test_ebnf.py 16.4 KB
Newer Older
1
2
#!/usr/bin/python3

3
"""test_ebnf.py - tests of the ebnf module of DHParser 
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
                             

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.
"""

Eckhart Arnold's avatar
Eckhart Arnold committed
23
import sys
Eckhart Arnold's avatar
Eckhart Arnold committed
24
from multiprocessing import Pool
Eckhart Arnold's avatar
Eckhart Arnold committed
25

Eckhart Arnold's avatar
Eckhart Arnold committed
26
27
sys.path.extend(['../', './'])

28
from DHParser.toolkit import compile_python_object, re
29
from DHParser.preprocess import nil_preprocessor
30
from DHParser import compile_source
31
32
from DHParser.error import has_errors
from DHParser.syntaxtree import WHITESPACE_PTYPE
33
from DHParser.ebnf import get_ebnf_grammar, get_ebnf_transformer, EBNFTransform, get_ebnf_compiler
34
from DHParser.dsl import CompilationError, compileDSL, DHPARSER_IMPORTS, grammar_provider
35
from DHParser.testing import grammar_unit
36
37


38
39
40
41
42
43
44
45
46
47
48
class TestDirectives:
    mini_language = """
        expression =  term  { ("+" | "-") term }
        term       =  factor  { ("*" | "/") factor }
        factor     =  constant | "("  expression  ")"
        constant   =  digit { digit } [ //~ ]
        digit      = /0/ | /1/ | /2/ | /3/ | /4/ | /5/ | /6/ | /7/ | /8/ | /9/ 
        """

    def test_whitespace_linefeed(self):
        lang = "@ whitespace = linefeed\n" + self.mini_language
49
        MinilangParser = grammar_provider(lang)
50
51
        parser = MinilangParser()
        assert parser
Eckhart Arnold's avatar
Eckhart Arnold committed
52
        syntax_tree = parser("3 + 4 * 12")
53
        # parser.log_parsing_history("WSP")
54
        assert not syntax_tree.collect_errors()
Eckhart Arnold's avatar
Eckhart Arnold committed
55
        syntax_tree = parser("3 + 4 \n * 12")
56
57
        # parser.log_parsing_history("WSPLF")
        assert not syntax_tree.collect_errors()
Eckhart Arnold's avatar
Eckhart Arnold committed
58
        syntax_tree = parser("3 + 4 \n \n * 12")
59
        assert syntax_tree.collect_errors()
Eckhart Arnold's avatar
Eckhart Arnold committed
60
        syntax_tree = parser("3 + 4 \n\n * 12")
61
62
63
64
        assert syntax_tree.collect_errors()

    def test_whitespace_vertical(self):
        lang = "@ whitespace = vertical\n" + self.mini_language
65
        parser = grammar_provider(lang)()
66
        assert parser
Eckhart Arnold's avatar
Eckhart Arnold committed
67
        syntax_tree = parser("3 + 4 * 12")
68
        assert not syntax_tree.collect_errors()
Eckhart Arnold's avatar
Eckhart Arnold committed
69
        syntax_tree = parser("3 + 4 \n * 12")
70
        assert not syntax_tree.collect_errors()
Eckhart Arnold's avatar
Eckhart Arnold committed
71
        syntax_tree = parser("3 + 4 \n \n * 12")
72
        assert not syntax_tree.collect_errors()
Eckhart Arnold's avatar
Eckhart Arnold committed
73
        syntax_tree = parser("3 + 4 \n\n * 12")
74
75
        assert not syntax_tree.collect_errors()

76
77
    def test_whitespace_horizontal(self):
        lang = "@ whitespace = horizontal\n" + self.mini_language
78
        parser = grammar_provider(lang)()
79
        assert parser
Eckhart Arnold's avatar
Eckhart Arnold committed
80
        syntax_tree = parser("3 + 4 * 12")
81
        assert not syntax_tree.collect_errors()
Eckhart Arnold's avatar
Eckhart Arnold committed
82
        syntax_tree = parser("3 + 4 \n * 12")
83
84
        assert syntax_tree.collect_errors()

85

Eckhart Arnold's avatar
Eckhart Arnold committed
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
class TestReservedSymbols:
    def test_comment_usage(self):
        lang = r"""
        @comment = /#.*(?:\n|$)/
        document = text [ COMMENT__ ]
        text = /[^#]+/
        """
        parser = grammar_provider(lang)()

    def test_whitespace(self):
        lang = r"""
        @whitespace = /\s*/
        document = WSP__ { word WSP__ }
        word = /\w+/ 
        """
        parser = grammar_provider(lang)()

    def test_mixin(self):
        lang = r"""
        @comment = /#.*(?:\n|$)/
        @whitespace = /\s*/
        document = WSP__ { word WSP__ }
        word = /\w+/ 
        """
        parser = grammar_provider(lang)()
        result = parser("test # kommentar")
        assert not result.error_flag, str(result.as_sxpr())


