test_parse.py 46.8 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
from typing import List
26

27
28
scriptpath = os.path.dirname(__file__) or '.'
sys.path.append(os.path.abspath(os.path.join(scriptpath, '..')))
29

eckhart's avatar
eckhart committed
30
from DHParser.configuration import get_config_value, set_config_value
31
from DHParser.toolkit import compile_python_object, re
32
from DHParser.log import is_logging, log_ST, log_parsing_history
33
from DHParser.error import Error, is_error
34
from DHParser.parse import ParserError, Parser, Grammar, Forward, TKN, ZeroOrMore, RE, \
eckhart's avatar
eckhart committed
35
    RegExp, Lookbehind, NegativeLookahead, OneOrMore, Series, Alternative, \
36
37
    Interleave, UnknownParserError, MetaParser, Token, EMPTY_NODE, Capture, Drop, Whitespace, \
    GrammarError
38
from DHParser import compile_source
39
40
from DHParser.ebnf import get_ebnf_grammar, get_ebnf_transformer, get_ebnf_compiler, \
    compile_ebnf, DHPARSER_IMPORTS
41
from DHParser.dsl import grammar_provider
42
from DHParser.syntaxtree import Node, parse_sxpr
43
from DHParser.stringview import StringView
44
from DHParser.trace import set_tracer, trace_history, resume_notices_on
45
46


47
48
49
50
51
52

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

Eckhart Arnold's avatar
Eckhart Arnold committed
53
    def test_non_empty_derivation(self):
54
55
        pass

56

57
58
class TestParserError:
    def test_parser_error_str(self):
59
        pe = ParserError(Node('TAG', 'test').with_pos(0), StringView('Beispiel'), None, True)
60
61
        assert str(pe).find('Beispiel') >= 0 and str(pe).find('TAG') >= 0

62
63
64
65
66
67
68
69
    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)()
70
71
        set_tracer(gr, trace_history)
        st = gr('hard-time')
72
        assert not st.errors
73
        st = gr('hard-')
74
75
        assert st.errors and not any(e.code == 1045 for e in st.errors)

eckhart's avatar
eckhart committed
76

eckhart's avatar
eckhart committed
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 = []
86
87
        def visitor(context: List[Parser]):
            p = context[-1]
di68kap's avatar
di68kap committed
88
            l.append(p.pname + p.ptype)
eckhart's avatar
eckhart committed
89
90
91
92
93
94
95
96
97
98
        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

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
107
        assert result == 'word', result
108

eckhart's avatar
eckhart committed
109

110
class TestInfiLoopsAndRecursion:
111
    def test_direct_left_recursion1(self):
eckhart's avatar
eckhart committed
112
        minilang = """
113
114
115
116
            expr = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
117
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
118
        parser = grammar_provider(minilang)()
119
        assert parser
Eckhart Arnold's avatar
Eckhart Arnold committed
120
        syntax_tree = parser(snippet)
121
122
        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
123
        if is_logging():
124
            log_ST(syntax_tree, "test_LeftRecursion_direct.cst")
125
            log_parsing_history(parser, "test_LeftRecursion_direct")
126

127
    def test_direct_left_recursion2(self):
128
129
130
131
132
133
        minilang = """
            expr = ex
            ex   = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
134
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
135
        parser = grammar_provider(minilang)()
136
        assert parser
137
        syntax_tree = parser(snippet)
138
139
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
140

141
    def test_indirect_left_recursion1(self):
142
143
        minilang = """
            Expr    = //~ (Product | Sum | Value)
144
145
            Product = Expr { ('*' | '/') Expr }+
            Sum     = Expr { ('+' | '-') Expr }+
146
            Value   = /[0-9.]+/~ | '(' §Expr ')'
147
            """
148
        parser = grammar_provider(minilang)()
149
        assert parser
150
        snippet = "8 * 4"
151
        syntax_tree = parser(snippet)
152
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
153
154
        snippet = "7 + 8 * 4"
        syntax_tree = parser(snippet)
155
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
156
157
        snippet = "9 + 8 * (4 + 3)"
        syntax_tree = parser(snippet)
158
159
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
160
        if is_logging():
161
            log_ST(syntax_tree, "test_LeftRecursion_indirect.cst")
162
            log_parsing_history(parser, "test_LeftRecursion_indirect")
163

