##// END OF EJS Templates
Removed some debug code in get_home_dir
Jorgen Stenarson -
Show More
@@ -1,2171 +1,2167 b''
1 1 # -*- coding: utf-8 -*-
2 2 """
3 3 General purpose utilities.
4 4
5 5 This is a grab-bag of stuff I find useful in most programs I write. Some of
6 6 these things are also convenient when working at the command line.
7 7
8 8 $Id: genutils.py 2998 2008-01-31 10:06:04Z vivainio $"""
9 9
10 10 #*****************************************************************************
11 11 # Copyright (C) 2001-2006 Fernando Perez. <fperez@colorado.edu>
12 12 #
13 13 # Distributed under the terms of the BSD License. The full license is in
14 14 # the file COPYING, distributed as part of this software.
15 15 #*****************************************************************************
16 16
17 17 from IPython import Release
18 18 __author__ = '%s <%s>' % Release.authors['Fernando']
19 19 __license__ = Release.license
20 20
21 21 #****************************************************************************
22 22 # required modules from the Python standard library
23 23 import __main__
24 24 import commands
25 25 try:
26 26 import doctest
27 27 except ImportError:
28 28 pass
29 29 import os
30 30 import platform
31 31 import re
32 32 import shlex
33 33 import shutil
34 34 import subprocess
35 35 import sys
36 36 import tempfile
37 37 import time
38 38 import types
39 39 import warnings
40 40
41 41 # Curses and termios are Unix-only modules
42 42 try:
43 43 import curses
44 44 # We need termios as well, so if its import happens to raise, we bail on
45 45 # using curses altogether.
46 46 import termios
47 47 except ImportError:
48 48 USE_CURSES = False
49 49 else:
50 50 # Curses on Solaris may not be complete, so we can't use it there
51 51 USE_CURSES = hasattr(curses,'initscr')
52 52
53 53 # Other IPython utilities
54 54 import IPython
55 55 from IPython.Itpl import Itpl,itpl,printpl
56 56 from IPython import DPyGetOpt, platutils
57 57 from IPython.generics import result_display
58 58 import IPython.ipapi
59 59 from IPython.external.path import path
60 60 if os.name == "nt":
61 61 from IPython.winconsole import get_console_size
62 62
63 63 try:
64 64 set
65 65 except:
66 66 from sets import Set as set
67 67
68 68
69 69 #****************************************************************************
70 70 # Exceptions
71 71 class Error(Exception):
72 72 """Base class for exceptions in this module."""
73 73 pass
74 74
75 75 #----------------------------------------------------------------------------
76 76 class IOStream:
77 77 def __init__(self,stream,fallback):
78 78 if not hasattr(stream,'write') or not hasattr(stream,'flush'):
79 79 stream = fallback
80 80 self.stream = stream
81 81 self._swrite = stream.write
82 82 self.flush = stream.flush
83 83
84 84 def write(self,data):
85 85 try:
86 86 self._swrite(data)
87 87 except:
88 88 try:
89 89 # print handles some unicode issues which may trip a plain
90 90 # write() call. Attempt to emulate write() by using a
91 91 # trailing comma
92 92 print >> self.stream, data,
93 93 except:
94 94 # if we get here, something is seriously broken.
95 95 print >> sys.stderr, \
96 96 'ERROR - failed to write data to stream:', self.stream
97 97
98 98 def close(self):
99 99 pass
100 100
101 101
102 102 class IOTerm:
103 103 """ Term holds the file or file-like objects for handling I/O operations.
104 104
105 105 These are normally just sys.stdin, sys.stdout and sys.stderr but for
106 106 Windows they can can replaced to allow editing the strings before they are
107 107 displayed."""
108 108
109 109 # In the future, having IPython channel all its I/O operations through
110 110 # this class will make it easier to embed it into other environments which
111 111 # are not a normal terminal (such as a GUI-based shell)
112 112 def __init__(self,cin=None,cout=None,cerr=None):
113 113 self.cin = IOStream(cin,sys.stdin)
114 114 self.cout = IOStream(cout,sys.stdout)
115 115 self.cerr = IOStream(cerr,sys.stderr)
116 116
117 117 # Global variable to be used for all I/O
118 118 Term = IOTerm()
119 119
120 120 import IPython.rlineimpl as readline
121 121 # Remake Term to use the readline i/o facilities
122 122 if sys.platform == 'win32' and readline.have_readline:
123 123
124 124 Term = IOTerm(cout=readline._outputfile,cerr=readline._outputfile)
125 125
126 126
127 127 #****************************************************************************
128 128 # Generic warning/error printer, used by everything else
129 129 def warn(msg,level=2,exit_val=1):
130 130 """Standard warning printer. Gives formatting consistency.
131 131
132 132 Output is sent to Term.cerr (sys.stderr by default).
133 133
134 134 Options:
135 135
136 136 -level(2): allows finer control:
137 137 0 -> Do nothing, dummy function.
138 138 1 -> Print message.
139 139 2 -> Print 'WARNING:' + message. (Default level).
140 140 3 -> Print 'ERROR:' + message.
141 141 4 -> Print 'FATAL ERROR:' + message and trigger a sys.exit(exit_val).
142 142
143 143 -exit_val (1): exit value returned by sys.exit() for a level 4
144 144 warning. Ignored for all other levels."""
145 145
146 146 if level>0:
147 147 header = ['','','WARNING: ','ERROR: ','FATAL ERROR: ']
148 148 print >> Term.cerr, '%s%s' % (header[level],msg)
149 149 if level == 4:
150 150 print >> Term.cerr,'Exiting.\n'
151 151 sys.exit(exit_val)
152 152
153 153 def info(msg):
154 154 """Equivalent to warn(msg,level=1)."""
155 155
156 156 warn(msg,level=1)
157 157
158 158 def error(msg):
159 159 """Equivalent to warn(msg,level=3)."""
160 160
161 161 warn(msg,level=3)
162 162
163 163 def fatal(msg,exit_val=1):
164 164 """Equivalent to warn(msg,exit_val=exit_val,level=4)."""
165 165
166 166 warn(msg,exit_val=exit_val,level=4)
167 167
168 168 #---------------------------------------------------------------------------
169 169 # Debugging routines
170 170 #
171 171 def debugx(expr,pre_msg=''):
172 172 """Print the value of an expression from the caller's frame.
173 173
174 174 Takes an expression, evaluates it in the caller's frame and prints both
175 175 the given expression and the resulting value (as well as a debug mark
176 176 indicating the name of the calling function. The input must be of a form
177 177 suitable for eval().
178 178
179 179 An optional message can be passed, which will be prepended to the printed
180 180 expr->value pair."""
181 181
182 182 cf = sys._getframe(1)
183 183 print '[DBG:%s] %s%s -> %r' % (cf.f_code.co_name,pre_msg,expr,
184 184 eval(expr,cf.f_globals,cf.f_locals))
185 185
186 186 # deactivate it by uncommenting the following line, which makes it a no-op
187 187 #def debugx(expr,pre_msg=''): pass
188 188
189 189 #----------------------------------------------------------------------------
190 190 StringTypes = types.StringTypes
191 191
192 192 # Basic timing functionality
193 193
194 194 # If possible (Unix), use the resource module instead of time.clock()
195 195 try:
196 196 import resource
197 197 def clocku():
198 198 """clocku() -> floating point number
199 199
200 200 Return the *USER* CPU time in seconds since the start of the process.
201 201 This is done via a call to resource.getrusage, so it avoids the
202 202 wraparound problems in time.clock()."""
203 203
204 204 return resource.getrusage(resource.RUSAGE_SELF)[0]
205 205
206 206 def clocks():
207 207 """clocks() -> floating point number
208 208
209 209 Return the *SYSTEM* CPU time in seconds since the start of the process.
210 210 This is done via a call to resource.getrusage, so it avoids the
211 211 wraparound problems in time.clock()."""
212 212
213 213 return resource.getrusage(resource.RUSAGE_SELF)[1]
214 214
215 215 def clock():
216 216 """clock() -> floating point number
217 217
218 218 Return the *TOTAL USER+SYSTEM* CPU time in seconds since the start of
219 219 the process. This is done via a call to resource.getrusage, so it
220 220 avoids the wraparound problems in time.clock()."""
221 221
222 222 u,s = resource.getrusage(resource.RUSAGE_SELF)[:2]
223 223 return u+s
224 224
225 225 def clock2():
226 226 """clock2() -> (t_user,t_system)
227 227
228 228 Similar to clock(), but return a tuple of user/system times."""
229 229 return resource.getrusage(resource.RUSAGE_SELF)[:2]
230 230
231 231 except ImportError:
232 232 # There is no distinction of user/system time under windows, so we just use
233 233 # time.clock() for everything...
234 234 clocku = clocks = clock = time.clock
235 235 def clock2():
236 236 """Under windows, system CPU time can't be measured.
237 237
238 238 This just returns clock() and zero."""
239 239 return time.clock(),0.0
240 240
241 241 def timings_out(reps,func,*args,**kw):
242 242 """timings_out(reps,func,*args,**kw) -> (t_total,t_per_call,output)
243 243
244 244 Execute a function reps times, return a tuple with the elapsed total
245 245 CPU time in seconds, the time per call and the function's output.
246 246
247 247 Under Unix, the return value is the sum of user+system time consumed by
248 248 the process, computed via the resource module. This prevents problems
249 249 related to the wraparound effect which the time.clock() function has.
250 250
251 251 Under Windows the return value is in wall clock seconds. See the
252 252 documentation for the time module for more details."""
253 253
254 254 reps = int(reps)
255 255 assert reps >=1, 'reps must be >= 1'
256 256 if reps==1:
257 257 start = clock()
258 258 out = func(*args,**kw)
259 259 tot_time = clock()-start
260 260 else:
261 261 rng = xrange(reps-1) # the last time is executed separately to store output
262 262 start = clock()
263 263 for dummy in rng: func(*args,**kw)
264 264 out = func(*args,**kw) # one last time
265 265 tot_time = clock()-start
266 266 av_time = tot_time / reps
267 267 return tot_time,av_time,out
268 268
269 269 def timings(reps,func,*args,**kw):
270 270 """timings(reps,func,*args,**kw) -> (t_total,t_per_call)
271 271
272 272 Execute a function reps times, return a tuple with the elapsed total CPU
273 273 time in seconds and the time per call. These are just the first two values
274 274 in timings_out()."""
275 275
276 276 return timings_out(reps,func,*args,**kw)[0:2]
277 277
278 278 def timing(func,*args,**kw):
279 279 """timing(func,*args,**kw) -> t_total
280 280
281 281 Execute a function once, return the elapsed total CPU time in
282 282 seconds. This is just the first value in timings_out()."""
283 283
284 284 return timings_out(1,func,*args,**kw)[0]
285 285
286 286 #****************************************************************************
287 287 # file and system
288 288
289 289 def arg_split(s,posix=False):
290 290 """Split a command line's arguments in a shell-like manner.
