test_parse.py 38 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
30
from DHParser.configuration import get_config_value, set_config_value
from DHParser.toolkit import compile_python_object
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

eckhart's avatar
eckhart committed
48

eckhart's avatar
eckhart committed
49
50
51
52
53
54
55
56
57
58
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
59
            l.append(p.pname + p.ptype)
eckhart's avatar
eckhart committed
60
61
62
63
64
65
66
67
68
69
70
        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


71
class TestInfiLoopsAndRecursion:
72
    def test_direct_left_recursion1(self):
eckhart's avatar
eckhart committed
73
        minilang = """
74
75
76
77
            expr = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
78
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
79
        parser = grammar_provider(minilang)()
80
        assert parser
Eckhart Arnold's avatar
Eckhart Arnold committed
81
        syntax_tree = parser(snippet)
82
83
        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
84
        if is_logging():
85
            log_ST(syntax_tree, "test_LeftRecursion_direct.cst")
86
            log_parsing_history(parser, "test_LeftRecursion_direct")
87

88
    def test_direct_left_recursion2(self):
89
90
91
92
93
94
        minilang = """
            expr = ex
            ex   = expr ("+"|"-") term | term
            term = term ("*"|"/") factor | factor
            factor = /[0-9]+/~
            """
95
        snippet = "9 + 8 + 7 + 6 + 5 + 3 * 4"
96
        parser = grammar_provider(minilang)()
97
        assert parser
98
        syntax_tree = parser(snippet)
99
100
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
101

102
    def test_indirect_left_recursion1(self):
103
104
        minilang = """
            Expr    = //~ (Product | Sum | Value)
105
106
            Product = Expr { ('*' | '/') Expr }+
            Sum     = Expr { ('+' | '-') Expr }+
107
            Value   = /[0-9.]+/~ | '(' §Expr ')'
108
            """
109
        parser = grammar_provider(minilang)()
110
        assert parser
111
        snippet = "8 * 4"
112
        syntax_tree = parser(snippet)
113
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
114
115
        snippet = "7 + 8 * 4"
        syntax_tree = parser(snippet)
116
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
117
118
        snippet = "9 + 8 * (4 + 3)"
        syntax_tree = parser(snippet)
119
120
        assert not is_error(syntax_tree.error_flag), syntax_tree.errors_sorted
        assert snippet == syntax_tree.content
121
        if is_logging():
122
            log_ST(syntax_tree, "test_LeftRecursion_indirect.cst")
123
            log_parsing_history(parser, "test_LeftRecursion_indirect")
124

125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
    # # 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
145

146
147
148
149
150
151
152
153
154
155
    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)

156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
    # 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)
181

eckhart's avatar
eckhart committed
182

Eckhart Arnold's avatar
Eckhart Arnold committed
183
class TestFlowControl:
184
185
    def setup(self):
        self.t1 = """
eckhart's avatar
eckhart committed
186
        All work and no play
187
188
189
190
191
        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
192
    def test_lookbehind(self):
193
        ws = RegExp(r'\s*')
Eckhart Arnold's avatar
Eckhart Arnold committed
194
        end = RegExp("END")
195
        doc_end = Lookbehind(RegExp('\\s*?\\n')) + end
196
        word = RegExp(r'\w+')
Eckhart Arnold's avatar
Eckhart Arnold committed
197
198
199
200
        sequence = OneOrMore(NegativeLookahead(end) + word + ws)
        document = ws + sequence + doc_end + ws

        parser = Grammar(document)
201
202
203
204
205
206
207
        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
208
            parser_initialization__ = ["upon instantiation"]
209
            ws = RegExp(r'\s*')
210
            end = RegExp('END')
211
            SUCC_LB = RegExp('\\s*?\\n')
212
            doc_end = Series(Lookbehind(SUCC_LB), end)
213
            word = RegExp(r'\w+')
214
215
216
217
218
219
            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
220
        assert not cst.error_flag, cst.as_sxpr()
221
        cst = parser(self.t2)
Eckhart Arnold's avatar
Eckhart Arnold committed
222
223
224
        assert cst.error_flag, cst.as_sxpr()


