Enhanced the photo collection script to look in multiple locations, in order to capture missing IAB and IESG member photos, and to select full and thumbnail photos from a preference list based on timestamp and size.

 - Legacy-Id: 11270

Author: levkowetz

ietf/bin/2016-05-25-collect-photos

@@ -1,6 +1,8 @@
 #!/usr/bin/env python
 
 import os, sys, shutil, pathlib
+from collections import namedtuple
+from PIL import Image
 
 # boilerplate
 basedir = os.path.abspath(os.path.join(os.path.dirname(__file__), "../../"))
@@ -10,35 +12,37 @@ os.environ.setdefault("DJANGO_SETTINGS_MODULE", "ietf.settings")
 import django
 django.setup()
 
+from django.conf import settings
 import debug
 
 from ietf.group.models import Role, Person
 
-old_images_dir = django.conf.settings.OLD_PHOTOS_DIR
-new_images_dir = django.conf.settings.PHOTOS_DIR
 
-if not os.path.exists(old_images_dir):
-    print("Old images directory does not exist: %s" % old_images_dir)
-    sys.exit(1)
+
+
+old_images_dir = ''
+new_images_dir = settings.PHOTOS_DIR
+
 if not os.path.exists(new_images_dir):
     print("New images directory does not exist: %s" % new_images_dir)
     sys.exit(1)
 
 old_image_files = []
+for dir in settings.OLD_PHOTO_DIRS:
+    if not os.path.exists(dir):
+        print("Old images directory does not exist: %s" % dir)
+        sys.exit(1)
+    old_image_files += [ f for f in pathlib.Path(dir).iterdir() if f.is_file() and f.suffix.lower() in ['.jpg', '.jpeg', '.png'] ]
 
+photo = namedtuple('photo', ['path', 'name', 'ext', 'width', 'height', 'time', 'file'])
 
-for (dirpath, dirnames, filenames) in os.walk(old_images_dir):
-    if len(filenames) == 0:
-        print("No image files found in %s" % old_images_dir)
-        sys.exit(2)
-    old_image_files.extend(filenames)
-    break # Only interested in the files in the top directory
+old_images = []
+for f in old_image_files:
+    path = str(f)
+    img = Image.open(path)
+    old_images.append(photo(path, f.stem.decode('utf8'), f.suffix, img.size[0], img.size[1], f.stat().st_mtime, f))
 
-old_image_files = [ f.name.decode('utf8') for f in pathlib.Path(old_images_dir).iterdir() if f.is_file() and not f.suffix.lower() in ['.lck'] ]
-
-interesting_persons = set()
-
-interesting_persons.update(list(Person.objects.all()))
+interesting_persons = set(Person.objects.all())
 
 name_alias = {
     "andy":     ["andrew", ],
@@ -121,62 +125,63 @@ for person in sorted(list(interesting_persons),key=lambda x:x.last_name()+x.asci
     if not substr_pattern:
         substr_pattern = u'-'.join(name_parts[-1:]+name_parts[0:1])
 
-    candidates = [x for x in old_image_files if x.lower().startswith(substr_pattern)]
+    candidates = [x for x in old_images if x.name.lower().startswith(substr_pattern)]
+    # Also check the reverse the name order (necessary for Deng Hui, for instance)
+    substr_pattern = u'-'.join(name_parts[0:1]+name_parts[-1:])
+    candidates += [x for x in old_images if x.name.lower().startswith(substr_pattern)]
+    if candidates:
+        print("   Used '%s %s' instead of '%s %s'" % (name_parts[-1], name_parts[0], name_parts[0], name_parts[-1], ))
     # If no joy, try a short name
-    if not candidates and name_parts[0] in name_alias:
+    if name_parts[0] in name_alias:
         for alias in name_alias[name_parts[0]]:
             substr_pattern = u'-'.join(name_parts[-1:]+[alias])
-            candidates += [x for x in old_image_files if x.lower().startswith(substr_pattern)]
+            candidates += [x for x in old_images if x.name.lower().startswith(substr_pattern)]
         if candidates:
             print("   Used '%s %s' instead of '%s %s'" % (alias, name_parts[-1], name_parts[0], name_parts[-1], ))
-    # If still no joy, reverse the name order (necessary for Deng Hui, for instance)
-    if not candidates:
-        substr_pattern = u'-'.join(name_parts[0:1]+name_parts[-1:])
-        candidates = [x for x in old_image_files if x.lower().startswith(substr_pattern)]
-        if candidates:
-            print("   Used '%s %s' instead of '%s %s'" % (name_parts[-1], name_parts[0], name_parts[0], name_parts[-1], ))
     # If still no joy, try with Person.plain_name() (necessary for Donald Eastlake)
     if not candidates:
         name_parts = person.plain_name().lower().split()
         substr_pattern = u'-'.join(name_parts[-1:]+name_parts[0:1])
-        candidates = [x for x in old_image_files if x.lower().startswith(substr_pattern)]
+        candidates = [x for x in old_images if x.name.lower().startswith(substr_pattern)]
         # If no joy, try a short name
         if not candidates and name_parts[0] in name_alias:
             for alias in name_alias[name_parts[0]]:
                 substr_pattern = u'-'.join(name_parts[-1:]+[alias])
-                candidates += [x for x in old_image_files if x.lower().startswith(substr_pattern)]
+                candidates += [x for x in old_images if x.name.lower().startswith(substr_pattern)]
             if candidates:
                 print("   Used '%s %s' instead of '%s %s'" % (alias, name_parts[-1], name_parts[0], name_parts[-1], ))
 