291 291
292 292 This is a modified version of the standard library's shlex.split()
293 293 function, but with a default of posix=False for splitting, so that quotes
294 294 in inputs are respected."""
295 295
296 296 # XXX - there may be unicode-related problems here!!! I'm not sure that
297 297 # shlex is truly unicode-safe, so it might be necessary to do
298 298 #
299 299 # s = s.encode(sys.stdin.encoding)
300 300 #
301 301 # first, to ensure that shlex gets a normal string. Input from anyone who
302 302 # knows more about unicode and shlex than I would be good to have here...
303 303 lex = shlex.shlex(s, posix=posix)
304 304 lex.whitespace_split = True
305 305 return list(lex)
306 306
307 307 def system(cmd,verbose=0,debug=0,header=''):
308 308 """Execute a system command, return its exit status.
309 309
310 310 Options:
311 311
312 312 - verbose (0): print the command to be executed.
313 313
314 314 - debug (0): only print, do not actually execute.
315 315
316 316 - header (''): Header to print on screen prior to the executed command (it
317 317 is only prepended to the command, no newlines are added).
318 318
319 319 Note: a stateful version of this function is available through the
320 320 SystemExec class."""
321 321
322 322 stat = 0
323 323 if verbose or debug: print header+cmd
324 324 sys.stdout.flush()
325 325 if not debug: stat = os.system(cmd)
326 326 return stat
327 327
328 328 def abbrev_cwd():
329 329 """ Return abbreviated version of cwd, e.g. d:mydir """
330 330 cwd = os.getcwd().replace('\\','/')
331 331 drivepart = ''
332 332 tail = cwd
333 333 if sys.platform == 'win32':
334 334 if len(cwd) < 4:
335 335 return cwd
336 336 drivepart,tail = os.path.splitdrive(cwd)
337 337
338 338
339 339 parts = tail.split('/')
340 340 if len(parts) > 2:
341 341 tail = '/'.join(parts[-2:])
342 342
343 343 return (drivepart + (
344 344 cwd == '/' and '/' or tail))
345 345
346 346
347 347 # This function is used by ipython in a lot of places to make system calls.
348 348 # We need it to be slightly different under win32, due to the vagaries of
349 349 # 'network shares'. A win32 override is below.
350 350
351 351 def shell(cmd,verbose=0,debug=0,header=''):
352 352 """Execute a command in the system shell, always return None.
353 353
354 354 Options:
355 355
356 356 - verbose (0): print the command to be executed.
357 357
358 358 - debug (0): only print, do not actually execute.
359 359
360 360 - header (''): Header to print on screen prior to the executed command (it
361 361 is only prepended to the command, no newlines are added).
362 362
363 363 Note: this is similar to genutils.system(), but it returns None so it can
364 364 be conveniently used in interactive loops without getting the return value
365 365 (typically 0) printed many times."""
366 366
367 367 stat = 0
368 368 if verbose or debug: print header+cmd
369 369 # flush stdout so we don't mangle python's buffering
370 370 sys.stdout.flush()
371 371
372 372 if not debug:
373 373 platutils.set_term_title("IPy " + cmd)
374 374 os.system(cmd)
375 375 platutils.set_term_title("IPy " + abbrev_cwd())
376 376
377 377 # override shell() for win32 to deal with network shares
378 378 if os.name in ('nt','dos'):
379 379
380 380 shell_ori = shell
381 381
382 382 def shell(cmd,verbose=0,debug=0,header=''):
383 383 if os.getcwd().startswith(r"\\"):
384 384 path = os.getcwd()
385 385 # change to c drive (cannot be on UNC-share when issuing os.system,
386 386 # as cmd.exe cannot handle UNC addresses)
387 387 os.chdir("c:")
388 388 # issue pushd to the UNC-share and then run the command
389 389 try:
390 390 shell_ori('"pushd %s&&"'%path+cmd,verbose,debug,header)
391 391 finally:
392 392 os.chdir(path)
393 393 else:
394 394 shell_ori(cmd,verbose,debug,header)
395 395
396 396 shell.__doc__ = shell_ori.__doc__
397 397
398 398 def getoutput(cmd,verbose=0,debug=0,header='',split=0):
399 399 """Dummy substitute for perl's backquotes.
400 400
401 401 Executes a command and returns the output.
402 402
403 403 Accepts the same arguments as system(), plus:
404 404
405 405 - split(0): if true, the output is returned as a list split on newlines.
406 406
407 407 Note: a stateful version of this function is available through the
408 408 SystemExec class.
409 409
410 410 This is pretty much deprecated and rarely used,
411 411 genutils.getoutputerror may be what you need.
412 412
413 413 """
414 414
415 415 if verbose or debug: print header+cmd
416 416 if not debug:
417 417 output = os.popen(cmd).read()
418 418 # stipping last \n is here for backwards compat.
419 419 if output.endswith('\n'):
420 420 output = output[:-1]
421 421 if split:
422 422 return output.split('\n')
423 423 else:
424 424 return output
425 425
426 426 def getoutputerror(cmd,verbose=0,debug=0,header='',split=0):
427 427 """Return (standard output,standard error) of executing cmd in a shell.
428 428
429 429 Accepts the same arguments as system(), plus:
430 430
431 431 - split(0): if true, each of stdout/err is returned as a list split on
432 432 newlines.
433 433
434 434 Note: a stateful version of this function is available through the
435 435 SystemExec class."""
436 436
437 437 if verbose or debug: print header+cmd
438 438 if not cmd:
439 439 if split:
440 440 return [],[]
441 441 else:
442 442 return '',''
443 443 if not debug:
444 444 pin,pout,perr = os.popen3(cmd)
445 445 tout = pout.read().rstrip()
446 446 terr = perr.read().rstrip()
447 447 pin.close()
448 448 pout.close()
449 449 perr.close()
450 450 if split:
451 451 return tout.split('\n'),terr.split('\n')
452 452 else:
453 453 return tout,terr
454 454
455 455 # for compatibility with older naming conventions
456 456 xsys = system
457 457 bq = getoutput
458 458
459 459 class SystemExec:
460 460 """Access the system and getoutput functions through a stateful interface.
461 461
462 462 Note: here we refer to the system and getoutput functions from this
463 463 library, not the ones from the standard python library.
464 464
465 465 This class offers the system and getoutput functions as methods, but the
466 466 verbose, debug and header parameters can be set for the instance (at
467 467 creation time or later) so that they don't need to be specified on each
468 468 call.
469 469
470 470 For efficiency reasons, there's no way to override the parameters on a
471 471 per-call basis other than by setting instance attributes. If you need
472 472 local overrides, it's best to directly call system() or getoutput().
473 473
474 474 The following names are provided as alternate options:
475 475 - xsys: alias to system
476 476 - bq: alias to getoutput
477 477
478 478 An instance can then be created as:
479 479 >>> sysexec = SystemExec(verbose=1,debug=0,header='Calling: ')
480 480 """
481 481
482 482 def __init__(self,verbose=0,debug=0,header='',split=0):
483 483 """Specify the instance's values for verbose, debug and header."""
484 484 setattr_list(self,'verbose debug header split')
485 485
486 486 def system(self,cmd):
487 487 """Stateful interface to system(), with the same keyword parameters."""
488 488
489 489 system(cmd,self.verbose,self.debug,self.header)
490 490
491 491 def shell(self,cmd):
492 492 """Stateful interface to shell(), with the same keyword parameters."""
493 493
494 494 shell(cmd,self.verbose,self.debug,self.header)
495 495
496 496 xsys = system # alias
497 497
498 498 def getoutput(self,cmd):
499 499 """Stateful interface to getoutput()."""
500 500
501 501 return getoutput(cmd,self.verbose,self.debug,self.header,self.split)
502 502
503 503 def getoutputerror(self,cmd):
504 504 """Stateful interface to getoutputerror()."""
505 505
506 506 return getoutputerror(cmd,self.verbose,self.debug,self.header,self.split)
507 507
508 508 bq = getoutput # alias
509 509
510 510 #-----------------------------------------------------------------------------
511 511 def mutex_opts(dict,ex_op):
512 512 """Check for presence of mutually exclusive keys in a dict.
513 513
514 514 Call: mutex_opts(dict,[[op1a,op1b],[op2a,op2b]...]"""
515 515 for op1,op2 in ex_op:
516 516 if op1 in dict and op2 in dict:
517 517 raise ValueError,'\n*** ERROR in Arguments *** '\
518 518 'Options '+op1+' and '+op2+' are mutually exclusive.'
519 519
520 520 #-----------------------------------------------------------------------------
521 521 def get_py_filename(name):
522 522 """Return a valid python filename in the current directory.
523 523
524 524 If the given name is not a file, it adds '.py' and searches again.
525 525 Raises IOError with an informative message if the file isn't found."""
526 526
527 527 name = os.path.expanduser(name)
528 528 if not os.path.isfile(name) and not name.endswith('.py'):
529 529 name += '.py'
530 530 if os.path.isfile(name):
531 531 return name
532 532 else:
533 533 raise IOError,'File `%s` not found.' % name
534 534
535 535 #-----------------------------------------------------------------------------
536 536 def filefind(fname,alt_dirs = None):
537 537 """Return the given filename either in the current directory, if it
538 538 exists, or in a specified list of directories.
539 539
540 540 ~ expansion is done on all file and directory names.
541 541
542 542 Upon an unsuccessful search, raise an IOError exception."""
543 543
544 544 if alt_dirs is None:
545 545 try:
546 546 alt_dirs = get_home_dir()
547 547 except HomeDirError:
548 548 alt_dirs = os.getcwd()
549 549 search = [fname] + list_strings(alt_dirs)
550 550 search = map(os.path.expanduser,search)
551 551 #print 'search list for',fname,'list:',search # dbg
552 552 fname = search[0]
553 553 if os.path.isfile(fname):
554 554 return fname
555 555 for direc in search[1:]:
556 556 testname = os.path.join(direc,fname)
557 557 #print 'testname',testname # dbg
558 558 if os.path.isfile(testname):
559 559 return testname
560 560 raise IOError,'File' + `fname` + \
561 561 ' not found in current or supplied directories:' + `alt_dirs`
562 562
563 563 #----------------------------------------------------------------------------
564 564 def file_read(filename):
565 565 """Read a file and close it. Returns the file source."""
566 566 fobj = open(filename,'r');
567 567 source = fobj.read();
568 568 fobj.close()
569 569 return source
570 570
571 571 def file_readlines(filename):
572 572 """Read a file and close it. Returns the file source using readlines()."""
573 573 fobj = open(filename,'r');
574 574 lines = fobj.readlines();
575 575 fobj.close()
576 576 return lines
577 577
578 578 #----------------------------------------------------------------------------
579 579 def target_outdated(target,deps):
580 580 """Determine whether a target is out of date.
581 581
582 582 target_outdated(target,deps) -> 1/0
583 583
584 584 deps: list of filenames which MUST exist.
585 585 target: single filename which may or may not exist.
586 586
587 587 If target doesn't exist or is older than any file listed in deps, return
588 588 true, otherwise return false.
