test_parse.py 38.5 KB
Newer Older
1
2
#!/usr/bin/python3

3
"""test_parse.py - tests of the parsers-module of DHParser
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

Author: Eckhart Arnold <arnold@badw.de>

Copyright 2017 Bavarian Academy of Sciences and Humanities

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

    http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
"""

22
import os
23
import sys
24
from functools import partial
25

26
27
scriptpath = os.path.dirname(__file__) or '.'
sys.path.append(os.path.abspath(os.path.join(scriptpath, '..')))
28

eckhart's avatar
eckhart committed
29
from DHParser.configuration import get_config_value, set_config_value
30
from DHParser.toolkit import compile_python_object, re
31
from DHParser.log import is_logging, log_ST, log_parsing_history
32
from DHParser.error import Error, is_error
33
from DHParser.parse import ParserError, Parser, Grammar, Forward, TKN, ZeroOrMore, RE, \
eckhart's avatar
eckhart committed
34
    RegExp, Lookbehind, NegativeLookahead, OneOrMore, Series, Alternative, AllOf, SomeOf, \
35
    UnknownParserError, MetaParser, GrammarError, EMPTY_NODE
36
from DHParser import compile_source
37
38
from DHParser.ebnf import get_ebnf_grammar, get_ebnf_transformer, get_ebnf_compiler, DHPARSER_IMPORTS
from DHParser.dsl import grammar_provider, CompilationError
39
from DHParser.syntaxtree import Node, parse_sxpr
40
from DHParser.stringview import StringView
41
42


43
44
class TestParserError:
    def test_parser_error_str(self):
45
        pe = ParserError(Node('TAG', 'test').with_pos(0), StringView('Beispiel'), None, True)
46
47
        assert str(pe).find('Beispiel') >= 0 and str(pe).find('TAG') >= 0

48
49
50
51
52
53
54
55
56
57
58
59
60
    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)()
        st = gr('hard-time', track_history=True)
        assert not st.errors
        st = gr('hard-', track_history=True)
        assert st.errors and not any(e.code == 1045 for e in st.errors)

eckhart's avatar
eckhart committed
61

eckhart's avatar
eckhart committed
62
63
64
65
66
67
68
69
70
71
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
72
            l.append(p.pname + p.ptype)
eckhart's avatar
eckhart committed
73
74
75
76
77
78
79
80
81
82
83
        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


84
class TestInfiLoopsAndRecursion:
85
    def test_direct_left_recursion1(self):
eckhart's avatar
eckhart committed
86
        minilang = """
87
88
89
90
            expr = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
91
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
92
        parser = grammar_provider(minilang)()
93
        assert parser
Eckhart Arnold's avatar
Eckhart Arnold committed
94
        syntax_tree = parser(snippet)
95
96
        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
97
        if is_logging():
98
            log_ST(syntax_tree, "test_LeftRecursion_direct.cst")
99
            log_parsing_history(parser, "test_LeftRecursion_direct")
100

101
    def test_direct_left_recursion2(self):
102
103
104
105
106
107
        minilang = """
            expr = ex
            ex   = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
108
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
109
        parser = grammar_provider(minilang)()
110
        assert parser
111
        syntax_tree = parser(snippet)
112
113
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
114

115
    def test_indirect_left_recursion1(self):
116
117
        minilang = """
            Expr    = //~ (Product | Sum | Value)
118
119
            Product = Expr { ('*' | '/') Expr }+
            Sum     = Expr { ('+' | '-') Expr }+
120
            Value   = /[0-9.]+/~ | '(' §Expr ')'
121
            """
122
        parser = grammar_provider(minilang)()
123
        assert parser
124
        snippet = "8 * 4"
125
        syntax_tree = parser(snippet)
126
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
127
128
        snippet = "7 + 8 * 4"
        syntax_tree = parser(snippet)
129
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
130
131
        snippet = "9 + 8 * (4 + 3)"
        syntax_tree = parser(snippet)
132
133
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
134
        if is_logging():
135
            log_ST(syntax_tree, "test_LeftRecursion_indirect.cst")
136
            log_parsing_history(parser, "test_LeftRecursion_indirect")
137

