def downloadResult(match, book):
""" match: best result from search providers
book: book we are downloading
return: True if already snatched, False if failed to snatch, >True if we snatched it
"""
try:
myDB = database.DBConnection()
resultTitle = match[1]
newValueDict = match[2]
controlValueDict = match[3]
if book['library'] == 'AudioBook':
auxinfo = 'AudioBook'
else: # elif book['library'] == 'eBook':
auxinfo = 'eBook'
if auxinfo == 'eBook':
snatchedbooks = myDB.match('SELECT BookID from books WHERE BookID=? and Status="Snatched"',
(newValueDict["BookID"],))
else:
snatchedbooks = myDB.match('SELECT BookID from books WHERE BookID=? and AudioStatus="Snatched"',
(newValueDict["BookID"],))
if snatchedbooks:
logger.debug('%s %s already marked snatched' % (book['authorName'], book['bookName']))
return True # someone else already found it
else:
myDB.upsert("wanted", newValueDict, controlValueDict)
if 'libgen' in newValueDict["NZBprov"]: # for libgen we use direct download links
snatch = DirectDownloadMethod(newValueDict["BookID"], newValueDict["NZBtitle"],
controlValueDict["NZBurl"], resultTitle, auxinfo)
elif newValueDict['NZBmode'] in ["torznab", "torrent", "magnet"]:
snatch = TORDownloadMethod(newValueDict["BookID"], newValueDict["NZBtitle"],
controlValueDict["NZBurl"], auxinfo)
elif newValueDict['NZBmode'] == 'nzb':
snatch = NZBDownloadMethod(newValueDict["BookID"], newValueDict["NZBtitle"],
controlValueDict["NZBurl"], auxinfo)
else:
logger.error('Unhandled NZBmode [%s] for %s' % (newValueDict['NZBmode'], controlValueDict["NZBurl"]))
snatch = False
if snatch:
logger.info('Downloading %s %s from %s' %
(auxinfo, newValueDict["NZBtitle"], newValueDict["NZBprov"]))
notify_snatch("%s %s from %s at %s" %
(auxinfo, newValueDict["NZBtitle"], newValueDict["NZBprov"], now()))
custom_notify_snatch(newValueDict["BookID"])
# at this point we could add NZBprov to the blocklist with a short timeout, a second or two?
# This would implement a round-robin search system. Blocklist with an incremental counter.
# If number of active providers == number blocklisted, so no unblocked providers are left,
# either sleep for a while, or unblock the one with the lowest counter.
scheduleJob(action='Start', target='processDir')
return True + True # we found it
return False
except Exception:
logger.error('Unhandled exception in downloadResult: %s' % traceback.format_exc())
def search_magazines(mags=None, reset=False):
# produce a list of magazines to search for, tor, nzb, torznab, rss
# noinspection PyBroadException
try:
threadname = threading.currentThread().name
if "Thread-" in threadname:
if mags is None:
threading.currentThread().name = "SEARCHALLMAG"
else:
threading.currentThread().name = "SEARCHMAG"
myDB = database.DBConnection()
searchlist = []
if mags is None: # backlog search
searchmags = myDB.select('SELECT Title, Regex, LastAcquired, \
IssueDate from magazines WHERE Status="Active"'
)
else:
searchmags = []
for magazine in mags:
searchmags_temp = myDB.select(
'SELECT Title, Regex, LastAcquired, IssueDate from magazines \
WHERE Title=? AND Status="Active"',
(magazine['bookid'], ))
for terms in searchmags_temp:
searchmags.append(terms)
if len(searchmags) == 0:
threading.currentThread().name = "WEBSERVER"
return
# should clear old search results as might not be available any more
# ie torrent not available, changed providers, out of news server retention etc.
# Only delete the "skipped" ones, not wanted/snatched/processed/ignored
logger.debug("Removing old magazine search results")
myDB.action('DELETE from pastissues WHERE Status="Skipped"')
logger.info('Searching for %i magazine%s' %
(len(searchmags), plural(len(searchmags))))
for searchmag in searchmags:
bookid = searchmag['Title']
searchterm = searchmag['Regex']
if not searchterm:
dic = {
'...': '',
' & ': ' ',
' = ': ' ',
'?': '',
'$': 's',
' + ': ' ',
'"': '',
',': '',
'*': ''
}
# strip accents from the magazine title for easier name-matching
searchterm = unaccented_str(searchmag['Title'])
if not searchterm:
# unless there are no ascii characters left
searchterm = searchmag['Title']
searchterm = replace_all(searchterm, dic)
searchterm = re.sub('[.\-/]', ' ', searchterm)
searchterm = searchterm.encode(lazylibrarian.SYS_ENCODING)
searchlist.append({"bookid": bookid, "searchterm": searchterm})
if not searchlist:
logger.warn(
'There is nothing to search for. Mark some magazines as active.'