589 589 """
590 590 try:
591 591 target_time = os.path.getmtime(target)
592 592 except os.error:
593 593 return 1
594 594 for dep in deps:
595 595 dep_time = os.path.getmtime(dep)
596 596 if dep_time > target_time:
597 597 #print "For target",target,"Dep failed:",dep # dbg
598 598 #print "times (dep,tar):",dep_time,target_time # dbg
599 599 return 1
600 600 return 0
601 601
602 602 #-----------------------------------------------------------------------------
603 603 def target_update(target,deps,cmd):
604 604 """Update a target with a given command given a list of dependencies.
605 605
606 606 target_update(target,deps,cmd) -> runs cmd if target is outdated.
607 607
608 608 This is just a wrapper around target_outdated() which calls the given
609 609 command if target is outdated."""
610 610
611 611 if target_outdated(target,deps):
612 612 xsys(cmd)
613 613
614 614 #----------------------------------------------------------------------------
615 615 def unquote_ends(istr):
616 616 """Remove a single pair of quotes from the endpoints of a string."""
617 617
618 618 if not istr:
619 619 return istr
620 620 if (istr[0]=="'" and istr[-1]=="'") or \
621 621 (istr[0]=='"' and istr[-1]=='"'):
622 622 return istr[1:-1]
623 623 else:
624 624 return istr
625 625
626 626 #----------------------------------------------------------------------------
627 627 def process_cmdline(argv,names=[],defaults={},usage=''):
628 628 """ Process command-line options and arguments.
629 629
630 630 Arguments:
631 631
632 632 - argv: list of arguments, typically sys.argv.
633 633
634 634 - names: list of option names. See DPyGetOpt docs for details on options
635 635 syntax.
636 636
637 637 - defaults: dict of default values.
638 638
639 639 - usage: optional usage notice to print if a wrong argument is passed.
640 640
641 641 Return a dict of options and a list of free arguments."""
642 642
643 643 getopt = DPyGetOpt.DPyGetOpt()
644 644 getopt.setIgnoreCase(0)
645 645 getopt.parseConfiguration(names)
646 646
647 647 try:
648 648 getopt.processArguments(argv)
649 649 except DPyGetOpt.ArgumentError, exc:
650 650 print usage
651 651 warn('"%s"' % exc,level=4)
652 652
653 653 defaults.update(getopt.optionValues)
654 654 args = getopt.freeValues
655 655
656 656 return defaults,args
657 657
658 658 #----------------------------------------------------------------------------
659 659 def optstr2types(ostr):
660 660 """Convert a string of option names to a dict of type mappings.
661 661
662 662 optstr2types(str) -> {None:'string_opts',int:'int_opts',float:'float_opts'}
663 663
664 664 This is used to get the types of all the options in a string formatted
665 665 with the conventions of DPyGetOpt. The 'type' None is used for options
666 666 which are strings (they need no further conversion). This function's main
667 667 use is to get a typemap for use with read_dict().
668 668 """
669 669
670 670 typeconv = {None:'',int:'',float:''}
671 671 typemap = {'s':None,'i':int,'f':float}
672 672 opt_re = re.compile(r'([\w]*)([^:=]*:?=?)([sif]?)')
673 673
674 674 for w in ostr.split():
675 675 oname,alias,otype = opt_re.match(w).groups()
676 676 if otype == '' or alias == '!': # simple switches are integers too
677 677 otype = 'i'
678 678 typeconv[typemap[otype]] += oname + ' '
679 679 return typeconv
680 680
681 681 #----------------------------------------------------------------------------
682 682 def read_dict(filename,type_conv=None,**opt):
683 683 r"""Read a dictionary of key=value pairs from an input file, optionally
684 684 performing conversions on the resulting values.
685 685
686 686 read_dict(filename,type_conv,**opt) -> dict
687 687
688 688 Only one value per line is accepted, the format should be
689 689 # optional comments are ignored
690 690 key value\n
691 691
692 692 Args:
693 693
694 694 - type_conv: A dictionary specifying which keys need to be converted to
695 695 which types. By default all keys are read as strings. This dictionary
696 696 should have as its keys valid conversion functions for strings
697 697 (int,long,float,complex, or your own). The value for each key
698 698 (converter) should be a whitespace separated string containing the names
699 699 of all the entries in the file to be converted using that function. For
700 700 keys to be left alone, use None as the conversion function (only needed
701 701 with purge=1, see below).
702 702
703 703 - opt: dictionary with extra options as below (default in parens)
704 704
705 705 purge(0): if set to 1, all keys *not* listed in type_conv are purged out
706 706 of the dictionary to be returned. If purge is going to be used, the
707 707 set of keys to be left as strings also has to be explicitly specified
708 708 using the (non-existent) conversion function None.
709 709
710 710 fs(None): field separator. This is the key/value separator to be used
711 711 when parsing the file. The None default means any whitespace [behavior
712 712 of string.split()].
713 713
714 714 strip(0): if 1, strip string values of leading/trailinig whitespace.
715 715
716 716 warn(1): warning level if requested keys are not found in file.
717 717 - 0: silently ignore.
718 718 - 1: inform but proceed.
719 719 - 2: raise KeyError exception.
720 720
721 721 no_empty(0): if 1, remove keys with whitespace strings as a value.
722 722
723 723 unique([]): list of keys (or space separated string) which can't be
724 724 repeated. If one such key is found in the file, each new instance
725 725 overwrites the previous one. For keys not listed here, the behavior is
726 726 to make a list of all appearances.
727 727
728 728 Example:
729 729
730 730 If the input file test.ini contains (we put it in a string to keep the test
731 731 self-contained):
732 732
733 733 >>> test_ini = '''\
734 734 ... i 3
735 735 ... x 4.5
736 736 ... y 5.5
737 737 ... s hi ho'''
738 738
739 739 Then we can use it as follows:
740 740 >>> type_conv={int:'i',float:'x',None:'s'}
741 741
742 742 >>> d = read_dict(test_ini)
743 743
744 744 >>> sorted(d.items())
745 745 [('i', '3'), ('s', 'hi ho'), ('x', '4.5'), ('y', '5.5')]
746 746
747 747 >>> d = read_dict(test_ini,type_conv)
748 748
749 749 >>> sorted(d.items())
750 750 [('i', 3), ('s', 'hi ho'), ('x', 4.5), ('y', '5.5')]
751 751
752 752 >>> d = read_dict(test_ini,type_conv,purge=True)
753 753
754 754 >>> sorted(d.items())
755 755 [('i', 3), ('s', 'hi ho'), ('x', 4.5)]
756 756 """
757 757
758 758 # starting config
759 759 opt.setdefault('purge',0)
760 760 opt.setdefault('fs',None) # field sep defaults to any whitespace
761 761 opt.setdefault('strip',0)
762 762 opt.setdefault('warn',1)
763 763 opt.setdefault('no_empty',0)
764 764 opt.setdefault('unique','')
765 765 if type(opt['unique']) in StringTypes:
766 766 unique_keys = qw(opt['unique'])
767 767 elif type(opt['unique']) in (types.TupleType,types.ListType):
768 768 unique_keys = opt['unique']
769 769 else:
770 770 raise ValueError, 'Unique keys must be given as a string, List or Tuple'
771 771
772 772 dict = {}
773 773
774 774 # first read in table of values as strings
775 775 if '\n' in filename:
776 776 lines = filename.splitlines()
777 777 file = None
778 778 else:
779 779 file = open(filename,'r')
780 780 lines = file.readlines()
781 781 for line in lines:
782 782 line = line.strip()
783 783 if len(line) and line[0]=='#': continue
784 784 if len(line)>0:
785 785 lsplit = line.split(opt['fs'],1)
786 786 try:
787 787 key,val = lsplit
788 788 except ValueError:
789 789 key,val = lsplit[0],''
790 790 key = key.strip()
791 791 if opt['strip']: val = val.strip()
792 792 if val == "''" or val == '""': val = ''
793 793 if opt['no_empty'] and (val=='' or val.isspace()):
794 794 continue
795 795 # if a key is found more than once in the file, build a list
796 796 # unless it's in the 'unique' list. In that case, last found in file
797 797 # takes precedence. User beware.
798 798 try:
799 799 if dict[key] and key in unique_keys:
800 800 dict[key] = val
801 801 elif type(dict[key]) is types.ListType:
802 802 dict[key].append(val)
803 803 else:
804 804 dict[key] = [dict[key],val]
805 805 except KeyError:
806 806 dict[key] = val
807 807 # purge if requested
808 808 if opt['purge']:
809 809 accepted_keys = qwflat(type_conv.values())
810 810 for key in dict.keys():
811 811 if key in accepted_keys: continue
812 812 del(dict[key])
813 813 # now convert if requested
814 814 if type_conv==None: return dict
815 815 conversions = type_conv.keys()
816 816 try: conversions.remove(None)
817 817 except: pass
818 818 for convert in conversions:
819 819 for val in qw(type_conv[convert]):
820 820 try:
821 821 dict[val] = convert(dict[val])
822 822 except KeyError,e:
823 823 if opt['warn'] == 0:
824 824 pass
825 825 elif opt['warn'] == 1:
826 826 print >>sys.stderr, 'Warning: key',val,\
827 827 'not found in file',filename
828 828 elif opt['warn'] == 2:
829 829 raise KeyError,e
830 830 else:
831 831 raise ValueError,'Warning level must be 0,1 or 2'
832 832
833 833 return dict
834 834
835 835 #----------------------------------------------------------------------------
836 836 def flag_calls(func):
837 837 """Wrap a function to detect and flag when it gets called.
838 838
839 839 This is a decorator which takes a function and wraps it in a function with
840 840 a 'called' attribute. wrapper.called is initialized to False.
841 841
842 842 The wrapper.called attribute is set to False right before each call to the
843 843 wrapped function, so if the call fails it remains False. After the call
844 844 completes, wrapper.called is set to True and the output is returned.
845 845
846 846 Testing for truth in wrapper.called allows you to determine if a call to
847 847 func() was attempted and succeeded."""
848 848
849 849 def wrapper(*args,**kw):
850 850 wrapper.called = False
851 851 out = func(*args,**kw)
852 852 wrapper.called = True
853 853 return out
854 854
855 855 wrapper.called = False
856 856 wrapper.__doc__ = func.__doc__
857 857 return wrapper
858 858
859 859 #----------------------------------------------------------------------------
860 860 def dhook_wrap(func,*a,**k):
861 861 """Wrap a function call in a sys.displayhook controller.
862 862
863 863 Returns a wrapper around func which calls func, with all its arguments and
864 864 keywords unmodified, using the default sys.displayhook. Since IPython
865 865 modifies sys.displayhook, it breaks the behavior of certain systems that
866 866 rely on the default behavior, notably doctest.
