Coverage for /wheeldirectory/casa-6.7.0-12-py3.10.el8/lib/py/lib/python3.10/site-packages/casatasks/private/task_sdfixscan.py: 90%
411 statements
« prev ^ index » next coverage.py v7.6.4, created at 2024-11-01 07:19 +0000
1# sd task for image processing (fft_mask or model)
2import os
3import time
5import numpy
6import numpy.fft as npfft
8from casatasks import casalog
9from casatools import ctsys
10from casatools import image as iatool
11from casatools import quanta
13from . import sdutil
16def create_4d_image(infile, outfile):
17 ia = iatool()
18 ia.open(infile)
19 image_shape = ia.shape()
20 try:
21 if len(image_shape) < 4:
22 # add degenerate axes
24 cs = ia.coordsys()
25 axistypes = cs.axiscoordinatetypes()
26 no_stokes = 'Stokes' not in axistypes
27 no_spectral = 'Spectral' not in axistypes
28 stokes_axis = 'I' if no_stokes else ''
29 outimage = ia.adddegaxes(outfile=outfile, spectral=no_spectral,
30 stokes=stokes_axis)
31 else:
32 # generage complete copy of input image using subimage
33 outimage = ia.subimage(outfile=outfile)
34 finally:
35 if len(image_shape) < 4:
36 cs.done()
37 ia.close()
39 return outimage
42@sdutil.sdtask_decorator
43def sdfixscan(infiles, mode, numpoly, beamsize, smoothsize, direction, maskwidth,
44 tmax, tmin, outfile, overwrite):
45 with sdfixscan_worker(**locals()) as worker:
46 worker.initialize()
47 worker.execute()
48 worker.finalize()
51class sdfixscan_worker(sdutil.sdtask_interface):
52 def __init__(self, **kwargs):
53 super(sdfixscan_worker, self).__init__(**kwargs)
55 def __del__(self, base=sdutil.sdtask_interface):
56 # cleanup method must be called when the instance is
57 # deleted
58 self.cleanup()
59 super(sdfixscan_worker, self).__del__()
61 def initialize(self):
62 self.parameter_check()
64 # temporary filename
65 tmpstr = time.ctime().replace(' ', '_').replace(':', '_')
66 self.tmpmskname = 'masked.' + tmpstr + '.im'
67 self.tmpconvname = 'convolve2d.' + tmpstr + '.im'
68 self.tmppolyname = 'polyfit.' + tmpstr + '.im'
69 # set tempolary filename
70 self.tmprealname = []
71 self.tmpimagname = []
72 self.image = None
73 self.convimage = None
74 self.polyimage = None
75 self.imageorg = None
76 self.realimage = None
77 self.imagimage = None
78 if type(self.infiles) == str:
79 self.tmprealname.append('fft.' + tmpstr + '.real..0.im')
80 self.tmpimagname.append('fft.' + tmpstr + '.imag.0.im')
81 else:
82 for i in range(len(self.infiles)):
83 self.tmprealname.append('fft.%s.%s.real.im' % (tmpstr, i))
84 self.tmpimagname.append('fft.%s.%s.imag.im' % (tmpstr, i))
86 # default output filename
87 if self.outfile == '':
88 self.outfile = 'sdfixscan.out.im'
89 casalog.post('outfile=%s' % self.outfile)
91 # threshold
92 self.nolimit = 'nolimit'
93 if self.tmin == 0.0 and self.tmax == 0.0:
94 self.thresh = []
95 elif self.tmin > self.tmax:
96 casalog.post('tmin > tmax. Swapped.')
97 self.thresh = [self.tmax, self.tmin]
98 elif self.tmin == self.tmax:
99 if self.tmin > 0.0:
100 casalog.post('tmin == tmax. Use tmin as minumum threshold.')
101 self.thresh = [self.tmin, self.nolimit]
102 else:
103 casalog.post('tmin == tmax. Use tmax as maximum threshold.')
104 self.thresh = [self.nolimit, self.tmin]
105 else:
106 self.thresh = [self.tmin, self.tmax]
108 def parameter_check(self):
109 if self.mode.lower() == 'model':
110 # Pressed-out method
111 # check input file
112 if type(self.infiles) == list:
113 if len(self.infiles) != 1:
114 raise Exception("infiles allows only one input file for pressed-out method.")
115 else:
116 self.infiles = self.infiles[0]
117 # check direction
118 if type(self.direction) == list:
119 if len(self.direction) != 1:
120 raise Exception("direction allows only one direction for pressed-out method.")
