2to3: Port to Python 3.x.

baselines/grabCutCropEval.py

@@ -11,7 +11,7 @@ def cropMask(imageDir,saveDir,imagePath, border):
     if mask is None or mask.shape[0]==0:
         img = cv2.imread(imageDir+imagePath)
         if img is None:
-            print "Could not read: "+imageDir+imagePath
+            print("Could not read: "+imageDir+imagePath)
         mask = np.zeros(img.shape[:2],np.uint8)
         bgdModel = np.zeros((1,65),np.float64)
         fgdModel = np.zeros((1,65),np.float64)
@@ -36,8 +36,8 @@ def cropMask(imageDir,saveDir,imagePath, border):
     return  (labels!=maxLabel).astype('uint8')
 
 if len(sys.argv)<5:
-    print 'This evalutes using grab cut'
-    print 'usage: '+sys.argv[0]+' gtFile.csv imageDir saveIntermDir numThreads' # [reverse]'
+    print('This evalutes using grab cut')
+    print('usage: '+sys.argv[0]+' gtFile.csv imageDir saveIntermDir numThreads') # [reverse]'
     exit(0)
 
 gtFile=sys.argv[1]
@@ -57,7 +57,7 @@ countIOU=0
 
 scale=1
 
-print 'eval on '+gtFile
+print('eval on '+gtFile)
 
 outFile = gtFile+'_fullgrab.res'
 #numLines=0
@@ -115,8 +115,8 @@ def worker(line):
             cv2.imwrite(saveDir+"ERR_"+imagePath.replace('/','_')+'.png',mask*255);
         return imagePath, float(intersection)/union
     except:
-        print 'Error: '
-        print sys.exc_info()
+        print('Error: ')
+        print(sys.exc_info())
 
 with open(gtFile) as f:
     pool = ThreadPool(numThreads)
@@ -129,6 +129,6 @@ with open(gtFile) as f:
 
 out.write('mean IOU for '+gtFile+': '+str(sumIOU/countIOU)+'\n')
 out.close()
-print 'mean IOU for '+gtFile+': '+str(sumIOU/countIOU)
+print('mean IOU for '+gtFile+': '+str(sumIOU/countIOU))
 
 

baselines/noCutMeanCutCropEval.py

@@ -21,9 +21,9 @@ def cropMaskPortion(imgH,imgW,xp1,yp1,xp2,yp2,xp3,yp3,xp4,yp4):
     return  mask
 
 if len(sys.argv)<3:
-    print 'Computes no-crop and mean-crop baselines, generates two files {gtFile}_fullno.res and {gtFile}_fullmean.res'
-    print 'usage: '+sys.argv[0]+' gtFile.csv imageDir [mean_x1 mean_y1 mean_x2 mean_y2 mean_x3 mean_y3 mean_x4 mean_y4]'
-    print '                                           (optional mean box)'
+    print('Computes no-crop and mean-crop baselines, generates two files {gtFile}_fullno.res and {gtFile}_fullmean.res')
+    print('usage: '+sys.argv[0]+' gtFile.csv imageDir [mean_x1 mean_y1 mean_x2 mean_y2 mean_x3 mean_y3 mean_x4 mean_y4]')
+    print('                                           (optional mean box)')
     exit()
 
 gtFile=sys.argv[1]
@@ -38,7 +38,7 @@ countIOU=0
 
 scale=1
 
-print 'eval on '+gtFile
+print('eval on '+gtFile)
 
 outFile_no = gtFile+'_fullno.res'
 outFile_mean = gtFile+'_fullmean.res'
@@ -120,7 +120,7 @@ with open(gtFile) as f:
         xp4=float(sys.argv[9])
         yp4=float(sys.argv[10])
     else:
-        print str(xp1)+' '+str(yp1)+' '+str(xp2)+' '+str(yp2)+' '+str(xp3)+' '+str(yp3)+' '+str(xp4)+' '+str(yp4)
+        print(str(xp1)+' '+str(yp1)+' '+str(xp2)+' '+str(yp2)+' '+str(xp3)+' '+str(yp3)+' '+str(xp4)+' '+str(yp4))
 
     cc=0
     for line in lines:

pageCropAnnotator.py

@@ -6,7 +6,7 @@ import re
 import xml.etree.ElementTree as ET
 import os
 import sys
-from StringIO import StringIO
+from io import StringIO
 import cv2
 