115
class TestEBNFParser:
Eckhart Arnold's avatar
Eckhart Arnold committed
116
    cases = {
117
118
119
120
121
122
123
        "list_": {
            "match": {
                1: "hund",
                2: "hund, katze,maus",
                3: "hund , katze"
            },
            "fail": {
124
125
126
                4: "123",
                5: '"literal"',
                6: "/regexp/"
127
128
129
130
            }
        }
    }

131
    def setup(self):
Eckhart Arnold's avatar
Eckhart Arnold committed
132
        self.EBNF = get_ebnf_grammar()
133

134
135
    def test_RE(self):
        gr = get_ebnf_grammar()
136
137
        print(gr.regexp.parsers)
        m = gr.regexp.parsers[0].regexp.match(r'/[\\\\]/ xxx /')
138
139
140
141
        rs = m.group()
        assert rs.find('x') < 0, rs.group()
        rx = re.compile(rs[1:-1])
        assert rx.match(r'\\')
142

143
    def test_literal(self):
144
        snippet = '"text" '
Eckhart Arnold's avatar
Eckhart Arnold committed
145
        result = self.EBNF(snippet, 'literal')
146
147
        assert not result.error_flag
        assert str(result) == snippet
148
        assert result.select(lambda node: node.parser.ptype == WHITESPACE_PTYPE)
149

150
151
152
        result = self.EBNF('"text" ', 'literal')
        assert not result.error_flag
        result = self.EBNF(' "text"', 'literal')
153
154
        assert result.error_flag  # literals catch following, but not leading whitespace

155
156
157
158
159
160
161
162
    def test_plaintext(self):
        result = self.EBNF('`plain`', 'plaintext')
        assert not result.error_flag

    def test_list(self):
        grammar_unit(self.cases, get_ebnf_grammar, get_ebnf_transformer)


163

164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
class TestParserNameOverwriteBug:
    def test_term_bug(self):
        grammar = get_ebnf_grammar()
        st = grammar('impossible = [§"an optional requirement"]')
        # print(st.as_sxpr())
        get_ebnf_transformer()(st)
        # print(st.as_sxpr())
        lang = """series = "A" "B" §"C" "D"
        """
        parser = get_ebnf_grammar()
        st = grammar(lang)
        # print(st.as_sxpr())
        get_ebnf_transformer()(st)
        # print(st.as_sxpr())
        result = get_ebnf_compiler()(st)
        messages = st.collect_errors()
        assert not has_errors(messages), str(messages)


183
184
class TestSemanticValidation:
    def check(self, minilang, bool_filter=lambda x: x):
Eckhart Arnold's avatar
Eckhart Arnold committed
185
        grammar = get_ebnf_grammar()
Eckhart Arnold's avatar
Eckhart Arnold committed
186
        st = grammar(minilang)
187
        assert not st.collect_errors()
188
        EBNFTransform()(st)
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
        assert bool_filter(st.collect_errors())

    def test_illegal_nesting(self):
        self.check('impossible = { [ "an optional requirement" ] }')

    def test_illegal_nesting_option_required(self):
        self.check('impossible = [ §"an optional requirement" ]')

    def test_illegal_nesting_oneormore_option(self):
        self.check('impossible = { [ "no use"] }+')

    def test_legal_nesting(self):
        self.check('possible = { [ "+" ] "1" }', lambda x: not x)


class TestCompilerErrors:
    def test_error_propagation(self):
206
        ebnf = "@ literalws = wrongvalue  # testing error propagation\n"
Eckhart Arnold's avatar
Eckhart Arnold committed
207
208
        result, messages, st = compile_source(ebnf, None, get_ebnf_grammar(),
            get_ebnf_transformer(), get_ebnf_compiler('ErrorPropagationTest'))
209
210
        assert messages

211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
    def test_undefined_symbols(self):
        """Use of undefined symbols should be reported.
        """
        ebnf = """syntax = { intermediary }
                  intermediary = "This symbol is " [ badly_spelled ] "!"
                  bedly_spilled = "wrong" """
        result, messages, st = compile_source(ebnf, None, get_ebnf_grammar(),
            get_ebnf_transformer(), get_ebnf_compiler('UndefinedSymbols'))
        assert messages

    def test_no_error(self):
        """But reserved symbols should not be repoted as undefined.
        """
        ebnf = """nothing =  WSP__ | COMMENT__\n"""
        result, messages, st = compile_source(ebnf, None, get_ebnf_grammar(),
            get_ebnf_transformer(), get_ebnf_compiler('UndefinedSymbols'))
        assert not bool(messages), messages

