summaryrefslogtreecommitdiffstats
path: root/bitbake/lib/bb/codeparser.py
blob: 0cec452c00527631883bc0f4d0d1b2e0cb805d2f (plain)
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
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
#
# SPDX-License-Identifier: GPL-2.0-only
#

"""
BitBake code parser

Parses actual code (i.e. python and shell) for functions and in-line
expressions. Used mainly to determine dependencies on other functions
and variables within the BitBake metadata. Also provides a cache for
this information in order to speed up processing.

(Not to be confused with the code that parses the metadata itself,
see lib/bb/parse/ for that).

NOTE: if you change how the parsers gather information you will almost
certainly need to increment CodeParserCache.CACHE_VERSION below so that
any existing codeparser cache gets invalidated. Additionally you'll need
to increment __cache_version__ in cache.py in order to ensure that old
recipe caches don't trigger "Taskhash mismatch" errors.

"""

import ast
import sys
import codegen
import logging
import bb.pysh as pysh
import bb.utils, bb.data
import hashlib
from itertools import chain
from bb.pysh import pyshyacc, pyshlex
from bb.cache import MultiProcessCache

logger = logging.getLogger('BitBake.CodeParser')

def bbhash(s):
    return hashlib.sha256(s.encode("utf-8")).hexdigest()

def check_indent(codestr):
    """If the code is indented, add a top level piece of code to 'remove' the indentation"""

    i = 0
    while codestr[i] in ["\n", "\t", " "]:
        i = i + 1

    if i == 0:
        return codestr

    if codestr[i-1] == "\t" or codestr[i-1] == " ":
        if codestr[0] == "\n":
            # Since we're adding a line, we need to remove one line of any empty padding
            # to ensure line numbers are correct
            codestr = codestr[1:]
        return "if 1:\n" + codestr

    return codestr

# A custom getstate/setstate using tuples is actually worth 15% cachesize by
# avoiding duplication of the attribute names!


class SetCache(object):
    def __init__(self):
        self.setcache = {}

    def internSet(self, items):
        
        new = []
        for i in items:
            new.append(sys.intern(i))
        s = frozenset(new)
        h = hash(s)
        if h in self.setcache:
            return self.setcache[h]
        self.setcache[h] = s
        return s

codecache = SetCache()

class pythonCacheLine(object):
    def __init__(self, refs, execs, contains):
        self.refs = codecache.internSet(refs)
        self.execs = codecache.internSet(execs)
        self.contains = {}
        for c in contains:
            self.contains[c] = codecache.internSet(contains[c])

    def __getstate__(self):
        return (self.refs, self.execs, self.contains)

    def __setstate__(self, state):
        (refs, execs, contains) = state
        self.__init__(refs, execs, contains)
    def __hash__(self):
        l = (hash(self.refs), hash(self.execs))
        for c in sorted(self.contains.keys()):
            l = l + (c, hash(self.contains[c]))
        return hash(l)
    def __repr__(self):
        return " ".join([str(self.refs), str(self.execs), str(self.contains)]) 


class shellCacheLine(object):
    def __init__(self, execs):
        self.execs = codecache.internSet(execs)

    def __getstate__(self):
        return (self.execs)

    def __setstate__(self, state):
        (execs) = state
        self.__init__(execs)
    def __hash__(self):
        return hash(self.execs)
    def __repr__(self):
        return str(self.execs)

class CodeParserCache(MultiProcessCache):
    cache_file_name = "bb_codeparser.dat"
    # NOTE: you must increment this if you change how the parsers gather information,
    # so that an existing cache gets invalidated. Additionally you'll need
    # to increment __cache_version__ in cache.py in order to ensure that old
    # recipe caches don't trigger "Taskhash mismatch" errors.
    CACHE_VERSION = 11

    def __init__(self):
        MultiProcessCache.__init__(self)
        self.pythoncache = self.cachedata[0]
        self.shellcache = self.cachedata[1]
        self.pythoncacheextras = self.cachedata_extras[0]
        self.shellcacheextras = self.cachedata_extras[1]

        # To avoid duplication in the codeparser cache, keep
        # a lookup of hashes of objects we already have
        self.pythoncachelines = {}
        self.shellcachelines = {}

    def newPythonCacheLine(self, refs, execs, contains):
        cacheline = pythonCacheLine(refs, execs, contains)
        h = hash(cacheline)
        if h in self.pythoncachelines:
            return self.pythoncachelines[h]
        self.pythoncachelines[h] = cacheline
        return cacheline

    def newShellCacheLine(self, execs):
        cacheline = shellCacheLine(execs)
        h = hash(cacheline)
        if h in self.shellcachelines:
            return self.shellcachelines[h]
        self.shellcachelines[h] = cacheline
        return cacheline

    def init_cache(self, d):
        # Check if we already have the caches
        if self.pythoncache:
            return

        MultiProcessCache.init_cache(self, d)