867 867 """
868 868
869 869 def f(*a,**k):
870 870
871 871 dhook_s = sys.displayhook
872 872 sys.displayhook = sys.__displayhook__
873 873 try:
874 874 out = func(*a,**k)
875 875 finally:
876 876 sys.displayhook = dhook_s
877 877
878 878 return out
879 879
880 880 f.__doc__ = func.__doc__
881 881 return f
882 882
883 883 #----------------------------------------------------------------------------
884 884 def doctest_reload():
885 885 """Properly reload doctest to reuse it interactively.
886 886
887 887 This routine:
888 888
889 889 - reloads doctest
890 890
891 891 - resets its global 'master' attribute to None, so that multiple uses of
892 892 the module interactively don't produce cumulative reports.
893 893
894 894 - Monkeypatches its core test runner method to protect it from IPython's
895 895 modified displayhook. Doctest expects the default displayhook behavior
896 896 deep down, so our modification breaks it completely. For this reason, a
897 897 hard monkeypatch seems like a reasonable solution rather than asking
898 898 users to manually use a different doctest runner when under IPython."""
899 899
900 900 import doctest
901 901 reload(doctest)
902 902 doctest.master=None
903 903
904 904 try:
905 905 doctest.DocTestRunner
906 906 except AttributeError:
907 907 # This is only for python 2.3 compatibility, remove once we move to
908 908 # 2.4 only.
909 909 pass
910 910 else:
911 911 doctest.DocTestRunner.run = dhook_wrap(doctest.DocTestRunner.run)
912 912
913 913 #----------------------------------------------------------------------------
914 914 class HomeDirError(Error):
915 915 pass
916 916
917 917 def get_home_dir():
918 918 """Return the closest possible equivalent to a 'home' directory.
919 919
920 920 We first try $HOME. Absent that, on NT it's $HOMEDRIVE\$HOMEPATH.
921 921
922 922 Currently only Posix and NT are implemented, a HomeDirError exception is
923 923 raised for all other OSes. """
924 924
925 925 isdir = os.path.isdir
926 926 env = os.environ
927 927
928 928 # first, check py2exe distribution root directory for _ipython.
929 929 # This overrides all. Normally does not exist.
930 930
931 931 if hasattr(sys, "frozen"): #Is frozen by py2exe
932 932 if '\\library.zip\\' in IPython.__file__.lower():#libraries compressed to zip-file
933 933 root, rest = IPython.__file__.lower().split('library.zip')
934 934 else:
935 935 root=os.path.join(os.path.split(IPython.__file__)[0],"../../")
936 936 root=os.path.abspath(root).rstrip('\\')
937 937 if isdir(os.path.join(root, '_ipython')):
938 938 os.environ["IPYKITROOT"] = root
939 939 return root
940 940 try:
941 941 homedir = env['HOME']
942 942 if not isdir(homedir):
943 943 # in case a user stuck some string which does NOT resolve to a
944 944 # valid path, it's as good as if we hadn't foud it
945
946 #raise KeyError # dbg
947 # dbg - figuring out what's going on here
948 pp = os.listdir(homedir+'/..')
949 raise ValueError('Wrong dir: %s\n%s' % (homedir,pp)) # dbg
945 raise KeyError
950 946 return homedir
951 947 except KeyError:
952 948 if os.name == 'posix':
953 949 raise HomeDirError,'undefined $HOME, IPython can not proceed.'
954 950 elif os.name == 'nt':
955 951 # For some strange reason, win9x returns 'nt' for os.name.
956 952 try:
957 953 homedir = os.path.join(env['HOMEDRIVE'],env['HOMEPATH'])
958 954 if not isdir(homedir):
959 955 homedir = os.path.join(env['USERPROFILE'])
960 956 if not isdir(homedir):
961 957 raise HomeDirError
962 958 return homedir
963 959 except KeyError:
964 960 try:
965 961 # Use the registry to get the 'My Documents' folder.
966 962 import _winreg as wreg
967 963 key = wreg.OpenKey(wreg.HKEY_CURRENT_USER,
968 964 "Software\Microsoft\Windows\CurrentVersion\Explorer\Shell Folders")
969 965 homedir = wreg.QueryValueEx(key,'Personal')[0]
970 966 key.Close()
971 967 if not isdir(homedir):
972 968 e = ('Invalid "Personal" folder registry key '
973 969 'typically "My Documents".\n'
974 970 'Value: %s\n'
975 971 'This is not a valid directory on your system.' %
976 972 homedir)
977 973 raise HomeDirError(e)
978 974 return homedir
979 975 except HomeDirError:
980 976 raise
981 977 except:
982 978 return 'C:\\'
983 979 elif os.name == 'dos':
984 980 # Desperate, may do absurd things in classic MacOS. May work under DOS.
985 981 return 'C:\\'
986 982 else:
987 983 raise HomeDirError,'support for your operating system not implemented.'
988 984
989 985
990 986 def get_ipython_dir():
991 987 """Get the IPython directory for this platform and user.
992 988
993 989 This uses the logic in `get_home_dir` to find the home directory
994 990 and the adds either .ipython or _ipython to the end of the path.
995 991 """
996 992 if os.name == 'posix':
997 993 ipdir_def = '.ipython'
998 994 else:
999 995 ipdir_def = '_ipython'
1000 996 home_dir = get_home_dir()
1001 997 ipdir = os.path.abspath(os.environ.get('IPYTHONDIR',
1002 998 os.path.join(home_dir, ipdir_def)))
1003 999 return ipdir.decode(sys.getfilesystemencoding())
1004 1000
1005 1001 def get_security_dir():
1006 1002 """Get the IPython security directory.
1007 1003
1008 1004 This directory is the default location for all security related files,
1009 1005 including SSL/TLS certificates and FURL files.
1010 1006
1011 1007 If the directory does not exist, it is created with 0700 permissions.
1012 1008 If it exists, permissions are set to 0700.
1013 1009 """
1014 1010 security_dir = os.path.join(get_ipython_dir(), 'security')
1015 1011 if not os.path.isdir(security_dir):
1016 1012 os.mkdir(security_dir, 0700)
1017 1013 else:
1018 1014 os.chmod(security_dir, 0700)
1019 1015 return security_dir
1020 1016
1021 1017 #****************************************************************************
1022 1018 # strings and text
1023 1019
1024 1020 class LSString(str):
1025 1021 """String derivative with a special access attributes.
1026 1022
1027 1023 These are normal strings, but with the special attributes:
1028 1024
1029 1025 .l (or .list) : value as list (split on newlines).
1030 1026 .n (or .nlstr): original value (the string itself).
1031 1027 .s (or .spstr): value as whitespace-separated string.
1032 1028 .p (or .paths): list of path objects
1033 1029
1034 1030 Any values which require transformations are computed only once and
1035 1031 cached.
1036 1032
1037 1033 Such strings are very useful to efficiently interact with the shell, which
1038 1034 typically only understands whitespace-separated options for commands."""
1039 1035
1040 1036 def get_list(self):
1041 1037 try:
1042 1038 return self.__list
1043 1039 except AttributeError:
1044 1040 self.__list = self.split('\n')
1045 1041 return self.__list
1046 1042
1047 1043 l = list = property(get_list)
1048 1044
1049 1045 def get_spstr(self):
1050 1046 try:
1051 1047 return self.__spstr
1052 1048 except AttributeError:
1053 1049 self.__spstr = self.replace('\n',' ')
1054 1050 return self.__spstr
1055 1051
1056 1052 s = spstr = property(get_spstr)
1057 1053
1058 1054 def get_nlstr(self):
1059 1055 return self
1060 1056
1061 1057 n = nlstr = property(get_nlstr)
1062 1058
1063 1059 def get_paths(self):
1064 1060 try:
1065 1061 return self.__paths
1066 1062 except AttributeError:
1067 1063 self.__paths = [path(p) for p in self.split('\n') if os.path.exists(p)]
1068 1064 return self.__paths
1069 1065
1070 1066 p = paths = property(get_paths)
1071 1067
1072 1068 def print_lsstring(arg):
1073 1069 """ Prettier (non-repr-like) and more informative printer for LSString """
1074 1070 print "LSString (.p, .n, .l, .s available). Value:"
1075 1071 print arg
1076 1072
1077 1073 print_lsstring = result_display.when_type(LSString)(print_lsstring)
1078 1074
1079 1075 #----------------------------------------------------------------------------
1080 1076 class SList(list):
1081 1077 """List derivative with a special access attributes.
1082 1078
1083 1079 These are normal lists, but with the special attributes:
1084 1080
1085 1081 .l (or .list) : value as list (the list itself).
1086 1082 .n (or .nlstr): value as a string, joined on newlines.
1087 1083 .s (or .spstr): value as a string, joined on spaces.
1088 1084 .p (or .paths): list of path objects
1089 1085
1090 1086 Any values which require transformations are computed only once and
1091 1087 cached."""
1092 1088
1093 1089 def get_list(self):
1094 1090 return self
1095 1091
1096 1092 l = list = property(get_list)
1097 1093
1098 1094 def get_spstr(self):
1099 1095 try:
1100 1096 return self.__spstr
1101 1097 except AttributeError:
1102 1098 self.__spstr = ' '.join(self)
1103 1099 return self.__spstr
1104 1100
1105 1101 s = spstr = property(get_spstr)
1106 1102
1107 1103 def get_nlstr(self):
1108 1104 try:
1109 1105 return self.__nlstr
1110 1106 except AttributeError:
1111 1107 self.__nlstr = '\n'.join(self)
1112 1108 return self.__nlstr
1113 1109
1114 1110 n = nlstr = property(get_nlstr)
1115 1111
1116 1112 def get_paths(self):
1117 1113 try:
1118 1114 return self.__paths
1119 1115 except AttributeError:
1120 1116 self.__paths = [path(p) for p in self if os.path.exists(p)]
1121 1117 return self.__paths
1122 1118
1123 1119 p = paths = property(get_paths)
1124 1120
1125 1121 def grep(self, pattern, prune = False, field = None):
1126 1122 """ Return all strings matching 'pattern' (a regex or callable)
1127 1123
1128 1124 This is case-insensitive. If prune is true, return all items
1129 1125 NOT matching the pattern.
1130 1126
1131 1127 If field is specified, the match must occur in the specified
1132 1128 whitespace-separated field.
1133 1129
1134 1130 Examples::
1135 1131
1136 1132 a.grep( lambda x: x.startswith('C') )
1137 1133 a.grep('Cha.*log', prune=1)
1138 1134 a.grep('chm', field=-1)
1139 1135 """
1140 1136
1141 1137 def match_target(s):
1142 1138 if field is None:
1143 1139 return s
1144 1140 parts = s.split()
1145 1141 try:
1146 1142 tgt = parts[field]
1147 1143 return tgt
1148 1144 except IndexError:
1149 1145 return ""
1150 1146
1151 1147 if isinstance(pattern, basestring):
1152 1148 pred = lambda x : re.search(pattern, x, re.IGNORECASE)
1153 1149 else:
1154 1150 pred = pattern
1155 1151 if not prune:
1156 1152 return SList([el for el in self if pred(match_target(el))])
1157 1153 else:
1158 1154 return SList([el for el in self if not pred(match_target(el))])
1159 1155 def fields(self, *fields):
1160 1156 """ Collect whitespace-separated fields from string list
1161 1157
1162 1158 Allows quick awk-like usage of string lists.