138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
    # # 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
158

159
160
161
162
163
164
165
166
167
168
    def test_break_inifnite_loop_ZeroOrMore(self):
        forever = ZeroOrMore(RegExp(''))
        result = Grammar(forever)('')  # infinite loops will automatically be broken
        assert repr(result) == "Node(:EMPTY__, )", repr(result)

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

169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
    # 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)
194

eckhart's avatar
eckhart committed
195

Eckhart Arnold's avatar
Eckhart Arnold committed
196
class TestFlowControl:
197
198
    def setup(self):
        self.t1 = """
eckhart's avatar
eckhart committed
199
        All work and no play
200
201
202
203
204
        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
205
    def test_lookbehind(self):
206
        ws = RegExp(r'\s*')
Eckhart Arnold's avatar
Eckhart Arnold committed
207
        end = RegExp("END")
208
        doc_end = Lookbehind(RegExp('\\s*?\\n')) + end
209
        word = RegExp(r'\w+')
Eckhart Arnold's avatar
Eckhart Arnold committed
210
211
212
213
        sequence = OneOrMore(NegativeLookahead(end) + word + ws)
        document = ws + sequence + doc_end + ws

        parser = Grammar(document)
214
215
216
217
218
219
220
        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
221
            parser_initialization__ = ["upon instantiation"]
222
            ws = RegExp(r'\s*')
223
            end = RegExp('END')
224
            SUCC_LB = RegExp('\\s*?\\n')
225
            doc_end = Series(Lookbehind(SUCC_LB), end)
226
            word = RegExp(r'\w+')
227
228
229
230
231
232
            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
233
        assert not cst.error_flag, cst.as_sxpr()
234
        cst = parser(self.t2)
Eckhart Arnold's avatar
Eckhart Arnold committed
235
236
237
        assert cst.error_flag, cst.as_sxpr()


238
239
240
241
242
243
244
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
245
246
        result, messages, syntax_tree = compile_source(mlregex, None, get_ebnf_grammar(),
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
247
        assert result
248
        assert not messages, str(messages)
249
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
eckhart's avatar
eckhart committed
250
251
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
252
        assert node.tag_name == "regex"
253
254
        assert str(node) == 'abc+def'

255
256
    def test_multilineRegex_wo_Comments(self):
        mlregex = r"""
eckhart's avatar
eckhart committed
257
258
        regex =  /\w+
                  [+]
259
260
261
262
263
264
                  \w* /
        """
        result, messages, syntax_tree = compile_source(mlregex, None, get_ebnf_grammar(),
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages, str(messages)
265
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
266
267
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
268
        assert node.tag_name == "regex"
269
270
        assert str(node) == 'abc+def'

271
272
273
274
275
276
277
278
279
    def text_ignore_case(self):
        mlregex = r"""
        @ ignorecase = True
        regex = /alpha/
        """
        result, messages, syntax_tree = compile_source(mlregex, None, get_ebnf_grammar(),
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages
280
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
281
282
283
284
        node, rest = parser.regex('Alpha')
        assert node
        assert not node.error_flag
        assert rest == ''
285
        assert node.tag_name == "regex"
286
287
288
289
290
291
292
293
294
295
        assert str(node) == 'Alpha'

        mlregex = r"""
        @ ignorecase = False
        regex = /alpha/
        """
        result, messages, syntax_tree = compile_source(mlregex, None, get_ebnf_grammar(),
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
        assert result
        assert not messages
296
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
297
298
299
300
        node, rest = parser.regex('Alpha')
        assert node.error_flag