)
for book in searchlist:
resultlist = []
if lazylibrarian.USE_NZB():
resultlist, nproviders = IterateOverNewzNabSites(book, 'mag')
if not nproviders:
# don't nag. Show warning message no more than every 20 mins
timenow = int(time.time())
if check_int(lazylibrarian.NO_NZB_MSG, 0) + 1200 < timenow:
logger.warn(
'No nzb providers are available. Check config and blocklist'
)
lazylibrarian.NO_NZB_MSG = timenow
if lazylibrarian.USE_DIRECT():
dir_resultlist, nproviders = IterateOverDirectSites(
book, 'mag')
if not nproviders:
# don't nag. Show warning message no more than every 20 mins
timenow = int(time.time())
if check_int(lazylibrarian.NO_DIRECT_MSG,
0) + 1200 < timenow:
logger.warn(
'No direct providers are available. Check config and blocklist'
)
lazylibrarian.NO_DIRECT_MSG = timenow
if dir_resultlist:
for item in dir_resultlist: # reformat the results so they look like nzbs
resultlist.append({
'bookid': item['bookid'],
'nzbprov': item['tor_prov'],
'nzbtitle': item['tor_title'],
'nzburl': item['tor_url'],
'nzbdate':
'Fri, 01 Jan 1970 00:00:00 +0100', # fake date as none returned
'nzbsize': item['tor_size'],
'nzbmode': 'torrent'
})
if lazylibrarian.USE_TOR():
tor_resultlist, nproviders = IterateOverTorrentSites(
book, 'mag')
if not nproviders:
# don't nag. Show warning message no more than every 20 mins
timenow = int(time.time())
if check_int(lazylibrarian.NO_TOR_MSG, 0) + 1200 < timenow:
logger.warn(
'No tor providers are available. Check config and blocklist'
)
lazylibrarian.NO_TOR_MSG = timenow
if tor_resultlist:
for item in tor_resultlist: # reformat the torrent results so they look like nzbs
resultlist.append({
'bookid': item['bookid'],
'nzbprov': item['tor_prov'],
'nzbtitle': item['tor_title'],
'nzburl': item['tor_url'],
'nzbdate':
'Fri, 01 Jan 1970 00:00:00 +0100', # fake date as none returned from torrents
'nzbsize': item['tor_size'],
'nzbmode': 'torrent'
})
if lazylibrarian.USE_RSS():
rss_resultlist, nproviders = IterateOverRSSSites()
if not nproviders:
# don't nag. Show warning message no more than every 20 mins
timenow = int(time.time())
if check_int(lazylibrarian.NO_RSS_MSG, 0) + 1200 < timenow:
logger.warn(
'No rss providers are available. Check config and blocklist'
)
lazylibrarian.NO_RSS_MSG = timenow
if rss_resultlist:
for item in rss_resultlist: # reformat the rss results so they look like nzbs
resultlist.append({
'bookid': book['bookid'],
'nzbprov': item['tor_prov'],
'nzbtitle': item['tor_title'],
'nzburl': item['tor_url'],
'nzbdate': item[
'tor_date'], # may be fake date as none returned from rss torrents, only rss nzb
'nzbsize': item['tor_size'],
'nzbmode': item['tor_type']
})
if not resultlist:
logger.debug("No results for magazine %s" % book['searchterm'])
else:
bad_name = 0
bad_date = 0
old_date = 0
rejects = 0
total_nzbs = 0
new_date = 0
maglist = []
issues = []
bookid = ''
for nzb in resultlist:
total_nzbs += 1
bookid = nzb['bookid']
# strip accents from the magazine title for easier name-matching
nzbtitle = unaccented_str(nzb['nzbtitle'])
if not nzbtitle:
# unless it's not a latin-1 encodable name
nzbtitle = nzb['nzbtitle']
nzbtitle = nzbtitle.replace('"', '').replace(
"'", "") # suppress " in titles
nzburl = nzb['nzburl']
nzbprov = nzb['nzbprov']
nzbdate_temp = nzb['nzbdate']
nzbsize_temp = nzb['nzbsize']
nzbsize_temp = check_int(
nzbsize_temp, 1000
) # not all torrents returned by torznab have a size
nzbsize = round(float(nzbsize_temp) / 1048576, 2)
nzbdate = nzbdate2format(nzbdate_temp)
nzbmode = nzb['nzbmode']
results = myDB.match(
'SELECT * from magazines WHERE Title=?', (bookid, ))
if not results:
logger.debug(
'Magazine [%s] does not match search term [%s].' %
(nzbtitle, bookid))
bad_name += 1
else:
rejected = False
maxsize = check_int(
lazylibrarian.CONFIG['REJECT_MAGSIZE'], 0)
if maxsize and nzbsize > maxsize:
logger.debug("Rejecting %s, too large" % nzbtitle)
rejected = True
if not rejected:
minsize = check_int(
lazylibrarian.CONFIG['REJECT_MAGMIN'], 0)
if minsize and nzbsize < minsize:
logger.debug("Rejecting %s, too small" %
nzbtitle)
rejected = True
if not rejected:
dic = {
'.': ' ',
'-': ' ',
'/': ' ',
'+': ' ',
'_': ' ',
'(': '',
')': ''
}
nzbtitle_formatted = replace_all(nzbtitle,
dic).strip()
# Need to make sure that substrings of magazine titles don't get found
# (e.g. Maxim USA will find Maximum PC USA)
# remove extra spaces if they're in a row
if nzbtitle_formatted and nzbtitle_formatted[
0] == '[' and nzbtitle_formatted[-1] == ']':
nzbtitle_formatted = nzbtitle_formatted[1:-1]
nzbtitle_exploded_temp = " ".join(
nzbtitle_formatted.split())
nzbtitle_exploded = nzbtitle_exploded_temp.split(
' ')
if ' ' in bookid:
bookid_exploded = bookid.split(' ')
else:
bookid_exploded = [bookid]