1163 1159
1164 1160 Example data (in var a, created by 'a = !ls -l')::
1165 1161 -rwxrwxrwx 1 ville None 18 Dec 14 2006 ChangeLog
1166 1162 drwxrwxrwx+ 6 ville None 0 Oct 24 18:05 IPython
1167 1163
1168 1164 a.fields(0) is ['-rwxrwxrwx', 'drwxrwxrwx+']
1169 1165 a.fields(1,0) is ['1 -rwxrwxrwx', '6 drwxrwxrwx+']
1170 1166 (note the joining by space).
1171 1167 a.fields(-1) is ['ChangeLog', 'IPython']
1172 1168
1173 1169 IndexErrors are ignored.
1174 1170
1175 1171 Without args, fields() just split()'s the strings.
1176 1172 """
1177 1173 if len(fields) == 0:
1178 1174 return [el.split() for el in self]
1179 1175
1180 1176 res = SList()
1181 1177 for el in [f.split() for f in self]:
1182 1178 lineparts = []
1183 1179
1184 1180 for fd in fields:
1185 1181 try:
1186 1182 lineparts.append(el[fd])
1187 1183 except IndexError:
1188 1184 pass
1189 1185 if lineparts:
1190 1186 res.append(" ".join(lineparts))
1191 1187
1192 1188 return res
1193 1189 def sort(self,field= None, nums = False):
1194 1190 """ sort by specified fields (see fields())
1195 1191
1196 1192 Example::
1197 1193 a.sort(1, nums = True)
1198 1194
1199 1195 Sorts a by second field, in numerical order (so that 21 > 3)
1200 1196
1201 1197 """
1202 1198
1203 1199 #decorate, sort, undecorate
1204 1200 if field is not None:
1205 1201 dsu = [[SList([line]).fields(field), line] for line in self]
1206 1202 else:
1207 1203 dsu = [[line, line] for line in self]
1208 1204 if nums:
1209 1205 for i in range(len(dsu)):
1210 1206 numstr = "".join([ch for ch in dsu[i][0] if ch.isdigit()])
1211 1207 try:
1212 1208 n = int(numstr)
1213 1209 except ValueError:
1214 1210 n = 0;
1215 1211 dsu[i][0] = n
1216 1212
1217 1213
1218 1214 dsu.sort()
1219 1215 return SList([t[1] for t in dsu])
1220 1216
1221 1217 def print_slist(arg):
1222 1218 """ Prettier (non-repr-like) and more informative printer for SList """
1223 1219 print "SList (.p, .n, .l, .s, .grep(), .fields(), sort() available):"
1224 1220 if hasattr(arg, 'hideonce') and arg.hideonce:
1225 1221 arg.hideonce = False
1226 1222 return
1227 1223
1228 1224 nlprint(arg)
1229 1225
1230 1226 print_slist = result_display.when_type(SList)(print_slist)
1231 1227
1232 1228
1233 1229
1234 1230 #----------------------------------------------------------------------------
1235 1231 def esc_quotes(strng):
1236 1232 """Return the input string with single and double quotes escaped out"""
1237 1233
1238 1234 return strng.replace('"','\\"').replace("'","\\'")
1239 1235
1240 1236 #----------------------------------------------------------------------------
1241 1237 def make_quoted_expr(s):
1242 1238 """Return string s in appropriate quotes, using raw string if possible.
1243 1239
1244 1240 Effectively this turns string: cd \ao\ao\
1245 1241 to: r"cd \ao\ao\_"[:-1]
1246 1242
1247 1243 Note the use of raw string and padding at the end to allow trailing backslash.
1248 1244
1249 1245 """
1250 1246
1251 1247 tail = ''
1252 1248 tailpadding = ''
1253 1249 raw = ''
1254 1250 if "\\" in s:
1255 1251 raw = 'r'
1256 1252 if s.endswith('\\'):
1257 1253 tail = '[:-1]'
1258 1254 tailpadding = '_'
1259 1255 if '"' not in s:
1260 1256 quote = '"'
1261 1257 elif "'" not in s:
1262 1258 quote = "'"
1263 1259 elif '"""' not in s and not s.endswith('"'):
1264 1260 quote = '"""'
1265 1261 elif "'''" not in s and not s.endswith("'"):
1266 1262 quote = "'''"
1267 1263 else:
1268 1264 # give up, backslash-escaped string will do
1269 1265 return '"%s"' % esc_quotes(s)
1270 1266 res = raw + quote + s + tailpadding + quote + tail
1271 1267 return res
1272 1268
1273 1269
1274 1270 #----------------------------------------------------------------------------
1275 1271 def raw_input_multi(header='', ps1='==> ', ps2='..> ',terminate_str = '.'):
1276 1272 """Take multiple lines of input.
1277 1273
1278 1274 A list with each line of input as a separate element is returned when a
1279 1275 termination string is entered (defaults to a single '.'). Input can also
1280 1276 terminate via EOF (^D in Unix, ^Z-RET in Windows).
1281 1277
1282 1278 Lines of input which end in \\ are joined into single entries (and a
1283 1279 secondary continuation prompt is issued as long as the user terminates
1284 1280 lines with \\). This allows entering very long strings which are still
1285 1281 meant to be treated as single entities.
1286 1282 """
1287 1283
1288 1284 try:
1289 1285 if header:
1290 1286 header += '\n'
1291 1287 lines = [raw_input(header + ps1)]
1292 1288 except EOFError:
1293 1289 return []
1294 1290 terminate = [terminate_str]
1295 1291 try:
1296 1292 while lines[-1:] != terminate:
1297 1293 new_line = raw_input(ps1)
1298 1294 while new_line.endswith('\\'):
1299 1295 new_line = new_line[:-1] + raw_input(ps2)
1300 1296 lines.append(new_line)
1301 1297
1302 1298 return lines[:-1] # don't return the termination command
1303 1299 except EOFError:
1304 1300 print
1305 1301 return lines
1306 1302
1307 1303 #----------------------------------------------------------------------------
1308 1304 def raw_input_ext(prompt='', ps2='... '):
1309 1305 """Similar to raw_input(), but accepts extended lines if input ends with \\."""
1310 1306
1311 1307 line = raw_input(prompt)
1312 1308 while line.endswith('\\'):
1313 1309 line = line[:-1] + raw_input(ps2)
1314 1310 return line
1315 1311
1316 1312 #----------------------------------------------------------------------------
1317 1313 def ask_yes_no(prompt,default=None):
1318 1314 """Asks a question and returns a boolean (y/n) answer.
1319 1315
1320 1316 If default is given (one of 'y','n'), it is used if the user input is
1321 1317 empty. Otherwise the question is repeated until an answer is given.
1322 1318
1323 1319 An EOF is treated as the default answer. If there is no default, an
1324 1320 exception is raised to prevent infinite loops.
1325 1321
1326 1322 Valid answers are: y/yes/n/no (match is not case sensitive)."""
1327 1323
1328 1324 answers = {'y':True,'n':False,'yes':True,'no':False}
1329 1325 ans = None
1330 1326 while ans not in answers.keys():
1331 1327 try:
1332 1328 ans = raw_input(prompt+' ').lower()
1333 1329 if not ans: # response was an empty string
1334 1330 ans = default
1335 1331 except KeyboardInterrupt:
1336 1332 pass
1337 1333 except EOFError:
1338 1334 if default in answers.keys():
1339 1335 ans = default
1340 1336 print
1341 1337 else:
1342 1338 raise
1343 1339
1344 1340 return answers[ans]
1345 1341
1346 1342 #----------------------------------------------------------------------------
1347 1343 def marquee(txt='',width=78,mark='*'):
1348 1344 """Return the input string centered in a 'marquee'."""
1349 1345 if not txt:
1350 1346 return (mark*width)[:width]
1351 1347 nmark = (width-len(txt)-2)/len(mark)/2
1352 1348 if nmark < 0: nmark =0
1353 1349 marks = mark*nmark
1354 1350 return '%s %s %s' % (marks,txt,marks)
1355 1351
1356 1352 #----------------------------------------------------------------------------
1357 1353 class EvalDict:
1358 1354 """
1359 1355 Emulate a dict which evaluates its contents in the caller's frame.
1360 1356
1361 1357 Usage:
1362 1358 >>> number = 19
1363 1359
1364 1360 >>> text = "python"
1365 1361
1366 1362 >>> print "%(text.capitalize())s %(number/9.0).1f rules!" % EvalDict()
1367 1363 Python 2.1 rules!
1368 1364 """
1369 1365
1370 1366 # This version is due to sismex01@hebmex.com on c.l.py, and is basically a
1371 1367 # modified (shorter) version of:
1372 1368 # http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/66018 by
1373 1369 # Skip Montanaro (skip@pobox.com).
1374 1370
1375 1371 def __getitem__(self, name):
1376 1372 frame = sys._getframe(1)
1377 1373 return eval(name, frame.f_globals, frame.f_locals)
1378 1374
1379 1375 EvalString = EvalDict # for backwards compatibility
1380 1376 #----------------------------------------------------------------------------
1381 1377 def qw(words,flat=0,sep=None,maxsplit=-1):
1382 1378 """Similar to Perl's qw() operator, but with some more options.
1383 1379
1384 1380 qw(words,flat=0,sep=' ',maxsplit=-1) -> words.split(sep,maxsplit)
1385 1381
1386 1382 words can also be a list itself, and with flat=1, the output will be
1387 1383 recursively flattened.
1388 1384
1389 1385 Examples:
1390 1386
1391 1387 >>> qw('1 2')
1392 1388 ['1', '2']
1393 1389
1394 1390 >>> qw(['a b','1 2',['m n','p q']])
1395 1391 [['a', 'b'], ['1', '2'], [['m', 'n'], ['p', 'q']]]
1396 1392
1397 1393 >>> qw(['a b','1 2',['m n','p q']],flat=1)
1398 1394 ['a', 'b', '1', '2', 'm', 'n', 'p', 'q']
1399 1395 """
1400 1396
1401 1397 if type(words) in StringTypes:
1402 1398 return [word.strip() for word in words.split(sep,maxsplit)
1403 1399 if word and not word.isspace() ]
1404 1400 if flat:
1405 1401 return flatten(map(qw,words,[1]*len(words)))
1406 1402 return map(qw,words)
1407 1403
1408 1404 #----------------------------------------------------------------------------
1409 1405 def qwflat(words,sep=None,maxsplit=-1):
1410 1406 """Calls qw(words) in flat mode. It's just a convenient shorthand."""
1411 1407 return qw(words,1,sep,maxsplit)
1412 1408
1413 1409 #----------------------------------------------------------------------------
1414 1410 def qw_lol(indata):
1415 1411 """qw_lol('a b') -> [['a','b']],
1416 1412 otherwise it's just a call to qw().
1417 1413
1418 1414 We need this to make sure the modules_some keys *always* end up as a
1419 1415 list of lists."""
