test_ebnf.py 11.8 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
28
from functools import partial
Eckhart Arnold's avatar
Eckhart Arnold committed
29
from multiprocessing import Pool
Eckhart Arnold's avatar
Eckhart Arnold committed
30

Eckhart Arnold's avatar
Eckhart Arnold committed
31
32
sys.path.extend(['../', './'])

33
from DHParser.toolkit import is_logging, compile_python_object
34
from DHParser.parsers import compile_source, Retrieve, WHITESPACE_PTYPE, nil_preprocessor
35
from DHParser.ebnf import get_ebnf_grammar, get_ebnf_transformer, EBNFTransformer, get_ebnf_compiler
36
from DHParser.dsl import CompilationError, compileDSL, DHPARSER_IMPORTS, parser_factory
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 = parser_factory(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 = parser_factory(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 = parser_factory(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

87
class TestEBNFParser:
Eckhart Arnold's avatar
Eckhart Arnold committed
88
    cases = {
89
90
91
92
93
94
95
96
97
98
99
100
101
102
        "list_": {
            "match": {
                1: "hund",
                2: "hund, katze,maus",
                3: "hund , katze"
            },
            "fail": {
                1: "123",
                2: '"literal"',
                3: "/regexp/"
            }
        }
    }

di68kap's avatar
di68kap committed
103

104
    def setup(self):
Eckhart Arnold's avatar
Eckhart Arnold committed
105
        self.EBNF = get_ebnf_grammar()
106

107
108
    def test_RE(self):
        gr = get_ebnf_grammar()
109
110
111
112
113
        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'\\')
114

115
116
    def test_literal(self):
        snippet = '"literal" '
Eckhart Arnold's avatar
Eckhart Arnold committed
117
        result = self.EBNF(snippet, 'literal')
118
119
        assert not result.error_flag
        assert str(result) == snippet
120
        assert result.find(lambda node: node.parser.ptype == WHITESPACE_PTYPE)
121

Eckhart Arnold's avatar
Eckhart Arnold committed
122
        result = self.EBNF(' "literal"', 'literal')
123
124
125
        assert result.error_flag  # literals catch following, but not leading whitespace


126
127
class TestSemanticValidation:
    def check(self, minilang, bool_filter=lambda x: x):
Eckhart Arnold's avatar
Eckhart Arnold committed
128
        grammar = get_ebnf_grammar()
Eckhart Arnold's avatar
Eckhart Arnold committed
129
        st = grammar(minilang)
130
        assert not st.collect_errors()
131
        EBNFTransformer(st)
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
        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):
149
        ebnf = "@ literalws = wrongvalue  # testing error propagation\n"
Eckhart Arnold's avatar
Eckhart Arnold committed
150
151
        result, messages, st = compile_source(ebnf, None, get_ebnf_grammar(),
            get_ebnf_transformer(), get_ebnf_compiler('ErrorPropagationTest'))
152
153
        assert messages

154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
    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

172

173
class TestSelfHosting:
Eckhart Arnold's avatar
Eckhart Arnold committed
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
    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 =  !/./        
        """

217
218
    def test_self(self):
        compiler_name = "EBNF"
Eckhart Arnold's avatar
Eckhart Arnold committed
219
220
221
222
        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)
223
224
225
        assert not errors, str(errors)
        # compile the grammar again using the result of the previous
        # compilation as parser
226
        compileDSL(self.grammar, nil_preprocessor, result, get_ebnf_transformer(), compiler)
227

Eckhart Arnold's avatar
Eckhart Arnold committed
228
229
230
231
232
233
234
235
236
237
238
    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)]
239
            errors = [r.get(timeout=10) for r in res]
Eckhart Arnold's avatar
Eckhart Arnold committed
240
241
        for i, e in enumerate(errors):
            assert not e, ("%i: " % i) + str(e)
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
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

268
269
270
271
272
    def test_unconnected_symbols(self):
        ebnf = """root = /.*/
                  unconnected = /.*/
        """
        try:
273
            grammar = parser_factory(ebnf)()
274
275
            assert False, "EBNF compiler should complain about unconnected rules."
        except CompilationError as err:
276
277
278
279
280
281
282
            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
283
            assert False, "Grammar objects should be able to cope with unconnected parsers!"
284
285
286
287
288
289
        try:
            nonexistant = grammar['nonexistant']
            assert False, "Grammar object shoul raise a KeyError if subscripted by " \
                          "a non-existant parser name!"
        except KeyError:
            pass
290
291
292
293
294
295
        ebnf_testing = "@testing = True\n" + ebnf
        try:
            grammar = parser_factory(ebnf_testing)()
        except CompilationError:
            assert False, "EBNF compiler should not complain about unconnected " \
                          "rules when directive @testing is set."
296
297
298
299
300
301
302
303
304
305


class TestSynonymDetection:
    def test_synonym_detection(self):
        ebnf = """a = b
                  b = /b/
        """
        grammar = parser_factory(ebnf)()
        assert grammar['a'].name == 'a', grammar['a'].name
        assert grammar['b'].name == 'b', grammar['b'].name
306
        assert grammar('b').as_sxpr().count('b') == 2
307

308

309
if __name__ == "__main__":
310
311
    from DHParser.testing import runner
    runner("", globals())