301
302
303
304
305
306
307
308
309
310
311
312
    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's avatar
eckhart committed
313
314
315
        result, messages, syntax_tree = compile_source(
            tokenlang, None, get_ebnf_grammar(), get_ebnf_transformer(),
            get_ebnf_compiler("TokenTest"))
316
        assert result
eckhart's avatar
eckhart committed
317
        assert not messages, str(messages)
318
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
Eckhart Arnold's avatar
Eckhart Arnold committed
319
        result = parser(testdoc)
320
        # log_parsing_history(parser, "test.log")
321
322
        assert not result.error_flag

323

324
class TestGrammar:
325
    def setup(self):
326
327
328
329
330
331
        grammar = r"""@whitespace = horizontal
        haupt        = textzeile LEERZEILE
        textzeile    = { WORT }+
        WORT         = /[^ \t]+/~
        LEERZEILE    = /\n[ \t]*(?=\n)/~
        """
332
333
334
        self.pyparser, messages, syntax_tree = compile_source(grammar, None, get_ebnf_grammar(),
                                                              get_ebnf_transformer(), get_ebnf_compiler("PosTest"))
        assert self.pyparser
335
        assert not messages
336
337
338
339

    def test_pos_values_initialized(self):
        # checks whether pos values in the parsing result and in the
        # history record have been initialized
340
341
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
        grammar("no_file_name*")
342
        for record in grammar.history__:
343
344
            assert not record.node or record.node.pos >= 0

345
    def test_select_parsing(self):
346
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
347
348
349
350
        grammar("wort", "WORT")
        grammar("eine Zeile", "textzeile")
        grammar("kein Haupt", "haupt")
        grammar("so ist es richtig", "haupt")
351

di68kap's avatar
di68kap committed
352
353
    def test_grammar_subclassing(self):
        class Arithmetic(Grammar):
354
            r'''
di68kap's avatar
di68kap committed
355
356
357
358
359
360
            expression =  term  { ("+" | "-") term }
            term       =  factor  { ("*" | "/") factor }
            factor     =  INTEGER | "("  expression  ")"
            INTEGER    =  /\d+/~
            '''
            expression = Forward()
361
362
363
364
            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
365
366
367
368
369
370
            root__ = expression

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

371
372
373
374
375
376
377
378
379
380
381
    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

382

383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
class TestSeries:
    def test_non_mandatory(self):
        lang = """
        document = series | /.*/
        series = "A" "B" "C" "D"
        """
        parser = grammar_provider(lang)()
        st = parser("ABCD");
        assert not st.error_flag
        st = parser("A_CD");
        assert not st.error_flag
        st = parser("AB_D");
        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)()
406
407
408
        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
409
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
410
        # transitivity of mandatory-operator
411
        st = parser("ABC_");  assert st.error_flag
412
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
413
414

    def test_series_composition(self):
415
        TA, TB, TC, TD, TE = (TKN(b) for b in "ABCDE")
416
417
418
419
420
        s1 = Series(TA, TB, TC, mandatory=2)
        s2 = Series(TD, TE)

        combined = Alternative(s1 + s2, RegExp('.*'))
        parser = Grammar(combined)
421
422
423
        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
424
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
425
        st = parser("ABC_E");  assert st.error_flag
426
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
427
428
429

        combined = Alternative(s2 + s1, RegExp('.*'))
        parser = Grammar(combined)
430
431
432
433
434
435
        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
436
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
437

eckhart's avatar
eckhart committed
438
439
440
441
442
443
444
445
446
447
    # 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
448

449

450
451
452
class TestAllOfSomeOf:
    def test_allOf_order(self):
        """Test that parsers of an AllOf-List can match in arbitrary order."""
453
        prefixes = AllOf(TKN("A"), TKN("B"))
454
455
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
456
        # aternative Form
457
        prefixes = AllOf(Series(TKN("B"), TKN("A")))
458
        assert Grammar(prefixes)('A B').content == 'A B'
459
460
461
462