 
@@ -156,7 +156,7 @@ def clicker(event, x, y, flags, param):
 
 def segmenter(imDir,imagePath,dispHeight):
     global image,tl,tr,bl,br,tm,bm,lastDidList,orig,abnorm
-    print 'opening '+imDir+imagePath
+    print('opening '+imDir+imagePath)
     orig = cv2.imread(imDir+imagePath)
     scale = orig.shape[0]/dispHeight
     orig = cv2.resize(orig,(0,0),None,1.0/scale,1.0/scale)
@@ -234,8 +234,8 @@ def segmenter(imDir,imagePath,dispHeight):
     #return newWords, newWordBoxes
 
 if len(sys.argv)<4:
-    print 'usage: '+sys.argv[0]+' imgDir imgList outAnn.csv [displayHeight]'
-    print 'output format: imageFile, tlx, tly, trx, try, brx, bry, blx, bly, type (,tmx, tmy, bmx, bmy)'
+    print('usage: '+sys.argv[0]+' imgDir imgList outAnn.csv [displayHeight]')
+    print('output format: imageFile, tlx, tly, trx, try, brx, bry, blx, bly, type (,tmx, tmy, bmx, bmy)')
     exit(0)
 
 inFile = sys.argv[2]
@@ -257,34 +257,34 @@ try:
     did = check.read().splitlines()
     didCount=len(did)
     check.close()
-    print 'found '+outFile+', appending. Note: this is sychronizing based on count alone, if '+inFile+' hash changed, but sure to align '+outFile
+    print('found '+outFile+', appending. Note: this is sychronizing based on count alone, if '+inFile+' hash changed, but sure to align '+outFile)
 except IOError:
-    print ('making new out:'+outFile)
+    print(('making new out:'+outFile))
 
 out = open(outFile,'w')
 
-print ' =============================================== '
-print ' !!! INSTRUCTIONS !!!'
-print ' If the page does not contain a single page, or  '
-print ' an open book, mark as abnormal with INSERT (e.g.'
-print ' two seperate pages).'
-print ' Click on the four corners to include all the    '
-print ' full pages in the image (including two pages if '
-print ' fully present).'
-print ' If two pages a full present also mark page seam '
-print ' (middle-click).'
-print ' On placing points, prioritize the following to  '
-print ' be included/discluded from the polygons in the  '
-print ' following order:'
-print '   1. Including the present page(s) content.'
-print '   2. Discluding other pages and background.'
-print '   3. Discluding the present page(s) boudary.'
-print '   4. Including the present page(s) white area.'
+print(' =============================================== ')
+print(' !!! INSTRUCTIONS !!!')
+print(' If the page does not contain a single page, or  ')
+print(' an open book, mark as abnormal with INSERT (e.g.')
+print(' two seperate pages).')
+print(' Click on the four corners to include all the    ')
+print(' full pages in the image (including two pages if ')
+print(' fully present).')
+print(' If two pages a full present also mark page seam ')
+print(' (middle-click).')
+print(' On placing points, prioritize the following to  ')
+print(' be included/discluded from the polygons in the  ')
+print(' following order:')
+print('   1. Including the present page(s) content.')
+print('   2. Discluding other pages and background.')
+print('   3. Discluding the present page(s) boudary.')
+print('   4. Including the present page(s) white area.')
 #print ' book). If a corner is torn, click where it ought'
 #print ' to be, based on page edges. The page seem on an '
 #print ' open book is the page edge.'
-print ' Use ESC to exit or the latest page you finished '
-print ' will be lost.'
+print(' Use ESC to exit or the latest page you finished ')
+print(' will be lost.')
 
 #i=didCount
 i=0
@@ -299,7 +299,7 @@ doneOne=False
 while i<len(images) and not end:
     if i%10==9:
         showControls()
-    print(str(i+1)+' of '+str(len(images)))
+    print((str(i+1)+' of '+str(len(images))))
 
     if len(did)>i:
         line = did[i].strip().split(',')
@@ -322,7 +322,7 @@ while i<len(images) and not end:
         
         if undo and i>0 and doneOne:
             prevSeg=''
-            print(str(i)+' of '+str(len(images)))
+            print((str(i)+' of '+str(len(images))))
             prevSeg, undo, end = segmenter(imDir, images[i-1],dispHeight)
         else:
             out.write(prevSeg)

process_pixel_labels.py

@@ -130,7 +130,7 @@ def fill_holes(img):
 
     # Floodfill from point (0, 0)
     if img[0,0] != 0:
-        print "WARNING: Filling something you shouldn't"
+        print("WARNING: Filling something you shouldn't")
     cv2.floodFill(im_floodfill, mask, (0,0), 255);
 