1420 1416
1421 1417 if type(indata) in StringTypes:
1422 1418 return [qw(indata)]
1423 1419 else:
1424 1420 return qw(indata)
1425 1421
1426 1422 #-----------------------------------------------------------------------------
1427 1423 def list_strings(arg):
1428 1424 """Always return a list of strings, given a string or list of strings
1429 1425 as input."""
1430 1426
1431 1427 if type(arg) in StringTypes: return [arg]
1432 1428 else: return arg
1433 1429
1434 1430 #----------------------------------------------------------------------------
1435 1431 def grep(pat,list,case=1):
1436 1432 """Simple minded grep-like function.
1437 1433 grep(pat,list) returns occurrences of pat in list, None on failure.
1438 1434
1439 1435 It only does simple string matching, with no support for regexps. Use the
1440 1436 option case=0 for case-insensitive matching."""
1441 1437
1442 1438 # This is pretty crude. At least it should implement copying only references
1443 1439 # to the original data in case it's big. Now it copies the data for output.
1444 1440 out=[]
1445 1441 if case:
1446 1442 for term in list:
1447 1443 if term.find(pat)>-1: out.append(term)
1448 1444 else:
1449 1445 lpat=pat.lower()
1450 1446 for term in list:
1451 1447 if term.lower().find(lpat)>-1: out.append(term)
1452 1448
1453 1449 if len(out): return out
1454 1450 else: return None
1455 1451
1456 1452 #----------------------------------------------------------------------------
1457 1453 def dgrep(pat,*opts):
1458 1454 """Return grep() on dir()+dir(__builtins__).
1459 1455
1460 1456 A very common use of grep() when working interactively."""
1461 1457
1462 1458 return grep(pat,dir(__main__)+dir(__main__.__builtins__),*opts)
1463 1459
1464 1460 #----------------------------------------------------------------------------
1465 1461 def idgrep(pat):
1466 1462 """Case-insensitive dgrep()"""
1467 1463
1468 1464 return dgrep(pat,0)
1469 1465
1470 1466 #----------------------------------------------------------------------------
1471 1467 def igrep(pat,list):
1472 1468 """Synonym for case-insensitive grep."""
1473 1469
1474 1470 return grep(pat,list,case=0)
1475 1471
1476 1472 #----------------------------------------------------------------------------
1477 1473 def indent(str,nspaces=4,ntabs=0):
1478 1474 """Indent a string a given number of spaces or tabstops.
1479 1475
1480 1476 indent(str,nspaces=4,ntabs=0) -> indent str by ntabs+nspaces.
1481 1477 """
1482 1478 if str is None:
1483 1479 return
1484 1480 ind = '\t'*ntabs+' '*nspaces
1485 1481 outstr = '%s%s' % (ind,str.replace(os.linesep,os.linesep+ind))
1486 1482 if outstr.endswith(os.linesep+ind):
1487 1483 return outstr[:-len(ind)]
1488 1484 else:
1489 1485 return outstr
1490 1486
1491 1487 #-----------------------------------------------------------------------------
1492 1488 def native_line_ends(filename,backup=1):
1493 1489 """Convert (in-place) a file to line-ends native to the current OS.
1494 1490
1495 1491 If the optional backup argument is given as false, no backup of the
1496 1492 original file is left. """
1497 1493
1498 1494 backup_suffixes = {'posix':'~','dos':'.bak','nt':'.bak','mac':'.bak'}
1499 1495
1500 1496 bak_filename = filename + backup_suffixes[os.name]
1501 1497
1502 1498 original = open(filename).read()
1503 1499 shutil.copy2(filename,bak_filename)
1504 1500 try:
1505 1501 new = open(filename,'wb')
1506 1502 new.write(os.linesep.join(original.splitlines()))
1507 1503 new.write(os.linesep) # ALWAYS put an eol at the end of the file
1508 1504 new.close()
1509 1505 except:
1510 1506 os.rename(bak_filename,filename)
1511 1507 if not backup:
1512 1508 try:
1513 1509 os.remove(bak_filename)
1514 1510 except:
1515 1511 pass
1516 1512
1517 1513 #----------------------------------------------------------------------------
1518 1514 def get_pager_cmd(pager_cmd = None):
1519 1515 """Return a pager command.
1520 1516
1521 1517 Makes some attempts at finding an OS-correct one."""
1522 1518
1523 1519 if os.name == 'posix':
1524 1520 default_pager_cmd = 'less -r' # -r for color control sequences
1525 1521 elif os.name in ['nt','dos']:
1526 1522 default_pager_cmd = 'type'
1527 1523
1528 1524 if pager_cmd is None:
1529 1525 try:
1530 1526 pager_cmd = os.environ['PAGER']
1531 1527 except:
1532 1528 pager_cmd = default_pager_cmd
1533 1529 return pager_cmd
1534 1530
1535 1531 #-----------------------------------------------------------------------------
1536 1532 def get_pager_start(pager,start):
1537 1533 """Return the string for paging files with an offset.
1538 1534
1539 1535 This is the '+N' argument which less and more (under Unix) accept.
1540 1536 """
1541 1537
1542 1538 if pager in ['less','more']:
1543 1539 if start:
1544 1540 start_string = '+' + str(start)
1545 1541 else:
1546 1542 start_string = ''
1547 1543 else:
1548 1544 start_string = ''
1549 1545 return start_string
1550 1546
1551 1547 #----------------------------------------------------------------------------
1552 1548 # (X)emacs on W32 doesn't like to be bypassed with msvcrt.getch()
1553 1549 if os.name == 'nt' and os.environ.get('TERM','dumb') != 'emacs':
1554 1550 import msvcrt
1555 1551 def page_more():
1556 1552 """ Smart pausing between pages
1557 1553
1558 1554 @return: True if need print more lines, False if quit
1559 1555 """
1560 1556 Term.cout.write('---Return to continue, q to quit--- ')
1561 1557 ans = msvcrt.getch()
1562 1558 if ans in ("q", "Q"):
1563 1559 result = False
1564 1560 else:
1565 1561 result = True
1566 1562 Term.cout.write("\b"*37 + " "*37 + "\b"*37)
1567 1563 return result
1568 1564 else:
1569 1565 def page_more():
1570 1566 ans = raw_input('---Return to continue, q to quit--- ')
1571 1567 if ans.lower().startswith('q'):
1572 1568 return False
1573 1569 else:
1574 1570 return True
1575 1571
1576 1572 esc_re = re.compile(r"(\x1b[^m]+m)")
1577 1573
1578 1574 def page_dumb(strng,start=0,screen_lines=25):
1579 1575 """Very dumb 'pager' in Python, for when nothing else works.
1580 1576
1581 1577 Only moves forward, same interface as page(), except for pager_cmd and
1582 1578 mode."""
1583 1579
1584 1580 out_ln = strng.splitlines()[start:]
1585 1581 screens = chop(out_ln,screen_lines-1)
1586 1582 if len(screens) == 1:
1587 1583 print >>Term.cout, os.linesep.join(screens[0])
1588 1584 else:
1589 1585 last_escape = ""
1590 1586 for scr in screens[0:-1]:
1591 1587 hunk = os.linesep.join(scr)
1592 1588 print >>Term.cout, last_escape + hunk
1593 1589 if not page_more():
1594 1590 return
1595 1591 esc_list = esc_re.findall(hunk)
1596 1592 if len(esc_list) > 0:
1597 1593 last_escape = esc_list[-1]
1598 1594 print >>Term.cout, last_escape + os.linesep.join(screens[-1])
1599 1595
1600 1596 #----------------------------------------------------------------------------
1601 1597 def page(strng,start=0,screen_lines=0,pager_cmd = None):
1602 1598 """Print a string, piping through a pager after a certain length.
1603 1599
1604 1600 The screen_lines parameter specifies the number of *usable* lines of your
1605 1601 terminal screen (total lines minus lines you need to reserve to show other
1606 1602 information).
1607 1603
1608 1604 If you set screen_lines to a number <=0, page() will try to auto-determine
1609 1605 your screen size and will only use up to (screen_size+screen_lines) for
1610 1606 printing, paging after that. That is, if you want auto-detection but need
1611 1607 to reserve the bottom 3 lines of the screen, use screen_lines = -3, and for
1612 1608 auto-detection without any lines reserved simply use screen_lines = 0.
1613 1609
1614 1610 If a string won't fit in the allowed lines, it is sent through the
1615 1611 specified pager command. If none given, look for PAGER in the environment,
1616 1612 and ultimately default to less.
1617 1613
1618 1614 If no system pager works, the string is sent through a 'dumb pager'
1619 1615 written in python, very simplistic.
1620 1616 """
1621 1617
1622 1618 # Some routines may auto-compute start offsets incorrectly and pass a
1623 1619 # negative value. Offset to 0 for robustness.
1624 1620 start = max(0,start)
1625 1621
1626 1622 # first, try the hook
1627 1623 ip = IPython.ipapi.get()
1628 1624 if ip:
1629 1625 try:
1630 1626 ip.IP.hooks.show_in_pager(strng)
1631 1627 return
1632 1628 except IPython.ipapi.TryNext:
1633 1629 pass
1634 1630
1635 1631 # Ugly kludge, but calling curses.initscr() flat out crashes in emacs
1636 1632 TERM = os.environ.get('TERM','dumb')
1637 1633 if TERM in ['dumb','emacs'] and os.name != 'nt':
1638 1634 print strng
1639 1635 return
1640 1636 # chop off the topmost part of the string we don't want to see
1641 1637 str_lines = strng.split(os.linesep)[start:]
1642 1638 str_toprint = os.linesep.join(str_lines)
1643 1639 num_newlines = len(str_lines)
1644 1640 len_str = len(str_toprint)
1645 1641
1646 1642 # Dumb heuristics to guesstimate number of on-screen lines the string
1647 1643 # takes. Very basic, but good enough for docstrings in reasonable
1648 1644 # terminals. If someone later feels like refining it, it's not hard.
1649 1645 numlines = max(num_newlines,int(len_str/80)+1)
1650 1646
1651 1647 if os.name == "nt":
1652 1648 screen_lines_def = get_console_size(defaulty=25)[1]
1653 1649 else:
1654 1650 screen_lines_def = 25 # default value if we can't auto-determine
1655 1651
1656 1652 # auto-determine screen size
1657 1653 if screen_lines <= 0:
1658 1654 if TERM=='xterm':
1659 1655 use_curses = USE_CURSES
1660 1656 else:
1661 1657 # curses causes problems on many terminals other than xterm.
1662 1658 use_curses = False
1663 1659 if use_curses:
1664 1660 # There is a bug in curses, where *sometimes* it fails to properly
1665 1661 # initialize, and then after the endwin() call is made, the
1666 1662 # terminal is left in an unusable state. Rather than trying to
1667 1663 # check everytime for this (by requesting and comparing termios
1668 1664 # flags each time), we just save the initial terminal state and
1669 1665 # unconditionally reset it every time. It's cheaper than making
1670 1666 # the checks.
