Ignore HttpOnly attribute in cookie file

I get the following when using the dcc CLI from master on Debian buster:

$ python3 -m dcc view -v E1200123
Found cookie file: /tmp/ecpcookie.u1000
Fetching contents at URL https://dcc.ligo.org/E1200123/of=xml
Authentication failure, retrieving new cookie...
Fetching cookies for: https://dcc.ligo.org/dcc
Found cookie file: /tmp/ecpcookie.u1000
Fetching contents at URL https://dcc.ligo.org/E1200123/of=xml
Traceback (most recent call last):
  File "/home/ccw/Repos/dcc/dcc/patterns.py", line 30, in validate
    self.root = ET.fromstring(self.content)
  File "/usr/lib/python3.7/xml/etree/ElementTree.py", line 1315, in XML
    parser.feed(text)
xml.etree.ElementTree.ParseError: mismatched tag: line 8, column 5
[...]

The cause is that the relevant entries in the cookie file have a prefix #HttpOnly_ (apparently to indicate that the HttpOnly cookie attribute was set). This MR changes the cookie file parsing so the prefix is ignored.

dcc/comms.py

@@ -258,6 +258,10 @@ class ECPCookieJar(http.cookiejar.MozillaCookieJar):
                 # last field may be absent, so keep any trailing tab
                 if line.endswith("\n"): line = line[:-1]
 
+                # remove #HttpOnly_ prefix
+                if line.strip().startswith("#HttpOnly_"):
+                    line = line[len("#HttpOnly_"):]
+
                 # skip comments and blank lines XXX what is $ for?
                 if (line.strip().startswith(("#", "$")) or
                     line.strip() == ""):