225
226
227
228
229
230
231
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
232
233
        result, messages, syntax_tree = compile_source(mlregex, None, get_ebnf_grammar(),
                        get_ebnf_transformer(), get_ebnf_compiler('MultilineRegexTest'))
234
        assert result
235
        assert not messages, str(messages)
236
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
eckhart's avatar
eckhart committed
237
238
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
239
        assert node.tag_name == "regex"
240
241
        assert str(node) == 'abc+def'

242
243
    def test_multilineRegex_wo_Comments(self):
        mlregex = r"""
eckhart's avatar
eckhart committed
244
245
        regex =  /\w+
                  [+]
246
247
248
249
250
251
                  \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)
252
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
253
254
        node = parser('abc+def', parser.regex)
        assert not node.error_flag
255
        assert node.tag_name == "regex"
256
257
        assert str(node) == 'abc+def'

258
259
260
261
262
263
264
265
266
    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
267
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
268
269
270
271
        node, rest = parser.regex('Alpha')
        assert node
        assert not node.error_flag
        assert rest == ''
272
        assert node.tag_name == "regex"
273
274
275
276
277
278
279
280
281
282
        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
283
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
284
285
286
287
        node, rest = parser.regex('Alpha')
        assert node.error_flag


288
289
290
291
292
293
294
295
296
297
298
299
    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
300
301
302
        result, messages, syntax_tree = compile_source(
            tokenlang, None, get_ebnf_grammar(), get_ebnf_transformer(),
            get_ebnf_compiler("TokenTest"))
303
        assert result
eckhart's avatar
eckhart committed
304
        assert not messages, str(messages)
305
        parser = compile_python_object(DHPARSER_IMPORTS + result, r'\w+Grammar$')()
Eckhart Arnold's avatar
Eckhart Arnold committed
306
        result = parser(testdoc)
307
        # log_parsing_history(parser, "test.log")
308
309
        assert not result.error_flag

310

311
class TestGrammar:
312
    def setup(self):
313
314
315
316
317
318
        grammar = r"""@whitespace = horizontal
        haupt        = textzeile LEERZEILE
        textzeile    = { WORT }+
        WORT         = /[^ \t]+/~
        LEERZEILE    = /\n[ \t]*(?=\n)/~
        """
319
320
321
        self.pyparser, messages, syntax_tree = compile_source(grammar, None, get_ebnf_grammar(),
                                                              get_ebnf_transformer(), get_ebnf_compiler("PosTest"))
        assert self.pyparser
322
        assert not messages
323
324
325
326

    def test_pos_values_initialized(self):
        # checks whether pos values in the parsing result and in the
        # history record have been initialized
327
328
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
        grammar("no_file_name*")
329
        for record in grammar.history__:
330
331
            assert not record.node or record.node.pos >= 0

332
    def test_select_parsing(self):
333
        grammar = compile_python_object(DHPARSER_IMPORTS + self.pyparser, r'\w+Grammar$')()
334
335
336
337
        grammar("wort", "WORT")
        grammar("eine Zeile", "textzeile")
        grammar("kein Haupt", "haupt")
        grammar("so ist es richtig", "haupt")
338

di68kap's avatar
di68kap committed
339
340
    def test_grammar_subclassing(self):
        class Arithmetic(Grammar):
341
            r'''
di68kap's avatar
di68kap committed
342
343
344
345
346
347
            expression =  term  { ("+" | "-") term }
            term       =  factor  { ("*" | "/") factor }
            factor     =  INTEGER | "("  expression  ")"
            INTEGER    =  /\d+/~
            '''
            expression = Forward()
348
349
350
351
            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
352
353
354
355
356
357
            root__ = expression

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

358
359
360
361
362
363
364
365
366
367
368
    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