1671 1667 term_flags = termios.tcgetattr(sys.stdout)
1672 1668 scr = curses.initscr()
1673 1669 screen_lines_real,screen_cols = scr.getmaxyx()
1674 1670 curses.endwin()
1675 1671 # Restore terminal state in case endwin() didn't.
1676 1672 termios.tcsetattr(sys.stdout,termios.TCSANOW,term_flags)
1677 1673 # Now we have what we needed: the screen size in rows/columns
1678 1674 screen_lines += screen_lines_real
1679 1675 #print '***Screen size:',screen_lines_real,'lines x',\
1680 1676 #screen_cols,'columns.' # dbg
1681 1677 else:
1682 1678 screen_lines += screen_lines_def
1683 1679
1684 1680 #print 'numlines',numlines,'screenlines',screen_lines # dbg
1685 1681 if numlines <= screen_lines :
1686 1682 #print '*** normal print' # dbg
1687 1683 print >>Term.cout, str_toprint
1688 1684 else:
1689 1685 # Try to open pager and default to internal one if that fails.
1690 1686 # All failure modes are tagged as 'retval=1', to match the return
1691 1687 # value of a failed system command. If any intermediate attempt
1692 1688 # sets retval to 1, at the end we resort to our own page_dumb() pager.
1693 1689 pager_cmd = get_pager_cmd(pager_cmd)
1694 1690 pager_cmd += ' ' + get_pager_start(pager_cmd,start)
1695 1691 if os.name == 'nt':
1696 1692 if pager_cmd.startswith('type'):
1697 1693 # The default WinXP 'type' command is failing on complex strings.
1698 1694 retval = 1
1699 1695 else:
1700 1696 tmpname = tempfile.mktemp('.txt')
1701 1697 tmpfile = file(tmpname,'wt')
1702 1698 tmpfile.write(strng)
1703 1699 tmpfile.close()
1704 1700 cmd = "%s < %s" % (pager_cmd,tmpname)
1705 1701 if os.system(cmd):
1706 1702 retval = 1
1707 1703 else:
1708 1704 retval = None
1709 1705 os.remove(tmpname)
1710 1706 else:
1711 1707 try:
1712 1708 retval = None
1713 1709 # if I use popen4, things hang. No idea why.
1714 1710 #pager,shell_out = os.popen4(pager_cmd)
1715 1711 pager = os.popen(pager_cmd,'w')
1716 1712 pager.write(strng)
1717 1713 pager.close()
1718 1714 retval = pager.close() # success returns None
1719 1715 except IOError,msg: # broken pipe when user quits
1720 1716 if msg.args == (32,'Broken pipe'):
1721 1717 retval = None
1722 1718 else:
1723 1719 retval = 1
1724 1720 except OSError:
1725 1721 # Other strange problems, sometimes seen in Win2k/cygwin
1726 1722 retval = 1
1727 1723 if retval is not None:
1728 1724 page_dumb(strng,screen_lines=screen_lines)
1729 1725
1730 1726 #----------------------------------------------------------------------------
1731 1727 def page_file(fname,start = 0, pager_cmd = None):
1732 1728 """Page a file, using an optional pager command and starting line.
1733 1729 """
1734 1730
1735 1731 pager_cmd = get_pager_cmd(pager_cmd)
1736 1732 pager_cmd += ' ' + get_pager_start(pager_cmd,start)
1737 1733
1738 1734 try:
1739 1735 if os.environ['TERM'] in ['emacs','dumb']:
1740 1736 raise EnvironmentError
1741 1737 xsys(pager_cmd + ' ' + fname)
1742 1738 except:
1743 1739 try:
1744 1740 if start > 0:
1745 1741 start -= 1
1746 1742 page(open(fname).read(),start)
1747 1743 except:
1748 1744 print 'Unable to show file',`fname`
1749 1745
1750 1746
1751 1747 #----------------------------------------------------------------------------
1752 1748 def snip_print(str,width = 75,print_full = 0,header = ''):
1753 1749 """Print a string snipping the midsection to fit in width.
1754 1750
1755 1751 print_full: mode control:
1756 1752 - 0: only snip long strings
1757 1753 - 1: send to page() directly.
1758 1754 - 2: snip long strings and ask for full length viewing with page()
1759 1755 Return 1 if snipping was necessary, 0 otherwise."""
1760 1756
1761 1757 if print_full == 1:
1762 1758 page(header+str)
1763 1759 return 0
1764 1760
1765 1761 print header,
1766 1762 if len(str) < width:
1767 1763 print str
1768 1764 snip = 0
1769 1765 else:
1770 1766 whalf = int((width -5)/2)
1771 1767 print str[:whalf] + ' <...> ' + str[-whalf:]
1772 1768 snip = 1
1773 1769 if snip and print_full == 2:
1774 1770 if raw_input(header+' Snipped. View (y/n)? [N]').lower() == 'y':
1775 1771 page(str)
1776 1772 return snip
1777 1773
1778 1774 #****************************************************************************
1779 1775 # lists, dicts and structures
1780 1776
1781 1777 def belong(candidates,checklist):
1782 1778 """Check whether a list of items appear in a given list of options.
1783 1779
1784 1780 Returns a list of 1 and 0, one for each candidate given."""
1785 1781
1786 1782 return [x in checklist for x in candidates]
1787 1783
1788 1784 #----------------------------------------------------------------------------
1789 1785 def uniq_stable(elems):
1790 1786 """uniq_stable(elems) -> list
1791 1787
1792 1788 Return from an iterable, a list of all the unique elements in the input,
1793 1789 but maintaining the order in which they first appear.
1794 1790
1795 1791 A naive solution to this problem which just makes a dictionary with the
1796 1792 elements as keys fails to respect the stability condition, since
1797 1793 dictionaries are unsorted by nature.
1798 1794
1799 1795 Note: All elements in the input must be valid dictionary keys for this
1800 1796 routine to work, as it internally uses a dictionary for efficiency
1801 1797 reasons."""
1802 1798
1803 1799 unique = []
1804 1800 unique_dict = {}
1805 1801 for nn in elems:
1806 1802 if nn not in unique_dict:
1807 1803 unique.append(nn)
1808 1804 unique_dict[nn] = None
1809 1805 return unique
1810 1806
1811 1807 #----------------------------------------------------------------------------
1812 1808 class NLprinter:
1813 1809 """Print an arbitrarily nested list, indicating index numbers.
1814 1810
1815 1811 An instance of this class called nlprint is available and callable as a
1816 1812 function.
1817 1813
1818 1814 nlprint(list,indent=' ',sep=': ') -> prints indenting each level by 'indent'
1819 1815 and using 'sep' to separate the index from the value. """
1820 1816
1821 1817 def __init__(self):
1822 1818 self.depth = 0
1823 1819
1824 1820 def __call__(self,lst,pos='',**kw):
1825 1821 """Prints the nested list numbering levels."""
1826 1822 kw.setdefault('indent',' ')
1827 1823 kw.setdefault('sep',': ')
1828 1824 kw.setdefault('start',0)
1829 1825 kw.setdefault('stop',len(lst))
1830 1826 # we need to remove start and stop from kw so they don't propagate
1831 1827 # into a recursive call for a nested list.
1832 1828 start = kw['start']; del kw['start']
1833 1829 stop = kw['stop']; del kw['stop']
1834 1830 if self.depth == 0 and 'header' in kw.keys():
1835 1831 print kw['header']
1836 1832
1837 1833 for idx in range(start,stop):
1838 1834 elem = lst[idx]
1839 1835 if type(elem)==type([]):
1840 1836 self.depth += 1
1841 1837 self.__call__(elem,itpl('$pos$idx,'),**kw)
1842 1838 self.depth -= 1
1843 1839 else:
1844 1840 printpl(kw['indent']*self.depth+'$pos$idx$kw["sep"]$elem')
1845 1841
1846 1842 nlprint = NLprinter()
1847 1843 #----------------------------------------------------------------------------
1848 1844 def all_belong(candidates,checklist):
1849 1845 """Check whether a list of items ALL appear in a given list of options.
1850 1846
1851 1847 Returns a single 1 or 0 value."""
1852 1848
1853 1849 return 1-(0 in [x in checklist for x in candidates])
1854 1850
1855 1851 #----------------------------------------------------------------------------
1856 1852 def sort_compare(lst1,lst2,inplace = 1):
1857 1853 """Sort and compare two lists.
1858 1854
1859 1855 By default it does it in place, thus modifying the lists. Use inplace = 0
1860 1856 to avoid that (at the cost of temporary copy creation)."""
1861 1857 if not inplace:
1862 1858 lst1 = lst1[:]
1863 1859 lst2 = lst2[:]
1864 1860 lst1.sort(); lst2.sort()
1865 1861 return lst1 == lst2
1866 1862
1867 1863 #----------------------------------------------------------------------------
1868 1864 def list2dict(lst):
1869 1865 """Takes a list of (key,value) pairs and turns it into a dict."""
1870 1866
1871 1867 dic = {}
1872 1868 for k,v in lst: dic[k] = v
1873 1869 return dic
1874 1870
1875 1871 #----------------------------------------------------------------------------
1876 1872 def list2dict2(lst,default=''):
1877 1873 """Takes a list and turns it into a dict.
1878 1874 Much slower than list2dict, but more versatile. This version can take
1879 1875 lists with sublists of arbitrary length (including sclars)."""
1880 1876
1881 1877 dic = {}
1882 1878 for elem in lst:
1883 1879 if type(elem) in (types.ListType,types.TupleType):
1884 1880 size = len(elem)
1885 1881 if size == 0:
1886 1882 pass
1887 1883 elif size == 1:
1888 1884 dic[elem] = default
1889 1885 else:
1890 1886 k,v = elem[0], elem[1:]
1891 1887 if len(v) == 1: v = v[0]
1892 1888 dic[k] = v
1893 1889 else:
1894 1890 dic[elem] = default
1895 1891 return dic
1896 1892
1897 1893 #----------------------------------------------------------------------------
1898 1894 def flatten(seq):
1899 1895 """Flatten a list of lists (NOT recursive, only works for 2d lists)."""
1900 1896
1901 1897 return [x for subseq in seq for x in subseq]
1902 1898
1903 1899 #----------------------------------------------------------------------------
1904 1900 def get_slice(seq,start=0,stop=None,step=1):
1905 1901 """Get a slice of a sequence with variable step. Specify start,stop,step."""
1906 1902 if stop == None:
1907 1903 stop = len(seq)
1908 1904 item = lambda i: seq[i]
1909 1905 return map(item,xrange(start,stop,step))
1910 1906
1911 1907 #----------------------------------------------------------------------------
1912 1908 def chop(seq,size):
1913 1909 """Chop a sequence into chunks of the given size."""
1914 1910 chunk = lambda i: seq[i:i+size]
1915 1911 return map(chunk,xrange(0,len(seq),size))
1916 1912
1917 1913 #----------------------------------------------------------------------------
1918 1914 # with is a keyword as of python 2.5, so this function is renamed to withobj
1919 1915 # from its old 'with' name.
