Hot-keys on this page

r m x p   toggle line displays

j k   next/prev highlighted chunk

0   (zero) top of page

1   (one) first highlighted chunk

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

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

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

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

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

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

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

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

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

# Natural Language Toolkit: Interface to the Mace4 Model Builder 

# 

# Author: Dan Garrette <dhgarrette@gmail.com> 

#         Ewan Klein <ewan@inf.ed.ac.uk> 

 

# URL: <http://www.nltk.org/> 

# For license information, see LICENSE.TXT 

 

""" 

A model builder that makes use of the external 'Mace4' package. 

""" 

from __future__ import print_function 

 

import os 

import tempfile 

 

from nltk.sem.logic import is_indvar 

from nltk.sem import Valuation, LogicParser 

 

from nltk.inference.api import ModelBuilder, BaseModelBuilderCommand 

from nltk.inference.prover9 import Prover9CommandParent, Prover9Parent 

 

 

class MaceCommand(Prover9CommandParent, BaseModelBuilderCommand): 

    """ 

    A ``MaceCommand`` specific to the ``Mace`` model builder.  It contains 

    a print_assumptions() method that is used to print the list 

    of assumptions in multiple formats. 

    """ 

    _interpformat_bin = None 

 

    def __init__(self, goal=None, assumptions=None, max_models=500, model_builder=None): 

        """ 

        :param goal: Input expression to prove 

        :type goal: sem.Expression 

        :param assumptions: Input expressions to use as assumptions in 

            the proof. 

        :type assumptions: list(sem.Expression) 

        :param max_models: The maximum number of models that Mace will try before 

            simply returning false. (Use 0 for no maximum.) 

        :type max_models: int 

        """ 

        if model_builder is not None: 

            assert isinstance(model_builder, Mace) 

        else: 

            model_builder = Mace(max_models) 

 

        BaseModelBuilderCommand.__init__(self, model_builder, goal, assumptions) 

 

    @property 

    def valuation(mbc): return mbc.model('valuation') 

 

    def _convert2val(self, valuation_str): 

        """ 

        Transform the output file into an NLTK-style Valuation. 

 

        :return: A model if one is generated; None otherwise. 

        :rtype: sem.Valuation 

        """ 

        valuation_standard_format = self._transform_output(valuation_str, 'standard') 

 

        val = [] 

        for line in valuation_standard_format.splitlines(False): 

            l = line.strip() 

 

            if l.startswith('interpretation'): 

                # find the number of entities in the model 

                num_entities = int(l[l.index('(')+1:l.index(',')].strip()) 

 

            elif l.startswith('function') and l.find('_') == -1: 

                # replace the integer identifier with a corresponding alphabetic character 

                name = l[l.index('(')+1:l.index(',')].strip() 

                if is_indvar(name): 

                    name = name.upper() 

                value = int(l[l.index('[')+1:l.index(']')].strip()) 

                val.append((name, MaceCommand._make_model_var(value))) 

 

            elif l.startswith('relation'): 

                l = l[l.index('(')+1:] 

                if '(' in l: 

                    #relation is not nullary 

                    name = l[:l.index('(')].strip() 

                    values = [int(v.strip()) for v in l[l.index('[')+1:l.index(']')].split(',')] 

                    val.append((name, MaceCommand._make_relation_set(num_entities, values))) 

                else: 

                    #relation is nullary 

                    name = l[:l.index(',')].strip() 

                    value = int(l[l.index('[')+1:l.index(']')].strip()) 

                    val.append((name, value == 1)) 

 

        return Valuation(val) 

 

    @staticmethod 

    def _make_relation_set(num_entities, values): 

        """ 

        Convert a Mace4-style relation table into a dictionary. 

 

        :param num_entities: the number of entities in the model; determines the row length in the table. 

        :type num_entities: int 

        :param values: a list of 1's and 0's that represent whether a relation holds in a Mace4 model. 

