Statistics
| Revision:

svn-gvsig-desktop / tags / v1_0_2_Build_905 / extensions / extScripting / scripts / jython / Lib / pyclbr.py @ 10767

History | View | Annotate | Download (11.3 KB)

1
"""Parse a Python file and retrieve classes and methods.
2

3
Parse enough of a Python file to recognize class and method
4
definitions and to find out the superclasses of a class.
5

6
The interface consists of a single function:
7
        readmodule(module, path)
8
module is the name of a Python module, path is an optional list of
9
directories where the module is to be searched.  If present, path is
10
prepended to the system search path sys.path.
11
The return value is a dictionary.  The keys of the dictionary are
12
the names of the classes defined in the module (including classes
13
that are defined via the from XXX import YYY construct).  The values
14
are class instances of the class Class defined here.
15

16
A class is described by the class Class in this module.  Instances
17
of this class have the following instance variables:
18
        name -- the name of the class
19
        super -- a list of super classes (Class instances)
20
        methods -- a dictionary of methods
21
        file -- the file in which the class was defined
22
        lineno -- the line in the file on which the class statement occurred
23
The dictionary of methods uses the method names as keys and the line
24
numbers on which the method was defined as values.
25
If the name of a super class is not recognized, the corresponding
26
entry in the list of super classes is not a class instance but a
27
string giving the name of the super class.  Since import statements
28
are recognized and imported modules are scanned as well, this
29
shouldn't happen often.
30

31
BUGS
32
- Continuation lines are not dealt with at all.
33
- While triple-quoted strings won't confuse it, lines that look like
34
  def, class, import or "from ... import" stmts inside backslash-continued
35
  single-quoted strings are treated like code.  The expense of stopping
36
  that isn't worth it.
37
- Code that doesn't pass tabnanny or python -t will confuse it, unless
38
  you set the module TABWIDTH vrbl (default 8) to the correct tab width
39
  for the file.
40

41
PACKAGE RELATED BUGS
42
- If you have a package and a module inside that or another package
43
  with the same name, module caching doesn't work properly since the
44
  key is the base name of the module/package.
45
- The only entry that is returned when you readmodule a package is a
46
  __path__ whose value is a list which confuses certain class browsers.
47
- When code does:
48
  from package import subpackage
49
  class MyClass(subpackage.SuperClass):
50
    ...
51
  It can't locate the parent.  It probably needs to have the same
52
  hairy logic that the import locator already does.  (This logic
53
  exists coded in Python in the freeze package.)
54
"""
55

    
56
import os
57
import sys
58
import imp
59
import re
60
import string
61

    
62
__all__ = ["readmodule"]
63

    
64
TABWIDTH = 8
65

    
66
_getnext = re.compile(r"""
67
    (?P<String>
68
       \""" [^"\\]* (?:
69
                        (?: \\. | "(?!"") )
70
                        [^"\\]*
71
                    )*
72
       \"""
73

74
    |   ''' [^'\\]* (?:
75
                        (?: \\. | '(?!'') )
76
                        [^'\\]*
77
                    )*
78
        '''
79
    )
80

81
|   (?P<Method>
82
        ^
83
        (?P<MethodIndent> [ \t]* )
84
        def [ \t]+
85
        (?P<MethodName> [a-zA-Z_] \w* )
86
        [ \t]* \(
87
    )
88

89
|   (?P<Class>
90
        ^
91
        (?P<ClassIndent> [ \t]* )
92
        class [ \t]+
93
        (?P<ClassName> [a-zA-Z_] \w* )
94
        [ \t]*
95
        (?P<ClassSupers> \( [^)\n]* \) )?
96
        [ \t]* :
97
    )
98

99
|   (?P<Import>
100
        ^ import [ \t]+
101
        (?P<ImportList> [^#;\n]+ )
102
    )
103