1920 1916 def with_obj(object, **args):
1921 1917 """Set multiple attributes for an object, similar to Pascal's with.
1922 1918
1923 1919 Example:
1924 1920 with_obj(jim,
1925 1921 born = 1960,
1926 1922 haircolour = 'Brown',
1927 1923 eyecolour = 'Green')
1928 1924
1929 1925 Credit: Greg Ewing, in
1930 1926 http://mail.python.org/pipermail/python-list/2001-May/040703.html.
1931 1927
1932 1928 NOTE: up until IPython 0.7.2, this was called simply 'with', but 'with'
1933 1929 has become a keyword for Python 2.5, so we had to rename it."""
1934 1930
1935 1931 object.__dict__.update(args)
1936 1932
1937 1933 #----------------------------------------------------------------------------
1938 1934 def setattr_list(obj,alist,nspace = None):
1939 1935 """Set a list of attributes for an object taken from a namespace.
1940 1936
1941 1937 setattr_list(obj,alist,nspace) -> sets in obj all the attributes listed in
1942 1938 alist with their values taken from nspace, which must be a dict (something
1943 1939 like locals() will often do) If nspace isn't given, locals() of the
1944 1940 *caller* is used, so in most cases you can omit it.
1945 1941
1946 1942 Note that alist can be given as a string, which will be automatically
1947 1943 split into a list on whitespace. If given as a list, it must be a list of
1948 1944 *strings* (the variable names themselves), not of variables."""
1949 1945
1950 1946 # this grabs the local variables from the *previous* call frame -- that is
1951 1947 # the locals from the function that called setattr_list().
1952 1948 # - snipped from weave.inline()
1953 1949 if nspace is None:
1954 1950 call_frame = sys._getframe().f_back
1955 1951 nspace = call_frame.f_locals
1956 1952
1957 1953 if type(alist) in StringTypes:
1958 1954 alist = alist.split()
1959 1955 for attr in alist:
1960 1956 val = eval(attr,nspace)
1961 1957 setattr(obj,attr,val)
1962 1958
1963 1959 #----------------------------------------------------------------------------
1964 1960 def getattr_list(obj,alist,*args):
1965 1961 """getattr_list(obj,alist[, default]) -> attribute list.
1966 1962
1967 1963 Get a list of named attributes for an object. When a default argument is
1968 1964 given, it is returned when the attribute doesn't exist; without it, an
1969 1965 exception is raised in that case.
1970 1966
1971 1967 Note that alist can be given as a string, which will be automatically
1972 1968 split into a list on whitespace. If given as a list, it must be a list of
1973 1969 *strings* (the variable names themselves), not of variables."""
1974 1970
1975 1971 if type(alist) in StringTypes:
1976 1972 alist = alist.split()
1977 1973 if args:
1978 1974 if len(args)==1:
1979 1975 default = args[0]
1980 1976 return map(lambda attr: getattr(obj,attr,default),alist)
1981 1977 else:
1982 1978 raise ValueError,'getattr_list() takes only one optional argument'
1983 1979 else:
1984 1980 return map(lambda attr: getattr(obj,attr),alist)
1985 1981
1986 1982 #----------------------------------------------------------------------------
1987 1983 def map_method(method,object_list,*argseq,**kw):
1988 1984 """map_method(method,object_list,*args,**kw) -> list
1989 1985
1990 1986 Return a list of the results of applying the methods to the items of the
1991 1987 argument sequence(s). If more than one sequence is given, the method is
1992 1988 called with an argument list consisting of the corresponding item of each
1993 1989 sequence. All sequences must be of the same length.
1994 1990
1995 1991 Keyword arguments are passed verbatim to all objects called.
1996 1992
1997 1993 This is Python code, so it's not nearly as fast as the builtin map()."""
1998 1994
1999 1995 out_list = []
2000 1996 idx = 0
2001 1997 for object in object_list:
2002 1998 try:
2003 1999 handler = getattr(object, method)
2004 2000 except AttributeError:
2005 2001 out_list.append(None)
2006 2002 else:
2007 2003 if argseq:
2008 2004 args = map(lambda lst:lst[idx],argseq)
2009 2005 #print 'ob',object,'hand',handler,'ar',args # dbg
2010 2006 out_list.append(handler(args,**kw))
2011 2007 else:
2012 2008 out_list.append(handler(**kw))
2013 2009 idx += 1
2014 2010 return out_list
2015 2011
2016 2012 #----------------------------------------------------------------------------
2017 2013 def get_class_members(cls):
2018 2014 ret = dir(cls)
2019 2015 if hasattr(cls,'__bases__'):
2020 2016 for base in cls.__bases__:
2021 2017 ret.extend(get_class_members(base))
2022 2018 return ret
2023 2019
2024 2020 #----------------------------------------------------------------------------
2025 2021 def dir2(obj):
2026 2022 """dir2(obj) -> list of strings
2027 2023
2028 2024 Extended version of the Python builtin dir(), which does a few extra
2029 2025 checks, and supports common objects with unusual internals that confuse
2030 2026 dir(), such as Traits and PyCrust.
2031 2027
2032 2028 This version is guaranteed to return only a list of true strings, whereas
2033 2029 dir() returns anything that objects inject into themselves, even if they
2034 2030 are later not really valid for attribute access (many extension libraries
2035 2031 have such bugs).
2036 2032 """
2037 2033
2038 2034 # Start building the attribute list via dir(), and then complete it
2039 2035 # with a few extra special-purpose calls.
2040 2036 words = dir(obj)
2041 2037
2042 2038 if hasattr(obj,'__class__'):
2043 2039 words.append('__class__')
2044 2040 words.extend(get_class_members(obj.__class__))
2045 2041 #if '__base__' in words: 1/0
2046 2042
2047 2043 # Some libraries (such as traits) may introduce duplicates, we want to
2048 2044 # track and clean this up if it happens
2049 2045 may_have_dupes = False
2050 2046
2051 2047 # this is the 'dir' function for objects with Enthought's traits
2052 2048 if hasattr(obj, 'trait_names'):
2053 2049 try:
2054 2050 words.extend(obj.trait_names())
2055 2051 may_have_dupes = True
2056 2052 except TypeError:
2057 2053 # This will happen if `obj` is a class and not an instance.
2058 2054 pass
2059 2055
2060 2056 # Support for PyCrust-style _getAttributeNames magic method.
2061 2057 if hasattr(obj, '_getAttributeNames'):
2062 2058 try:
2063 2059 words.extend(obj._getAttributeNames())
2064 2060 may_have_dupes = True
2065 2061 except TypeError:
2066 2062 # `obj` is a class and not an instance. Ignore
2067 2063 # this error.
2068 2064 pass
2069 2065
2070 2066 if may_have_dupes:
2071 2067 # eliminate possible duplicates, as some traits may also
2072 2068 # appear as normal attributes in the dir() call.
2073 2069 words = list(set(words))
2074 2070 words.sort()
2075 2071
2076 2072 # filter out non-string attributes which may be stuffed by dir() calls
2077 2073 # and poor coding in third-party modules
2078 2074 return [w for w in words if isinstance(w, basestring)]
2079 2075
2080 2076 #----------------------------------------------------------------------------
2081 2077 def import_fail_info(mod_name,fns=None):
2082 2078 """Inform load failure for a module."""
2083 2079
2084 2080 if fns == None:
2085 2081 warn("Loading of %s failed.\n" % (mod_name,))
2086 2082 else:
2087 2083 warn("Loading of %s from %s failed.\n" % (fns,mod_name))
2088 2084
2089 2085 #----------------------------------------------------------------------------
2090 2086 # Proposed popitem() extension, written as a method
2091 2087
2092 2088
2093 2089 class NotGiven: pass
2094 2090
2095 2091 def popkey(dct,key,default=NotGiven):
2096 2092 """Return dct[key] and delete dct[key].
2097 2093
2098 2094 If default is given, return it if dct[key] doesn't exist, otherwise raise
2099 2095 KeyError. """
2100 2096
2101 2097 try:
2102 2098 val = dct[key]
2103 2099 except KeyError:
2104 2100 if default is NotGiven:
2105 2101 raise
2106 2102 else:
2107 2103 return default
2108 2104 else:
2109 2105 del dct[key]
2110 2106 return val
2111 2107
2112 2108 def wrap_deprecated(func, suggest = '<nothing>'):
2113 2109 def newFunc(*args, **kwargs):
2114 2110 warnings.warn("Call to deprecated function %s, use %s instead" %
2115 2111 ( func.__name__, suggest),
2116 2112 category=DeprecationWarning,
2117 2113 stacklevel = 2)
2118 2114 return func(*args, **kwargs)
2119 2115 return newFunc
2120 2116
2121 2117
2122 2118 def _num_cpus_unix():
2123 2119 """Return the number of active CPUs on a Unix system."""
2124 2120 return os.sysconf("SC_NPROCESSORS_ONLN")
2125 2121
2126 2122
2127 2123 def _num_cpus_darwin():
2128 2124 """Return the number of active CPUs on a Darwin system."""
2129 2125 p = subprocess.Popen(['sysctl','-n','hw.ncpu'],stdout=subprocess.PIPE)
2130 2126 return p.stdout.read()
2131 2127
2132 2128
2133 2129 def _num_cpus_windows():
2134 2130 """Return the number of active CPUs on a Windows system."""
2135 2131 return os.environ.get("NUMBER_OF_PROCESSORS")
2136 2132
2137 2133
2138 2134 def num_cpus():
2139 2135 """Return the effective number of CPUs in the system as an integer.
2140 2136
2141 2137 This cross-platform function makes an attempt at finding the total number of
2142 2138 available CPUs in the system, as returned by various underlying system and
2143 2139 python calls.
2144 2140
2145 2141 If it can't find a sensible answer, it returns 1 (though an error *may* make
2146 2142 it return a large positive number that's actually incorrect).
2147 2143 """
2148 2144
2149 2145 # Many thanks to the Parallel Python project (http://www.parallelpython.com)
2150 2146 # for the names of the keys we needed to look up for this function. This
2151 2147 # code was inspired by their equivalent function.
2152 2148
2153 2149 ncpufuncs = {'Linux':_num_cpus_unix,
2154 2150 'Darwin':_num_cpus_darwin,
2155 2151 'Windows':_num_cpus_windows,
2156 2152 # On Vista, python < 2.5.2 has a bug and returns 'Microsoft'
2157 2153 # See http://bugs.python.org/issue1082 for details.
2158 2154 'Microsoft':_num_cpus_windows,
2159 2155 }
2160 2156
2161 2157 ncpufunc = ncpufuncs.get(platform.system(),
2162 2158 # default to unix version (Solaris, AIX, etc)
2163 2159 _num_cpus_unix)
2164 2160
2165 2161 try:
2166 2162 ncpus = max(1,int(ncpufunc()))
2167 2163 except:
2168 2164 ncpus = 1
2169 2165 return ncpus
2170 2166
2171 2167 #*************************** end of file <genutils.py> **********************
General Comments 0
You need to be logged in to leave comments. Login now