164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
    # # 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
184

185
186
187
    def test_break_inifnite_loop_ZeroOrMore(self):
        forever = ZeroOrMore(RegExp(''))
        result = Grammar(forever)('')  # infinite loops will automatically be broken
188
        assert repr(result) == "Node(':EMPTY', '')", repr(result)
189
190
191
192

    def test_break_inifnite_loop_OneOrMore(self):
        forever = OneOrMore(RegExp(''))
        result = Grammar(forever)('')  # infinite loops will automatically be broken
193
        assert repr(result) == "Node(':EMPTY', '')", repr(result)
194

195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
    # 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)
220

eckhart's avatar
eckhart committed
221

Eckhart Arnold's avatar
Eckhart Arnold committed
222
class TestFlowControl:
223
224
    def setup(self):
        self.t1 = """
eckhart's avatar
eckhart committed
225
        All work and no play
226
227
228
229
230
        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
231
    def test_lookbehind(self):
232
        ws = RegExp(r'\s*')
Eckhart Arnold's avatar
Eckhart Arnold committed
233
        end = RegExp("END")
234
        doc_end = Lookbehind(RegExp('\\s*?\\n')) + end
235
        word = RegExp(r'\w+')
Eckhart Arnold's avatar
Eckhart Arnold committed
236
237
238
239
        sequence = OneOrMore(NegativeLookahead(end) + word + ws)
        document = ws + sequence + doc_end + ws

        parser = Grammar(document)
240
241
242
243
244
245
246
        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
247
            parser_initialization__ = ["upon instantiation"]
248
            ws = RegExp(r'\s*')
249
            end = RegExp('END')
250
            SUCC_LB = RegExp('\\s*?\\n')
251
            doc_end = Series(Lookbehind(SUCC_LB), end)
252
            word = RegExp(r'\w+')
253
254
255
256
257
258
            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
259
        assert not cst.error_flag, cst.as_sxpr()
260
        cst = parser(self.t2)
Eckhart Arnold's avatar
Eckhart Arnold committed
261
262
263
        assert cst.error_flag, cst.as_sxpr()


264
265
266
267
268
269
270
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
271
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
Eckhart Arnold's avatar
Eckhart Arnold committed
272
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
273
        assert result
274
        assert not messages, str(messages)
275
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
eckhart's avatar
eckhart committed
276
277
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
278
        assert node.tag_name == "regex"
279
280
        assert str(node) == 'abc+def'

281
282
    def test_multilineRegex_wo_Comments(self):
        mlregex = r"""
eckhart's avatar
eckhart committed
283
284
        regex =  /\w+
                  [+]
285
286
                  \w* /
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
287
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
288
289
290
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages, str(messages)
291
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
292
293
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
294
        assert node.tag_name == "regex"
295
296
        assert str(node) == 'abc+def'

297
    def test_ignore_case(self):
298
299
300
301
        mlregex = r"""
        @ ignorecase = True
        regex = /alpha/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
302
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
303
304
305
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages
306
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
307
        node, rest = parser.regex(StringView('Alpha'))
308
309
        assert node
        assert rest == ''
310
        assert node.tag_name == "regex"
311
312
313
314
315
316
        assert str(node) == 'Alpha'

        mlregex = r"""
        @ ignorecase = False
        regex = /alpha/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
317
        result, messages, _ = compile_source(mlregex, None, get_ebnf_grammar(),
318
319
320
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages
321
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
322
323
        node, rest = parser.regex(StringView('Alpha'))
        assert node is None
324

325
326
327
328
329
330
331
332
333
334
335
336
    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
337
        result, messages, _ = compile_source(
eckhart's avatar
eckhart committed
338
339
            tokenlang, None, get_ebnf_grammar(), get_ebnf_transformer(),
            get_ebnf_compiler("TokenTest"))
340
        assert result
eckhart's avatar
eckhart committed
341
        assert not messages, str(messages)
342
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
Eckhart Arnold's avatar
Eckhart Arnold committed
343
        result = parser(testdoc)
344
        # log_parsing_history(parser, "test.log")
345
346
        assert not result.error_flag