121 else:
122 self.direction = self.direction[0]
123 elif self.mode.lower() == 'fft_mask':
124 # FFT-based basket-weaving method
125 # check input file
126 if type(self.infiles) == str or \
127 (type(self.infiles) == list and len(self.infiles) < 2):
128 raise Exception("infiles should be a list of input images for Basket-Weaving.")
130 # check direction
131 if type(self.direction) == float:
132 raise Exception('direction must have at least two different direction.')
133 else:
134 if len(self.direction) < 2:
135 raise Exception('direction must have at least two different direction.')
136 else:
137 raise Exception('Unsupported processing mode: %s' % (self.mode))
139 def execute(self):
140 if self.mode.lower() == 'model':
141 self.__execute_press()
142 elif self.mode.lower() == 'fft_mask':
143 self.__execute_basket_weaving()
145 def __execute_press(self):
146 ###
147 # Pressed-out method (Sofue & Reich 1979)
148 ###
149 casalog.post('Apply Pressed-out method')
151 # CAS-5410 Use private tools inside task scripts
152 ia = iatool()
154 # mask
155 self.image = ia.newimagefromimage(infile=self.infiles, outfile=self.tmpmskname)
156 # back-up original mask name
157 is_initial_mask = (self.image.maskhandler('default')[0] != '')
158 temp_maskname = "temporal"
159 # imshape = self.image.shape()
160 # ndim = len(imshape)
161 # nx = imshape[0]
162 # ny = imshape[1]
163 if len(self.thresh) == 0:
164 casalog.post('Use whole region')
165 else:
166 # mask pixels beyond thresholds
167 maskstr = ("mask('%s')" % self.tmpmskname)
168 if self.thresh[0] != self.nolimit:
169 maskstr += (" && '%s'>=%f" % (self.tmpmskname, self.thresh[0]))
170 if self.thresh[1] != self.nolimit:
171 maskstr += (" && '%s'<=%f" % (self.tmpmskname, self.thresh[1]))
172 # Need to flush to image once to calcmask ... sigh
173 self.image.done()
174 self.image = ia.newimage(self.tmpmskname)
175 self.image.calcmask(mask=maskstr, name=temp_maskname, asdefault=True)
177 # smoothing
178 # bmajor = 0.0
179 # bminor = 0.0
180 # CAS-5410 Use private tools inside task scripts
181 qa = quanta()
182 if type(self.beamsize) == str:
183 qbeamsize = qa.quantity(self.beamsize)
184 else:
185 qbeamsize = qa.quantity(self.beamsize, 'arcsec')
186 if type(self.smoothsize) == str:
187 # bmajor = smoothsize
188 # bminor = smoothsize
189 qsmoothsize = qa.quantity(self.smoothsize)
190 else:
191 # bmajor = '%sarcsec' % (beamsize*smoothsize)
192 # bminor = '%sarcsec' % (beamsize*smoothsize)
193 qsmoothsize = qa.mul(qbeamsize, self.smoothsize)
194 bmajor = qsmoothsize
195 bminor = qsmoothsize
196 pa = qa.quantity(0.0, 'deg')
197 # masked channels are replaced by zero and convolved here.
198 self.convimage = self.image.convolve2d(outfile=self.tmppolyname,
199 major=bmajor, minor=bminor, pa=pa,
200 overwrite=True)
201 self.convimage.done()
203 # get dTij (original - smoothed)
204 self.convimage = ia.imagecalc(outfile=self.tmpconvname,
205 pixels='"{org}" - "{conv}"'.format(org=self.tmpmskname,
206 conv=self.tmppolyname),
207 overwrite=True)
209 # polynomial fit
210 fitaxis = 0
211 if self.direction == 0.0:
212 fitaxis = 0
213 elif self.direction == 90.0:
214 fitaxis = 1
215 else:
216 raise Exception(
217 "Sorry, the task don't support inclined scan "
218 "with respect to horizontal or vertical axis, right now.")
219 if os.path.exists(self.tmppolyname):
220 # CAS-5410 Use private tools inside task scripts
221 ctsys.removetable([self.tmppolyname])
222 self.convimage.setbrightnessunit('K')
223 # Unfortunately, ia.fitprofile is very fragile.
