def _get_info_for_launchpad_username(identifier):
"""This figures out what the named person has been involved with.
It returns a dictionary like this:
{
'F-Spot': {
'url': 'http://launchpad.net/f-spot',
'involvement_types': ['Bug Management', 'Bazaar Branches'],
'languages' : ['python', 'shell script']
}
}
"""
# Maybe identifier is an email address? ...
username = None
# If the OpenHatch user entered an email address,
# try to find the corresponding Launchpad username first.
if '@' in identifier:
maybe_email_address = identifier
username = get_launchpad_username_by_email(maybe_email_address)
# ^ This method returns None
# if we couldn't find anybody by that email address.
# Maybe it's not an email address...
if username is None:
username = identifier
try:
b = mechanize_get('https://launchpad.net/~%s' % urllib.quote(username))
except urllib2.HTTPError, e:
if str(e.code) == '404':
return {}
else:
raise # not a 404? Bubble-up the explosion.
def project2languages(project_name):
"""Find the Launchpad URL for the given project. Scrape launchpad page for languages."""
if project_name.startswith('http://') or project_name.startswith('https://'):
project_url = project_name
else:
# Normalize project name to be lowercase
project_name = project_name.lower()
project_url = 'https://launchpad.net/%s' % (
urllib.quote(project_name))
# Now grab the project page and parse it
b = mechanize_get(project_url)
doc = b.response().read()
doc_u = unicode(doc, 'utf-8')
tree = lxml.html.document_fromstring(doc_u)
try:
langs_text = tree.cssselect('#programminglang')[0].text_content()
return map(lambda s: s.strip(), langs_text.split(','))
except IndexError:
return []