Check-in [4e95040e40]
Overview
| Comment: | Move internal commentary out of plugin description block. Try/Catch processing failures in update_rows() loop. Use quieter=1 option for image downloads. Reimplement custom base_url+href favicon patching. |
|---|---|
| Downloads: | Tarball | ZIP archive | SQL archive |
| Timelines: | family | ancestors | descendants | both | trunk |
| Files: | files | file ages | folders |
| SHA1: |
4e95040e40f5f000c74a983e32184549 |
| User & Date: | mario on 2015-05-10 22:11:49 |
| Other Links: | manifest | tags |
Context
|
2015-05-10
| ||
| 22:34 | Fix URL patching (after actually testing it). And provide fallback for image resizing (.thumbnail in PIL 2.3.0 is broken). check-in: 12e7646cbf user: mario tags: trunk | |
| 22:11 | Move internal commentary out of plugin description block. Try/Catch processing failures in update_rows() loop. Use quieter=1 option for image downloads. Reimplement custom base_url+href favicon patching. check-in: 4e95040e40 user: mario tags: trunk | |
| 22:08 | Add ahttp.get( quieter= ) option for less log.HTTP notices. check-in: 529222eb9b user: mario tags: trunk | |
Changes
Modified channels/favicon.py from [608e7fc3f8] to [453360923b].
| ︙ | ︙ | |||
21 22 23 24 25 26 27 | # agreement with the station list display logic. Either uses station # row["homepage"] or row["img"] URLs from any entry. # # While it can often discover favicons directly from station homepages, # it's often speedier to use the Google PNG conversion service. Both # depend on a recent Pillow2 python module (superseding the PIL module). # Else may display images with fragments if converted from ICO files. | < < < < < < < < < < < < < < > > > > > > > > > > > > > > > > > > > > > > > > | 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 |
# agreement with the station list display logic. Either uses station
# row["homepage"] or row["img"] URLs from any entry.
#
# While it can often discover favicons directly from station homepages,
# it's often speedier to use the Google PNG conversion service. Both
# depend on a recent Pillow2 python module (superseding the PIL module).
# Else may display images with fragments if converted from ICO files.
import os, os.path
from io import BytesIO
import re
from config import *
import ahttp
from PIL import Image
from uikit import gtk
# Ensure that we don't try to download a single favicon twice per session.
# If it's not available the first time, we won't get it after switching
# stations back and forth either. So URLs are skipped simply.
tried_urls = []
# Has recently been rewritten, is somewhat less entangled with other
# modules now:
#
# · GenericChannel presets row["favicon"] with cache image filename
# in any case. It uses row["homepage"] or row["img"] as template.
#
# · The url-to-filename shortening functionality is therefore shared.
# GenericChannel.prepare() duplicates all row_to_fn() logic.
#
# · uikit.columns() merely checks row["favicon"] for file existence
# when redrawing a station list.
#
# · main.play() only calls .update_playing() or .update_all()
#
# · urllib is no longer required. Using just ahttp/requests API now.
#
# · Might need unhtml() utility from channels/__init__ later..
#
# · Still need to consolidate config options → Move main favicon
# options here?
#
# Hook up as feature plugin
#
class favicon(object):
# plugin attributes
|
| ︙ | ︙ | |||
115 116 117 118 119 120 121 |
# Cache image filename: have or can't have
favicon_fn = row_to_fn(row)
if not favicon_fn:
continue
if os.path.exists(favicon_fn):
continue
| > | | | | | | | | | | | | | | > > > > | 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 |
# Cache image filename: have or can't have
favicon_fn = row_to_fn(row)
if not favicon_fn:
continue
if os.path.exists(favicon_fn):
continue
try:
# Custom "img" banner/logo as favicon
if row.get("img"):
tried_urls.append(row["img"])
ok = banner_localcopy(row["img"], favicon_fn)
# Homepage to favicon
elif row.get("homepage"):
tried_urls.append(row["homepage"])
if conf.favicon_google_first:
ok = fav_google_ico2png(row["homepage"], favicon_fn)
else:
ok = fav_from_homepage(row["homepage"], favicon_fn)
# Update TreeView
if ok:
self.update_pixstore(row, pixstore, i)
# catch HTTP Timeouts etc., so update_all() row processing just continues..
except Exception as e:
log.WARN("favicon.update_rows():", e)
pass
# Update favicon in treeview/liststore
def update_pixstore(self, row, pixstore=None, row_i=None):
log.FAVICON_UPDATE_PIXSTORE(pixstore, row_i)
if not pixstore:
|
| ︙ | ︙ | |||
259 260 261 262 263 264 265 |
except Exception as e:
log.ERR("favicon.store_image() failure:", e)
# PNG via Google ico2png
def fav_google_ico2png(url, fn):
| < | | | | | | > > > > > > > > | < | 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 |
except Exception as e:
log.ERR("favicon.store_image() failure:", e)
# PNG via Google ico2png
def fav_google_ico2png(url, fn):
# Download from service
domain = re.sub("^\w+://|/.*$", "", url).lower()
geturl = "http://www.google.com/s2/favicons?domain={}".format(domain)
imgdata = ahttp.get(geturl, binary=1, timeout=3.5, quieter=1)
# Check for stub sizes
if conf.favicon_delete_stub and len(imgdata) in (726,896): # google_placeholder_filesizes
log.FAVICON("placeholder size, skipping")
return False
# Save
else:
return store_image(imgdata, fn)
# Peek at homepage URL, download favicon.ico <link rel>, convert to PNG file, resize to 16x16
def fav_from_homepage(url, fn):
# Check for <link rel=icon>
img = html_link_icon(url)
if not img:
return False
# Fetch image, verify MIME type
r = ahttp.get(img, binary=1, content=0, timeout=4.25, quieter=1)
if not re.match('image/(png|jpe?g|png|ico|x-ico|vnd.microsoft.ico)', r.headers["content-type"], re.I):
log.WARN("content-type wrong", r.headers)
return False
# Convert, resize and save
return store_image(r.content, fn, resize=16)
# Download HTML, look for favicon name in <link rel=shortcut icon>.
#
# Very rough, doesn't respect any <base href=> and manually patches
# icon path to homepage url; nor does any entity decoding.
#
def html_link_icon(url, href="/favicon.png"):
html = ahttp.get(url, encoding="iso-8859-1", timeout=4.5, quieter=1)
# Extract
for link in re.findall(r""" <link ([^<>]+) > """, html, re.X):
pair = re.findall(r""" \b(rel|href) \s*=\s* ["']? ([^<>"']+) ["']? """, link, re.X)
pair = { name: val for name, val in pair }
for name in ("shortcut icon", "favicon", "icon", "icon shortcut"):
if name == pair.get("rel", "ignore").lower() and pair.get("href"):
href = pair["href"].replace("&", "&") # unhtml()
break
# Patch URL together
if re.match("^https?://", href): # absolute URL
return href
elif re.startswith("//", href): # proto-absolute
return "http:" + href
elif re.startswith("/", href): # root path
return re.sub("(https?://[^/]).*$", "\g<1>") + href
else: # relative path references xyz/../
href = re.sub("[^/]+$", "", url) + href
return re.sub("[^/]+/../", "/", href)
#-- test
if __name__ == "__main__":
import sys
favicon(None).download(sys.argv[1])
|