224 # Using numpy instead for fitting with masked pixels (KS, 2014/07/02)
225 # resultdic = self.convimage.fitprofile(model=self.tmppolyname, axis=fitaxis,
226 # poly=self.numpoly, ngauss=0, multifit=True, gmncomps=0 )
227 self.__polynomial_fit_model(image=self.tmpmskname,
228 model=self.tmppolyname, axis=fitaxis, order=self.numpoly)
229 polyimage = ia.newimage(self.tmppolyname)
230 # set back defalut mask (need to get from self.image)
231 avail_mask = polyimage.maskhandler('get')
232 if is_initial_mask:
233 casalog.post("copying mask from %s" % (self.infiles))
234 polyimage.calcmask("mask('%s')" % self.infiles, asdefault=True)
235 else: # no mask in the original image
236 polyimage.calcmask('T', asdefault=True)
237 if temp_maskname in avail_mask:
238 polyimage.maskhandler('delete', name=temp_maskname)
240 # subtract fitted image from original map
241 subtracted = ia.imagecalc(outfile=self.outfile,
242 pixels='"{org}" - "{fit}"'.format(org=self.infiles,
243 fit=self.tmppolyname),
244 overwrite=self.overwrite)
245 subtracted.done()
247 # finalization
248 polyimage.done(remove=True)
249 self.convimage.done(remove=True)
250 self.image.done()
252 def __polynomial_fit_model(self, image=None, model=None, axis=0, order=2):
253 if not image or not os.path.exists(image):
254 raise RuntimeError("No image found to fit.")
255 if os.path.exists(model):
256 # CAS-5410 Use private tools inside task scripts
257 ctsys.removetable([model])
258 tmpia = iatool()
259 modelimg = tmpia.newimagefromimage(infile=image, outfile=model)
260 try:
261 if tmpia.isopen():
262 tmpia.close()
263 imshape = modelimg.shape()
264 # the axis order of [ra, dec, chan(, pol)] is assumed throughout the task.
265 ndim = len(imshape)
266 nx = imshape[0]
267 ny = imshape[1]
268 # an xy-plane can be fit simultaneously (doing per plane to save memory)
269 if ndim == 3:
270 def get_blc(i, j):
271 return [0, 0, i]
273 def get_trc(i, j):
274 return [nx - 1, ny - 1, i]
276 imshape2 = imshape[2]
277 imshape3 = 1
278 elif ndim == 4:
279 def get_blc(i, j):
280 return [0, 0, i, j]
282 def get_trc(i, j):
283 return [nx - 1, ny - 1, i, j]
285 imshape2 = imshape[2]
286 imshape3 = imshape[3]
287 else: # ndim == 2
288 def get_blc(i, j):
289 return [0, 0]
291 def get_trc(i, j):
292 return [nx - 1, ny - 1]
294 imshape2 = 1
295 imshape3 = 1
297 for i3 in range(imshape3):
298 for i2 in range(imshape2):
299 blc = get_blc(i2, i3)
300 trc = get_trc(i2, i3)
301 dslice = modelimg.getchunk(blc, trc)
302 mslice = modelimg.getchunk(blc, trc, getmask=True)
303 model = self._get_polyfit_model_array(dslice.reshape(nx, ny),
304 mslice.reshape(nx, ny),
305 axis, order)
306 modelimg.putchunk(model, blc)
308 # the fit model image itself is free from invalid pixels
309 modelimg.calcmask('T', asdefault=True)
310 except Exception:
311 raise
312 finally:
313 modelimg.close()
315 def _get_polyfit_model_array(self, data, mask, axis, order):
316 if axis == 1:
317 tmp = data.transpose()
318 data = tmp
319 tmp = mask.transpose()
320 mask = tmp
321 del tmp
322 nx = data.shape[0]
323 ny = data.shape[1]
324 x = range(nx)
325 flag = mask ^ True # invert mask for masked array
326 mdata = numpy.ma.masked_array(data, flag)
327 retc = numpy.ma.polyfit(x, mdata, order)
328 del flag
329 coeffs = retc.transpose()
330 tmpmodel = numpy.zeros([nx, ny])
331 for iy in range(ny):
332 tmpmodel[:, iy] = numpy.poly1d(coeffs[iy])(x)
333 if axis == 1:
334 return tmpmodel.transpose()
335 return tmpmodel
337 def __execute_basket_weaving(self):
338 ###
339 # Basket-Weaving (Emerson & Grave 1988)
340 ###
341 casalog.post('Apply Basket-Weaving')
343 # CAS-5410 Use private tools inside task scripts
344 ia = iatool()
346 # initial setup
347 outimage = ia.newimagefromimage(
348 infile=self.infiles[0], outfile=self.outfile, overwrite=self.overwrite)
349 imshape_out = outimage.shape()
350 # ndim_out = len(imshape_out)
351 coordsys = outimage.coordsys()
352 axis_types = coordsys.axiscoordinatetypes()
353 # direction axis should always exist
354 try:
355 direction_axis0 = axis_types.index('Direction')
356 direction_axis1 = axis_types[direction_axis0 + 1:].index('Direction') \
357 + direction_axis0 + 1
358 except IndexError:
359 raise RuntimeError('Direction axes don\'t exist.')