# check nzb has magazine title and a date/issue nr
# eg The MagPI July 2015
if len(nzbtitle_exploded) > len(bookid_exploded):
# needs to be longer as it has to include a date
# check all the words in the mag title are in the nzbtitle
rejected = False
wlist = []
for word in nzbtitle_exploded:
wlist.append(unaccented(word).lower())
for word in bookid_exploded:
if unaccented(word).lower() not in wlist:
rejected = True
break
if rejected:
logger.debug(
u"Magazine title match failed " +
bookid + " for " + nzbtitle_formatted)
else:
logger.debug(u"Magazine matched " +
bookid + " for " +
nzbtitle_formatted)
else:
logger.debug("Magazine name too short (%s)" %
len(nzbtitle_exploded))
rejected = True
if not rejected:
blocked = myDB.match(
'SELECT * from wanted WHERE NZBurl=? and Status="Failed"',
(nzburl, ))
if blocked:
logger.debug(
"Rejecting %s, blacklisted at %s" %
(nzbtitle_formatted, blocked['NZBprov']))
rejected = True
if not rejected:
reject_list = getList(
str(results['Reject']).lower())
reject_list += getList(
lazylibrarian.CONFIG['REJECT_MAGS'])
lower_title = unaccented(
nzbtitle_formatted).lower()
lower_bookid = unaccented(bookid).lower()
if reject_list:
if lazylibrarian.LOGLEVEL > 2:
logger.debug('Reject: %s' %
str(reject_list))
logger.debug('Title: %s' % lower_title)
logger.debug('Bookid: %s' % lower_bookid)
for word in reject_list:
if word in lower_title and word not in lower_bookid:
rejected = True
logger.debug("Rejecting %s, contains %s" %
(nzbtitle_formatted, word))
break
regex_pass = 0
if not rejected:
# Magazine names have many different styles of date
# DD MonthName YYYY OR MonthName YYYY or Issue nn, MonthName YYYY
# MonthName DD YYYY or MonthName DD, YYYY
# YYYY MM or YYYY MM DD
# Issue/No/Nr/Vol nn, YYYY or Issue/No/Nr/Vol nn
# nn YYYY issue number without "Nr" before it
# issue and year as a single 6 digit string eg 222015
newdatish = "none"
# DD MonthName YYYY OR MonthName YYYY or Issue nn, MonthName YYYY
pos = 0
while pos < len(nzbtitle_exploded):
year = check_year(nzbtitle_exploded[pos])
if year and pos:
month = month2num(nzbtitle_exploded[pos -
1])
if month:
if pos - 1:
day = check_int(
nzbtitle_exploded[pos - 2], 1)
if day > 31: # probably issue number nn
day = 1
else:
day = 1
newdatish = "%04d-%02d-%02d" % (
year, month, day)
try:
_ = datetime.date(year, month, day)
regex_pass = 1
break
except ValueError:
regex_pass = 0
pos += 1
# MonthName DD YYYY or MonthName DD, YYYY
if not regex_pass:
pos = 0
while pos < len(nzbtitle_exploded):
year = check_year(nzbtitle_exploded[pos])
if year and (pos - 1):
month = month2num(
nzbtitle_exploded[pos - 2])
if month:
day = check_int(
nzbtitle_exploded[
pos - 1].rstrip(','), 1)
try:
_ = datetime.date(
year, month, day)
newdatish = "%04d-%02d-%02d" % (
year, month, day)
regex_pass = 2
break
except ValueError:
regex_pass = 0
pos += 1
# YYYY MM or YYYY MM DD
if not regex_pass:
pos = 0
while pos < len(nzbtitle_exploded):
year = check_year(nzbtitle_exploded[pos])
if year and pos + 1 < len(
nzbtitle_exploded):
month = check_int(
nzbtitle_exploded[pos + 1], 0)
if month:
if pos + 2 < len(
nzbtitle_exploded):
day = check_int(
nzbtitle_exploded[pos + 2],
1)
else:
day = 1
try:
_ = datetime.date(
year, month, day)
newdatish = "%04d-%02d-%02d" % (
year, month, day)
regex_pass = 3
break
except ValueError:
regex_pass = 0
pos += 1
# Issue/No/Nr/Vol nn, YYYY or Issue/No/Nr/Vol nn
if not regex_pass:
pos = 0
while pos < len(nzbtitle_exploded):
if nzbtitle_exploded[pos].lower() in [
"issue", "no", "nr", "vol"
]:
if pos + 1 < len(nzbtitle_exploded):
issue = check_int(
nzbtitle_exploded[pos + 1], 0)
if issue:
newdatish = str(
issue) # 4 == 04 == 004
if pos + 2 < len(
nzbtitle_exploded):
year = check_year(
nzbtitle_exploded[pos +
2])
if year and year < int(
datetime.date.
today().year):
newdatish = '0' # it's old
regex_pass = 4 # Issue/No/Nr/Vol nn, YYYY
else:
regex_pass = 5 # Issue/No/Nr/Vol nn
break
pos += 1
# nn YYYY issue number without "Nr" before it
if not regex_pass:
pos = 1
while pos < len(nzbtitle_exploded):
year = check_year(nzbtitle_exploded[pos])
if year:
issue = check_int(
nzbtitle_exploded[pos - 1], 0)
if issue:
newdatish = str(
issue) # 4 == 04 == 004
regex_pass = 6
if year < int(datetime.date.today(
).year):
newdatish = '0' # it's old
break
pos += 1
# issue and year as a single 6 digit string eg 222015
if not regex_pass:
pos = 0
while pos < len(nzbtitle_exploded):
issue = nzbtitle_exploded[pos]
if issue.isdigit() and len(issue) == 6:
year = int(issue[2:])
issue = int(issue[:2])
newdatish = str(
issue) # 4 == 04 == 004
regex_pass = 7
if year < int(
datetime.date.today().year):
newdatish = '0' # it's old
break
pos += 1
if not regex_pass:
logger.debug(
'Magazine %s not in a recognised date format.'