369

370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
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)()
393
394
395
        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
396
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
397
        # transitivity of mandatory-operator
398
        st = parser("ABC_");  assert st.error_flag
399
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
400
401

    def test_series_composition(self):
402
        TA, TB, TC, TD, TE = (TKN(b) for b in "ABCDE")
403
404
405
406
407
        s1 = Series(TA, TB, TC, mandatory=2)
        s2 = Series(TD, TE)

        combined = Alternative(s1 + s2, RegExp('.*'))
        parser = Grammar(combined)
408
409
410
        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
411
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
412
        st = parser("ABC_E");  assert st.error_flag
413
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
414
415
416

        combined = Alternative(s2 + s1, RegExp('.*'))
        parser = Grammar(combined)
417
418
419
420
421
422
        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
423
        assert st.errors_sorted[0].code == Error.MANDATORY_CONTINUATION
424

eckhart's avatar
eckhart committed
425
426
427
428
429
430
431
432
433
434
    # 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
435

436

437
438
439
class TestAllOfSomeOf:
    def test_allOf_order(self):
        """Test that parsers of an AllOf-List can match in arbitrary order."""
440
        prefixes = AllOf(TKN("A"), TKN("B"))
441
442
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
443
        # aternative Form
444
        prefixes = AllOf(Series(TKN("B"), TKN("A")))
445
        assert Grammar(prefixes)('A B').content == 'A B'
446
447
448
449

    def test_allOf_completeness(self):
        """Test that an error is raised if not  all parsers of an AllOf-List
        match."""
450
        prefixes = AllOf(TKN("A"), TKN("B"))
451
452
453
454
455
        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."""
456
        prefixes = AllOf(TKN("A"), TKN("B"), TKN("A"))
457
458
459
        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'
460
461
462
463
        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."""
464
        prefixes = SomeOf(TKN("A"), TKN("B"))
465
466
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B A').content == 'B A'
467
        # aternative Form
468
        prefixes = SomeOf(Alternative(TKN("B"), TKN("A")))
469
470
        assert Grammar(prefixes)('A B').content == 'A B'
        assert Grammar(prefixes)('B').content == 'B'
471
472
473
474

    def test_someOf_redundance(self):
        """Test that one and the same parser may be listed several times
        and must be matched several times accordingly."""
475
        prefixes = SomeOf(TKN("A"), TKN("B"), TKN("A"))
476
477
478
        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'
479
480
481
        assert Grammar(prefixes)('A B B').error_flag


482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
class TestErrorRecovery:
    def test_series_skip(self):
        lang = """
        document = series | /.*/
        @series_skip = /[A-Z]/
        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"


518
class TestPopRetrieve:
519
    mini_language = r"""
520
521
        document       = { text | codeblock }
        codeblock      = delimiter { text | (!:delimiter delimiter_sign) } ::delimiter
522
        delimiter      = delimiter_sign  # never use delimiter between capture and pop except for retrival!
523
        delimiter_sign = /`+/
eckhart's avatar
eckhart committed
524
        text           = /[^`]+/
525
        """
526
    mini_lang2 = r"""
527
528
529
530
531
532
533
534
        @braces_filter=counterpart
        document       = { text | codeblock }
        codeblock      = braces { text | opening_braces | (!:braces closing_braces) } ::braces
        braces         = opening_braces
        opening_braces = /\{+/
        closing_braces = /\}+/
        text           = /[^{}]+/
        """
535
    mini_lang3 = r"""
536
537
538
539
        document       = { text | env }
        env            = (specialtag | opentag) text [closespecial | closetag]
        opentag        = "<" name ">"
        specialtag     = "<" /ABC/ !name ">"
eckhart's avatar
eckhart committed
540
        closetag       = close_slash | close_star
541
542
543
544
545
546
547
548
        close_slash    = "<" ::name "/>"
        close_star     = "<" ::name "*>"
        closespecial   = "<" /ABC/~ ">"
        name           = /\w+/~
        text           = /[^<>]+/
        """

    def setup(self):
549
550
551
        self.minilang_parser = grammar_provider(self.mini_language)()
        self.minilang_parser2 = grammar_provider(self.mini_lang2)()
        self.minilang_parser3 = grammar_provider(self.mini_lang3)()
552
553
554