104
|   (?P<ImportFrom>
105
        ^ from [ \t]+
106
        (?P<ImportFromPath>
107
            [a-zA-Z_] \w*
108
            (?:
109
                [ \t]* \. [ \t]* [a-zA-Z_] \w*
110
            )*
111
        )
112
        [ \t]+
113
        import [ \t]+
114
        (?P<ImportFromList> [^#;\n]+ )
115
    )
116
""", re.VERBOSE | re.DOTALL | re.MULTILINE).search
117

    
118
_modules = {}                           # cache of modules we've seen
119

    
120
# each Python class is represented by an instance of this class
121
class Class:
122
    '''Class to represent a Python class.'''
123
    def __init__(self, module, name, super, file, lineno):
124
        self.module = module
125
        self.name = name
126
        if super is None:
127
            super = []
128
        self.super = super
129
        self.methods = {}
130
        self.file = file
131
        self.lineno = lineno
132

    
133
    def _addmethod(self, name, lineno):
134
        self.methods[name] = lineno
135

    
136
class Function(Class):
137
    '''Class to represent a top-level Python function'''
138
    def __init__(self, module, name, file, lineno):
139
        Class.__init__(self, module, name, None, file, lineno)
140
    def _addmethod(self, name, lineno):
141
        assert 0, "Function._addmethod() shouldn't be called"
142

    
143
def readmodule(module, path=[], inpackage=0):
144
    '''Backwards compatible interface.
145

146
    Like readmodule_ex() but strips Function objects from the
147
    resulting dictionary.'''
148

    
149
    dict = readmodule_ex(module, path, inpackage)
150
    res = {}
151
    for key, value in dict.items():
152
        if not isinstance(value, Function):
153
            res[key] = value
154
    return res
155

    
156
def readmodule_ex(module, path=[], inpackage=0):
157
    '''Read a module file and return a dictionary of classes.
158

159
    Search for MODULE in PATH and sys.path, read and parse the
160
    module and return a dictionary with one entry for each class
161
    found in the module.'''
162

    
163
    dict = {}
164

    
165
    i = module.rfind('.')
166
    if i >= 0:
167
        # Dotted module name
168
        package = module[:i].strip()
169
        submodule = module[i+1:].strip()
170
        parent = readmodule(package, path, inpackage)
171
        child = readmodule(submodule, parent['__path__'], 1)
172
        return child
173

    
174
    if _modules.has_key(module):
175
        # we've seen this module before...
176
        return _modules[module]
177
    if module in sys.builtin_module_names:
178
        # this is a built-in module
179
        _modules[module] = dict
180
        return dict
181

    
182
    # search the path for the module
183
    f = None
184
    if inpackage:
185
        try:
186
            f, file, (suff, mode, type) = \
187
                    imp.find_module(module, path)
188
        except ImportError:
189
            f = None
190
    if f is None:
191
        fullpath = list(path) + sys.path
192
        f, file, (suff, mode, type) = imp.find_module(module, fullpath)
193
    if type == imp.PKG_DIRECTORY:
194
        dict['__path__'] = [file]
195
        _modules[module] = dict
196
        path = [file] + path
197
        f, file, (suff, mode, type) = \
198
                        imp.find_module('__init__', [file])
199
    if type != imp.PY_SOURCE:
200
        # not Python source, can't do anything with this module
201
        f.close()
202
        _modules[module] = dict
203
        return dict
204

    
205
    _modules[module] = dict
206
    imports = []
207
    classstack = [] # stack of (class, indent) pairs
208
    src = f.read()
209
    f.close()
210

    
211
    # To avoid having to stop the regexp at each newline, instead
212
    # when we need a line number we simply string.count the number of
213
    # newlines in the string since the last time we did this; i.e.,
214
    #    lineno = lineno + \
215
    #             string.count(src, '\n', last_lineno_pos, here)
216
    #    last_lineno_pos = here
217
    countnl = string.count
218
    lineno, last_lineno_pos = 1, 0
219
    i = 0
220
    while 1:
221
        m = _getnext(src, i)
222
        if not m:
223
            break
224
        start, i = m.span()
225

