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
|
try:
# Python 2
import commands as cmdstatus
except ImportError:
# Python 3
import subprocess as cmdstatus
def read_file(filename):
try:
f = open( filename, "r" )
except IOError as reason:
return "" # WARNING: can't raise an error now because of the new RDEPENDS handling. This is a bit ugly. :M:
else:
data = f.read().strip()
f.close()
return data
return None
def ifelse(condition, iftrue = True, iffalse = False):
if condition:
return iftrue
else:
return iffalse
def conditional(variable, checkvalue, truevalue, falsevalue, d):
if d.getVar(variable,1) == checkvalue:
return truevalue
else:
return falsevalue
def less_or_equal(variable, checkvalue, truevalue, falsevalue, d):
if float(d.getVar(variable,1)) <= float(checkvalue):
return truevalue
else:
return falsevalue
def version_less_or_equal(variable, checkvalue, truevalue, falsevalue, d):
result = bb.utils.vercmp_string(d.getVar(variable,True), checkvalue)
if result <= 0:
return truevalue
else:
return falsevalue
def both_contain(variable1, variable2, checkvalue, d):
val1 = d.getVar(variable1, True)
val2 = d.getVar(variable2, True)
val1 = set(val1.split())
val2 = set(val2.split())
if isinstance(checkvalue, basestring):
checkvalue = set(checkvalue.split())
else:
checkvalue = set(checkvalue)
if checkvalue.issubset(val1) and checkvalue.issubset(val2):
return " ".join(checkvalue)
else:
return ""
def set_intersect(variable1, variable2, d):
"""
Expand both variables, interpret them as lists of strings, and return the
intersection as a flattened string.
For example:
s1 = "a b c"
s2 = "b c d"
s3 = set_intersect(s1, s2)
=> s3 = "b c"
"""
val1 = set(d.getVar(variable1, True).split())
val2 = set(d.getVar(variable2, True).split())
return " ".join(val1 & val2)
def prune_suffix(var, suffixes, d):
# See if var ends with any of the suffixes listed and
# remove it if found
for suffix in suffixes:
if var.endswith(suffix):
var = var.replace(suffix, "")
prefix = d.getVar("MLPREFIX", True)
if prefix and var.startswith(prefix):
var = var.replace(prefix, "")
return var
def str_filter(f, str, d):
from re import match
return " ".join(filter(lambda x: match(f, x, 0), str.split()))
def str_filter_out(f, str, d):
from re import match
return " ".join(filter(lambda x: not match(f, x, 0), str.split()))
def param_bool(cfg, field, dflt = None):
"""Lookup <field> in <cfg> map and convert it to a boolean; take
<dflt> when this <field> does not exist"""
value = cfg.get(field, dflt)
strvalue = str(value).lower()
if strvalue in ('yes', 'y', 'true', 't', '1'):
return True
elif strvalue in ('no', 'n', 'false', 'f', '0'):
return False
raise ValueError("invalid value for boolean parameter '%s': '%s'" % (field, value))
def inherits(d, *classes):
"""Return True if the metadata inherits any of the specified classes"""
return any(bb.data.inherits_class(cls, d) for cls in classes)
def features_backfill(var,d):
# This construct allows the addition of new features to variable specified
# as var
# Example for var = "DISTRO_FEATURES"
# This construct allows the addition of new features to DISTRO_FEATURES
# that if not present would disable existing functionality, without
# disturbing distributions that have already set DISTRO_FEATURES.
# Distributions wanting to elide a value in DISTRO_FEATURES_BACKFILL should
# add the feature to DISTRO_FEATURES_BACKFILL_CONSIDERED
features = (d.getVar(var, True) or "").split()
backfill = (d.getVar(var+"_BACKFILL", True) or "").split()
considered = (d.getVar(var+"_BACKFILL_CONSIDERED", True) or "").split()
addfeatures = []
for feature in backfill:
if feature not in features and feature not in considered:
addfeatures.append(feature)
if addfeatures:
d.appendVar(var, " " + " ".join(addfeatures))
def packages_filter_out_system(d):
"""
Return a list of packages from PACKAGES with the "system" packages such as
PN-dbg PN-doc PN-locale-eb-gb removed.