    @staticmethod
    def opening_delimiter(node, name):
555
        return node.tag_name == name # and not isinstance(node.parser, Retrieve)
556
557
558

    @staticmethod
    def closing_delimiter(node):
559
560
        return node.tag_name in {':Pop', ':Retrieve'}
        # return isinstance(node.parser, Retrieve)
561
562
563
564
565
566
567
568
569

    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)
570
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
571
572
573

        ambigous_opening = "<ABCnormal> normal tag <ABCnormal/>"
        syntax_tree = self.minilang_parser3(ambigous_opening)
574
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
575
576
577

        forgot_closing_tag = "<em> where is the closing tag?"
        syntax_tree = self.minilang_parser3(forgot_closing_tag)
578
        assert syntax_tree.error_flag, str(syntax_tree.errors_sorted)
579
580
581

        proper = "<em> has closing tag <em/>"
        syntax_tree = self.minilang_parser3(proper)
582
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
583
584
585

        proper = "<em> has closing tag <em*>"
        syntax_tree = self.minilang_parser3(proper)
586
        assert not syntax_tree.error_flag, str(syntax_tree.errors_sorted)
587

588
    def test_cache_neutrality(self):
589
590
        """Test that packrat-caching does not interfere with the variable-
        changing parsers: Capture and Retrieve."""
591
        lang = r"""
592
593
594
595
596
597
598
599
600
            text = opening closing
            opening = (unmarked_package | marked_package)
            closing = ::variable
            unmarked_package = package "."
            marked_package = package "*" "."
            package = "(" variable ")"
            variable = /\w+/~
            """
        case = "(secret)*. secret"
601
        gr = grammar_provider(lang)()
602
        st = gr(case)
603
        assert not st.error_flag, str(st.errors_sorted)
604
605
606
607

    def test_single_line(self):
        teststr = "Anfang ```code block `` <- keine Ende-Zeichen ! ``` Ende"
        syntax_tree = self.minilang_parser(teststr)
608
        assert not syntax_tree.errors_sorted
609
610
        delim = str(next(syntax_tree.select_if(partial(self.opening_delimiter, name="delimiter"))))
        pop = str(next(syntax_tree.select_if(self.closing_delimiter)))
611
612
        assert delim == pop
        if is_logging():
613
            log_ST(syntax_tree, "test_PopRetrieve_single_line.cst")
614
615
616
617
618
619
620
621
622
623
624

    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)
625
        assert not syntax_tree.errors_sorted
626
627
        delim = str(next(syntax_tree.select_if(partial(self.opening_delimiter, name="delimiter"))))
        pop = str(next(syntax_tree.select_if(self.closing_delimiter)))
628
629
        assert delim == pop
        if is_logging():
630
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
631
632
633
634

    def test_single_line_complement(self):
        teststr = "Anfang {{{code block }} <- keine Ende-Zeichen ! }}} Ende"
        syntax_tree = self.minilang_parser2(teststr)
635
        assert not syntax_tree.errors_sorted
636
637
        delim = str(next(syntax_tree.select_if(partial(self.opening_delimiter, name="braces"))))
        pop = str(next(syntax_tree.select_if(self.closing_delimiter)))
638
639
        assert len(delim) == len(pop) and delim != pop
        if is_logging():
640
            log_ST(syntax_tree, "test_PopRetrieve_single_line.cst")
641
642
643
644
645
646
647
648
649
650
651

    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)
652
        assert not syntax_tree.errors_sorted
653
654
        delim = str(next(syntax_tree.select_if(partial(self.opening_delimiter, name="braces"))))
        pop = str(next(syntax_tree.select_if(self.closing_delimiter)))
655
656
        assert len(delim) == len(pop) and delim != pop
        if is_logging():
657
            log_ST(syntax_tree, "test_PopRetrieve_multi_line.cst")
658
659


