Merged [5902],[5904] from markus.stenberg@iki.fi:

Added support and scripts for generation of wg- and draft-aliases.  Fixes issue #713.
 - Legacy-Id: 5937
Note: SVN reference [5902] has been migrated to Git commit e5b551f4ce

Note: SVN reference [5904] has been migrated to Git commit e68e51cc55

ietf/bin/generate-draft-aliases

@@ -0,0 +1,147 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+# -*- Python -*-
+#
+# $Id: generate-draft-aliases $
+#
+# Author: Markus Stenberg <markus.stenberg@iki.fi>
+#
+"""
+
+This code dumps Django model InternetDraft's contents as postfix email
+aliases
+
+<no suffix> (same as -authors)
+.authors (list of authors)
+.chairs (WG chairs)
+.notify (notify emails(?))
+.ad (sponsoring AD)
+.all (all of the above)
+
+TODO:
+
+- results somewhat inconsistent with the results from the old tool;
+  should examine why (ask me for diff tool if interested in fixing it)
+
+"""
+
+DRAFT_EMAIL_SUFFIX='@tools.ietf.org'
+
+# boilerplate (from various other ietf/bin scripts)
+import os, sys
+basedir = os.path.abspath(os.path.join(os.path.dirname(__file__), "../.."))
+sys.path = [ basedir ] + sys.path
+
+from ietf import settings
+from django.core import management
+management.setup_environ(settings)
+
+from ietf.doc.models import Document
+from ietf.group.utils import get_group_chairs_emails, get_group_ads_emails
+from ietf.utils.aliases import *
+import time
+
+def get_draft_ad_emails(draft):
+    " Get AD email for the given draft, if any. "
+    # If working group document, return current WG ADs
+    wg = draft.group
+    if wg and wg.acronym != 'none' and wg.parent and wg.parent.acronym != 'none':
+        return get_group_ads_emails(wg)
+    # If not, return explicit AD set (whether up to date or not)
+    ad = draft.ad
+    #return [ad and ad.user and ad.user.email]
+    return [ad and ad.email_address()]
+
+def get_draft_authors_emails(draft):
+    " Get list of authors for the given draft."
+
+    # This feels 'correct'; however, it creates fairly large delta
+    return [email.email_address() for email in draft.authors.all()]
+
+    # This gives fairly small delta compared to current state,
+    # however, it seems to be wrong (doesn't check for emails being
+    # active etc).
+    #return [email.address for email in draft.authors.all()]
+
+def get_draft_notify_emails(draft):
+    " Get list of email addresses to notify for the given draft."
+    n = draft.notify
+    if not n:
+        return
+    l = []
+    draft_email = draft.name + DRAFT_EMAIL_SUFFIX
+    for e in n.split(','):
+        # If the draft name itself is listed as notify list element, we
+        # expand it (to make results better verifiable with the old ones)
+        e = e.strip()
+        if e == draft_email:
+            l.extend(get_draft_authors_emails(draft))
+        else:
+            l.append(e)
+    # Alternative: if we don't want to do expansion, just this would be
+    # perhaps better (MTA can do expansion too):
+    # l = n.split(',')
+    return l
+
+if __name__ == '__main__':
+    import datetime
+    import time
+
+    # Year ago?
+    show_since = datetime.datetime.now() - datetime.timedelta(365)
+    # 10 years ago?
+    #show_since = datetime.datetime.now() - datetime.timedelta(10 * 365)
+
+    modname = 'ietf.generate_draft_aliases'
+    date = time.strftime("%Y-%m-%d_%H:%M:%S")
+    print '# Generated by python -m %s at %s' % (modname, date)
+
+    drafts = Document.objects.all()
+
+    # Drafts with active status
+    active_drafts = drafts.filter(states__slug='active')
+
+    # Drafts that expired within year
+    inactive_recent_drafts = drafts.exclude(states__slug='active').filter(expires__gte=show_since)
+
+    interesting_drafts = active_drafts | inactive_recent_drafts
+
+    for draft in interesting_drafts.distinct().iterator():
+        # Omit RFCs, we care only about drafts
+        if draft.docalias_set.filter(name__startswith='rfc'):
+            continue
+        name = draft.name
+        done = []
+        all = []
+        def handle_sublist(name, f, o, is_ad=False):
+            r = dump_sublist(name, f, o, is_ad)
+            if r:
+                done.append(name)
+                all.extend(r)
+            return r
+        #.authors (/and no suffix) = authors
+
+        # First, do no suffix case
+        # If no authors, don't generate list either
+        r = dump_sublist(name, get_draft_authors_emails, draft)
+        if not r:
+            continue
+        handle_sublist('%s%s' % (name, '.authors'), get_draft_authors_emails, draft)
+        wg = draft.group
+
+        if wg:
+            # .chairs = WG chairs
+            handle_sublist('%s%s' % (name, '.chairs'), get_group_chairs_emails, wg)
+
+        # .ad = sponsoring AD / WG AD (WG document)
+        handle_sublist('%s%s' % (name, '.ad'), get_draft_ad_emails, draft, True)
+
+        # .notify = notify email list from the Document
+        handle_sublist('%s%s' % (name, '.notify'), get_draft_notify_emails, draft)
+
+        # .all = everything on 'done' (recursive aliases)
+        #dump_sublist('%s%s' % (name, '.all'), None, done)
+        # .all = everything on 'all' (expanded aliases)
+        dump_sublist('%s%s' % (name, '.all'), None, all)
+
+