        # cachedata gets re-assigned in the parent
        self.pythoncache = self.cachedata[0]
        self.shellcache = self.cachedata[1]

    def create_cachedata(self):
        data = [{}, {}]
        return data

codeparsercache = CodeParserCache()

def parser_cache_init(d):
    codeparsercache.init_cache(d)

def parser_cache_save():
    codeparsercache.save_extras()

def parser_cache_savemerge():
    codeparsercache.save_merge()

Logger = logging.getLoggerClass()
class BufferedLogger(Logger):
    def __init__(self, name, level=0, target=None):
        Logger.__init__(self, name)
        self.setLevel(level)
        self.buffer = []
        self.target = target

    def handle(self, record):
        self.buffer.append(record)

    def flush(self):
        for record in self.buffer:
            if self.target.isEnabledFor(record.levelno):
                self.target.handle(record)
        self.buffer = []

class PythonParser():
    getvars = (".getVar", ".appendVar", ".prependVar", "oe.utils.conditional")
    getvarflags = (".getVarFlag", ".appendVarFlag", ".prependVarFlag")
    containsfuncs = ("bb.utils.contains", "base_contains")
    containsanyfuncs = ("bb.utils.contains_any",  "bb.utils.filter")
    execfuncs = ("bb.build.exec_func", "bb.build.exec_task")

    def warn(self, func, arg):
        """Warn about calls of bitbake APIs which pass a non-literal
        argument for the variable name, as we're not able to track such
        a reference.
        """

        try:
            funcstr = codegen.to_source(func)
            argstr = codegen.to_source(arg)
        except TypeError:
            self.log.debug2('Failed to convert function and argument to source form')
        else:
            self.log.debug(self.unhandled_message % (funcstr, argstr))

    def visit_Call(self, node):
        name = self.called_node_name(node.func)
        if name and (name.endswith(self.getvars) or name.endswith(self.getvarflags) or name in self.containsfuncs or name in self.containsanyfuncs):
            if isinstance(node.args[0], ast.Str):
                varname = node.args[0].s
                if name in self.containsfuncs and isinstance(node.args[1], ast.Str):
                    if varname not in self.contains:
                        self.contains[varname] = set()
                    self.contains[varname].add(node.args[1].s)
                elif name in self.containsanyfuncs and isinstance(node.args[1], ast.Str):
                    if varname not in self.contains:
                        self.contains[varname] = set()
                    self.contains[varname].update(node.args[1].s.split())
                elif name.endswith(self.getvarflags):
                    if isinstance(node.args[1], ast.Str):
                        self.references.add('%s[%s]' % (varname, node.args[1].s))
                    else:
                        self.warn(node.func, node.args[1])
                else:
                    self.references.add(varname)
            else:
                self.warn(node.func, node.args[0])
        elif name and name.endswith(".expand"):
            if isinstance(node.args[0], ast.Str):
                value = node.args[0].s
                d = bb.data.init()
                parser = d.expandWithRefs(value, self.name)
                self.references |= parser.references
                self.execs |= parser.execs
                for varname in parser.contains:
                    if varname not in self.contains:
                        self.contains[varname] = set()
                    self.contains[varname] |= parser.contains[varname]
        elif name in self.execfuncs:
            if isinstance(node.args[0], ast.Str):
                self.var_execs.add(node.args[0].s)
            else:
                self.warn(node.func, node.args[0])
        elif name and isinstance(node.func, (ast.Name, ast.Attribute)):
            self.execs.add(name)

    def called_node_name(self, node):
        """Given a called node, return its original string form"""
        components = []
        while node:
            if isinstance(node, ast.Attribute):
                components.append(node.attr)
                node = node.value
            elif isinstance(node, ast.Name):
                components.append(node.id)
                return '.'.join(reversed(components))
            else:
                break

    def __init__(self, name, log):
        self.name = name
        self.var_execs = set()
        self.contains = {}
        self.execs = set()
        self.references = set()
        self.log = BufferedLogger('BitBake.Data.PythonParser', logging.DEBUG, log)

        self.unhandled_message = "in call of %s, argument '%s' is not a string literal"
        self.unhandled_message = "while parsing %s, %s" % (name, self.unhandled_message)

    def parse_python(self, node, lineno=0, filename="<string>"):
        if not node or not node.strip():
            return

        h = bbhash(str(node))

        if h in codeparsercache.pythoncache:
            self.references = set(codeparsercache.pythoncache[h].refs)
            self.execs = set(codeparsercache.pythoncache[h].execs)
            self.contains = {}
            for i in codeparsercache.pythoncache[h].contains:
                self.contains[i] = set(codeparsercache.pythoncache[h].contains[i])
            return

        if h in codeparsercache.pythoncacheextras:
            self.references = set(codeparsercache.pythoncacheextras[h].refs)
            self.execs = set(codeparsercache.pythoncacheextras[h].execs)
            self.contains = {}
            for i in codeparsercache.pythoncacheextras[h].contains:
                self.contains[i] = set(codeparsercache.pythoncacheextras[h].contains[i])
            return