360 finally:
361 coordsys.done()
362 nx = imshape_out[direction_axis0]
363 ny = imshape_out[direction_axis1]
364 tmp = []
365 nfile = len(self.infiles)
366 for i in range(nfile):
367 tmp.append(numpy.zeros(imshape_out, dtype=float))
368 maskedpixel = numpy.array(tmp)
369 del tmp
371 # direction
372 dirs = []
373 if len(self.direction) == nfile:
374 dirs = self.direction
375 else:
376 casalog.post('direction information is extrapolated.')
377 for i in range(nfile):
378 dirs.append(self.direction[i % len(self.direction)])
380 # maskwidth
381 masks = []
382 if isinstance(self.maskwidth, int) or isinstance(self.maskwidth, float):
383 for i in range(nfile):
384 masks.append(self.maskwidth)
385 elif isinstance(self.maskwidth, list): # and nfile != len(self.maskwidth):
386 for i in range(nfile):
387 masks.append(self.maskwidth[i % len(self.maskwidth)])
388 for i in range(len(masks)):
389 masks[i] = 0.01 * masks[i]
391 # mask
392 for i in range(nfile):
393 self.realimage = create_4d_image(self.infiles[i], self.tmprealname[i])
394 self.imagimage = self.realimage.subimage(outfile=self.tmpimagname[i])
396 # replace masked pixels with 0.0
397 if not self.realimage.getchunk(getmask=True).all():
398 casalog.post("Replacing masked pixels with 0.0 in %d-th image" % (i))
399 self.realimage.replacemaskedpixels(0.0)
400 self.realimage.close()
401 self.imagimage.close()
403 # Below working images are all 4D regardless of dimension of input images
404 # image shape for temporary images (always 4D)
405 ia.open(self.tmprealname[0])
406 imshape = ia.shape()
407 # ndim = len(imshape)
408 ia.close()
410 if len(self.thresh) == 0:
411 casalog.post('Use whole region')
412 else:
413 for i in range(nfile):
414 self.realimage = ia.newimage(self.tmprealname[i])
415 for iaxis2 in range(imshape[2]):
416 for iaxis3 in range(imshape[3]):
417 pixmsk = self.realimage.getchunk([0, 0, iaxis2, iaxis3],
418 [nx - 1, ny - 1, iaxis2, iaxis3])
419 for ix in range(pixmsk.shape[0]):
420 for iy in range(pixmsk.shape[1]):
421 if self.thresh[0] == self.nolimit:
422 if pixmsk[ix][iy] > self.thresh[1]:
423 maskedpixel[i][ix][iy][iaxis2][iaxis3] = pixmsk[ix][iy]
424 pixmsk[ix][iy] = 0.0
425 elif self.thresh[1] == self.nolimit:
426 if pixmsk[ix][iy] < self.thresh[0]:
427 maskedpixel[i][ix][iy][iaxis2][iaxis3] = pixmsk[ix][iy]
428 pixmsk[ix][iy] = 0.0
429 else:
430 if pixmsk[ix][iy] < self.thresh[0] \
431 or pixmsk[ix][iy] > self.thresh[1]:
432 maskedpixel[i][ix][iy][iaxis2][iaxis3] = pixmsk[ix][iy]
433 pixmsk[ix][iy] = 0.0
434 self.realimage.putchunk(pixmsk, [0, 0, iaxis2, iaxis3])
435 self.realimage.close()
436 maskedvalue = None
437 if any(maskedpixel.flatten() != 0.0):
438 maskedvalue = maskedpixel.mean(axis=0)
439 del maskedpixel
441 # set weight factor
442 weights = numpy.ones(shape=(nfile, nx, ny), dtype=float)
443 eps = 1.0e-5
444 dtor = numpy.pi / 180.0
445 for i in range(nfile):
446 scan_direction = ''
447 if abs(numpy.sin(dirs[i] * dtor)) < eps: # direction is around 0 deg
448 maskw = 0.5 * nx * masks[i]
449 scan_direction = 'horizontal'
450 elif abs(numpy.cos(dirs[i] * dtor)) < eps: # direction is around 90 deg