-    # Fixup for other exceptional cases
-
-    if person.ascii=="David Oran":
-        candidates = ['oran-dave-th.jpg','oran-david.jpg']
-
-    if person.ascii=="Susan Hares":
-        candidates = ['hares-sue-th.jpg','hares-susan.JPG']
-
-    if person.ascii=="Mahesh Jethanandani":
-        candidates = ['Mahesh-Jethanandani-th.jpg','Jethanandani-Mahesh.jpg']
-
-    processed_files += [ c for c in candidates ]
-
-    if len(candidates) not in [0,1,2]:
-        candidates = [x for x in candidates if not '00' in x]
-
-    if len(candidates) == 1:
-        candidates = candidates + candidates
-
-    if len(candidates) not in [0,2]:
-        thumb = [ c for c in candidates if '-th.' in c ][0]
-        photo = [ c for c in candidates if '-th.' not in c ][0]
-        trunc = [thumb, photo]
-        print("   Truncating %s to %s" % (candidates, trunc))
-        candidates = trunc
-        
-    if candidates and '-th' in candidates[1]:
-        candidates.reverse()
-        
+#     # Fixup for other exceptional cases
+#     if person.ascii=="David Oran":
+#         candidates = ['oran-dave-th.jpg','oran-david.jpg']
+# 
+#     if person.ascii=="Susan Hares":
+#         candidates = ['hares-sue-th.jpg','hares-susan.JPG']
+# 
+#     if person.ascii=="Mahesh Jethanandani":
+#         candidates = ['Mahesh-Jethanandani-th.jpg','Jethanandani-Mahesh.jpg']
+
+    processed_files += [ c.path for c in candidates ]
+
+    # We now have a list of candidate photos.
+    # * Consider anything less than 200x200 a thumbnail
+    # * For the full photo, sort by size (width) and time
+    # * For the thumbnail:
+    #    - first look for a square photo less than 200x200
+    #    - if none found, then for the first in the sorted list less than 200x200
+    #    - if none found, then the smallest photo
+    if candidates:
+        candidates.sort(key=lambda x: "%04d-%d" % (x.width, x.time))
+        full = candidates[-1]
+        thumbs = [ c for c in candidates if c.width==c.height and c.width <= 200 ]
+        if not thumbs:
+            thumbs = [ c for c in candidates if c.width==c.height ]
+        if not thumbs:
+            thumbs = [ c for c in candidates if c.width <= 200 ]
+        if not thumbs:
+            thumbs = candidates[:1]
+        thumb = thumbs[-1]
+        candidates = [ thumb, full ]
 
     # At this point we either have no candidates or two. If two, the first will be the thumb
 
@@ -186,33 +191,31 @@ for person in sorted(list(interesting_persons),key=lambda x:x.last_name()+x.asci
             shutil.copy(old, new)
             shutil.copystat(old, new)
 
+    assert(len(candidates) in [0,2])
     if len(candidates)==2:
-        old_name = candidates[1]
-        old_thumb_name = candidates[0]
-        old_name_ext = os.path.splitext(old_name)[1]
-        old_thumb_name_ext = os.path.splitext(old_thumb_name)[1]
+        thumb, full = candidates
 
-        new_name = person.photo_name(thumb=False)+old_name_ext.lower()
-        new_thumb_name = person.photo_name(thumb=True)+old_thumb_name_ext.lower()
+        new_name = person.photo_name(thumb=False)+full.ext.lower()
+        new_thumb_name = person.photo_name(thumb=True)+thumb.ext.lower()
 
-        copy( os.path.join(old_images_dir,old_name), os.path.join(new_images_dir,new_name) )
+        copy( full.path, os.path.join(new_images_dir,new_name) )
 
         #
-        copy( os.path.join(old_images_dir,old_thumb_name),  os.path.join(new_images_dir,new_thumb_name) )
+        copy( thumb.path,  os.path.join(new_images_dir,new_thumb_name) )
 
 print("")
 not_processed = 0
-for file in pathlib.Path(old_images_dir).iterdir():
+for file in old_image_files:
     if ( file.is_file()
         and not file.suffix.lower() in ['.txt', '.lck', '.html',]
         and not file.name.startswith('index.')
         and not file.name.startswith('milestoneupdate')
         and not file.name.startswith('nopicture')
         and not file.name.startswith('robots.txt')
     ):
-        if not file.name.decode('utf8') in processed_files:
+        if not str(file).decode('utf8') in processed_files:
             not_processed += 1
-            print(u"Not processed: "+file.name.decode('utf8'))
+            print(u"Not processed: "+str(file).decode('utf8'))
 print("")
 print("")
 print("Not processed: %s files" % not_processed)

ietf/settings.py

@@ -105,6 +105,14 @@
 PHOTOS_DIRNAME = 'photo'
 PHOTOS_DIR = MEDIA_ROOT + PHOTOS_DIRNAME
 
+OLD_PHOTO_DIRS = [
+    '/a/www/www6/wg/images',
+    '/a/www/www6/iesg/bio/photo',
+    '/a/www/iab/wp-content/IAB-uploads/2015/03/',
+    '/a/www/iab/wp-content/IAB-uploads/2015/08/',
+    '/a/www/iab/wp-content/IAB-uploads/2016/03/',
+]
+
 IETF_HOST_URL = 'https://www.ietf.org/'
 IETF_ID_URL = IETF_HOST_URL + 'id/'
 IETF_ID_ARCHIVE_URL = IETF_HOST_URL + 'archive/id/'