        :type values: list of int 

        """ 

        r = set() 

        for position in [pos for (pos,v) in enumerate(values) if v == 1]: 

            r.add(tuple(MaceCommand._make_relation_tuple(position, values, num_entities))) 

        return r 

 

    @staticmethod 

    def _make_relation_tuple(position, values, num_entities): 

        if len(values) == 1: 

            return [] 

        else: 

            sublist_size = len(values) / num_entities 

            sublist_start = position / sublist_size 

            sublist_position = position % sublist_size 

 

            sublist = values[sublist_start*sublist_size:(sublist_start+1)*sublist_size] 

            return [MaceCommand._make_model_var(sublist_start)] + \ 

                   MaceCommand._make_relation_tuple(sublist_position, 

                                                    sublist, 

                                                    num_entities) 

 

    @staticmethod 

    def _make_model_var(value): 

        """ 

        Pick an alphabetic character as identifier for an entity in the model. 

 

        :param value: where to index into the list of characters 

        :type value: int 

        """ 

        letter = ['a','b','c','d','e','f','g','h','i','j','k','l','m','n', 

                  'o','p','q','r','s','t','u','v','w','x','y','z'][value] 

        num = int(value) / 26 

        if num > 0: 

            return letter + str(num) 

        else: 

            return letter 

 

    def _decorate_model(self, valuation_str, format): 

        """ 

        Print out a Mace4 model using any Mace4 ``interpformat`` format. 

        See http://www.cs.unm.edu/~mccune/mace4/manual/ for details. 

 

        :param valuation_str: str with the model builder's output 

        :param format: str indicating the format for displaying 

        models. Defaults to 'standard' format. 

        :return: str 

        """ 

        if not format: 

            return valuation_str 

        elif format == 'valuation': 

            return self._convert2val(valuation_str) 

        else: 

            return self._transform_output(valuation_str, format) 

 

    def _transform_output(self, valuation_str, format): 

        """ 

        Transform the output file into any Mace4 ``interpformat`` format. 

 

        :param format: Output format for displaying models. 

        :type format: str 

        """ 

        if format in ['standard', 'standard2', 'portable', 'tabular', 

                      'raw', 'cooked', 'xml', 'tex']: 

            return self._call_interpformat(valuation_str, [format])[0] 

        else: 

            raise LookupError("The specified format does not exist") 

 

    def _call_interpformat(self, input_str, args=[], verbose=False): 

        """ 

        Call the ``interpformat`` binary with the given input. 

 

        :param input_str: A string whose contents are used as stdin. 

        :param args: A list of command-line arguments. 

        :return: A tuple (stdout, returncode) 

        :see: ``config_prover9`` 

        """ 

        if self._interpformat_bin is None: 

            self._interpformat_bin = self._modelbuilder._find_binary( 

                                                'interpformat', verbose) 

 

        return self._modelbuilder._call(input_str, self._interpformat_bin, 

                                        args, verbose) 

 

 

class Mace(Prover9Parent, ModelBuilder): 

    _mace4_bin = None 

 

    def __init__(self, end_size=500): 

        self._end_size = end_size 

        """The maximum model size that Mace will try before 

           simply returning false. (Use -1 for no maximum.)""" 

 

    def _build_model(self, goal=None, assumptions=None, verbose=False): 

        """ 

        Use Mace4 to build a first order model. 

 

        :return: ``True`` if a model was found (i.e. Mace returns value of 0), 

        else ``False`` 

        """ 

        if not assumptions: 

            assumptions = [] 

 

        stdout, returncode = self._call_mace4(self.prover9_input(goal, assumptions), 

                                              verbose=verbose) 

        return (returncode == 0, stdout) 

 

    def _call_mace4(self, input_str, args=[], verbose=False): 

        """ 

        Call the ``mace4`` binary with the given input. 

 

        :param input_str: A string whose contents are used as stdin. 

        :param args: A list of command-line arguments. 

