test_ebnf.py 14.2 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.
"""

23
24
25
26
try:
    import regex as re
except ImportError:
    import re
Eckhart Arnold's avatar
Eckhart Arnold committed
27
import sys
Eckhart Arnold's avatar
Eckhart Arnold committed
28
from multiprocessing import Pool
Eckhart Arnold's avatar
Eckhart Arnold committed
29

Eckhart Arnold's avatar
Eckhart Arnold committed
30
31
sys.path.extend(['../', './'])

32
from DHParser.toolkit import compile_python_object
Eckhart Arnold's avatar
Eckhart Arnold committed
33
from DHParser.syntaxtree import has_errors
34
from DHParser.parser import compile_source, WHITESPACE_PTYPE, nil_preprocessor
35
from DHParser.ebnf import get_ebnf_grammar, get_ebnf_transformer, EBNFTransform, get_ebnf_compiler
36
from DHParser.dsl import CompilationError, compileDSL, DHPARSER_IMPORTS, grammar_provider
37
38


39
40
41
42
43
44
45
46
47
48
49
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
50
        MinilangParser = grammar_provider(lang)
51
52
        parser = MinilangParser()
        assert parser
Eckhart Arnold's avatar
Eckhart Arnold committed
53
        syntax_tree = parser("3 + 4 * 12")
54
        # parser.log_parsing_history("WSP")
55
        assert not syntax_tree.collect_errors()
Eckhart Arnold's avatar
Eckhart Arnold committed
56
        syntax_tree = parser("3 + 4 \n * 12")
57
58
        # parser.log_parsing_history("WSPLF")
        assert not syntax_tree.collect_errors()
Eckhart Arnold's avatar
Eckhart Arnold committed
59
        syntax_tree = parser("3 + 4 \n \n * 12")
60
        assert syntax_tree.collect_errors()
Eckhart Arnold's avatar
Eckhart Arnold committed
61
        syntax_tree = parser("3 + 4 \n\n * 12")
62
63
64
65
        assert syntax_tree.collect_errors()

    def test_whitespace_vertical(self):
        lang = "@ whitespace = vertical\n" + self.mini_language
66
        parser = grammar_provider(lang)()
67
        assert parser
Eckhart Arnold's avatar
Eckhart Arnold committed
68
        syntax_tree = parser("3 + 4 * 12")
69
        assert not syntax_tree.collect_errors()
Eckhart Arnold's avatar
Eckhart Arnold committed
70
        syntax_tree = parser("3 + 4 \n * 12")
71
        assert not syntax_tree.collect_errors()
Eckhart Arnold's avatar
Eckhart Arnold committed
72
        syntax_tree = parser("3 + 4 \n \n * 12")
73
        assert not syntax_tree.collect_errors()
Eckhart Arnold's avatar
Eckhart Arnold committed
74
        syntax_tree = parser("3 + 4 \n\n * 12")
75
76
        assert not syntax_tree.collect_errors()

77
78
    def test_whitespace_horizontal(self):
        lang = "@ whitespace = horizontal\n" + self.mini_language
79
        parser = grammar_provider(lang)()
80
        assert parser
Eckhart Arnold's avatar
Eckhart Arnold committed
81
        syntax_tree = parser("3 + 4 * 12")
82
        assert not syntax_tree.collect_errors()
Eckhart Arnold's avatar
Eckhart Arnold committed
83
        syntax_tree = parser("3 + 4 \n * 12")
84
85
        assert syntax_tree.collect_errors()

86

Eckhart Arnold's avatar
Eckhart Arnold committed
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
115
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())


116
class TestEBNFParser:
Eckhart Arnold's avatar
Eckhart Arnold committed
117
    cases = {
118
119
120
121
122
123
124
125
126
127
128
129
130
131
        "list_": {
            "match": {
                1: "hund",
                2: "hund, katze,maus",
                3: "hund , katze"
            },
            "fail": {
                1: "123",
                2: '"literal"',
                3: "/regexp/"
            }
        }
    }

di68kap's avatar
di68kap committed
132

133
    def setup(self):
Eckhart Arnold's avatar
Eckhart Arnold committed
134
        self.EBNF = get_ebnf_grammar()
135

136
137
    def test_RE(self):
        gr = get_ebnf_grammar()
138
139
140
141
142
        m = gr.regexp.main.regexp.match(r'/[\\\\]/ xxx /')
        rs = m.group()
        assert rs.find('x') < 0, rs.group()
        rx = re.compile(rs[1:-1])
        assert rx.match(r'\\')
143