347

348
class TestGrammar:
349
    def setup(self):
350
351
352
353
354
355
        grammar = r"""@whitespace = horizontal
        haupt        = textzeile LEERZEILE
        textzeile    = { WORT }+
        WORT         = /[^ \t]+/~
        LEERZEILE    = /\n[ \t]*(?=\n)/~
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
356
357
        self.pyparser, messages, _ = compile_source(grammar, None, get_ebnf_grammar(),
                                    get_ebnf_transformer(), get_ebnf_compiler("PosTest"))
358
        assert self.pyparser
359
        assert not messages
360
361
362
363

    def test_pos_values_initialized(self):
        # checks whether pos values in the parsing result and in the
        # history record have been initialized
364
365
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
        grammar("no_file_name*")
366
        for record in grammar.history__:
367
368
            assert not record.node or record.node.pos >= 0

369
    def test_select_parsing(self):
370
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
371
372
373
374
        grammar("wort", "WORT")
        grammar("eine Zeile", "textzeile")
        grammar("kein Haupt", "haupt")
        grammar("so ist es richtig", "haupt")
375

di68kap's avatar
di68kap committed
376
377
    def test_grammar_subclassing(self):
        class Arithmetic(Grammar):
378
            r'''
di68kap's avatar
di68kap committed
379
380
381
382
383
384
            expression =  term  { ("+" | "-") term }
            term       =  factor  { ("*" | "/") factor }
            factor     =  INTEGER | "("  expression  ")"
            INTEGER    =  /\d+/~
            '''
            expression = Forward()
385
386
387
388
            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
389
390
391
392
393
394
            root__ = expression

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

395
396
397
398
399
400
401
402
403
404
405
    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

406
    def test_synonym(self):
di68kap's avatar
di68kap committed
407
        lang = r"""
408
409
410
411
412
413
414
415
416
417
418
419
420
            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')

421

422
423
424
425
426
427
428
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
429
        st = parser("ABCD")
430
        assert not st.error_flag
Eckhart Arnold's avatar
Eckhart Arnold committed
431
        st = parser("A_CD")
432
        assert not st.error_flag
Eckhart Arnold's avatar
Eckhart Arnold committed
433
        st = parser("AB_D")
434
435
436
437
438
439
440
441
442
443
444
        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)()
445
446
447
        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
448
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
449
        # transitivity of mandatory-operator
450
        st = parser("ABC_");  assert st.error_flag
451
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
452
453

    def test_series_composition(self):
454
        TA, TB, TC, TD, TE = (TKN(b) for b in "ABCDE")
455
456
457
458
459
        s1 = Series(TA, TB, TC, mandatory=2)
        s2 = Series(TD, TE)

        combined = Alternative(s1 + s2, RegExp('.*'))
        parser = Grammar(combined)
460
461
462
        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
463
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
464
        st = parser("ABC_E");  assert st.error_flag
465
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
466
467
468

        combined = Alternative(s2 + s1, RegExp('.*'))
        parser = Grammar(combined)
469
470
471
472
473
474
        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
475
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
476

eckhart's avatar
eckhart committed
477
478
479
480
481
482
483
484
485
486
    # 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
487

488
489
    def test_ebnf_serialization(self):
        ebnf_grammar = get_ebnf_grammar()
eckhart's avatar
eckhart committed
490
        # TODO: Add test here
491
        ebnf = ebnf_grammar.as_ebnf()
492
        # print(ebnf)
493
494


495
496
497
class TestAllOfSomeOf:
    def test_allOf_order(self):
        """Test that parsers of an AllOf-List can match in arbitrary order."""
eckhart's avatar
eckhart committed
498
        prefixes = Interleave(TKN("A"), TKN("B"))
499
500
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
501
502
503
504

    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
505
        prefixes = Interleave(TKN("A"), TKN("B"))
506
507
508
509
510
        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
511
        prefixes = Interleave(TKN("A"), TKN("B"), TKN("A"))
512
513
514
        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'
515
516
517
518
        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
519
        prefixes = Interleave(TKN("A"), TKN("B"))
520
521
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
eckhart's avatar
eckhart committed
522
523
524
        st = Grammar(prefixes)('B')
        assert st.error_flag
        prefixes = Interleave(TKN("B"), TKN("A"), repetitions=((0, 1), (0, 1)))
525
        assert Grammar(prefixes)('A B').content == 'A B'
eckhart's avatar
eckhart committed
526
527
528
        st = Grammar(prefixes)('B')
        assert not st.error_flag
        assert st.content == 'B'
529
530
531
532