% nzbtitle_formatted)
bad_date += 1
# allow issues with good name but bad date to be included
# so user can manually select them, incl those with issue numbers
newdatish = "1970-01-01" # provide a fake date for bad-date issues
regex_pass = 99
if rejected:
rejects += 1
else:
if lazylibrarian.LOGLEVEL > 2:
logger.debug("regex %s [%s] %s" %
(regex_pass, nzbtitle_formatted,
newdatish))
# wanted issues go into wanted table marked "Wanted"
# the rest into pastissues table marked "Skipped"
insert_table = "pastissues"
insert_status = "Skipped"
control_date = results['IssueDate']
if control_date is None: # we haven't got any copies of this magazine yet
# get a rough time just over a month ago to compare to, in format yyyy-mm-dd
# could perhaps calc differently for weekly, biweekly etc
# or for magazines with only an issue number, use zero
if str(newdatish).isdigit():
logger.debug(
'Magazine comparing issue numbers (%s)'
% newdatish)
control_date = 0
elif re.match('\d+-\d\d-\d\d', str(newdatish)):
start_time = time.time()
start_time -= int(
lazylibrarian.CONFIG['MAG_AGE']
) * 24 * 60 * 60 # number of seconds in days
if start_time < 0: # limit of unixtime (1st Jan 1970)
start_time = 0
control_date = time.strftime(
"%Y-%m-%d", time.localtime(start_time))
logger.debug(
'Magazine date comparing to %s' %
control_date)
else:
logger.debug(
'Magazine unable to find comparison type [%s]'
% newdatish)
control_date = 0
if str(control_date).isdigit() and str(
newdatish).isdigit():
# for issue numbers, check if later than last one we have
comp_date = int(newdatish) - int(control_date)
newdatish = "%s" % newdatish
newdatish = newdatish.zfill(
4) # pad so we sort correctly
elif re.match('\d+-\d\d-\d\d', str(control_date)) and \
re.match('\d+-\d\d-\d\d', str(newdatish)):
# only grab a copy if it's newer than the most recent we have,
# or newer than a month ago if we have none
comp_date = datecompare(
newdatish, control_date)
else:
# invalid comparison of date and issue number
if re.match('\d+-\d\d-\d\d',
str(control_date)):
logger.debug(
'Magazine %s failed: Expecting a date'
% nzbtitle_formatted)
else:
logger.debug(
'Magazine %s failed: Expecting issue number'
% nzbtitle_formatted)
bad_date += 1
newdatish = "1970-01-01" # this is our fake date for ones we can't decipher
comp_date = 0
if comp_date > 0:
# keep track of what we're going to download so we don't download dupes
new_date += 1
issue = bookid + ',' + newdatish
if issue not in issues:
maglist.append({
'bookid': bookid,
'nzbprov': nzbprov,
'nzbtitle': nzbtitle,
'nzburl': nzburl,
'nzbmode': nzbmode
})
logger.debug(
'This issue of %s is new, downloading'
% nzbtitle_formatted)
issues.append(issue)
logger.debug('Magazine request number %s' %
len(issues))
if lazylibrarian.LOGLEVEL > 2:
logger.debug(str(issues))
insert_table = "wanted"
insert_status = "Wanted"
nzbdate = now() # when we asked for it
else:
logger.debug(
'This issue of %s is already flagged for download'
% issue)
else:
if newdatish != "1970-01-01": # this is our fake date for ones we can't decipher
logger.debug(
'This issue of %s is old; skipping.' %
nzbtitle_formatted)
old_date += 1
# store only the _new_ matching results
# Don't add a new entry if this issue has been found on an earlier search
# and status has been user-set ( we only delete the "Skipped" ones )
# In "wanted" table it might be already snatched/downloading/processing
mag_entry = myDB.match(
'SELECT * from %s WHERE NZBtitle=? and NZBprov=?'
% insert_table, (nzbtitle, nzbprov))
if mag_entry:
if lazylibrarian.LOGLEVEL > 2:
logger.debug(
'%s is already in %s marked %s' %
(nzbtitle, insert_table,
insert_status))
else:
controlValueDict = {
"NZBtitle": nzbtitle,
"NZBprov": nzbprov
}
newValueDict = {
"NZBurl": nzburl,
"BookID": bookid,
"NZBdate": nzbdate,
"AuxInfo": newdatish,
"Status": insert_status,
"NZBsize": nzbsize,
"NZBmode": nzbmode
}
myDB.upsert(insert_table, newValueDict,
controlValueDict)
if lazylibrarian.LOGLEVEL > 2:
logger.debug('Added %s to %s marked %s' %
(nzbtitle, insert_table,
insert_status))
msg = 'Found %i result%s for %s. %i new,' % (
total_nzbs, plural(total_nzbs), bookid, new_date)
msg += ' %i old, %i fail date, %i fail name,' % (
old_date, bad_date, bad_name)
msg += ' %i rejected: %i to download' % (rejects, len(maglist))
logger.info(msg)
for magazine in maglist:
if magazine['nzbmode'] in ["torznab", "torrent", "magnet"]:
snatch = TORDownloadMethod(magazine['bookid'],
magazine['nzbtitle'],
magazine['nzburl'],
'magazine')
else:
snatch = NZBDownloadMethod(magazine['bookid'],
magazine['nzbtitle'],
magazine['nzburl'],
'magazine')
if snatch:
logger.info(
'Downloading %s from %s' %
(magazine['nzbtitle'], magazine["nzbprov"]))
notify_snatch("Magazine %s from %s at %s" %
(unaccented(magazine['nzbtitle']),
magazine["nzbprov"], now()))
custom_notify_snatch(magazine['bookid'])
scheduleJob(action='Start', target='processDir')
if reset:
scheduleJob(action='Restart', target='search_magazines')
logger.info("Search for magazines complete")
except Exception:
logger.error('Unhandled exception in search_magazines: %s' %
traceback.format_exc())
finally:
