.errors_sorted)
di68kap's avatar
di68kap committed
902
903
904
905
        cst = self.parser(testcode2)
        assert not cst.error_flag
        cst = self.parser(testcode3)
        assert cst.error_flag
906
907


di68kap's avatar
di68kap committed
908
909
class TestBorderlineCases:
    def test_not_matching(self):
910
        minilang = """parser = /X/\n"""
di68kap's avatar
di68kap committed
911
912
913
914
        gr = grammar_provider(minilang)()
        cst = gr('X', 'parser')
        assert not cst.error_flag
        cst = gr(' ', 'parser')
915
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
916
        cst = gr('', 'parser')
917
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
918
919
920
921
922
923
924

    def test_matching(self):
        minilang = """parser = /.?/"""
        gr = grammar_provider(minilang)()
        cst = gr(' ', 'parser')
        assert not cst.error_flag
        cst = gr('  ', 'parser')
925
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_STOPPED_BEFORE_END
di68kap's avatar
di68kap committed
926
927
928
929
        cst = gr('', 'parser')
        assert not cst.error_flag


Eckhart Arnold's avatar
Eckhart Arnold committed
930
class TestReentryAfterError:
931
    def setup(self):
Eckhart Arnold's avatar
Eckhart Arnold committed
932
933
934
935
936
        lang = """
        document = alpha [beta] gamma "."
          alpha = "ALPHA" abc
            abc = §"a" "b" "c"
          beta = "BETA" (bac | bca)
937
938
            bac = "b" "a" §"c"
            bca = "b" "c" §"a"
Eckhart Arnold's avatar
Eckhart Arnold committed
939
940
941
942
          gamma = "GAMMA" §(cab | cba)
            cab = "c" "a" §"b"
            cba = "c" "b" §"a"
        """
943
        self.gr = grammar_provider(lang)()
944

945
946
    def test_no_resume_rules(self):
        gr = self.gr;  gr.resume_rules = dict()
947
948
        content = 'ALPHA acb BETA bac GAMMA cab .'
        cst = gr(content)
949
950
951
952
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')

eckhart's avatar
eckhart committed
953
954
955
956
957
958
959
960
    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')

961
962
    def test_simple_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
963
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)')]
964
        content = 'ALPHA acb BETA bac GAMMA cab .'
965
966
967
968
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
969
        # because of resuming, there should be only one error message
Eckhart Arnold's avatar
Eckhart Arnold committed
970
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 1
971

972
973
    def test_failing_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
974
        gr.resume_rules__['alpha'] = [re.compile(r'(?=XXX)')]
975
        content = 'ALPHA acb BETA bac GAMMA cab .'
976
977
978
979
980
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        # assert cst.pick('alpha').content.startswith('ALPHA')

eckhart's avatar
eckhart committed
981
    def test_several_reentry_points(self):
982
        gr = self.gr;  gr.resume_rules = dict()
983
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
984
        content = 'ALPHA acb BETA bac GAMMA cab .'
985
986
987
988
        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
    def test_several_reentry_points_second_point_matching(self):
        gr = self.gr;  gr.resume_rules = dict()
994
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
995
996
        content = 'ALPHA acb GAMMA cab .'
        cst = gr(content)
Eckhart Arnold's avatar
Eckhart Arnold committed
997
        assert cst.error_flag
998
        assert cst.content == content
999
        assert cst.pick('alpha').content.startswith('ALPHA')
1000
        # because of resuming, there should be only one error message