    def test_allOf_completeness(self):
        """Test that an error is raised if not  all parsers of an AllOf-List
        match."""
463
        prefixes = AllOf(TKN("A"), TKN("B"))
464
465
466
467
468
        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."""
469
        prefixes = AllOf(TKN("A"), TKN("B"), TKN("A"))
470
471
472
        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'
473
474
475
476
        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."""
477
        prefixes = SomeOf(TKN("A"), TKN("B"))
478
479
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
480
        # aternative Form
481
        prefixes = SomeOf(Alternative(TKN("B"), TKN("A")))
482
483
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B').content == 'B'
484
485
486
487

    def test_someOf_redundance(self):
        """Test that one and the same parser may be listed several times
        and must be matched several times accordingly."""
488
        prefixes = SomeOf(TKN("A"), TKN("B"), TKN("A"))
489
490
491
        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'
492
493
494
        assert Grammar(prefixes)('A B B').error_flag


495
496
497
498
class TestErrorRecovery:
    def test_series_skip(self):
        lang = """
        document = series | /.*/
499
        @series_skip = /(?=[A-Z])/
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
        series = "A" "B" §"C" "D"
        """
        parser = grammar_provider(lang)()
        st = parser('AB_D')
        assert len(st.errors) == 1  # no additional "stopped before end"-error!

    def test_AllOf_skip(self):
        lang = """
        document = allof | /.*/
        @allof_skip = /[A-Z]/
        allof = < "A" §"B" "C" "D" >
        """
        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"
        st = parser('BC_A')
        assert st['allof'].content == "BC_A"


531
class TestPopRetrieve:
532
    mini_language = r"""
533
534
        document       = { text | codeblock }
        codeblock      = delimiter { text | (!:delimiter delimiter_sign) } ::delimiter
535
        delimiter      = delimiter_sign  # never use delimiter between capture and pop except for retrival!
536
        delimiter_sign = /`+/
eckhart's avatar
eckhart committed
537
        text           = /[^`]+/
538
        """
539
    mini_lang2 = r"""
540
541
542
543
544
545
546
547
        @braces_filter=counterpart
        document       = { text | codeblock }
        codeblock      = braces { text | opening_braces | (!:braces closing_braces) } ::braces
        braces         = opening_braces
        opening_braces = /\{+/
        closing_braces = /\}+/
        text           = /[^{}]+/
        """
548
    mini_lang3 = r"""
549
550
551
552
        document       = { text | env }
        env            = (specialtag | opentag) text [closespecial | closetag]
        opentag        = "<" name ">"
        specialtag     = "<" /ABC/ !name ">"
eckhart's avatar
eckhart committed
553
        closetag       = close_slash | close_star
554
555
556
557
558
559
560
561
        close_slash    = "<" ::name "/>"
        close_star     = "<" ::name "*>"
        closespecial   = "<" /ABC/~ ">"
        name           = /\w+/~
        text           = /[^<>]+/
        """

    def setup(self):
562
563
564
        self.minilang_parser = grammar_provider(self.mini_language)()
        self.minilang_parser2 = grammar_provider(self.mini_lang2)()
        self.minilang_parser3 = grammar_provider(self.mini_lang3)()
565
566
567

    @staticmethod
    def opening_delimiter(node, name):
568
        return node.tag_name == name # and not isinstance(node.parser, Retrieve)
569
570
571

    @staticmethod
    def closing_delimiter(node):
572
573
        return node.tag_name in {':Pop', ':Retrieve'}
        # return isinstance(node.parser, Retrieve)
574
575
576
577
578
579
580
581
582

    def test_compile_mini_language(self):
        assert self.minilang_parser
        assert self.minilang_parser2
        assert self.minilang_parser3

    def test_stackhandling(self):
        ambigous_opening = "<ABCnormal> normal tag <ABCnormal*>"
        syntax_tree = self.minilang_parser3(ambigous_opening)
583
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
584
585
586

        ambigous_opening = "<ABCnormal> normal tag <ABCnormal/>"
        syntax_tree = self.minilang_parser3(ambigous_opening)
587
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
588
589
590

        forgot_closing_tag = "<em> where is the closing tag?"
        syntax_tree = self.minilang_parser3(forgot_closing_tag)
591
        assert syntax_tree.error_flag, str(syntax_tree.errors_sorted)
592
593
594

        proper = "<em> has closing tag <em/>"
        syntax_tree = self.minilang_parser3(proper)
595
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
596
597
598

        proper = "<em> has closing tag <em*>"
        syntax_tree = self.minilang_parser3(proper)
599
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
600