threading.currentThread().name = "WEBSERVER"
def downloadResult(match, book):
""" match: best result from search providers
book: book we are downloading (needed for reporting author name)
return: 0 if failed to snatch
1 if already snatched
2 if we snatched it
"""
# noinspection PyBroadException
try:
myDB = database.DBConnection()
newValueDict = match[1]
controlValueDict = match[2]
# It's possible to get book and wanted tables "Snatched" status out of sync
# for example if a user marks a book as "Wanted" after a search task snatches it and before postprocessor runs
# so check status in both tables here
snatched = myDB.match(
'SELECT BookID from wanted WHERE BookID=? and AuxInfo=? and Status="Snatched"',
(newValueDict["BookID"], newValueDict["AuxInfo"]))
if snatched:
logger.debug('%s %s %s already marked snatched in wanted table' %
(newValueDict["AuxInfo"], book['authorName'],
book['bookName']))
return 1 # someone else already found it
if newValueDict["AuxInfo"] == 'eBook':
snatched = myDB.match(
'SELECT BookID from books WHERE BookID=? and Status="Snatched"',
(newValueDict["BookID"], ))
else:
snatched = myDB.match(
'SELECT BookID from books WHERE BookID=? and AudioStatus="Snatched"',
(newValueDict["BookID"], ))
if snatched:
logger.debug('%s %s %s already marked snatched in book table' %
(newValueDict["AuxInfo"], book['authorName'],
book['bookName']))
return 1 # someone else already found it
myDB.upsert("wanted", newValueDict, controlValueDict)
if newValueDict['NZBmode'] == 'direct':
snatch = DirectDownloadMethod(newValueDict["BookID"],
newValueDict["NZBtitle"],
controlValueDict["NZBurl"],
newValueDict["AuxInfo"])
elif newValueDict['NZBmode'] in ["torznab", "torrent", "magnet"]:
snatch = TORDownloadMethod(newValueDict["BookID"],
newValueDict["NZBtitle"],
controlValueDict["NZBurl"],
newValueDict["AuxInfo"])
elif newValueDict['NZBmode'] == 'nzb':
snatch = NZBDownloadMethod(newValueDict["BookID"],
newValueDict["NZBtitle"],
controlValueDict["NZBurl"],
newValueDict["AuxInfo"])
else:
logger.error('Unhandled NZBmode [%s] for %s' %
(newValueDict['NZBmode'], controlValueDict["NZBurl"]))
snatch = 0
if snatch:
logger.info('Downloading %s %s from %s' %
(newValueDict["AuxInfo"], newValueDict["NZBtitle"],
newValueDict["NZBprov"]))
custom_notify_snatch(
"%s %s" % (newValueDict["BookID"], newValueDict['AuxInfo']))
notify_snatch("%s %s from %s at %s" %
(newValueDict["AuxInfo"], newValueDict["NZBtitle"],
newValueDict["NZBprov"], now()))
# at this point we could add NZBprov to the blocklist with a short timeout, a second or two?
# This would implement a round-robin search system. Blocklist with an incremental counter.
# If number of active providers == number blocklisted, so no unblocked providers are left,
# either sleep for a while, or unblock the one with the lowest counter.
scheduleJob(action='Start', target='processDir')
return 2 # we found it
return 0
except Exception:
logger.error('Unhandled exception in downloadResult: %s' %
traceback.format_exc())
return 0
def search_magazines(mags=None, reset=False):
# produce a list of magazines to search for, tor, nzb, torznab, rss
# noinspection PyBroadException
try:
threadname = threading.currentThread().name
if "Thread-" in threadname:
if not mags:
threading.currentThread().name = "SEARCHALLMAG"
else:
threading.currentThread().name = "SEARCHMAG"
myDB = database.DBConnection()
searchlist = []
if not mags: # backlog search
searchmags = myDB.select('SELECT Title, Regex, DateType, LastAcquired, \
IssueDate from magazines WHERE Status="Active"')
else:
searchmags = []
for magazine in mags:
searchmags_temp = myDB.select('SELECT Title,Regex,DateType,LastAcquired,IssueDate from magazines \
WHERE Title=? AND Status="Active"', (magazine['bookid'],))
for terms in searchmags_temp:
searchmags.append(terms)
if len(searchmags) == 0:
threading.currentThread().name = "WEBSERVER"
return
# should clear old search results as might not be available any more
# ie torrent not available, changed providers, out of news server retention etc.
# Only delete the "skipped" ones, not wanted/snatched/processed/ignored
# logger.debug("Removing old magazine search results")
# myDB.action('DELETE from pastissues WHERE Status="Skipped"')
logger.info('Searching for %i magazine%s' % (len(searchmags), plural(len(searchmags))))
for searchmag in searchmags:
bookid = searchmag['Title']
searchterm = searchmag['Regex']
datetype = searchmag['DateType']
if not datetype:
datetype = ''
if not searchterm:
dic = {'...': '', ' & ': ' ', ' = ': ' ', '?': '', '$': 's', ' + ': ' ', '"': '', ',': '', '*': ''}
# strip accents from the magazine title for easier name-matching
searchterm = unaccented_str(searchmag['Title'])
if not searchterm:
# unless there are no ascii characters left
searchterm = searchmag['Title']
searchterm = replace_all(searchterm, dic)
searchterm = re.sub('[.\-/]', ' ', searchterm)
if PY2:
searchterm = searchterm.encode(lazylibrarian.SYS_ENCODING)
searchlist.append({"bookid": bookid, "searchterm": searchterm, "datetype": datetype})
if not searchlist:
logger.warn('There is nothing to search for. Mark some magazines as active.')