    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
533
        prefixes = Interleave(TKN("A"), TKN("B"), TKN("A"))
534
535
536
        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'
537
538
539
        assert Grammar(prefixes)('A B B').error_flag


540
541
542
543
544
class TestInterleave:
    def test_interleave_most_simple(self):
        letterset = Interleave(Token("A"), Token("B"), Token("C"))
        gr = Grammar(letterset)
        st = gr('ABC')
545
        assert not st.errors, str(st.errors)
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
        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
561
562
563
564
565
566
567
568
569
570
        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


571
572
573
574
class TestErrorRecovery:
    def test_series_skip(self):
        lang = """
        document = series | /.*/
575
        @series_skip = /(?=[A-Z])/
576
577
578
579
580
        series = "A" "B" §"C" "D"
        """
        parser = grammar_provider(lang)()
        st = parser('AB_D')
        assert len(st.errors) == 1  # no additional "stopped before end"-error!
581
        resume_notices_on(parser)
582
583
        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
584
        assert 'Skipping' in str(st.errors_sorted[1])
585

586
    def test_Interleave_skip(self):
587
588
589
        lang = """
        document = allof | /.*/
        @allof_skip = /[A-Z]/
590
        allof = "A" ° §"B" ° "C" ° "D"
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
        """
        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"
607
608
        st = parser('A_CB')
        assert st['allof'].content == "A_CB"
609
        st = parser('BC_A')
610
        assert 'allof' not in st
611
612


613
class TestPopRetrieve:
614
    mini_language = r"""
615
616
        document       = { text | codeblock }
        codeblock      = delimiter { text | (!:delimiter delimiter_sign) } ::delimiter
617
        delimiter      = delimiter_sign  # never use delimiter between capture and pop except for retrival!
618
        delimiter_sign = /`+/
eckhart's avatar
eckhart committed
619
        text           = /[^`]+/
620
        """
621
    mini_lang2 = r"""
622
        @braces_filter=matching_bracket
623
624
625
626
627
628
629
        document       = { text | codeblock }
        codeblock      = braces { text | opening_braces | (!:braces closing_braces) } ::braces
        braces         = opening_braces
        opening_braces = /\{+/
        closing_braces = /\}+/
        text           = /[^{}]+/
        """
630
    mini_lang3 = r"""
631
632
633
634
        document       = { text | env }
        env            = (specialtag | opentag) text [closespecial | closetag]
        opentag        = "<" name ">"
        specialtag     = "<" /ABC/ !name ">"
eckhart's avatar
eckhart committed
635
        closetag       = close_slash | close_star
636
637
638
639
640
641
        close_slash    = "<" ::name "/>"
        close_star     = "<" ::name "*>"
        closespecial   = "<" /ABC/~ ">"
        name           = /\w+/~
        text           = /[^<>]+/
        """
642
643
644
645
646
647
648
649
    mini_lang4 = r"""
        document       = { text | env }
        env            = opentag document closetag
        opentag        = "<" name ">"
        closetag       = "</" :?name ">"
        name           = /\w+/~
        text           = /[^<>]+/        
    """
650
651

    def setup(self):
652
653
654
        self.minilang_parser = grammar_provider(self.mini_language)()
        self.minilang_parser2 = grammar_provider(self.mini_lang2)()
        self.minilang_parser3 = grammar_provider(self.mini_lang3)()
655
        self.minilang_parser4 = grammar_provider(self.mini_lang4)()
656
657

    @staticmethod
658
    def has_tag_name(node, name):
659
        return node.tag_name == name # and not isinstance(node.parser, Retrieve)
660

661
662
    def test_capture_assertions(self):
        try:
663
664
665
            _ = Grammar(Capture(Drop(Whitespace(r'\s*'))))
            assert False, "GrammarError expected!"
        except GrammarError as ge:
666
667
            pass
        try:
668
            _ = Grammar(Capture(Series(Token(' '), Drop(Whitespace(r'\s*')))))
669
            assert False, "ValueError expected!"
670
        except GrammarError:
671
            pass
672
673
674
        cp = Capture(RegExp(r'\w+'))
        cp.pname = "capture"
        _ = Grammar(cp)
675