601
    def test_cache_neutrality(self):
602
603
        """Test that packrat-caching does not interfere with the variable-
        changing parsers: Capture and Retrieve."""
604
        lang = r"""
605
606
607
608
609
610
611
612
613
            text = opening closing
            opening = (unmarked_package | marked_package)
            closing = ::variable
            unmarked_package = package "."
            marked_package = package "*" "."
            package = "(" variable ")"
            variable = /\w+/~
            """
        case = "(secret)*. secret"
614
        gr = grammar_provider(lang)()
615
        st = gr(case)
616
        assert not st.error_flag, str(st.errors_sorted)
617
618
619
620

    def test_single_line(self):
        teststr = "Anfang ```code block `` <- keine Ende-Zeichen ! ``` Ende"
        syntax_tree = self.minilang_parser(teststr)
621
        assert not syntax_tree.errors_sorted
622
623
        delim = str(next(syntax_tree.select_if(partial(self.opening_delimiter, name="delimiter"))))
        pop = str(next(syntax_tree.select_if(self.closing_delimiter)))
624
625
        assert delim == pop
        if is_logging():
626
            log_ST(syntax_tree, "test_PopRetrieve_single_line.cst")
627
628
629
630
631
632
633
634
635
636
637

    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)
638
        assert not syntax_tree.errors_sorted
639
640
        delim = str(next(syntax_tree.select_if(partial(self.opening_delimiter, name="delimiter"))))
        pop = str(next(syntax_tree.select_if(self.closing_delimiter)))
641
642
        assert delim == pop
        if is_logging():
643
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
644
645
646
647

    def test_single_line_complement(self):
        teststr = "Anfang {{{code block }} <- keine Ende-Zeichen ! }}} Ende"
        syntax_tree = self.minilang_parser2(teststr)
648
        assert not syntax_tree.errors_sorted
649
650
        delim = str(next(syntax_tree.select_if(partial(self.opening_delimiter, name="braces"))))
        pop = str(next(syntax_tree.select_if(self.closing_delimiter)))
651
652
        assert len(delim) == len(pop) and delim != pop
        if is_logging():
653
            log_ST(syntax_tree, "test_PopRetrieve_single_line.cst")
654
655
656
657
658
659
660
661
662
663
664

    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)
665
        assert not syntax_tree.errors_sorted
666
667
        delim = str(next(syntax_tree.select_if(partial(self.opening_delimiter, name="braces"))))
        pop = str(next(syntax_tree.select_if(self.closing_delimiter)))
668
669
        assert len(delim) == len(pop) and delim != pop
        if is_logging():
670
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
671
672


673
class TestWhitespaceHandling:
Eckhart Arnold's avatar
Eckhart Arnold committed
674
    minilang = """
675
676
677
678
679
680
681
682
683
        doc = A B
        A = "A"
        B = "B"
        Rdoc = ar br
        ar = /A/
        br = /B/
        """

    def setup(self):
684
        self.gr = grammar_provider(self.minilang)()
685
686
687
688
689
690
691
692
693
694
695
696

    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
697
698
699
700
701


class TestErrorReporting:
    grammar = """
        root      = series alpha | anything
eckhart's avatar
eckhart committed
702
        series    = subseries &alpha
di68kap's avatar
di68kap committed
703
704
705
706
707
708
709
710
711
712
713
714
715
716
        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)
717
        assert not cst.error_flag, str(cst.errors_sorted)
di68kap's avatar
di68kap committed
718
719
720
721
        cst = self.parser(testcode2)
        assert not cst.error_flag
        cst = self.parser(testcode3)
        assert cst.error_flag
722
723