for book in searchlist:
resultlist = []
if lazylibrarian.USE_NZB():
resultlist, nproviders = IterateOverNewzNabSites(book, 'mag')
if not nproviders:
# don't nag. Show warning message no more than every 20 mins
timenow = int(time.time())
if check_int(lazylibrarian.NO_NZB_MSG, 0) + 1200 < timenow:
logger.warn('No nzb providers are available. Check config and blocklist')
lazylibrarian.NO_NZB_MSG = timenow
if lazylibrarian.USE_DIRECT():
dir_resultlist, nproviders = IterateOverDirectSites(book, 'mag')
if not nproviders:
# don't nag. Show warning message no more than every 20 mins
timenow = int(time.time())
if check_int(lazylibrarian.NO_DIRECT_MSG, 0) + 1200 < timenow:
logger.warn('No direct providers are available. Check config and blocklist')
lazylibrarian.NO_DIRECT_MSG = timenow
if dir_resultlist:
for item in dir_resultlist: # reformat the results so they look like nzbs
resultlist.append({
'bookid': item['bookid'],
'nzbprov': item['tor_prov'],
'nzbtitle': item['tor_title'],
'nzburl': item['tor_url'],
'nzbdate': 'Fri, 01 Jan 1970 00:00:00 +0100', # fake date as none returned
'nzbsize': item['tor_size'],
'nzbmode': 'direct'
})
if lazylibrarian.USE_TOR():
tor_resultlist, nproviders = IterateOverTorrentSites(book, 'mag')
if not nproviders:
# don't nag. Show warning message no more than every 20 mins
timenow = int(time.time())
if check_int(lazylibrarian.NO_TOR_MSG, 0) + 1200 < timenow:
logger.warn('No tor providers are available. Check config and blocklist')
lazylibrarian.NO_TOR_MSG = timenow
if tor_resultlist:
for item in tor_resultlist: # reformat the torrent results so they look like nzbs
resultlist.append({
'bookid': item['bookid'],
'nzbprov': item['tor_prov'],
'nzbtitle': item['tor_title'],
'nzburl': item['tor_url'],
'nzbdate': 'Fri, 01 Jan 1970 00:00:00 +0100', # fake date as none returned from torrents
'nzbsize': item['tor_size'],
'nzbmode': 'torrent'
})
if lazylibrarian.USE_RSS():
rss_resultlist, nproviders, dltypes = IterateOverRSSSites()
if not nproviders or 'M' not in dltypes:
# don't nag. Show warning message no more than every 20 mins
timenow = int(time.time())
if check_int(lazylibrarian.NO_RSS_MSG, 0) + 1200 < timenow:
logger.warn('No rss providers are available. Check config and blocklist')
lazylibrarian.NO_RSS_MSG = timenow
if rss_resultlist:
for item in rss_resultlist: # reformat the rss results so they look like nzbs
if 'M' in item['types']:
resultlist.append({
'bookid': book['bookid'],
'nzbprov': item['tor_prov'],
'nzbtitle': item['tor_title'],
'nzburl': item['tor_url'],
'nzbdate': item['tor_date'], # may be fake date as none returned from rss torrents
'nzbsize': item['tor_size'],
'nzbmode': item['tor_type']
})
if not resultlist:
logger.debug("No results for magazine %s" % book['searchterm'])
else:
bad_name = 0
bad_date = 0
old_date = 0
rejects = 0
total_nzbs = 0
new_date = 0
maglist = []
issues = []
bookid = ''
for nzb in resultlist:
total_nzbs += 1
bookid = nzb['bookid']
# strip accents from the magazine title for easier name-matching
nzbtitle = unaccented_str(nzb['nzbtitle'])
if not nzbtitle:
# unless it's not a latin-1 encodable name
nzbtitle = nzb['nzbtitle']
nzbtitle = nzbtitle.replace('"', '').replace("'", "") # suppress " in titles
nzburl = nzb['nzburl']
nzbprov = nzb['nzbprov']
nzbdate_temp = nzb['nzbdate']
nzbsize_temp = nzb['nzbsize']
nzbsize_temp = check_int(nzbsize_temp, 1000) # not all torrents returned by torznab have a size
nzbsize = round(float(nzbsize_temp) / 1048576, 2)
nzbdate = nzbdate2format(nzbdate_temp)
nzbmode = nzb['nzbmode']
# Need to make sure that substrings of magazine titles don't get found
# (e.g. Maxim USA will find Maximum PC USA) so split into "words"
dic = {'.': ' ', '-': ' ', '/': ' ', '+': ' ', '_': ' ', '(': '', ')': '', '[': ' ', ']': ' ',
'#': '# '}
nzbtitle_formatted = replace_all(nzbtitle, dic)
# remove extra spaces if they're in a row
nzbtitle_formatted = " ".join(nzbtitle_formatted.split())
nzbtitle_exploded = nzbtitle_formatted.split()
results = myDB.match('SELECT * from magazines WHERE Title=?', (bookid,))
if not results:
logger.debug('Magazine [%s] does not match search term [%s].' % (nzbtitle, bookid))
bad_name += 1
else:
rejected = False
maxsize = check_int(lazylibrarian.CONFIG['REJECT_MAGSIZE'], 0)
if maxsize and nzbsize > maxsize:
logger.debug("Rejecting %s, too large" % nzbtitle)
rejected = True
if not rejected:
minsize = check_int(lazylibrarian.CONFIG['REJECT_MAGMIN'], 0)
if minsize and nzbsize < minsize:
logger.debug("Rejecting %s, too small" % nzbtitle)
rejected = True
if not rejected:
bookid_exploded = replace_all(bookid, dic).split()
# Check nzb has magazine title and a date/issue nr
# eg The MagPI July 2015
if len(nzbtitle_exploded) > len(bookid_exploded):
