def downloadCurrentFileTask(self, task):
""" Continues downloading the URL in self.url and self.filename. """
if self.ch.run():
return task.cont
if self.ch.getStatusCode() == 304:
# This file is still cached from before. We don't need to
# download it again. Move on to the next file.
self.notify.info("already cached: %s" % (self.filename))
return self.downloadNextFile(task)
localFilename = Filename(self.newsDir, self.filename)
if not self.ch.isValid():
self.notify.warning("Unable to download %s" % (self.url))
localFilename.unlink()
if self.filename in self.newsCache:
del self.newsCache[self.filename]
self.saveNewsCache()
# Might as well see if we can get the next file.
return self.downloadNextFile(task)
# Successfully downloaded.
self.notify.info("downloaded %s" % (self.filename))
# The HTTP "Entity Tag" appears to be useless with our CDN:
# different CDN servers will serve up different etag values
# for the same file. We rely on file size and date instead.
size = self.ch.getFileSize()
doc = self.ch.getDocumentSpec()
date = ''
if doc.hasDate():
date = doc.getDate().getString()
self.newsCache[self.filename] = (size, date)
self.saveNewsCache()
# Continue downloading files.
return self.downloadNextFile(task)
def downloadIndexTask(self, task):
""" Get the initial index file from the HTTP server. """
if self.ch.run():
return task.cont
if not self.ch.isValid():
self.notify.warning("Unable to download %s" % (self.url))
self.redownloadingNews = False
return task.done
# OK, now we've got the list of files hosted by the server.
# Parse the list.
self.newsFiles = []
filename = self.rf.readline()
while filename:
filename = filename.strip()
if filename:
self.newsFiles.append(filename)
filename = self.rf.readline()
del self.rf
self.newsFiles.sort()
self.newsIndexEntries = list(self.newsFiles)
self.notify.info("Server lists %s news files" % (len(self.newsFiles)))
self.notify.debug("self.newsIndexEntries=%s" % self.newsIndexEntries)
# Now see if we already have copies of these files we
# downloaded previously.
self.readNewsCache()
# Clean up any unexpected files in this directory--they might
# be old news files, or partial failed downloads from before.
for basename in os.listdir(self.newsDir.toOsSpecific()):
if basename != self.CacheIndexFilename and basename not in self.newsCache:
junk = Filename(self.newsDir, basename)
self.notify.info("Removing %s" % (junk))
junk.unlink()
# And start downloading the files.
self.nextNewsFile = 0
return self.downloadNextFile(task)