144
145
    def test_literal(self):
        snippet = '"literal" '
Eckhart Arnold's avatar
Eckhart Arnold committed
146
        result = self.EBNF(snippet, 'literal')
147
148
        assert not result.error_flag
        assert str(result) == snippet
149
        assert result.find(lambda node: node.parser.ptype == WHITESPACE_PTYPE)
150

Eckhart Arnold's avatar
Eckhart Arnold committed
151
        result = self.EBNF(' "literal"', 'literal')
152
153
154
        assert result.error_flag  # literals catch following, but not leading whitespace


155
156
class TestSemanticValidation:
    def check(self, minilang, bool_filter=lambda x: x):
Eckhart Arnold's avatar
Eckhart Arnold committed
157
        grammar = get_ebnf_grammar()
Eckhart Arnold's avatar
Eckhart Arnold committed
158
        st = grammar(minilang)
159
        assert not st.collect_errors()
160
        EBNFTransform()(st)
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
        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):
178
        ebnf = "@ literalws = wrongvalue  # testing error propagation\n"
Eckhart Arnold's avatar
Eckhart Arnold committed
179
180
        result, messages, st = compile_source(ebnf, None, get_ebnf_grammar(),
            get_ebnf_transformer(), get_ebnf_compiler('ErrorPropagationTest'))
181
182
        assert messages

183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
    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

201

202
class TestSelfHosting:
Eckhart Arnold's avatar
Eckhart Arnold committed
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
    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
        directive  =  "@" §symbol §"=" ( regexp | literal | list_ )

        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
        EOF =  !/./        
        """

246
247
    def test_self(self):
        compiler_name = "EBNF"
Eckhart Arnold's avatar
Eckhart Arnold committed
248
249
250
251
        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)
252
253
254
        assert not errors, str(errors)
        # compile the grammar again using the result of the previous
        # compilation as parser
255
        compileDSL(self.grammar, nil_preprocessor, result, get_ebnf_transformer(), compiler)
256

Eckhart Arnold's avatar
Eckhart Arnold committed
257
258
259
260
261
262
263
264
265
266
267
    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):
        with Pool(processes=2) as pool:
            res = [pool.apply_async(self.multiprocessing_task, ()) for i in range(4)]
268
            errors = [r.get(timeout=10) for r in res]
Eckhart Arnold's avatar
Eckhart Arnold committed
269
270
        for i, e in enumerate(errors):
            assert not e, ("%i: " % i) + str(e)
271
272


273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
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

297
298
299
300
    def test_unconnected_symbols(self):
        ebnf = """root = /.*/
                  unconnected = /.*/
        """
Eckhart Arnold's avatar
Eckhart Arnold committed
301
302
303
304
305
306
307
308
        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
309
310
            grammar = compile_python_object(DHPARSER_IMPORTS + grammar_src,
                                            'get_(?:\w+_)?grammar$')()
Eckhart Arnold's avatar
Eckhart Arnold committed
311
312
313
        else:
            assert False, "EBNF compiler should warn about unconnected rules."

314
315
316
317
        assert grammar['root'], "Grammar objects should be subscriptable by parser names!"
        try:
            unconnected = grammar['unconnected']
        except KeyError:
Eckhart Arnold's avatar
Eckhart Arnold committed
318
            assert False, "Grammar objects should be able to cope with unconnected parsers!"
319
320
321
322
323
324
        try:
            nonexistant = grammar['nonexistant']
            assert False, "Grammar object shoul raise a KeyError if subscripted by " \
                          "a non-existant parser name!"
        except KeyError:
            pass
325
326
327
328
329
330
331


class TestSynonymDetection:
    def test_synonym_detection(self):
        ebnf = """a = b
                  b = /b/
        """
332
        grammar = grammar_provider(ebnf)()
333
334
        assert grammar['a'].name == 'a', grammar['a'].name
        assert grammar['b'].name == 'b', grammar['b'].name
335
        assert grammar('b').as_sxpr().count('b') == 2
336

337

338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
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
356
            indirection = /\s*?\n/
357
358
359
360
361
        """
        parser = grammar_provider(lang)()
        cst = parser(self.t1)
        assert not cst.error_flag, cst.as_sxpr()
        cst = parser(self.t2)
362
        # this should fail, because 'END' is not preceded by a line feed
363
364
        assert cst.error_flag, cst.as_sxpr()

365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
    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
            # print(error)
        try:
            parser_class = grammar_provider(lang2)
            assert False, "Compilation error expected."
        except CompilationError as error:
            pass
            # print(error)

384

385

386
if __name__ == "__main__":
387
388
    from DHParser.testing import runner
    runner("", globals())