di68kap's avatar
di68kap committed
724
725
class TestBorderlineCases:
    def test_not_matching(self):
726
        minilang = """parser = /X/\n"""
di68kap's avatar
di68kap committed
727
728
729
730
        gr = grammar_provider(minilang)()
        cst = gr('X', 'parser')
        assert not cst.error_flag
        cst = gr(' ', 'parser')
731
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
732
        cst = gr('', 'parser')
733
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
734
735
736
737
738
739
740

    def test_matching(self):
        minilang = """parser = /.?/"""
        gr = grammar_provider(minilang)()
        cst = gr(' ', 'parser')
        assert not cst.error_flag
        cst = gr('  ', 'parser')
741
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_STOPPED_BEFORE_END
di68kap's avatar
di68kap committed
742
743
744
745
        cst = gr('', 'parser')
        assert not cst.error_flag


Eckhart Arnold's avatar
Eckhart Arnold committed
746
class TestReentryAfterError:
747
    def setup(self):
Eckhart Arnold's avatar
Eckhart Arnold committed
748
749
750
751
752
        lang = """
        document = alpha [beta] gamma "."
          alpha = "ALPHA" abc
            abc = §"a" "b" "c"
          beta = "BETA" (bac | bca)
753
754
            bac = "b" "a" §"c"
            bca = "b" "c" §"a"
Eckhart Arnold's avatar
Eckhart Arnold committed
755
756
757
758
          gamma = "GAMMA" §(cab | cba)
            cab = "c" "a" §"b"
            cba = "c" "b" §"a"
        """
759
        self.gr = grammar_provider(lang)()
760

761
762
    def test_no_resume_rules(self):
        gr = self.gr;  gr.resume_rules = dict()
763
764
        content = 'ALPHA acb BETA bac GAMMA cab .'
        cst = gr(content)
765
766
767
768
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')

eckhart's avatar
eckhart committed
769
770
771
772
773
774
775
776
    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')

777
778
    def test_simple_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
779
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)')]
780
        content = 'ALPHA acb BETA bac GAMMA cab .'
781
782
783
784
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
785
        # because of resuming, there should be only one error message
786
        assert len(cst.errors_sorted) == 1
787

788
789
    def test_failing_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
790
        gr.resume_rules__['alpha'] = [re.compile(r'(?=XXX)')]
791
        content = 'ALPHA acb BETA bac GAMMA cab .'
792
793
794
795
796
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        # assert cst.pick('alpha').content.startswith('ALPHA')

797
798
    def test_severl_reentry_points(self):
        gr = self.gr;  gr.resume_rules = dict()
799
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
800
        content = 'ALPHA acb BETA bac GAMMA cab .'
801
802
803
804
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
805
        # because of resuming, there should be only one error message
806
        assert len(cst.errors_sorted) == 1
807

808
809
    def test_several_reentry_points_second_point_matching(self):
        gr = self.gr;  gr.resume_rules = dict()
810
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
811
812
        content = 'ALPHA acb GAMMA cab .'
        cst = gr(content)
Eckhart Arnold's avatar
Eckhart Arnold committed
813
        assert cst.error_flag
814
        assert cst.content == content
815
        assert cst.pick('alpha').content.startswith('ALPHA')
816
        # because of resuming, there should be only one error message
817
        assert len(cst.errors_sorted) == 1
818

819
820
    def test_several_resume_rules_innermost_rule_matching(self):
        gr = self.gr;  gr.resume_rules = dict()
821
822
823
        gr.resume_rules__['alpha'] = [re.compile(r'(?=BETA)'), re.compile(r'(?=GAMMA)')]
        gr.resume_rules__['beta'] = [re.compile(r'(?=GAMMA)')]
        gr.resume_rules__['bac'] = [re.compile(r'(?=GAMMA)')]
824
825
826
827
828
        content = 'ALPHA abc BETA bad GAMMA cab .'
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
829
        # because of resuming, there should be only one error message
830
        assert len(cst.errors_sorted) == 1
831
832
833
834
835
836
        # multiple failures
        content = 'ALPHA acb BETA bad GAMMA cab .'
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
837
        # there should be only two error messages
838
        assert len(cst.errors_sorted) == 2
839