# needs to be longer as it has to include a date
# check all the words in the mag title are in the nzbtitle
rejected = False
wlist = []
for word in nzbtitle_exploded:
word = unaccented(word).lower()
if word:
wlist.append(word)
for word in bookid_exploded:
word = unaccented(word).lower()
if word and word not in wlist:
logger.debug("Rejecting %s, missing %s" % (nzbtitle, word))
rejected = True
break
if rejected:
logger.debug(
"Magazine title match failed " + bookid + " for " + nzbtitle_formatted)
else:
logger.debug(
"Magazine title matched " + bookid + " for " + nzbtitle_formatted)
else:
logger.debug("Magazine name too short (%s)" % len(nzbtitle_exploded))
rejected = True
if not rejected and lazylibrarian.CONFIG['BLACKLIST_FAILED']:
blocked = myDB.match('SELECT * from wanted WHERE NZBurl=? and Status="Failed"', (nzburl,))
if blocked:
logger.debug("Rejecting %s, blacklisted at %s" %
(nzbtitle_formatted, blocked['NZBprov']))
rejected = True
if not rejected and lazylibrarian.CONFIG['BLACKLIST_PROCESSED']:
blocked = myDB.match('SELECT * from wanted WHERE NZBurl=?', (nzburl,))
if blocked:
logger.debug("Rejecting %s, blacklisted at %s" %
(nzbtitle_formatted, blocked['NZBprov']))
rejected = True
if not rejected:
reject_list = getList(str(results['Reject']).lower())
reject_list += getList(lazylibrarian.CONFIG['REJECT_MAGS'], ',')
lower_title = unaccented(nzbtitle_formatted).lower()
lower_bookid = unaccented(bookid).lower()
if reject_list:
if lazylibrarian.LOGLEVEL & lazylibrarian.log_searchmag:
logger.debug('Reject: %s' % str(reject_list))
logger.debug('Title: %s' % lower_title)
logger.debug('Bookid: %s' % lower_bookid)
for word in reject_list:
if word in lower_title and word not in lower_bookid:
rejected = True
logger.debug("Rejecting %s, contains %s" % (nzbtitle_formatted, word))
break
if rejected:
rejects += 1
else:
regex_pass, issuedate, year = get_issue_date(nzbtitle_exploded)
if regex_pass:
logger.debug('Issue %s (regex %s) for %s ' %
(issuedate, regex_pass, nzbtitle_formatted))
datetype_ok = True
datetype = book['datetype']
if datetype:
# check all wanted parts are in the regex result
# Day Month Year Vol Iss (MM needs two months)
if 'M' in datetype and regex_pass not in [1, 2, 3, 4, 5, 6, 7, 12]:
datetype_ok = False
elif 'D' in datetype and regex_pass not in [3, 5, 6]:
datetype_ok = False
elif 'MM' in datetype and regex_pass not in [1]: # bi monthly
datetype_ok = False
elif 'V' in datetype and 'I' in datetype and regex_pass not in [8, 9, 17, 18]:
datetype_ok = False
elif 'V' in datetype and regex_pass not in [2, 10, 11, 12, 13, 14, 17, 18]:
datetype_ok = False
elif 'I' in datetype and regex_pass not in [2, 10, 11, 12, 13, 14, 16, 17, 18]:
datetype_ok = False
elif 'Y' in datetype and regex_pass not in [1, 2, 3, 4, 5, 6, 7, 8, 10,
12, 13, 15, 16, 18]:
datetype_ok = False
else:
datetype_ok = False
logger.debug('Magazine %s not in a recognised date format.' % nzbtitle_formatted)
bad_date += 1
# allow issues with good name but bad date to be included
# so user can manually select them, incl those with issue numbers
issuedate = "1970-01-01" # provide a fake date for bad-date issues
# wanted issues go into wanted table marked "Wanted"
# the rest into pastissues table marked "Skipped" or "Have"
insert_table = "pastissues"
comp_date = 0
if datetype_ok:
control_date = results['IssueDate']
logger.debug("Control date: [%s]" % control_date)
if not control_date: # we haven't got any copies of this magazine yet
# get a rough time just over MAX_AGE days ago to compare to, in format yyyy-mm-dd
# could perhaps calc differently for weekly, biweekly etc
# For magazines with only an issue number use zero as we can't tell age
if str(issuedate).isdigit():
logger.debug('Magazine comparing issue numbers (%s)' % issuedate)
control_date = 0
elif re.match('\d+-\d\d-\d\d', str(issuedate)):
start_time = time.time()
start_time -= int(
lazylibrarian.CONFIG['MAG_AGE']) * 24 * 60 * 60 # number of seconds in days
if start_time < 0: # limit of unixtime (1st Jan 1970)
start_time = 0
control_date = time.strftime("%Y-%m-%d", time.localtime(start_time))
logger.debug('Magazine date comparing to %s' % control_date)
else:
logger.debug('Magazine unable to find comparison type [%s]' % issuedate)
control_date = 0
if str(control_date).isdigit() and str(issuedate).isdigit():
# for issue numbers, check if later than last one we have
if regex_pass in [10, 12, 13] and year:
issuedate = "%s%04d" % (year, int(issuedate))
else:
issuedate = str(issuedate).zfill(4)
if not control_date:
comp_date = 1
else:
comp_date = int(issuedate) - int(control_date)
elif re.match('\d+-\d\d-\d\d', str(control_date)) and \