676
677
678
679
    def test_compile_mini_language(self):
        assert self.minilang_parser
        assert self.minilang_parser2
        assert self.minilang_parser3
680
        assert self.minilang_parser4
681
682
683
684

    def test_stackhandling(self):
        ambigous_opening = "<ABCnormal> normal tag <ABCnormal*>"
        syntax_tree = self.minilang_parser3(ambigous_opening)
685
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
686
687
688

        ambigous_opening = "<ABCnormal> normal tag <ABCnormal/>"
        syntax_tree = self.minilang_parser3(ambigous_opening)
689
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
690
691
692

        forgot_closing_tag = "<em> where is the closing tag?"
        syntax_tree = self.minilang_parser3(forgot_closing_tag)
693
        assert syntax_tree.error_flag, str(syntax_tree.errors_sorted)
694
695
696

        proper = "<em> has closing tag <em/>"
        syntax_tree = self.minilang_parser3(proper)
697
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
698
699
700

        proper = "<em> has closing tag <em*>"
        syntax_tree = self.minilang_parser3(proper)
701
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
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
728
729
    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

730
    def test_cache_neutrality(self):
731
732
        """Test that packrat-caching does not interfere with the variable-
        changing parsers: Capture and Retrieve."""
733
        lang = r"""
734
735
736
737
738
739
740
741
742
            text = opening closing
            opening = (unmarked_package | marked_package)
            closing = ::variable
            unmarked_package = package "."
            marked_package = package "*" "."
            package = "(" variable ")"
            variable = /\w+/~
            """
        case = "(secret)*. secret"
743
        gr = grammar_provider(lang)()
744
        st = gr(case)
745
        assert not st.error_flag, str(st.errors_sorted)
746
747
748
749

    def test_single_line(self):
        teststr = "Anfang ```code block `` <- keine Ende-Zeichen ! ``` Ende"
        syntax_tree = self.minilang_parser(teststr)
750
751
        assert not syntax_tree.errors_sorted, \
            ''.join(str(error) for error in syntax_tree.errors_sorted)
752
753
754
        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)))
755
756
        assert delim == pop
        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(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)
769
        assert not syntax_tree.errors_sorted
770
771
772
        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)))
773
774
        assert delim == pop
        if is_logging():
775
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
776
777
778
779

    def test_single_line_complement(self):
        teststr = "Anfang {{{code block }} <- keine Ende-Zeichen ! }}} Ende"
        syntax_tree = self.minilang_parser2(teststr)
780
        assert not syntax_tree.errors_sorted
781
782
783
784
785
        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
786
        if is_logging():
787
            log_ST(syntax_tree, "test_PopRetrieve_single_line.cst")
788
789
790
791
792
793
794
795
796
797
798

    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)
799
        assert not syntax_tree.errors_sorted
800
801
802
        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)))
803
804
        assert len(delim) == len(pop) and delim != pop
        if is_logging():
805
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
806

807
    def test_autoretrieve(self):
808
809
        lang = r"""
            document   = { definition } § EOF
810
            definition = symbol :defsign value
811
            symbol     = /\w+/~                      
812
813
            defsign    = "=" | ":="
            value      = /\d+/~
814
            EOF        = !/./ [:?defsign]   # eat up captured defsigns
815
        """
816
817
        # code, _, _ = compile_ebnf(lang)
        # print(code)
818
        parser = grammar_provider(lang)()
819
820
821
        st = parser("X := 1")
        assert not st.error_flag
        st1 = st
822
823
        st = parser("")
        assert not st.error_flag
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847

        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+/~
848
            EOF        = !/./ :?defsign   # eat up captured defsign, only if it has been retrieved
849
850
851
852
853
854
            definition = symbol :defsign value
        """
        parser = grammar_provider(lang_variant)()
        st = parser("X := 1")
        assert not st.errors
        assert st.equals(st1)
855
856
        st = parser('')
        assert "EOF expected" in str(st.errors)
di68kap's avatar
di68kap committed
857