"""
pn = d.getVar('PN', True)
blacklist = map(lambda suffix: pn + suffix, ('', '-dbg', '-dev', '-doc', '-locale', '-staticdev'))
localepkg = pn + "-locale-"
pkgs = []
for pkg in d.getVar('PACKAGES', True).split():
if pkg not in blacklist and localepkg not in pkg:
pkgs.append(pkg)
return pkgs
def getstatusoutput(cmd):
return cmdstatus.getstatusoutput(cmd)
def trim_version(version, num_parts=2):
"""
Return just the first <num_parts> of <version>, split by periods. For
example, trim_version("1.2.3", 2) will return "1.2".
"""
if type(version) is not str:
raise TypeError("Version should be a string")
if num_parts < 1:
raise ValueError("Cannot split to parts < 1")
parts = version.split(".")
trimmed = ".".join(parts[:num_parts])
return trimmed
def cpu_count():
import multiprocessing
return multiprocessing.cpu_count()
def execute_pre_post_process(d, cmds):
if cmds is None:
return
for cmd in cmds.strip().split(';'):
cmd = cmd.strip()
if cmd != '':
bb.note("Executing %s ..." % cmd)
bb.build.exec_func(cmd, d)
def multiprocess_exec(commands, function):
import signal
import multiprocessing
if not commands:
return []
def init_worker():
signal.signal(signal.SIGINT, signal.SIG_IGN)
nproc = min(multiprocessing.cpu_count(), len(commands))
pool = bb.utils.multiprocessingpool(nproc, init_worker)
imap = pool.imap(function, commands)
try:
res = list(imap)
pool.close()
pool.join()
results = []
for result in res:
if result is not None:
results.append(result)
return results
except KeyboardInterrupt:
pool.terminate()
pool.join()
raise
def squashspaces(string):
import re
return re.sub("\s+", " ", string).strip()
def format_pkg_list(pkg_dict, ret_format=None):
output = []
if ret_format == "arch":
for pkg in sorted(pkg_dict):
output.append("%s %s" % (pkg, pkg_dict[pkg]["arch"]))
elif ret_format == "file":
for pkg in sorted(pkg_dict):
output.append("%s %s %s" % (pkg, pkg_dict[pkg]["filename"], pkg_dict[pkg]["arch"]))
elif ret_format == "ver":
for pkg in sorted(pkg_dict):
output.append("%s %s %s" % (pkg, pkg_dict[pkg]["arch"], pkg_dict[pkg]["ver"]))
elif ret_format == "deps":
for pkg in sorted(pkg_dict):
for dep in pkg_dict[pkg]["deps"]:
output.append("%s|%s" % (pkg, dep))
else:
for pkg in sorted(pkg_dict):
output.append(pkg)
return '\n'.join(output)
#
# Python 2.7 doesn't have threaded pools (just multiprocessing)
# so implement a version here
#
from Queue import Queue
from threading import Thread
class ThreadedWorker(Thread):
"""Thread executing tasks from a given tasks queue"""
def __init__(self, tasks, worker_init, worker_end):
Thread.__init__(self)
self.tasks = tasks
self.daemon = True
self.worker_init = worker_init
self.worker_end = worker_end
def run(self):
from Queue import Empty
if self.worker_init is not None:
self.worker_init(self)
while True:
try:
func, args, kargs = self.tasks.get(block=False)
except Empty:
if self.worker_end is not None:
self.worker_end(self)
break
try:
func(self, *args, **kargs)
except Exception, e:
print e
finally:
self.tasks.task_done()
class ThreadedPool:
"""Pool of threads consuming tasks from a queue"""
def __init__(self, num_workers, num_tasks, worker_init=None,
worker_end=None):
self.tasks = Queue(num_tasks)
self.workers = []
for _ in range(num_workers):
worker = ThreadedWorker(self.tasks, worker_init, worker_end)
self.workers.append(worker)
def start(self):
for worker in self.workers:
worker.start()
def add_task(self, func, *args, **kargs):
"""Add a task to the queue"""
self.tasks.put((func, args, kargs))
def wait_completion(self):
"""Wait for completion of all the tasks in the queue"""
self.tasks.join()
for worker in self.workers:
worker.join()
def write_ld_so_conf(d):
# Some utils like prelink may not have the correct target library paths
# so write an ld.so.conf to help them
ldsoconf = d.expand("${STAGING_DIR_TARGET}${sysconfdir}/ld.so.conf")
if os.path.exists(ldsoconf):
bb.utils.remove(ldsoconf)
bb.utils.mkdirhier(os.path.dirname(ldsoconf))
with open(ldsoconf, "w") as f:
f.write(d.getVar("base_libdir", True) + '\n')
f.write(d.getVar("libdir", True) + '\n')
|