Add baked-in favicons for default popular sites on NTP

components/ntp_tiles/update_default_sites_resources.py

Index: components/ntp_tiles/update_default_sites_resources.py
diff --git a/components/ntp_tiles/update_default_sites_resources.py b/components/ntp_tiles/update_default_sites_resources.py
new file mode 100755
index 0000000000000000000000000000000000000000..a3ad1bfa5bd612b9504c4ca708a545486c53e39c
--- /dev/null
+++ b/components/ntp_tiles/update_default_sites_resources.py
@@ -0,0 +1,134 @@
+#!/usr/bin/env python2
+# Copyright 2017 The Chromium Authors. All rights reserved.
+# Use of this source code is governed by a BSD-style license that can be
+# found in the LICENSE file.
+
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+
+import argparse
+import contextlib
+import glob
+import io
+import json
+import os
+import sys
+import urllib
+import Image
+
+# This script downloads the default popular sites and large icons associated
+# with it. If an icon is too large, it will get resized in the process.

[sfiera, 2017/02/16 18:54:16]

This here would be a good use of a file-level """ comment.

[fhorschig, 2017/02/17 16:24:04]

Okay.

+
+DEFAULT_POPULAR_SITES = ("https://www.gstatic.com/chrome/ntp/"
+                         "suggested_sites_DEFAULT_5.json")
+LARGE_ICON_KEY = "large_icon_url"
+SITE_TITLE_KEY = "title"
+MAXIMAL_SIZE = 144
+SITE_ICON_DELETE = "icon[0-9].png"
+SITE_ICON_FORMAT = "icon%d.png"
+NTP_TILES_RESOURCE_PATH = os.path.join(
+    os.path.dirname(os.path.realpath(__file__)), "resources")
+DEFAULT_POPULAR_SITES_PATH = os.path.join(NTP_TILES_RESOURCE_PATH,
+                                          "default_popular_sites.json")
+
+def download_popular_sites(url):
+  print("Downloading popular sites... (" + url + ")")
+  with contextlib.closing(urllib.urlopen(url=url)) as url_data:
+    data = json.load(url_data)
+  print("... done. (%d sites found)" % len(data))
+  return data
+
+def write_to_json(data, out_path, pretty_print):
+  separators = (",", ":")
+  indent = None
+  sort = False
+  if pretty_print:
+    separators = (",", ": ")
+    indent = 4
+    sort = True
+  with open(out_path, "w") as f:
+    json.dump(data, f, sort_keys=sort, indent=indent, separators=separators)
+  print("JSON was written to " + out_path)
+
+def delete_old_icons():
+  print("Deleting old icons..")
+  for f in glob.glob(os.path.join(NTP_TILES_RESOURCE_PATH, SITE_ICON_DELETE)):
+    os.remove(os.path.join(f))
+  print("... done.")
+
+def download_image_for_popular_site(site):
+  with contextlib.closing(urllib.urlopen(url=site[LARGE_ICON_KEY])) as url_data:
+    image_response_data = url_data.read()
+  return Image.open(io.BytesIO(image_response_data))

[sfiera, 2017/02/16 18:54:16]

I'm still concerned that you might be un-optimizing the compression here.

[fhorschig, 2017/02/17 16:24:04]

As discussed, I noew introduced the highest compression level of PIL.
(Used to default to 6, now it's 9)

I will have a look at PNGOUT as the current downscaling seems to have
disadvantages for logos like Wikipedia's.

+
+def resize_if_too_large(image, max_size):
+  if image.size[0] > max_size:  # Width is sufficient to compare squared icons.
+    print("... and resizing image from %s to %s ..." %
+          (image.size, (max_size, max_size)));
+    image.thumbnail((max_size, max_size), Image.ANTIALIAS)
+
+
+def lacks_required_keys(site):
+  return not SITE_TITLE_KEY in site or not LARGE_ICON_KEY in site
+
+
+def main():
+  parser = argparse.ArgumentParser(
+      description="Downloads the latest popular sites and their icons. \n\n"
+          "It is possible to customize the default like this:\n"
+          "  1. python " + __file__ + " -o temp.json --skip_icons "
+          "--pretty_print\n"
+          "  2. Adjust the downloaded temp.json\n"
+          "  3. python " + __file__ + " -f temp.json -s 96\n\n"
+          "The result would be a minified version of your customized JSON "
+          "and all icons would be downloaded as you specified.\n The icons "
+          "had a max size of 96x96.",
+          formatter_class=argparse.RawTextHelpFormatter)
+  parser.add_argument("-s", "--size", metavar="size_in_px", type=int,
+                      default=MAXIMAL_SIZE,
+                      help="size to scale too large icons down to; defaults "
+                           "to 144px")
+  parser.add_argument("-u", "--url", metavar="url", type=str,

[sfiera, 2017/02/16 18:54:16]

If you have a long flag name, the metavar should default to it.

[fhorschig, 2017/02/17 16:24:04]

Yes but it's CAPS. But in this case it's maybe more appropriate.

+                      default=DEFAULT_POPULAR_SITES,
+                      help="the endpoint to query for json of sites")
+  parser.add_argument("--skip_icons", action="store_true",

[sfiera, 2017/02/16 18:54:16]

In Chromium, I think we typically use --skip-icons, not --skip_icons style flag
naming. (still accessed as args.skip_icons later)

I would also say "--no-icons" "--no-resizing". I think the --no pattern is more
common, at least at Google.

[fhorschig, 2017/02/17 16:24:04]

Done.

+                      help="do not download icons")
+  parser.add_argument("--skip_resizing", action="store_true",
+                      help="do not resize any icons")
+  parser.add_argument("-f", "--in_file", metavar="path_to_json_file",
+                      type=str,

[sfiera, 2017/02/16 18:54:16]

See also argparse.FileType.

[fhorschig, 2017/02/17 16:24:04]

There is little advantage using that as I would need to 
distinguish source (file or urllib response) later.

+                      help="skip download and load icons for a local json")
+  parser.add_argument("-o", "--out_path", metavar="path_to_out_file",
+                      type=str, default=DEFAULT_POPULAR_SITES_PATH,
+                      help="skip download and load icons for a local json")
+  parser.add_argument("-p", "--pretty_print", action="store_true",
+                      help="pretty_print instead of minifying the JSON")
+  args = parser.parse_args()
+
+  if args.in_file:
+    with open(args.in_file) as f:
+      popular_sites = json.load(f)
+  else:
+    popular_sites = download_popular_sites(args.url)
+  write_to_json(popular_sites, args.out_path, args.pretty_print)
+
+  if args.skip_icons:
+    return
+
+  delete_old_icons()
+  for i, site in enumerate(popular_sites):
+    if lacks_required_keys(site):
+      print("Could not download large image for site: %r" % site)
+      continue
+    print("Downloading icon for '%r'..." % site[SITE_TITLE_KEY])
+    image = download_image_for_popular_site(site)
+    if not args.skip_resizing:
+      resize_if_too_large(image, args.size)
+    image_name = SITE_ICON_FORMAT % i
+    image.save(os.path.join(NTP_TILES_RESOURCE_PATH, image_name), "PNG")
+    print("... done. (Stored as " + image_name + ")");
+
+
+if __name__ == "__main__":
+  main()