840
    def test_skip_comment_on_resume(self):
841
842
        lang = r"""
            @ comment =  /(?:\/\/.*)|(?:\/\*(?:.|\n)*?\*\/)/  # Kommentare im C++-Stil
843
            document = block_A block_B
844
            @ block_A_resume = /(?=x)/
845
846
            block_A = "a" §"b" "c"
            block_B = "x" "y" "z"
847
        """
848
849
850
851
852
853
854
855
856
857
858
859
        def mini_suite(grammar):
            tree = grammar('abc/*x*/xyz')
            assert not tree.errors
            tree = grammar('abDxyz')
            mandatory_cont = (Error.MANDATORY_CONTINUATION, Error.MANDATORY_CONTINUATION_AT_EOF)
            assert len(tree.errors) == 1 and tree.errors[0].code in mandatory_cont
            tree = grammar('abD/*x*/xyz')
            assert len(tree.errors) == 1 and tree.errors[0].code in mandatory_cont
            tree = grammar('aD /*x*/ c /* a */ /*x*/xyz')
            assert len(tree.errors) == 1 and tree.errors[0].code in mandatory_cont

        # test regex-defined resume rule
860
        grammar = grammar_provider(lang)()
861
        mini_suite(grammar)
862

863

864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
    def test_unambiguous_error_location(self):
        lang = r"""
            @ drop        = whitespace, token  # drop tokens and whitespace early
           
            @object_resume = /(?<=\})/
           
            json       = ~ value EOF
            value      = object | string 
            object     = "{" [member { "," §member }] "}"
            member     = string §":" value
            string     = `"` CHARACTERS `"` ~

            CHARACTERS = { /[^"\\]+/ }                  
            EOF      =  !/./        # no more characters ahead, end of file reached
            """
        test_case = """{
                "missing member": "abcdef",
            }"""
        gr = grammar_provider(lang)()
        cst = gr(test_case)
        assert any(err.code == Error.MANDATORY_CONTINUATION for err in cst.errors)


887
class TestConfiguredErrorMessages:
888
    def test_configured_error_message(self):
889
890
891
        lang = """
            document = series | /.*/
            @series_error = "a badly configured error message {5}"
892
            series = /X/ | head §"C" "D"
893
894
895
896
            head = "A" "B"
            """
        parser = grammar_provider(lang)()
        st = parser("AB_D");  assert st.error_flag
897
898
        assert st.errors_sorted[0].code == Error.MALFORMED_ERROR_STRING
        assert st.errors_sorted[1].code == Error.MANDATORY_CONTINUATION
899
900


901
902
903
904
905
906
907
908
909
910
class TestUnknownParserError:
    def test_unknown_parser_error(self):
        gr = Grammar()
        try:
            gr("", "NonExistantParser")
            assert False, "UnknownParserError expected!"
        except UnknownParserError:
            pass


911
912
class TestEarlyTokenWhitespaceDrop:
    def setup(self):
913
        self.lang = r"""
914
915
916
            @ drop = token, whitespace
            expression = term  { ("+" | "-") term}
            term       = factor  { ("*"|"/") factor}
eckhart's avatar
eckhart committed
917
            factor     = number | variable | "("  expression  ")"
918
                       | constant | fixed
919
            variable   = /[a-z]/~
920
921
            number     = /\d+/~
            constant   = "A" | "B"
eckhart's avatar
eckhart committed
922
            fixed      = "X"
923
924
925
            """
        self.gr = grammar_provider(self.lang)()

926
    def test_drop(self):
927
928
929
        cst = self.gr('4 + 3 * 5')
        assert not cst.pick(':Token')
        assert not cst.pick(':Whitespace')
930
931
        cst = self.gr('A + B')
        try:
932
            _ = next(cst.select_if(lambda node: node.content == 'A'))
933
934
935
936
            assert False, "Tokens in compound expressions should be dropped!"
        except StopIteration:
            pass
        cst = self.gr('X * y')
937
        assert next(cst.select_if(lambda node: node.content == 'X'))
938

