from django.core.management.base import BaseCommand, CommandError
from outside_spending.models import unprocessed_filing, processing_memo, Committee_Overlay, Filing_Scrape_Time
from outside_spending.read_FEC_settings import FILECACHE_DIRECTORY, USER_AGENT, FEC_DOWNLOAD, DELAY_TIME, MAX_FILING_KEY
from outside_spending.utils.fec_logging import fec_logger
# parser regexes -- it's cleaner to use regexes than html parsing
download_area_re = re.compile("<DL>(.+)<BR><P ALIGN=CENTER>", re.S)
committee_name_re = "<H4><A HREF='\/cgi-bin\/dcdev\/forms/(C\d+)\/'>(.*?)\s+-\s+C\d+\s*<\/A><\/H4>"
line_re = " (.+?)\n"
form_re = "<A HREF='\/cgi-bin\/dcdev\/forms\/(C\d+)\/(\d+)\/'>View<\/A> <A HREF='\/cgi-bin\/dcdev\/forms/DL\/\d+\/'>Download</A> \s+FEC-\d+\s+Form\s+(F.*?)\s+"
period_re = "-\s*period (\d\d\/\d\d\/\d\d\d\d)-(\d\d\/\d\d\/\d\d\d\d),"
filed_re = "filed\s+(\d\d\/\d\d\/\d\d\d\d)\s+"
my_logger = fec_logger()
def enter_filing(data_hash):
print "\tentering %s" % (data_hash['filing_number'])
is_superpac = False
filing_created = False
try:
Committee_Overlay.objects.get(fec_id=data_hash['committee_id'],
is_superpac=True)
is_superpac = True
except Committee_Overlay.DoesNotExist:
pass
try:
unprocessed_filing.objects.get(
def __init__(self):
self.fp = form_parser()
self.FEClogger=fec_logger()