        # We can't add to the linenumbers for compile, we can pad to the correct number of blank lines though
        node = "\n" * int(lineno) + node
        code = compile(check_indent(str(node)), filename, "exec",
                       ast.PyCF_ONLY_AST)

        for n in ast.walk(code):
            if n.__class__.__name__ == "Call":
                self.visit_Call(n)

        self.execs.update(self.var_execs)

        codeparsercache.pythoncacheextras[h] = codeparsercache.newPythonCacheLine(self.references, self.execs, self.contains)

class ShellParser():
    def __init__(self, name, log):
        self.funcdefs = set()
        self.allexecs = set()
        self.execs = set()
        self.log = BufferedLogger('BitBake.Data.%s' % name, logging.DEBUG, log)
        self.unhandled_template = "unable to handle non-literal command '%s'"
        self.unhandled_template = "while parsing %s, %s" % (name, self.unhandled_template)

    def parse_shell(self, value):
        """Parse the supplied shell code in a string, returning the external
        commands it executes.
        """

        h = bbhash(str(value))

        if h in codeparsercache.shellcache:
            self.execs = set(codeparsercache.shellcache[h].execs)
            return self.execs

        if h in codeparsercache.shellcacheextras:
            self.execs = set(codeparsercache.shellcacheextras[h].execs)
            return self.execs

        self._parse_shell(value)
        self.execs = set(cmd for cmd in self.allexecs if cmd not in self.funcdefs)

        codeparsercache.shellcacheextras[h] = codeparsercache.newShellCacheLine(self.execs)

        return self.execs

    def _parse_shell(self, value):
        try:
            tokens, _ = pyshyacc.parse(value, eof=True, debug=False)
        except Exception:
            bb.error('Error during parse shell code, the last 5 lines are:\n%s' % '\n'.join(value.split('\n')[-5:]))
            raise

        self.process_tokens(tokens)

    def process_tokens(self, tokens):
        """Process a supplied portion of the syntax tree as returned by
        pyshyacc.parse.
        """

        def function_definition(value):
            self.funcdefs.add(value.name)
            return [value.body], None

        def case_clause(value):
            # Element 0 of each item in the case is the list of patterns, and
            # Element 1 of each item in the case is the list of commands to be
            # executed when that pattern matches.
            words = chain(*[item[0] for item in value.items])
            cmds  = chain(*[item[1] for item in value.items])
            return cmds, words

        def if_clause(value):
            main = chain(value.cond, value.if_cmds)
            rest = value.else_cmds
            if isinstance(rest, tuple) and rest[0] == "elif":
                return chain(main, if_clause(rest[1]))
            else:
                return chain(main, rest)

        def simple_command(value):
            return None, chain(value.words, (assign[1] for assign in value.assigns))

        token_handlers = {
            "and_or": lambda x: ((x.left, x.right), None),
            "async": lambda x: ([x], None),
            "brace_group": lambda x: (x.cmds, None),
            "for_clause": lambda x: (x.cmds, x.items),
            "function_definition": function_definition,
            "if_clause": lambda x: (if_clause(x), None),
            "pipeline": lambda x: (x.commands, None),
            "redirect_list": lambda x: ([x.cmd], None),
            "subshell": lambda x: (x.cmds, None),
            "while_clause": lambda x: (chain(x.condition, x.cmds), None),
            "until_clause": lambda x: (chain(x.condition, x.cmds), None),
            "simple_command": simple_command,
            "case_clause": case_clause,
        }

        def process_token_list(tokens):
            for token in tokens:
                if isinstance(token, list):
                    process_token_list(token)
                    continue
                name, value = token
                try:
                    more_tokens, words = token_handlers[name](value)
                except KeyError:
                    raise NotImplementedError("Unsupported token type " + name)

                if more_tokens:
                    self.process_tokens(more_tokens)

                if words:
                    self.process_words(words)

        process_token_list(tokens)

    def process_words(self, words):
        """Process a set of 'words' in pyshyacc parlance, which includes
        extraction of executed commands from $() blocks, as well as grabbing
        the command name argument.
        """

        words = list(words)
        for word in list(words):
            wtree = pyshlex.make_wordtree(word[1])
            for part in wtree:
                if not isinstance(part, list):
                    continue

                if part[0] in ('`', '$('):
                    command = pyshlex.wordtree_as_string(part[1:-1])
                    self._parse_shell(command)

                    if word[0] in ("cmd_name", "cmd_word"):
                        if word in words:
                            words.remove(word)

        usetoken = False
        for word in words:
            if word[0] in ("cmd_name", "cmd_word") or \
               (usetoken and word[0] == "TOKEN"):
                if "=" in word[1]:
                    usetoken = True
                    continue

                cmd = word[1]
                if cmd.startswith("$"):
                    self.log.debug(self.unhandled_template % cmd)
                elif cmd == "eval":
                    command = " ".join(word for _, word in words[1:])
                    self._parse_shell(command)
                else:
                    self.allexecs.add(cmd)
                break