451 maskw = 0.5 * ny * masks[i]
452 scan_direction = 'vertical'
453 else:
454 maskw = 0.5 * numpy.sqrt(nx * ny) * masks[i]
455 for ix in range(nx):
456 halfwx = (nx - 1) // 2
457 for iy in range(ny):
458 halfwy = (ny - 1) // 2
459 if scan_direction == 'horizontal':
460 # dd = abs(float(ix) - 0.5*(nx-1))
461 dd = abs(float(ix) - halfwx) # for CAS-9434
462 elif scan_direction == 'vertical':
463 # dd = abs(float(iy) - 0.5*(ny-1))
464 dd = abs(float(iy) - halfwy) # for CAS-9434
465 else:
466 tand = numpy.tan((dirs[i] - 90.0) * dtor)
467 # dd = abs((float(ix) - 0.5*(nx-1)) * tand - (float(iy) - 0.5*(ny-1)))
468 dd = abs((float(ix) - halfwx) * tand - (float(iy) - halfwy)) # for CAS-9434
469 dd = dd / numpy.sqrt(1.0 + tand * tand)
470 if dd < maskw:
471 cosd = numpy.cos(0.5 * numpy.pi * dd / maskw)
472 weights[i][ix][iy] = 1.0 - cosd * cosd
473 if weights[i][ix][iy] == 0.0:
474 weights[i][ix][iy] += eps * 0.01
475 """
476 if abs(numpy.sin(dirs[i]*dtor)) < eps:
477 # direction is around 0 deg
478 maskw = 0.5 * nx * masks[i]
479 for ix in range(nx):
480 for iy in range(ny):
481 dd = abs( float(ix) - 0.5 * (nx-1) )
482 if dd < maskw:
483 cosd = numpy.cos(0.5*numpy.pi*dd/maskw)
484 weights[i][ix][iy] = 1.0 - cosd * cosd
485 if weights[i][ix][iy] == 0.0:
486 weights[i][ix][iy] += eps*0.01
487 elif abs(numpy.cos(dirs[i]*dtor)) < eps:
488 # direction is around 90 deg
489 maskw = 0.5 * ny * masks[i]
490 for ix in range(nx):
491 for iy in range(ny):
492 dd = abs( float(iy) - 0.5 * (ny-1) )
493 if dd < maskw:
494 cosd = numpy.cos(0.5*numpy.pi*dd/maskw)
495 weights[i][ix][iy] = 1.0 - cosd * cosd
496 if weights[i][ix][iy] == 0.0:
497 weights[i][ix][iy] += eps*0.01
498 else:
499 maskw = 0.5 * numpy.sqrt( nx * ny ) * masks[i]
500 for ix in range(nx):
501 for iy in range(ny):
502 tand = numpy.tan((dirs[i]-90.0)*dtor)
503 dd = abs( ix * tand - iy - 0.5 * (nx-1) * tand + 0.5 * (ny-1) )
504 dd = dd / numpy.sqrt( 1.0 + tand * tand )
505 if dd < maskw:
506 cosd = numpy.cos(0.5*numpy.pi*dd/maskw)
507 weights[i][ix][iy] = 1.0 - cosd * cosd
508 if weights[i][ix][iy] == 0.0:
509 weights[i][ix][iy] += eps*0.01
510 """
511 # shift
512 xshift = -((ny - 1) // 2)
513 yshift = -((nx - 1) // 2)
514 for ix in range(int(xshift), 0, 1):
515 tmp = weights[i, :, 0].copy()
516 weights[i, :, 0:ny - 1] = weights[i, :, 1:ny].copy()
517 weights[i, :, ny - 1] = tmp
518 for iy in range(int(yshift), 0, 1):
519 tmp = weights[i, 0:1].copy()
520 weights[i, 0:nx - 1] = weights[i, 1:nx].copy()
521 weights[i, nx - 1:nx] = tmp
523 # FFT
524 for i in range(nfile):
525 self.realimage = ia.newimage(self.tmprealname[i])
526 self.imagimage = ia.newimage(self.tmpimagname[i])
527 for iaxis2 in range(imshape[2]):
528 for iaxis3 in range(imshape[3]):
529 pixval = self.realimage.getchunk([0, 0, iaxis2, iaxis3],
530 [nx - 1, ny - 1, iaxis2, iaxis3])
531 pixval = pixval.reshape((nx, ny))
532 pixfft = npfft.fft2(pixval)
533 pixfft = pixfft.reshape((nx, ny, 1, 1))
534 self.realimage.putchunk(pixfft.real, [0, 0, iaxis2, iaxis3])
535 self.imagimage.putchunk(pixfft.imag, [0, 0, iaxis2, iaxis3])
536 del pixval, pixfft
537 self.realimage.close()