939

Eckhart Arnold's avatar
Eckhart Arnold committed
940
class TestMetaParser:
941
942
943
944
    def setup(self):
        self.mp = MetaParser()
        self.mp.grammar = Grammar()  # override placeholder warning
        self.mp.pname = "named"
945
        self.mp.anonymous = False
946
947
948
        self.mp.tag_name = self.mp.pname

    def test_return_value(self):
949
950
        save = get_config_value('flatten_tree_while_parsing')
        set_config_value('flatten_tree_while_parsing', True)
951
        nd = self.mp._return_value(Node('tagged', 'non-empty'))
Eckhart Arnold's avatar
Eckhart Arnold committed
952
953
954
955
        assert nd.tag_name == 'named', nd.as_sxpr()
        assert len(nd.children) == 1
        assert nd.children[0].tag_name == 'tagged'
        assert nd.children[0].result == "non-empty"
956
        nd = self.mp._return_value(Node('tagged', ''))
Eckhart Arnold's avatar
Eckhart Arnold committed
957
958
959
960
        assert nd.tag_name == 'named', nd.as_sxpr()
        assert len(nd.children) == 1
        assert nd.children[0].tag_name == 'tagged'
        assert not nd.children[0].result
961
        nd = self.mp._return_value(Node(':anonymous', 'content'))
Eckhart Arnold's avatar
Eckhart Arnold committed
962
963
964
        assert nd.tag_name == 'named', nd.as_sxpr()
        assert not nd.children
        assert nd.result == 'content'
965
        nd = self.mp._return_value(Node(':anonymous', ''))
Eckhart Arnold's avatar
Eckhart Arnold committed
966
967
968
        assert nd.tag_name == 'named', nd.as_sxpr()
        assert not nd.children
        assert not nd.content
969
970
971
        nd = self.mp._return_value(EMPTY_NODE)
        assert nd.tag_name == 'named' and not nd.children, nd.as_sxpr()
        self.mp.pname = ''
972
        self.mp.anonymous = True
973
974
        self.mp.tag_name = ':unnamed'
        nd = self.mp._return_value(Node('tagged', 'non-empty'))
Eckhart Arnold's avatar
Eckhart Arnold committed
975
976
977
        assert nd.tag_name == 'tagged', nd.as_sxpr()
        assert len(nd.children) == 0
        assert nd.content == 'non-empty'
978
        nd = self.mp._return_value(Node('tagged', ''))
Eckhart Arnold's avatar
Eckhart Arnold committed
979
980
981
        assert nd.tag_name == 'tagged', nd.as_sxpr()
        assert len(nd.children) == 0
        assert not nd.content
982
        nd = self.mp._return_value(Node(':anonymous', 'content'))
Eckhart Arnold's avatar
Eckhart Arnold committed
983
984
985
        assert nd.tag_name == ':anonymous', nd.as_sxpr()
        assert not nd.children
        assert nd.result == 'content'
986
        nd = self.mp._return_value(Node('', ''))
Eckhart Arnold's avatar
Eckhart Arnold committed
987
988
989
        assert nd.tag_name == '', nd.as_sxpr()
        assert not nd.children
        assert not nd.content
990
991
        assert self.mp._return_value(None) == EMPTY_NODE
        assert self.mp._return_value(EMPTY_NODE) == EMPTY_NODE
992
        set_config_value('flatten_tree_while_parsing', save)
Eckhart Arnold's avatar
Eckhart Arnold committed
993

994
995
996
997
998
999
    def test_return_values(self):
        self.mp.pname = "named"
        self.mp.tag_name = self.mp.pname
        rv = self.mp._return_values((Node('tag', 'content'), EMPTY_NODE))
        assert rv[-1].tag_name != EMPTY_NODE.tag_name, rv[-1].tag_name

1000
    def test_in_context(self):