     # Invert floodfilled image

render_quads.py

@@ -5,7 +5,7 @@ import cv2
 import os
 
 if len(sys.argv) < 4:
-	print "python %s manifest.txt dataset_dir out_dir" % __file__
+	print("python %s manifest.txt dataset_dir out_dir" % __file__)
 	exit()
 
 manifest_file = sys.argv[1]
@@ -17,11 +17,11 @@ try:
 except:
 	pass
 
-file_list = map(lambda s: s.strip(), open(manifest_file, 'r').readlines())
+file_list = [s.strip() for s in open(manifest_file, 'r').readlines()]
 for line in file_list:
 	tokens = line.split(',')
 	f = tokens[0]
-	coords = map(float, tokens[1:9])
+	coords = list(map(float, tokens[1:9]))
 
 	resolved = os.path.join(dataset_dir, f)
 	im = cv2.imread(resolved, 0)

test.py

@@ -34,11 +34,11 @@ def main(args):
 	net = caffe.Net(args.net_file, args.weight_file, caffe.TEST)
 	presolve(net, args)
 
-	file_list = map(lambda s: s.strip(), open(args.test_manifest, 'r').readlines())
+	file_list = [s.strip() for s in open(args.test_manifest, 'r').readlines()]
 	fd = open(args.out_file, 'w')
 	for idx, line in enumerate(file_list):
 		if idx % args.print_count == 0:
-			print "Processed %d/%d Images" % (idx, len(file_list))
+			print("Processed %d/%d Images" % (idx, len(file_list)))
 		tokens = line.split(',')
 		f = tokens[0]
 		resolved = os.path.join(args.dataset_dir, f)
@@ -94,7 +94,7 @@ def get_args():
 				help="Print interval")
 
 	args = parser.parse_args()
-	print args
+	print(args)
 
 	return args
 			

test_pretrained.py

@@ -45,11 +45,11 @@ def main(args):
 	net = caffe.Net(NET_FILE, WEIGHT_FILE, caffe.TEST)
 	presolve(net, args)
 
-	file_list = map(lambda s: s.strip(), open(args.manifest, 'r').readlines())
+	file_list = [s.strip() for s in open(args.manifest, 'r').readlines()]
 	fd = open(args.out_file, 'w')
 	for idx, line in enumerate(file_list):
 		if idx % args.print_count == 0:
-			print "Processed %d/%d Images" % (idx, len(file_list))
+			print("Processed %d/%d Images" % (idx, len(file_list)))
 		tokens = line.split(',')
 		f = tokens[0]
 		resolved = os.path.join(args.image_dir, f)
@@ -97,7 +97,7 @@ def get_args():
 				help="Print interval")
 
 	args = parser.parse_args()
-	print args
+	print(args)
 
 	return args
 			

train.py

@@ -25,7 +25,7 @@ def dump_debug(out_dir, data, dump_images=False):
 	pred_image_dir = os.path.join(out_dir, 'pred_images')
 	safe_mkdir(pred_image_dir)
 
-	for idx in xrange(len(data['images'])):
+	for idx in range(len(data['images'])):
 		fn = data['filenames'][idx]
 		preds = data['predictions'][idx] 
 
@@ -61,13 +61,13 @@ def prf(im1, im2):
 	
 
 def update_predictions(net, data, args):
-	print "Starting Predictions"
+	print("Starting Predictions")
 
 	total_iou = 0
 	total_p = 0
 	total_r = 0
 	total_f = 0
-	for idx in xrange(len(data['images'])):
+	for idx in range(len(data['images'])):
 		im = cv2.resize(data['images'][idx], (args.image_size, args.image_size))
 
 		outputs = predict(net, im, 'out', args)
@@ -84,7 +84,7 @@ def update_predictions(net, data, args):
 
 
 		if idx and idx % args.print_count == 0:
-			print "\tPredicted %d/%d" % (idx, len(data['images']))
+			print("\tPredicted %d/%d" % (idx, len(data['images'])))
 	avg_iou = total_iou / len(data['images'])
 	avg_p = total_p / len(data['images'])
 	avg_r = total_r / len(data['images'])
@@ -94,11 +94,11 @@ def update_predictions(net, data, args):
 
 def load_data(manifest, _dir, size, color=False):
 	dataset = collections.defaultdict(list)
-	file_list = map(lambda s: s.strip(), open(manifest, 'r').readlines())
+	file_list = [s.strip() for s in open(manifest, 'r').readlines()]
 	for line in file_list:
 		tokens = line.split(',')
 		f = tokens[0]
-		coords = map(float, tokens[1:9])
+		coords = list(map(float, tokens[1:9]))
 
 		dataset['filenames'].append(f)
 