660
class TestWhitespaceHandling:
Eckhart Arnold's avatar
Eckhart Arnold committed
661
    minilang = """
662
663
664
665
666
667
668
669
670
        doc = A B
        A = "A"
        B = "B"
        Rdoc = ar br
        ar = /A/
        br = /B/
        """

    def setup(self):
671
        self.gr = grammar_provider(self.minilang)()
672
673
674
675
676
677
678
679
680
681
682
683

    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
684
685
686
687
688


class TestErrorReporting:
    grammar = """
        root      = series alpha | anything
eckhart's avatar
eckhart committed
689
        series    = subseries &alpha
di68kap's avatar
di68kap committed
690
691
692
693
694
695
696
697
698
699
700
701
702
703
        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)
704
        assert not cst.error_flag, str(cst.errors_sorted)
di68kap's avatar
di68kap committed
705
706
707
708
        cst = self.parser(testcode2)
        assert not cst.error_flag
        cst = self.parser(testcode3)
        assert cst.error_flag
709
710


di68kap's avatar
di68kap committed
711
712
class TestBorderlineCases:
    def test_not_matching(self):
713
        minilang = """parser = /X/\n"""
di68kap's avatar
di68kap committed
714
715
716
717
        gr = grammar_provider(minilang)()
        cst = gr('X', 'parser')
        assert not cst.error_flag
        cst = gr(' ', 'parser')
718
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
719
        cst = gr('', 'parser')
720
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_DID_NOT_MATCH
di68kap's avatar
di68kap committed
721
722
723
724
725
726
727

    def test_matching(self):
        minilang = """parser = /.?/"""
        gr = grammar_provider(minilang)()
        cst = gr(' ', 'parser')
        assert not cst.error_flag
        cst = gr('  ', 'parser')
728
        assert cst.error_flag and cst.errors_sorted[0].code == Error.PARSER_STOPPED_BEFORE_END
di68kap's avatar
di68kap committed
729
730
731
732
        cst = gr('', 'parser')
        assert not cst.error_flag


Eckhart Arnold's avatar
Eckhart Arnold committed
733
class TestReentryAfterError:
734
    def setup(self):
Eckhart Arnold's avatar
Eckhart Arnold committed
735
736
737
738
739
        lang = """
        document = alpha [beta] gamma "."
          alpha = "ALPHA" abc
            abc = §"a" "b" "c"
          beta = "BETA" (bac | bca)
740
741
            bac = "b" "a" §"c"
            bca = "b" "c" §"a"
Eckhart Arnold's avatar
Eckhart Arnold committed
742
743
744
745
          gamma = "GAMMA" §(cab | cba)
            cab = "c" "a" §"b"
            cba = "c" "b" §"a"
        """
746
        self.gr = grammar_provider(lang)()
747

748
749
    def test_no_resume_rules(self):
        gr = self.gr;  gr.resume_rules = dict()
750
751
        content = 'ALPHA acb BETA bac GAMMA cab .'
        cst = gr(content)
752
753
754
755
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')

eckhart's avatar
eckhart committed
756
757
758
759
760
761
762
763
    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')

764
765
    def test_simple_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
766
        gr.resume_rules__['alpha'] = ['BETA']
767
        content = 'ALPHA acb BETA bac GAMMA cab .'
768
769
770
771
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
772
        # because of resuming, there should be only one error message
773
        assert len(cst.errors_sorted) == 1
774

775
776
    def test_failing_resume_rule(self):
        gr = self.gr;  gr.resume_rules = dict()
777
        gr.resume_rules__['alpha'] = ['XXX']
778
        content = 'ALPHA acb BETA bac GAMMA cab .'
779
780
781
782
783
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        # assert cst.pick('alpha').content.startswith('ALPHA')