229

230
class TestSelfHosting:
Eckhart Arnold's avatar
Eckhart Arnold committed
231
232
233
234
235
236
237
238
239
    grammar = r"""
        # EBNF-Grammar in EBNF

        @ comment    =  /#.*(?:\n|$)/                    # comments start with '#' and eat all chars up to and including '\n'
        @ whitespace =  /\s*/                            # whitespace includes linefeed
        @ literalws  =  right                            # trailing whitespace of literals will be ignored tacitly

        syntax     =  [~//] { definition | directive } §EOF
        definition =  symbol §"=" expression
240
        directive  =  "@" §symbol "=" ( regexp | literal | list_ )
Eckhart Arnold's avatar
Eckhart Arnold committed
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270

        expression =  term { "|" term }
        term       =  { factor }+
        factor     =  [flowmarker] [retrieveop] symbol !"="   # negative lookahead to be sure it's not a definition
                    | [flowmarker] literal
                    | [flowmarker] regexp
                    | [flowmarker] group
                    | [flowmarker] regexchain
                    | [flowmarker] oneormore
                    | repetition
                    | option

        flowmarker =  "!"  | "&"  | "§" |                # '!' negative lookahead, '&' positive lookahead, '§' required
                      "-!" | "-&"                        # '-' negative lookbehind, '-&' positive lookbehind
        retrieveop =  "::" | ":"                         # '::' pop, ':' retrieve

        group      =  "(" expression §")"
        regexchain =  ">" expression §"<"                # compiles "expression" into a singular regular expression
        oneormore  =  "{" expression "}+"
        repetition =  "{" expression §"}"
        option     =  "[" expression §"]"

        symbol     =  /(?!\d)\w+/~                       # e.g. expression, factor, parameter_list
        literal    =  /"(?:[^"]|\\")*?"/~                # e.g. "(", '+', 'while'
                    | /'(?:[^']|\\')*?'/~                # whitespace following literals will be ignored tacitly.
        regexp     =  /~?\/(?:[^\/]|(?<=\\)\/)*\/~?/~    # e.g. /\w+/, ~/#.*(?:\n|$)/~
                                                         # '~' is a whitespace-marker, if present leading or trailing
                                                         # whitespace of a regular expression will be ignored tacitly.
        list_      =  /\w+/~ { "," /\w+/~ }              # comma separated list of symbols, e.g. BEGIN_LIST, END_LIST,
                                                         # BEGIN_QUOTE, END_QUOTE ; see CommonMark/markdown.py for an exmaple
271
        EOF =  !/./
Eckhart Arnold's avatar
Eckhart Arnold committed
272
273
        """

274
275
    def test_self(self):
        compiler_name = "EBNF"
Eckhart Arnold's avatar
Eckhart Arnold committed
276
277
278
279
        compiler = get_ebnf_compiler(compiler_name, self.grammar)
        parser = get_ebnf_grammar()
        result, errors, syntax_tree = compile_source(self.grammar, None, parser,
                                            get_ebnf_transformer(), compiler)
280
281
282
        assert not errors, str(errors)
        # compile the grammar again using the result of the previous
        # compilation as parser
283
        compileDSL(self.grammar, nil_preprocessor, result, get_ebnf_transformer(), compiler)
284

Eckhart Arnold's avatar
Eckhart Arnold committed
285
286
287
288
289
290
291
292
293
    def multiprocessing_task(self):
        compiler_name = "EBNF"
        compiler = get_ebnf_compiler(compiler_name, self.grammar)
        parser = get_ebnf_grammar()
        result, errors, syntax_tree = compile_source(self.grammar, None, parser,
                                            get_ebnf_transformer(), compiler)
        return errors

    def test_multiprocessing(self):
Eckhart Arnold's avatar
Eckhart Arnold committed
294
        with Pool() as pool:
Eckhart Arnold's avatar
Eckhart Arnold committed
295
            res = [pool.apply_async(self.multiprocessing_task, ()) for i in range(4)]
296
            errors = [r.get(timeout=10) for r in res]
Eckhart Arnold's avatar
Eckhart Arnold committed
297
298
        for i, e in enumerate(errors):
            assert not e, ("%i: " % i) + str(e)
299
300


301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
class TestBoundaryCases:
    def setup(self):
        self.gr = get_ebnf_grammar()
        self.tr = get_ebnf_transformer()
        self.cp = get_ebnf_compiler()

    def test_empty_grammar(self):
        t = self.gr("")
        self.tr(t)
        r = self.cp(t)
        assert r

    def test_single_statement_grammar(self):
        t = self.gr("i = /i/")
        self.tr(t)
        r = self.cp(t)
        assert r

    def test_two_statement_grammar(self):
        t = self.gr("i = k {k}\nk = /k/")
        self.tr(t)
        r = self.cp(t)
        assert r