        :return: A tuple (stdout, returncode) 

        :see: ``config_prover9`` 

        """ 

        if self._mace4_bin is None: 

            self._mace4_bin = self._find_binary('mace4', verbose) 

 

        updated_input_str = '' 

        if self._end_size > 0: 

            updated_input_str += 'assign(end_size, %d).\n\n' % self._end_size 

        updated_input_str += input_str 

 

        return self._call(updated_input_str, self._mace4_bin, args, verbose) 

 

 

def spacer(num=30): 

    print('-' * num) 

 

def decode_result(found): 

    """ 

    Decode the result of model_found() 

 

    :param found: The output of model_found() 

    :type found: bool 

    """ 

    return {True: 'Countermodel found', False: 'No countermodel found', None: 'None'}[found] 

 

def test_model_found(arguments): 

    """ 

    Try some proofs and exhibit the results. 

    """ 

    lp = LogicParser() 

    for (goal, assumptions) in arguments: 

        g = lp.parse(goal) 

        alist = [lp.parse(a) for a in assumptions] 

        m = MaceCommand(g, assumptions=alist, end_size=50) 

        found = m.build_model() 

        for a in alist: 

            print('   %s' % a) 

        print('|- %s: %s\n' % (g, decode_result(found))) 

 

 

def test_build_model(arguments): 

    """ 

    Try to build a ``nltk.sem.Valuation``. 

    """ 

    lp = LogicParser() 

    g = lp.parse('all x.man(x)') 

    alist = [lp.parse(a) for a in ['man(John)', 

                                   'man(Socrates)', 

                                   'man(Bill)', 

                                   'some x.(-(x = John) & man(x) & sees(John,x))', 

                                   'some x.(-(x = Bill) & man(x))', 

                                   'all x.some y.(man(x) -> gives(Socrates,x,y))']] 

 

    m = MaceCommand(g, assumptions=alist) 

    m.build_model() 

    spacer() 

    print("Assumptions and Goal") 

    spacer() 

    for a in alist: 

        print('   %s' % a) 

    print('|- %s: %s\n' % (g, decode_result(m.build_model()))) 

    spacer() 

    #print m.model('standard') 

    #print m.model('cooked') 

    print("Valuation") 

    spacer() 

    print(m.valuation, '\n') 

 

def test_transform_output(argument_pair): 

    """ 

    Transform the model into various Mace4 ``interpformat`` formats. 

    """ 

    lp = LogicParser() 

    g = lp.parse(argument_pair[0]) 

    alist = [lp.parse(a) for a in argument_pair[1]] 

    m = MaceCommand(g, assumptions=alist) 

    m.build_model() 

    for a in alist: 

        print('   %s' % a) 

    print('|- %s: %s\n' % (g, m.build_model())) 

    for format in ['standard', 'portable', 'xml', 'cooked']: 

        spacer() 

        print("Using '%s' format" % format) 

        spacer() 

        print(m.model(format=format)) 

 

def test_make_relation_set(): 

    print(MaceCommand._make_relation_set(num_entities=3, values=[1,0,1]) == set([('c',), ('a',)])) 

    print(MaceCommand._make_relation_set(num_entities=3, values=[0,0,0,0,0,0,1,0,0]) == set([('c', 'a')])) 

    print(MaceCommand._make_relation_set(num_entities=2, values=[0,0,1,0,0,0,1,0]) == set([('a', 'b', 'a'), ('b', 'b', 'a')])) 

 

arguments = [ 

    ('mortal(Socrates)', ['all x.(man(x) -> mortal(x))', 'man(Socrates)']), 

    ('(not mortal(Socrates))', ['all x.(man(x) -> mortal(x))', 'man(Socrates)']) 

] 

 

def demo(): 

    test_model_found(arguments) 

    test_build_model(arguments) 

    test_transform_output(arguments[1]) 

 

if __name__ == '__main__': 

    demo()