ietf/bin/generate-wg-aliases

@@ -0,0 +1,91 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+# -*- Python -*-
+#
+# $Id: generate-wg-aliases $
+#
+# Author: Markus Stenberg <markus.stenberg@iki.fi>
+#
+"""
+
+This code dumps Django model IETFWG's contents as two sets of postfix
+mail lists: -ads, and -chairs
+
+"""
+
+# boilerplate (from various other ietf/bin scripts)
+import os, sys
+basedir = os.path.abspath(os.path.join(os.path.dirname(__file__), "../.."))
+sys.path = [ basedir ] + sys.path
+
+from ietf import settings
+from django.core import management
+management.setup_environ(settings)
+
+from ietf.group.models import Group
+from ietf.group.utils import get_group_ads_emails, get_group_chairs_emails, get_area_ads_emails, get_area_chairs_emails
+from ietf.person.models import Email
+from ietf.utils.aliases import *
+
+# from secr/utils/group.py..
+ACTIVE_STATES=['active', 'bof', 'proposed']
+
+if __name__ == '__main__':
+    import datetime
+    import time
+
+    # Year ago?
+    #show_since = datetime.datetime.now() - datetime.timedelta(365)
+
+    # 2 years ago?
+    #show_since = datetime.datetime.now() - datetime.timedelta(2 * 365)
+
+    # 3 years ago?
+    #show_since = datetime.datetime.now() - datetime.timedelta(3 * 365)
+
+    # 5 years ago?
+    show_since = datetime.datetime.now() - datetime.timedelta(5 * 365)
+
+    modname = 'ietf.generate_wg_aliases'
+    date = time.strftime("%Y-%m-%d_%H:%M:%S")
+    print '# Generated by python -m %s at %s' % (modname, date)
+    wgs = Group.objects.filter(type='wg').all()
+
+    print '# WGs'
+    # - status = Active
+    active_wgs = wgs.filter(state__in=ACTIVE_STATES)
+
+    # - activity within last year? (use concluded_date)
+    inactive_recent_wgs = wgs.exclude(state__in=ACTIVE_STATES).filter(time__gte=show_since)
+    interesting_wgs = active_wgs | inactive_recent_wgs
+
+    for wg in interesting_wgs.distinct().iterator():
+        name = wg.acronym
+        dump_sublist('%s%s' % (name, '-ads'), get_group_ads_emails, wg, True)
+        dump_sublist('%s%s' % (name, '-chairs'), get_group_chairs_emails, wg)
+
+    print '# RGs'
+    # - status = Active
+    rgs = Group.objects.filter(type='rg').all()
+    active_rgs = rgs.filter(state__in=ACTIVE_STATES)
+
+    # - activity within last year? (use concluded_date)
+    inactive_recent_rgs = rgs.exclude(state__in=ACTIVE_STATES).filter(time__gte=show_since)
+    interesting_rgs = active_rgs | inactive_recent_rgs
+
+    for rg in interesting_rgs.distinct().iterator():
+        name = rg.acronym
+        #dump_sublist('%s%s' % (name, '-ads'), get_group_ads_emails, rg, True)
+        dump_sublist('%s%s' % (name, '-chairs'), get_group_chairs_emails, rg)
+
+    # Additionally, for areaz, we should list -ads and -chairs
+    # (for every chair in active groups within the area).
+    print '# Areas'
+    areas = Group.objects.filter(type='area').all()
+    active_areas = areas.filter(state__in=ACTIVE_STATES)
+    for area in active_areas:
+        name = area.acronym
+        dump_sublist('%s%s' % (name, '-ads'), get_area_ads_emails, area, True)
+        dump_sublist('%s%s' % (name, '-chairs'), get_area_chairs_emails, area)
+
+

ietf/group/utils.py

@@ -45,10 +45,50 @@ def get_charter_text(group):
             filename = os.path.join(settings.IETFWG_DESCRIPTIONS_PATH, group.acronym) + ".desc.txt"
             desc_file = open(filename)
             desc = desc_file.read()
-        except BaseException:    
+        except BaseException:
             desc = 'Error Loading Work Group Description'
         return desc
 