538 self.imagimage.close()
540 # weighted mean
541 for ichan in range(imshape[2]):
542 for iaxis3 in range(imshape[3]):
543 pixout = numpy.zeros(shape=(nx, ny), dtype=complex)
544 denom = numpy.zeros(shape=(nx, ny), dtype=float)
545 for i in range(nfile):
546 self.realimage = ia.newimage(self.tmprealname[i])
547 self.imagimage = ia.newimage(self.tmpimagname[i])
548 pixval = self.realimage.getchunk(
549 [0, 0, ichan, iaxis3], [nx - 1, ny - 1, ichan, iaxis3]) \
550 + self.imagimage.getchunk([0, 0, ichan, iaxis3],
551 [nx - 1, ny - 1, ichan, iaxis3]) * 1.0j
552 pixval = pixval.reshape((nx, ny))
553 pixout = pixout + pixval * weights[i]
554 denom = denom + weights[i]
555 self.realimage.close()
556 self.imagimage.close()
557 pixout = pixout / denom
558 pixout = pixout.reshape((nx, ny, 1, 1))
559 self.realimage = ia.newimage(self.tmprealname[0])
560 self.imagimage = ia.newimage(self.tmpimagname[0])
561 self.realimage.putchunk(pixout.real, [0, 0, ichan, iaxis3])
562 self.imagimage.putchunk(pixout.imag, [0, 0, ichan, iaxis3])
563 self.realimage.close()
564 self.imagimage.close()
566 # inverse FFT
567 self.realimage = ia.newimage(self.tmprealname[0])
568 self.imagimage = ia.newimage(self.tmpimagname[0])
569 for ichan in range(imshape[2]):
570 for iaxis3 in range(imshape[3]):
571 pixval = self.realimage.getchunk([0, 0, ichan, iaxis3],
572 [nx - 1, ny - 1, ichan, iaxis3]) \
573 + self.imagimage.getchunk([0, 0, ichan, iaxis3],
574 [nx - 1, ny - 1, ichan, iaxis3]) * 1.0j
575 pixval = pixval.reshape((nx, ny))
576 pixifft = npfft.ifft2(pixval)
577 pixifft = pixifft.reshape((nx, ny, 1, 1))
578 self.realimage.putchunk(pixifft.real, blc=[0, 0, ichan, iaxis3])
579 del pixval, pixifft
580 if maskedvalue is not None:
581 self.realimage.putchunk(self.realimage.getchunk() + maskedvalue)
583 # put result into outimage
584 chunk = self.realimage.getchunk()
585 outimage.putchunk(chunk.reshape(imshape_out))
586 # handling of output image mask
587 maskstr = ""
588 for name in self.infiles:
589 if len(maskstr) > 0:
590 maskstr += " || "
591 maskstr += ("mask('%s')" % (name))
592 outimage.calcmask(maskstr, name="basketweaving", asdefault=True)
594 self.realimage.close()
595 self.imagimage.close()
596 outimage.close()
598 def finalize(self):
599 pass
601 def cleanup(self):
602 # finalize image analysis tool
603 if hasattr(self, 'image') and self.image is not None:
604 if self.image.isopen():
605 self.image.done()
606 tools = ['convimage', 'imageorg', 'realimage', 'imagimage']
607 for t in tools:
608 if hasattr(self, t):
609 v = getattr(self, t)
610 if v and v.isopen():
611 v.done(remove=True)
613 # remove tempolary files
614 filelist = ['tmpmskname', 'tmpconvname', 'tmppolyname',
615 'tmprealname', 'tmpimagname']
616 existing_files = []
617 for s in filelist:
618 if hasattr(self, s):
619 f = getattr(self, s)
620 if isinstance(f, list):
621 for g in f:
622 if os.path.exists(g):
623 existing_files.append(g)
624 else:
625 if os.path.exists(f):
626 existing_files.append(f)
627 # CAS-5410 Use private tools inside task scripts
628 if len(existing_files) > 0:
629 ctsys.removetable(existing_files)