django-helpdesk/lib.py
Ross Poulton 762f48a59e Resolves issue #62 - adds spam filtering via akismet.py using either
Akismet or TypePad AntiSpam.

See the README for configuration instructions.
2009-06-25 11:22:53 +00:00

344 lines
11 KiB
Python

"""
Jutda Helpdesk - A Django powered ticket tracker for small enterprise.
(c) Copyright 2008 Jutda. All Rights Reserved. See LICENSE for details.
lib.py - Common functions (eg multipart e-mail)
"""
chart_colours = ('80C65A', '990066', 'FF9900', '3399CC', 'BBCCED', '3399CC', 'FFCC33')
try:
from base64 import urlsafe_b64encode as b64encode
except ImportError:
from base64 import encodestring as b64encode
try:
from base64 import urlsafe_b64decode as b64decode
except ImportError:
from base64 import decodestring as b64decode
def send_templated_mail(template_name, email_context, recipients, sender=None, bcc=None, fail_silently=False, files=None):
"""
send_templated_mail() is a warpper around Django's e-mail routines that
allows us to easily send multipart (text/plain & text/html) e-mails using
templates that are stored in the database. This lets the admin provide
both a text and a HTML template for each message.
template_name is the slug of the template to use for this message (see
models.EmailTemplate)
email_context is a dictionary to be used when rendering the template
recipients can be either a string, eg 'a@b.com', or a list of strings.
sender should contain a string, eg 'My Site <me@z.com>'. If you leave it
blank, it'll use settings.DEFAULT_FROM_EMAIL as a fallback.
bcc is an optional list of addresses that will receive this message as a
blind carbon copy.
fail_silently is passed to Django's mail routine. Set to 'True' to ignore
any errors at send time.
files can be a list of file paths to be attached, or it can be left blank.
eg ('/tmp/file1.txt', '/tmp/image.png')
"""
from django.conf import settings
from django.core.mail import EmailMultiAlternatives
from django.template import loader, Context
from helpdesk.models import EmailTemplate
import os
context = Context(email_context)
locale = getattr(context['queue'], 'locale', 'en')
if locale:
template_localized = template_name + ':' + locale
else:
template_localized = None
t = None
if template_localized:
try:
t = EmailTemplate.objects.get(template_name__iexact=template_localized)
except EmailTemplate.DoesNotExist:
pass
if not t:
t = EmailTemplate.objects.get(template_name__iexact=template_name)
if not sender:
sender = settings.DEFAULT_FROM_EMAIL
footer_file = os.path.join('helpdesk', locale, 'email_text_footer.txt')
text_part = loader.get_template_from_string(
"%s{%% include '%s' %%}" % (t.plain_text, footer_file)
).render(context)
email_html_base_file = os.path.join('helpdesk', locale, 'email_html_base.html')
html_part = loader.get_template_from_string(
"{%% extends '%s' %%}{%% block title %%}%s{%% endblock %%}{%% block content %%}%s{%% endblock %%}" % (email_html_base_file, t.heading, t.html)
).render(context)
subject_part = loader.get_template_from_string(
"{{ ticket.ticket }} {{ ticket.title }} %s" % t.subject
).render(context)
if type(recipients) != list:
recipients = [recipients,]
msg = EmailMultiAlternatives( subject_part,
text_part,
sender,
recipients,
bcc=bcc)
msg.attach_alternative(html_part, "text/html")
if files:
if type(files) != list:
files = [files,]
for file in files:
msg.attach_file(file)
return msg.send(fail_silently)
def normalise_data(data, to=100):
"""
Used for normalising data prior to graphing with Google charting API. EG:
[1, 4, 10] becomes [10, 40, 100]
[36, 54, 240] becomes [15, 23, 100]
"""
max_value = max(data)
if max_value > to:
new_data = []
for d in data:
new_data.append(int(d/float(max_value)*to))
data = new_data
return data
def line_chart(data):
"""
'data' is a list of lists making a table.
Row 1, columns 2-n are data headings (the time periods)
Rows 2-n are data, with column 1 being the line labels
"""
column_headings = data[0][1:]
max = 0
for row in data[1:]:
for field in row[1:]:
if field > max:
max = field
# Set width to '65px * number of months + 100 for headings.'.
chart_url = 'http://chart.apis.google.com/chart?cht=lc&chs=%sx150&chd=t:' % (len(column_headings)*65+100)
first_row = True
row_headings = []
for row in data[1:]:
# Add data to URL, normalised to the maximum for all lines on this chart
norm = normalise_data(row[1:], max)
if not first_row:
chart_url += '|'
chart_url += ','.join([str(num) for num in norm])
row_headings.append(row[0])
first_row = False
chart_url += '&chds='
rows = len(data)-1
first = True
for row in range(rows):
# Set maximum data ranges to '0:x' where 'x' is the maximum number in use.
if not first:
chart_url += ','
chart_url += '0,%s' % max
first = False
chart_url += '&chdl=%s' % '|'.join(row_headings) # Display legend/labels
chart_url += '&chco=%s' % ','.join(chart_colours) # Default colour set
chart_url += '&chxt=x,y' # Turn on axis labels