    
226
        if m.start("Method") >= 0:
227
            # found a method definition or function
228
            thisindent = _indent(m.group("MethodIndent"))
229
            meth_name = m.group("MethodName")
230
            lineno = lineno + \
231
                     countnl(src, '\n',
232
                             last_lineno_pos, start)
233
            last_lineno_pos = start
234
            # close all classes indented at least as much
235
            while classstack and \
236
                  classstack[-1][1] >= thisindent:
237
                del classstack[-1]
238
            if classstack:
239
                # it's a class method
240
                cur_class = classstack[-1][0]
241
                cur_class._addmethod(meth_name, lineno)
242
            else:
243
                # it's a function
244
                f = Function(module, meth_name,
245
                             file, lineno)
246
                dict[meth_name] = f
247

    
248
        elif m.start("String") >= 0:
249
            pass
250

    
251
        elif m.start("Class") >= 0:
252
            # we found a class definition
253
            thisindent = _indent(m.group("ClassIndent"))
254
            # close all classes indented at least as much
255
            while classstack and \
256
                  classstack[-1][1] >= thisindent:
257
                del classstack[-1]
258
            lineno = lineno + \
259
                     countnl(src, '\n', last_lineno_pos, start)
260
            last_lineno_pos = start
261
            class_name = m.group("ClassName")
262
            inherit = m.group("ClassSupers")
263
            if inherit:
264
                # the class inherits from other classes
265
                inherit = inherit[1:-1].strip()
266
                names = []
267
                for n in inherit.split(','):
268
                    n = n.strip()
269
                    if dict.has_key(n):
270
                        # we know this super class
271
                        n = dict[n]
272
                    else:
273
                        c = n.split('.')
274
                        if len(c) > 1:
275
                            # super class
276
                            # is of the
277
                            # form module.class:
278
                            # look in
279
                            # module for class
280
                            m = c[-2]
281
                            c = c[-1]
282
                            if _modules.has_key(m):
283
                                d = _modules[m]
284
                                if d.has_key(c):
285
                                    n = d[c]
286
                    names.append(n)
287
                inherit = names
288
            # remember this class
289
            cur_class = Class(module, class_name, inherit,
290
                              file, lineno)
291
            dict[class_name] = cur_class
292
            classstack.append((cur_class, thisindent))
293

    
294
        elif m.start("Import") >= 0:
295
            # import module
296
            for n in m.group("ImportList").split(','):
297
                n = n.strip()
298
                try:
299
                    # recursively read the imported module
300
                    d = readmodule(n, path, inpackage)
301
                except:
302
                    ##print 'module', n, 'not found'
303
                    pass
304

    
305
        elif m.start("ImportFrom") >= 0:
306
            # from module import stuff
307
            mod = m.group("ImportFromPath")
308
            names = m.group("ImportFromList").split(',')
309
            try:
310
                # recursively read the imported module
311
                d = readmodule(mod, path, inpackage)
312
            except:
313
                ##print 'module', mod, 'not found'
314
                continue
315
            # add any classes that were defined in the
316
            # imported module to our name space if they
317
            # were mentioned in the list
318
            for n in names:
319
                n = n.strip()
320
                if d.has_key(n):
321
                    dict[n] = d[n]
322
                elif n == '*':
323
                    # only add a name if not
324
                    # already there (to mimic what
325
                    # Python does internally)
326
                    # also don't add names that
327
                    # start with _
328
                    for n in d.keys():
329
                        if n[0] != '_' and \
330
                           not dict.has_key(n):
331
                            dict[n] = d[n]
332
        else:
333
            assert 0, "regexp _getnext found something unexpected"
334

    
335
    return dict
336

    
337
def _indent(ws, _expandtabs=string.expandtabs):
338
    return len(_expandtabs(ws, TABWIDTH))