325
326
327
328
    def test_unconnected_symbols(self):
        ebnf = """root = /.*/
                  unconnected = /.*/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
329
330
331
332
333
334
335
336
        result, messages, AST = compile_source(ebnf, nil_preprocessor,
                                               get_ebnf_grammar(),
                                               get_ebnf_transformer(),
                                               get_ebnf_compiler())
        if messages:
            assert not has_errors(messages), "Unconnected rules should result in a warning, " \
                "not an error: " + str(messages)
            grammar_src = result
337
338
            grammar = compile_python_object(DHPARSER_IMPORTS + grammar_src,
                                            'get_(?:\w+_)?grammar$')()
Eckhart Arnold's avatar
Eckhart Arnold committed
339
340
341
        else:
            assert False, "EBNF compiler should warn about unconnected rules."

342
343
344
345
        assert grammar['root'], "Grammar objects should be subscriptable by parser names!"
        try:
            unconnected = grammar['unconnected']
        except KeyError:
Eckhart Arnold's avatar
Eckhart Arnold committed
346
            assert False, "Grammar objects should be able to cope with unconnected parsers!"
347
348
349
350
351
352
        try:
            nonexistant = grammar['nonexistant']
            assert False, "Grammar object shoul raise a KeyError if subscripted by " \
                          "a non-existant parser name!"
        except KeyError:
            pass
353
354
355
356
357
358
359


class TestSynonymDetection:
    def test_synonym_detection(self):
        ebnf = """a = b
                  b = /b/
        """
360
        grammar = grammar_provider(ebnf)()
361
362
        assert grammar['a'].name == 'a', grammar['a'].name
        assert grammar['b'].name == 'b', grammar['b'].name
363
        assert grammar('b').as_sxpr().count('b') == 2
364

365

366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
class TestFlowControlOperators:
    def setup(self):
        self.t1 = """
        All work and no play 
        makes Jack a dull boy
        END
        """
        self.t2 = "All word and not play makes Jack a dull boy END\n"

    def test_lookbehind_indirect(self):
        lang = r"""
            document = ws sequence doc_end ws         
            sequence = { !end word ws }+
            doc_end  = -&SUCC_LB end        
            ws       = /\s*/
            end      = /END/
            word     = /\w+/
            SUCC_LB  = indirection
384
            indirection = /\s*?\n/
385
386
387
388
389
        """
        parser = grammar_provider(lang)()
        cst = parser(self.t1)
        assert not cst.error_flag, cst.as_sxpr()
        cst = parser(self.t2)
390
        # this should fail, because 'END' is not preceded by a line feed
391
392
        assert cst.error_flag, cst.as_sxpr()

393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
    def test_required_error_reporting(self):
        """Tests whether failures to comply with the required operator '§'
        are correctly reported as such.
        """
        lang1 = "nonsense == /\w+/~  # wrong_equal_sign"
        lang2 = "nonsense = [^{}%]+  # someone forgot the '/'-delimiters for regular expressions"
        try:
            parser_class = grammar_provider(lang1)
            assert False, "Compilation error expected."
        except CompilationError as error:
            pass
        try:
            parser_class = grammar_provider(lang2)
            assert False, "Compilation error expected."
        except CompilationError as error:
            pass

410

411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
class TestWhitespace:
    def test_whitespace(self):
        tail = r"""
            WORD     =  /\w+/~
            EOF      =  !/./
        """
        lang1 = r'document = "DOC" { WORD } EOF' + tail
        parser = grammar_provider(lang1)()
        cst = parser("DOC Wörter Wörter Wörter")
        assert not cst.error_flag
        cst = parser("DOCWörter Wörter Wörter")
        assert not cst.error_flag

        lang2 = r'document = `DOC` { WORD } EOF' + tail
        parser = grammar_provider(lang2)()
        cst = parser("DOC Wörter Wörter Wörter")
        assert cst.error_flag
        cst = parser("DOCWörter Wörter Wörter")
        assert not cst.error_flag

        lang3 = r'document = `DOC` ~ { WORD } EOF' + tail
        parser = grammar_provider(lang3)()
        cst = parser("DOC Wörter Wörter Wörter")
        assert not cst.error_flag
        cst = parser("DOCWörter Wörter Wörter")
        assert not cst.error_flag


439
440
441
442
class TestAllSome:
    def test_all(self):
        ebnf = 'prefix = <"A" "B">'
        grammar = grammar_provider(ebnf)()
443
        assert grammar('B A').content == 'B A'
444
445
446
447

    def test_some(self):
        ebnf = 'prefix = <"A" | "B">'
        grammar = grammar_provider(ebnf)()
448
449
        assert grammar('B A').content == 'B A'
        assert grammar('B').content == 'B'
450

451

452
if __name__ == "__main__":
453
    from DHParser.testing import runner
454

455
    runner("", globals())