@@ -120,7 +120,7 @@ def load_data(manifest, _dir, size, color=False):
 
 
 def preprocess_data(data, args):
-	for idx in xrange(len(data['images'])):
+	for idx in range(len(data['images'])):
 		im = data['images'][idx]
 		im = args.scale * (im - args.mean)
 		data['images'][idx] = im
@@ -148,7 +148,7 @@ def presolve(net, args):
 
 
 def set_input_data(net, data, args):
-	for batch_idx in xrange(args.batch_size):
+	for batch_idx in range(args.batch_size):
 		im_idx = random.randint(0, len(data['images']) - 1)
 		im = data['images'][im_idx]
 		gt = data['gt'][im_idx]
@@ -168,7 +168,7 @@ def main(args):
 	preprocess_data(train_data, args)
 	preprocess_data(val_data, args)
 
-	print "Done loading data"
+	print("Done loading data")
 
 	solver = caffe.SGDSolver(args.solver_file)
 	max_iters, snapshot_interval = get_solver_params(args.solver_file)
@@ -179,56 +179,56 @@ def main(args):
 	train_r, val_r = [], []
 	train_f, val_f = [], []
 
-	for iter_num in xrange(max_iters + 1):
+	for iter_num in range(max_iters + 1):
 		set_input_data(solver.net, train_data, args)
 		solver.step(1)
 
 		if iter_num and iter_num % snapshot_interval == 0:
-			print "Validation Prediction: %d" % iter_num
+			print("Validation Prediction: %d" % iter_num)
 			avg_iou, avg_p, avg_r, avg_f = update_predictions(solver.net, val_data, args)
 			val_iou.append((iter_num, avg_iou))
 			val_p.append((iter_num, avg_p))
 			val_r.append((iter_num, avg_r))
 			val_f.append((iter_num, avg_f))
 			if args.debug_dir:
-				print "Dumping images"
+				print("Dumping images")
 				out_dir = os.path.join(args.debug_dir, 'val_%d' % iter_num)
 				dump_debug(out_dir, val_data)
 
 		if iter_num >= args.min_interval and iter_num % args.gt_interval == 0:
 
-			print "Train Prediction: %d" % iter_num
+			print("Train Prediction: %d" % iter_num)
 			avg_iou, avg_p, avg_r, avg_f = update_predictions(solver.net, train_data, args)
 			train_iou.append((iter_num, avg_iou))
 			train_p.append((iter_num, avg_p))
 			train_r.append((iter_num, avg_r))
 			train_f.append((iter_num, avg_f))
 				
-	print "Train IOU: ", train_iou
-	print
-	print "Val IOU: ", val_iou
+	print("Train IOU: ", train_iou)
+	print()
+	print("Val IOU: ", val_iou)
 	if args.debug_dir:
-		plt.plot(*zip(*train_iou), label='train')
-		plt.plot(*zip(*val_iou), label='val')
+		plt.plot(*list(zip(*train_iou)), label='train')
+		plt.plot(*list(zip(*val_iou)), label='val')
 		plt.legend()
 		plt.savefig(os.path.join(args.debug_dir, 'iou.png'))
 
 
 		plt.clf()
-		plt.plot(*zip(*train_p), label='train')
-		plt.plot(*zip(*val_p), label='val')
+		plt.plot(*list(zip(*train_p)), label='train')
+		plt.plot(*list(zip(*val_p)), label='val')
 		plt.legend()
 		plt.savefig(os.path.join(args.debug_dir, 'precision.png'))
 
 		plt.clf()
-		plt.plot(*zip(*train_r), label='train')
-		plt.plot(*zip(*val_r), label='val')
+		plt.plot(*list(zip(*train_r)), label='train')
+		plt.plot(*list(zip(*val_r)), label='val')
 		plt.legend()
 		plt.savefig(os.path.join(args.debug_dir, 'recall.png'))
 
 		plt.clf()
-		plt.plot(*zip(*train_f), label='train')
-		plt.plot(*zip(*val_f), label='val')
+		plt.plot(*list(zip(*train_f)), label='train')
+		plt.plot(*list(zip(*val_f)), label='val')
 		plt.legend()
 		plt.savefig(os.path.join(args.debug_dir, 'fmeasure.png'))
 
@@ -287,7 +287,7 @@ def get_args():
 				help="How often to print progress")
 
 	args = parser.parse_args()
-	print args
+	print(args)
 
 	return args