858

859
class TestWhitespaceHandling:
Eckhart Arnold's avatar
Eckhart Arnold committed
860
    minilang = """
861
862
863
864
865
866
867
868
869
        doc = A B
        A = "A"
        B = "B"
        Rdoc = ar br
        ar = /A/
        br = /B/
        """

    def setup(self):
870
        self.gr = grammar_provider(self.minilang)()
871
872
873
874
875
876
877
878
879
880
881
882

    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
883
884
885
886
887


class TestErrorReporting:
    grammar = """
        root      = series alpha | anything
eckhart's avatar
eckhart committed
888
        series    = subseries &alpha
di68kap's avatar
di68kap committed
889
890
891
892
893
894
895
896
897
898
899
900
901
902
        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)
903
        assert not cst.error_flag, str(cst.errors_sorted)
di68kap's avatar
di68kap committed
904
905
906
907
        cst = self.parser(testcode2)
        assert not cst.error_flag
        cst = self.parser(testcode3)
        assert cst.error_flag
908
909


di68kap's avatar
di68kap committed
910
911
class TestBorderlineCases:
    def test_not_matching(self):
912
        minilang = """parser = /X/\n"""
di68kap's avatar
di68kap committed
913
914
915
916
        gr = grammar_provider(minilang)()
        cst = gr('X', 'parser')
        assert not cst.error_flag
        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
        cst = gr('', 'parser')
919
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
920
921
922
923
924
925
926

    def test_matching(self):
        minilang = """parser = /.?/"""
        gr = grammar_provider(minilang)()
        cst = gr(' ', 'parser')
        assert not cst.error_flag
        cst = gr('  ', 'parser')
927
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_STOPPED_BEFORE_END
di68kap's avatar
di68kap committed
928
929
930
931
        cst = gr('', 'parser')
        assert not cst.error_flag


Eckhart Arnold's avatar
Eckhart Arnold committed
932
class TestReentryAfterError:
933
    def setup(self):
Eckhart Arnold's avatar
Eckhart Arnold committed
934
935
936
937
938
        lang = """
        document = alpha [beta] gamma "."
          alpha = "ALPHA" abc
            abc = §"a" "b" "c"
          beta = "BETA" (bac | bca)
939
940
            bac = "b" "a" §"c"
            bca = "b" "c" §"a"
Eckhart Arnold's avatar
Eckhart Arnold committed
941
942
943
944
          gamma = "GAMMA" §(cab | cba)
            cab = "c" "a" §"b"
            cba = "c" "b" §"a"
        """
945
        self.gr = grammar_provider(lang)()
946

947
948
    def test_no_resume_rules(self):
        gr = self.gr;  gr.resume_rules = dict()
949
950
        content = 'ALPHA acb BETA bac GAMMA cab .'
        cst = gr(content)
951
952
953
954
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')

eckhart's avatar
eckhart committed
955
956
957
958
959
960
961
962
    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')

963
964
    def test_simple_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
965
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)')]
966
        content = 'ALPHA acb BETA bac GAMMA cab .'
967
968
969
970
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
971
        # because of resuming, there should be only one error message
Eckhart Arnold's avatar
Eckhart Arnold committed
972
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 1
973

974
975
    def test_failing_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
976
        gr.resume_rules__['alpha'] = [re.compile(r'(?=XXX)')]
977
        content = 'ALPHA acb BETA bac GAMMA cab .'
978
979
980
981
982
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        # assert cst.pick('alpha').content.startswith('ALPHA')

eckhart's avatar
eckhart committed
983
    def test_several_reentry_points(self):
984
        gr = self.gr;  gr.resume_rules = dict()
985
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
986
        content = 'ALPHA acb BETA bac GAMMA cab .'
987
988
989
990
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
991
        # because of resuming, there should be only one error message
Eckhart Arnold's avatar
Eckhart Arnold committed
992
        assert len([err for err in cst.errors_sorted if err.code >= 1000]) == 1
993

994
995
    def test_several_reentry_points_second_point_matching(self):
        gr = self.gr;  gr.resume_rules = dict()
996
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
997
998
        content = 'ALPHA acb GAMMA cab .'
        cst = gr(content)
Eckhart Arnold's avatar
Eckhart Arnold committed
999
        assert cst.error_flag
1000
        assert cst.content == content