chart_url += '&chxl=0:|%s|1:|0|%s' % ('|'.join(column_headings), max) # Axis Label Text
return chart_url
def bar_chart(data):
"""
'data' is a list of lists making a table.
Row 1, columns 2-n are data headings
Rows 2-n are data, with column 1 being the line labels
"""
column_headings = data[0][1:]
max = 0
for row in data[1:]:
for field in row[1:]:
if field > max:
max = field
# Set width to '220px * number of months'.
chart_url = 'http://chart.apis.google.com/chart?cht=bvg&chs=%sx150&chd=t:' % (len(column_headings) * 220)
first_row = True
row_headings = []
for row in data[1:]:
# Add data to URL, normalised to the maximum for all lines on this chart
norm = normalise_data(row[1:], max)
if not first_row:
chart_url += '|'
chart_url += ','.join([str(num) for num in norm])
row_headings.append(row[0])
first_row = False
chart_url += '&chds=0,%s' % max
chart_url += '&chdl=%s' % '|'.join(row_headings) # Display legend/labels
chart_url += '&chco=%s' % ','.join(chart_colours) # Default colour set
chart_url += '&chxt=x,y' # Turn on axis labels
chart_url += '&chxl=0:|%s|1:|0|%s' % ('|'.join(column_headings), max) # Axis Label Text
return chart_url
def query_to_dict(results, descriptions):
"""
Replacement method for cursor.dictfetchall() as that method no longer
exists in psycopg2, and I'm guessing in other backends too.
Converts the results of a raw SQL query into a list of dictionaries, suitable
for use in templates etc.
"""
output = []
for data in results:
row = {}
i = 0
for column in descriptions:
row[column[0]] = data[i]
i += 1
output.append(row)
return output
def apply_query(queryset, params):
"""
Apply a dict-based set of filters & paramaters to a queryset.
queryset is a Django queryset, eg MyModel.objects.all() or
MyModel.objects.filter(user=request.user)
params is a dictionary that contains the following:
filtering: A dict of Django ORM filters, eg:
{'user__id__in': [1, 3, 103], 'title__contains': 'foo'}
other_filter: Another filter of some type, most likely a
set of Q() objects.
sorting: The name of the column to sort by
"""
for key in params['filtering'].keys():
filter = {key: params['filtering'][key]}
queryset = queryset.filter(**filter)
if params.get('other_filter', None):
# eg a Q() set
queryset = queryset.filter(params['other_filter'])
if params.get('sorting', None):
if params.get('sortreverse', None):
params['sorting'] = "-%s" % params['sorting']
queryset = queryset.order_by(params['sorting'])
return queryset
def safe_template_context(ticket):
"""
Return a dictionary that can be used as a template context to render
comments and other details with ticket or queue paramaters. Note that
we don't just provide the Ticket & Queue objects to the template as
they could reveal confidential information. Just imagine these two options:
* {{ ticket.queue.email_box_password }}
* {{ ticket.assigned_to.password }}
Ouch!
The downside to this is that if we make changes to the model, we will also
have to update this code. Perhaps we can find a better way in the future.
"""
context = {
'queue': {},
'ticket': {},
}
queue = ticket.queue
for field in ( 'title', 'slug', 'email_address', 'from_address'):
attr = getattr(queue, field, None)
if callable(attr):
context['queue'][field] = attr()
else:
context['queue'][field] = attr
for field in ( 'title', 'created', 'modified', 'submitter_email',
'status', 'get_status_display', 'on_hold', 'description',
'resolution', 'priority', 'get_priority_display',
'last_escalation', 'ticket', 'ticket_for_url',
'get_status', 'ticket_url', 'staff_url', '_get_assigned_to'
):
attr = getattr(ticket, field, None)
if callable(attr):
context['ticket'][field] = '%s' % attr()
else:
context['ticket'][field] = attr
context['ticket']['queue'] = context['queue']
context['ticket']['assigned_to'] = context['ticket']['_get_assigned_to']
return context
def text_is_spam(text, request):
# Based on a blog post by 'sciyoshi':
# http://sciyoshi.com/blog/2008/aug/27/using-akismet-djangos-new-comments-framework/
# This will return 'True' is the given text is deemed to be spam, or
# False if it is not spam. If it cannot be checked for some reason, we
# assume it isn't spam.
from django.contrib.sites.models import Site
from django.conf import settings
try:
from helpdesk.akismet import Akismet
except:
return False
ak = Akismet(
blog_url='http://%s/' % Site.objects.get(pk=settings.SITE_ID).domain,
agent='Jutda Helpdesk',
)
if hasattr(settings, 'TYPEPAD_ANTISPAM_API_KEY'):
ak.setAPIKey(key = settings.TYPEPAD_ANTISPAM_API_KEY)
ak.baseurl = 'api.antispam.typepad.com/1.1/'
elif hasattr(settings, 'AKISMET_API_KEY'):
ak.setAPIKey(key = settings.AKISMET_API_KEY)
else:
return False
if ak.verify_key():
ak_data = {
'user_ip': request.META.get('REMOTE_ADDR', '127.0.0.1'),
'user_agent': request.META.get('HTTP_USER_AGENT', ''),
'referrer': request.META.get('HTTP_REFERER', ''),
'comment_type': 'comment',
'comment_author': '',
}
return ak.comment_check(text, data=ak_data)
return False