+def get_area_ads_emails(area):
+    if area.acronym == 'none':
+        return []
+    emails = [r.email.email_address()
+              for r in area.role_set.filter(name__in=('ad', 'chair'))]
+    return filter(None, emails)
+
+def get_group_ads_emails(wg):
+    " Get list of area directors' emails for a given WG "
+    if wg.acronym == 'none':
+        return []
+
+    if wg.parent and wg.parent.acronym != 'none':
+        # By default, we should use _current_ list of ads!
+        return get_area_ads_emails(wg.parent)
+
+    # As fallback, just return the single ad within the wg
+    return [wg.ad and wg.ad.email_address()]
+
+def get_group_chairs_emails(wg):
+    " Get list of area chairs' emails for a given WG "
+    if wg.acronym == 'none':
+        return []
+    emails = Email.objects.filter(role__group=wg,
+                                  role__name='chair')
+    if not emails:
+        return
+    emails = [e.email_address() for e in emails]
+    emails = filter(None, emails)
+    return emails
+
+def get_area_chairs_emails(area):
+    emails = {}
+    # XXX - should we filter these by validity? Or not?
+    wgs = Group.objects.filter(parent=area, type="wg", state="active")
+    for wg in wgs:
+        for e in get_group_chairs_emails(wg):
+            emails[e] = True
+    return emails.keys()
+
 def save_milestone_in_history(milestone):
     h = get_history_object_for(milestone)
     h.milestone = milestone

ietf/person/models.py

@@ -73,14 +73,14 @@ class PersonInfo(models.Model):
 
 class Person(PersonInfo):
     user = models.OneToOneField(User, blank=True, null=True)
-    
+
     def person(self): # little temporary wrapper to help porting to new schema
         return self
 
 class PersonHistory(PersonInfo):
     person = models.ForeignKey(Person, related_name="history_set")
     user = models.ForeignKey(User, blank=True, null=True)
-    
+
 class Alias(models.Model):
     """This is used for alternative forms of a name.  This is the
     primary lookup point for names, and should always contain the
@@ -115,3 +115,14 @@ class Email(models.Model):
     def invalid_address(self):
         # we have some legacy authors with unknown email addresses
         return self.address.startswith("unknown-email") and "@" not in self.address
+
+    def email_address(self):
+        """Get valid, current email address; in practise, for active,
+        non-invalid addresses it is just the address field. In other
+        cases, we default to person's email address."""
+        if self.invalid_address() or not self.active:
+            if self.person:
+                return self.person.email_address()
+            return
+        return self.address
+

ietf/utils/aliases.py

@@ -0,0 +1,69 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+# -*- Python -*-
+#
+# $Id: aliasutil.py $
+#
+# Author: Markus Stenberg <mstenber@cisco.com>
+#
+"""
+
+Mailing list alias dumping utilities
+
+"""
+
+def rewrite_email_address(email, is_ad):
+    """ Prettify the email address (and if it's empty, skip it by
+    returning None). """
+    if not email:
+        return
+    email = email.strip()
+    if not email:
+        return
+    if email[0]=='<' and email[-1] == '>':
+        email = email[1:-1]
+    # If it doesn't look like email, skip
+    if '@' not in email and '?' not in email:
+        return
+    return email
+
+def rewrite_address_list(l):
+    """ This utility function makes sure there is exactly one instance
+    of an address within the result list, and preserves order
+    (although it may not be relevant to start with) """
+    h = {}
+    for address in l:
+        #address = address.strip()
+        if h.has_key(address): continue
+        h[address] = True
+        yield address
+
+def dump_sublist(alias, f, wg, is_adlist=False):
+    if f:
+        l = f(wg)
+    else:
+        l = wg
+    if not l:
+        return
+    # Nones in the list should be skipped
+    l = filter(None, l)
+
+    # Make sure emails are sane and eliminate the Nones again for
+    # non-sane ones
+    l = [rewrite_email_address(e, is_adlist) for e in l]
+    l = filter(None, l)
+
+    # And we'll eliminate the duplicates too but preserve order
+    l = list(rewrite_address_list(l))
+    if not l:
+        return
+    try:
+        print '%s: %s' % (alias, ', '.join(l))
+    except UnicodeEncodeError:
+        # If there's unicode in email address, something is badly
+        # wrong and we just silently punt
+        # XXX - is there better approach?
+        print '# Error encoding', alias, repr(l)
+        return
+    return l
+