re.match('\d+-\d\d-\d\d', str(issuedate)):
# only grab a copy if it's newer than the most recent we have,
# or newer than a month ago if we have none
comp_date = datecompare(issuedate, control_date)
else:
# invalid comparison of date and issue number
comp_date = 0
if re.match('\d+-\d\d-\d\d', str(control_date)):
if regex_pass > 9 and year:
# we assumed it was an issue number, but it could be a date
year = check_int(year, 0)
if regex_pass in [10, 12, 13]:
issuedate = int(issuedate[:4])
issuenum = check_int(issuedate, 0)
if year and 1 <= issuenum <= 12:
issuedate = "%04d-%02d-01" % (year, issuenum)
comp_date = datecompare(issuedate, control_date)
if not comp_date:
logger.debug('Magazine %s failed: Expecting a date' % nzbtitle_formatted)
else:
logger.debug('Magazine %s failed: Expecting issue number' % nzbtitle_formatted)
if not comp_date:
bad_date += 1
issuedate = "1970-01-01"
if issuedate == "1970-01-01":
logger.debug('This issue of %s is unknown age; skipping.' % nzbtitle_formatted)
elif not datetype_ok:
logger.debug('This issue of %s not in a wanted date format.' % nzbtitle_formatted)
elif comp_date > 0:
# keep track of what we're going to download so we don't download dupes
new_date += 1
issue = bookid + ',' + issuedate
if issue not in issues:
maglist.append({
'bookid': bookid,
'nzbprov': nzbprov,
'nzbtitle': nzbtitle,
'nzburl': nzburl,
'nzbmode': nzbmode
})
logger.debug('This issue of %s is new, downloading' % nzbtitle_formatted)
issues.append(issue)
logger.debug('Magazine request number %s' % len(issues))
if lazylibrarian.LOGLEVEL & lazylibrarian.log_searchmag:
logger.debug(str(issues))
insert_table = "wanted"
nzbdate = now() # when we asked for it
else:
logger.debug('This issue of %s is already flagged for download' % issue)
else:
if lazylibrarian.LOGLEVEL & lazylibrarian.log_searchmag:
logger.debug('This issue of %s is old; skipping.' % nzbtitle_formatted)
old_date += 1
# store only the _new_ matching results
# Don't add a new entry if this issue has been found on an earlier search
# and status has been user-set ( we only delete the "Skipped" ones )
# In "wanted" table it might be already snatched/downloading/processing
mag_entry = myDB.match('SELECT Status from %s WHERE NZBtitle=? and NZBprov=?' %
insert_table, (nzbtitle, nzbprov))
if mag_entry:
if lazylibrarian.LOGLEVEL & lazylibrarian.log_searchmag:
logger.debug('%s is already in %s marked %s' %
(nzbtitle, insert_table, mag_entry['Status']))
else:
controlValueDict = {
"NZBtitle": nzbtitle,
"NZBprov": nzbprov
}
if insert_table == 'pastissues':
# try to mark ones we've already got
match = myDB.match("SELECT * from issues WHERE Title=? AND IssueDate=?",
(bookid, issuedate))
if match:
insert_status = "Have"
else:
insert_status = "Skipped"
else:
insert_status = "Wanted"
newValueDict = {
"NZBurl": nzburl,
"BookID": bookid,
"NZBdate": nzbdate,
"AuxInfo": issuedate,
"Status": insert_status,
"NZBsize": nzbsize,
"NZBmode": nzbmode
}
myDB.upsert(insert_table, newValueDict, controlValueDict)
if lazylibrarian.LOGLEVEL & lazylibrarian.log_searchmag:
logger.debug('Added %s to %s marked %s' % (nzbtitle, insert_table, insert_status))
msg = 'Found %i result%s for %s. %i new,' % (total_nzbs, plural(total_nzbs), bookid, new_date)
msg += ' %i old, %i fail date, %i fail name,' % (old_date, bad_date, bad_name)
msg += ' %i rejected: %i to download' % (rejects, len(maglist))
logger.info(msg)
for magazine in maglist:
if magazine['nzbmode'] in ["torznab", "torrent", "magnet"]:
snatch, res = TORDownloadMethod(
magazine['bookid'],
magazine['nzbtitle'],
magazine['nzburl'],
'magazine')
elif magazine['nzbmode'] == 'direct':
snatch, res = DirectDownloadMethod(
magazine['bookid'],
magazine['nzbtitle'],
magazine['nzburl'],
'magazine')
elif magazine['nzbmode'] == 'nzb':
snatch, res = NZBDownloadMethod(
magazine['bookid'],
magazine['nzbtitle'],
magazine['nzburl'],
'magazine')
else:
res = 'Unhandled NZBmode [%s] for %s' % (magazine['nzbmode'], magazine["nzburl"])
logger.error(res)
snatch = 0
if snatch:
logger.info('Downloading %s from %s' % (magazine['nzbtitle'], magazine["nzbprov"]))
custom_notify_snatch("%s %s" % (magazine['bookid'], magazine['nzburl']))
notify_snatch("Magazine %s from %s at %s" %
(unaccented(magazine['nzbtitle']), magazine["nzbprov"], now()))
scheduleJob(action='Start', target='PostProcessor')
else:
myDB.action('UPDATE wanted SET status="Failed",DLResult=? WHERE NZBurl=?',
(res, magazine["nzburl"]))
if reset:
scheduleJob(action='Restart', target='search_magazines')
logger.info("Search for magazines complete")
except Exception:
logger.error('Unhandled exception in search_magazines: %s' % traceback.format_exc())
finally:
threading.currentThread().name = "WEBSERVER"