784
785
    def test_severl_reentry_points(self):
        gr = self.gr;  gr.resume_rules = dict()
786
        gr.resume_rules__['alpha'] = ['BETA', 'GAMMA']
787
        content = 'ALPHA acb BETA bac GAMMA cab .'
788
789
790
791
        cst = gr(content)
        assert cst.error_flag
        assert cst.content == content
        assert cst.pick('alpha').content.startswith('ALPHA')
792
        # because of resuming, there should be only one error message
793
        assert len(cst.errors_sorted) == 1
794

795
796
    def test_several_reentry_points_second_point_matching(self):
        gr = self.gr;  gr.resume_rules = dict()
797
798
799
        gr.resume_rules__['alpha'] = ['BETA', 'GAMMA']
        content = 'ALPHA acb GAMMA cab .'
        cst = gr(content)
Eckhart Arnold's avatar
Eckhart Arnold committed
800
        assert cst.error_flag
801
        assert cst.content == content
802
        assert cst.pick('alpha').content.startswith('ALPHA')
803
        # because of resuming, there should be only one error message
804
        assert len(cst.errors_sorted) == 1
805

806
807
808
809
810
811
812
813
814
815
    def test_several_resume_rules_innermost_rule_matching(self):
        gr = self.gr;  gr.resume_rules = dict()
        gr.resume_rules__['alpha'] = ['BETA', 'GAMMA']
        gr.resume_rules__['beta'] = ['GAMMA']
        gr.resume_rules__['bac'] = ['GAMMA']
        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')
816
        # because of resuming, there should be only one error message
817
        assert len(cst.errors_sorted) == 1
818
819
820
821
822
823
        # 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')
824
        # there should be only two error messages
825
        assert len(cst.errors_sorted) == 2
826

827
    def test_skip_comment_on_resume(self):
828
829
        lang = r"""
            @ comment =  /(?:\/\/.*)|(?:\/\*(?:.|\n)*?\*\/)/  # Kommentare im C++-Stil
830
831
832
833
            document = block_A block_B
            @ block_A_resume = /x/
            block_A = "a" §"b" "c"
            block_B = "x" "y" "z"
834
        """
835
836
837
838
839
840
841
842
843
844
845
846
        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
847
        grammar = grammar_provider(lang)()
848
        mini_suite(grammar)
849

850
851
852
853
854
        # test string-defined resume rule
        alt_lang = lang.replace('@ block_A_resume = /x/',
                                '@ block_A_resume = "x"')
        grammar = grammar_provider(alt_lang)()
        mini_suite(grammar)
855

856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
    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)


879
class TestConfiguredErrorMessages:
880
    def test_configured_error_message(self):
881
882
883
        lang = """
            document = series | /.*/
            @series_error = "a badly configured error message {5}"
884
            series = /X/ | head §"C" "D"
885
886
887
888
            head = "A" "B"
            """
        parser = grammar_provider(lang)()
        st = parser("AB_D");  assert st.error_flag
889
890
        assert st.errors_sorted[0].code == Error.MALFORMED_ERROR_STRING
        assert st.errors_sorted[1].code == Error.MANDATORY_CONTINUATION
891
892


893
894
895
896
897
898
899
900
901
902
class TestUnknownParserError:
    def test_unknown_parser_error(self):
        gr = Grammar()
        try:
            gr("", "NonExistantParser")
            assert False, "UnknownParserError expected!"
        except UnknownParserError:
            pass


903
904
class TestEarlyTokenWhitespaceDrop:
    def setup(self):
905
        self.lang = r"""
906
907
908
            @ drop = token, whitespace
            expression = term  { ("+" | "-") term}
            term       = factor  { ("*"|"/") factor}
eckhart's avatar
eckhart committed
909
            factor     = number | variable | "("  expression  ")"
910
                       | constant | fixed
911
            variable   = /[a-z]/~
912
913
            number     = /\d+/~
            constant   = "A" | "B"
eckhart's avatar
eckhart committed
914
            fixed      = "X"
915
916
917
            """
        self.gr = grammar_provider(self.lang)()

