test_ebnf.py 13.6 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
33
from DHParser.toolkit import compile_python_object
from DHParser.parser import compile_source, WHITESPACE_PTYPE, nil_preprocessor
34
from DHParser.ebnf import get_ebnf_grammar, get_ebnf_transformer, EBNFTransform, get_ebnf_compiler
35
from DHParser.dsl import CompilationError, compileDSL, DHPARSER_IMPORTS, grammar_provider
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
124
125
126
127
128
129
130
        "list_": {
            "match": {
                1: "hund",
                2: "hund, katze,maus",
                3: "hund , katze"
            },
            "fail": {
                1: "123",
                2: '"literal"',
                3: "/regexp/"
            }
        }
    }

di68kap's avatar
di68kap committed
131

132
    def setup(self):
Eckhart Arnold's avatar
Eckhart Arnold committed
133
        self.EBNF = get_ebnf_grammar()
134

135
136
    def test_RE(self):
        gr = get_ebnf_grammar()
137
138
139
140
141
        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'\\')
142

143
144
    def test_literal(self):
        snippet = '"literal" '
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.find(lambda node: node.parser.ptype == WHITESPACE_PTYPE)
149

Eckhart Arnold's avatar
Eckhart Arnold committed
150
        result = self.EBNF(' "literal"', 'literal')
151
152
153
        assert result.error_flag  # literals catch following, but not leading whitespace


154
155
class TestSemanticValidation:
    def check(self, minilang, bool_filter=lambda x: x):
Eckhart Arnold's avatar
Eckhart Arnold committed
156
        grammar = get_ebnf_grammar()
Eckhart Arnold's avatar
Eckhart Arnold committed
157
        st = grammar(minilang)
158
        assert not st.collect_errors()
159
        EBNFTransform()(st)
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
        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):
177
        ebnf = "@ literalws = wrongvalue  # testing error propagation\n"
Eckhart Arnold's avatar
Eckhart Arnold committed
178
179
        result, messages, st = compile_source(ebnf, None, get_ebnf_grammar(),
            get_ebnf_transformer(), get_ebnf_compiler('ErrorPropagationTest'))
180
181
        assert messages

182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
    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

200

201
class TestSelfHosting:
Eckhart Arnold's avatar
Eckhart Arnold committed
202
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
    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 =  !/./        
        """

245
246
    def test_self(self):
        compiler_name = "EBNF"
Eckhart Arnold's avatar
Eckhart Arnold committed
247
248
249
250
        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)
251
252
253
        assert not errors, str(errors)
        # compile the grammar again using the result of the previous
        # compilation as parser
254
        compileDSL(self.grammar, nil_preprocessor, result, get_ebnf_transformer(), compiler)
255

Eckhart Arnold's avatar
Eckhart Arnold committed
256
257
258
259
260
261
262
263
264
265
266
    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)]
267
            errors = [r.get(timeout=10) for r in res]
Eckhart Arnold's avatar
Eckhart Arnold committed
268
269
        for i, e in enumerate(errors):
            assert not e, ("%i: " % i) + str(e)
270
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
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

296
297
298
299
300
    def test_unconnected_symbols(self):
        ebnf = """root = /.*/
                  unconnected = /.*/
        """
        try:
301
            grammar = grammar_provider(ebnf)()
302
303
            assert False, "EBNF compiler should complain about unconnected rules."
        except CompilationError as err:
304
305
306
307
308
309
310
            grammar_src = err.result
            grammar = compile_python_object(DHPARSER_IMPORTS + grammar_src,
                                            'get_(?:\w+_)?grammar$')()
        assert grammar['root'], "Grammar objects should be subscriptable by parser names!"
        try:
            unconnected = grammar['unconnected']
        except KeyError:
Eckhart Arnold's avatar
Eckhart Arnold committed
311
            assert False, "Grammar objects should be able to cope with unconnected parsers!"
312
313
314
315
316
317
        try:
            nonexistant = grammar['nonexistant']
            assert False, "Grammar object shoul raise a KeyError if subscripted by " \
                          "a non-existant parser name!"
        except KeyError:
            pass
318
319
        ebnf_testing = "@testing = True\n" + ebnf
        try:
320
            grammar = grammar_provider(ebnf_testing)()
321
322
323
        except CompilationError:
            assert False, "EBNF compiler should not complain about unconnected " \
                          "rules when directive @testing is set."
324
325
326
327
328
329
330


class TestSynonymDetection:
    def test_synonym_detection(self):
        ebnf = """a = b
                  b = /b/
        """
331
        grammar = grammar_provider(ebnf)()
332
333
        assert grammar['a'].name == 'a', grammar['a'].name
        assert grammar['b'].name == 'b', grammar['b'].name
334
        assert grammar('b').as_sxpr().count('b') == 2
335

336

337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
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
355
            indirection = /\s*?\n/
356
357
358
359
360
361
362
363
364
365
366
367
        """
        # result, messages, syntax_tree = compile_source(lang, None, get_ebnf_grammar(),
        #                 get_ebnf_transformer(), get_ebnf_compiler('LookbehindTest'))
        # print(result)
        parser = grammar_provider(lang)()
        cst = parser(self.t1)
        assert not cst.error_flag, cst.as_sxpr()
        cst = parser(self.t2)
        # this should fail, because 'END' is not preceeded by a line feed
        assert cst.error_flag, cst.as_sxpr()


368
if __name__ == "__main__":
369
370
    from DHParser.testing import runner
    runner("", globals())