918
    def test_drop(self):
919
920
921
        cst = self.gr('4 + 3 * 5')
        assert not cst.pick(':Token')
        assert not cst.pick(':Whitespace')
922
923
        cst = self.gr('A + B')
        try:
924
            _ = next(cst.select_if(lambda node: node.content == 'A'))
925
926
927
928
            assert False, "Tokens in compound expressions should be dropped!"
        except StopIteration:
            pass
        cst = self.gr('X * y')
929
        assert next(cst.select_if(lambda node: node.content == 'X'))
930

931

Eckhart Arnold's avatar
Eckhart Arnold committed
932
class TestMetaParser:
933
934
935
936
937
938
939
    def setup(self):
        self.mp = MetaParser()
        self.mp.grammar = Grammar()  # override placeholder warning
        self.mp.pname = "named"
        self.mp.tag_name = self.mp.pname

    def test_return_value(self):
940
941
        save = get_config_value('flatten_tree_while_parsing')
        set_config_value('flatten_tree_while_parsing', True)
942
        nd = self.mp._return_value(Node('tagged', 'non-empty'))
Eckhart Arnold's avatar
Eckhart Arnold committed
943
944
945
946
        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"
947
        nd = self.mp._return_value(Node('tagged', ''))
Eckhart Arnold's avatar
Eckhart Arnold committed
948
949
950
951
        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
952
        nd = self.mp._return_value(Node(':anonymous', 'content'))
Eckhart Arnold's avatar
Eckhart Arnold committed
953
954
955
        assert nd.tag_name == 'named', nd.as_sxpr()
        assert not nd.children
        assert nd.result == 'content'
956
        nd = self.mp._return_value(Node(':anonymous', ''))
Eckhart Arnold's avatar
Eckhart Arnold committed
957
958
959
        assert nd.tag_name == 'named', nd.as_sxpr()
        assert not nd.children
        assert not nd.content
960
961
962
963
964
        nd = self.mp._return_value(EMPTY_NODE)
        assert nd.tag_name == 'named' and not nd.children, nd.as_sxpr()
        self.mp.pname = ''
        self.mp.tag_name = ':unnamed'
        nd = self.mp._return_value(Node('tagged', 'non-empty'))
Eckhart Arnold's avatar
Eckhart Arnold committed
965
966
967
        assert nd.tag_name == 'tagged', nd.as_sxpr()
        assert len(nd.children) == 0
        assert nd.content == 'non-empty'
968
        nd = self.mp._return_value(Node('tagged', ''))
Eckhart Arnold's avatar
Eckhart Arnold committed
969
970
971
        assert nd.tag_name == 'tagged', nd.as_sxpr()
        assert len(nd.children) == 0
        assert not nd.content
972
        nd = self.mp._return_value(Node(':anonymous', 'content'))
Eckhart Arnold's avatar
Eckhart Arnold committed
973
974
975
        assert nd.tag_name == ':anonymous', nd.as_sxpr()
        assert not nd.children
        assert nd.result == 'content'
976
        nd = self.mp._return_value(Node('', ''))
Eckhart Arnold's avatar
Eckhart Arnold committed
977
978
979
        assert nd.tag_name == '', nd.as_sxpr()
        assert not nd.children
        assert not nd.content
980
981
        assert self.mp._return_value(None) == EMPTY_NODE
        assert self.mp._return_value(EMPTY_NODE) == EMPTY_NODE
982
        set_config_value('flatten_tree_while_parsing', save)
Eckhart Arnold's avatar
Eckhart Arnold committed
983

984
985
986
987
988
989
    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

990
    def test_in_context(self):
991
        minilang = r"""
992
993
994
            term       = factor  { (DIV|MUL) factor}
            factor     = NUMBER | VARIABLE
            MUL        = "*" | &factor
eckhart's avatar
eckhart committed
995
            DIV        = "/"
996
997
998
999
1000
            NUMBER     = /(?:0|(?:[1-9]\d*))(?:\.\d+)?/~
            VARIABLE   = /[A-Za-z]/~
            """
        gr = grammar_provider(minilang)()
        cst = gr("2x")