implement parsing account request Comment field for peer email addresses in reqacceptpeer.py and add new -E AUTO flag to enable it. Added a few sanity checks to avoid users vouching for themselves and expired or suspended accounts being used as peers in createuser while at it

git-svn-id: svn+ssh://svn.code.sf.net/p/migrid/code/trunk@5092 b75ad72c-e7d7-11dd-a971-7dbc132099af

Author: jonasbardino

mig/server/reqacceptpeer.py

@@ -4,7 +4,7 @@
 # --- BEGIN_HEADER ---
 #
 # reqacceptpeer - Forward account request from peer to existing user(s)
-# Copyright (C) 2003-2020  The MiG Project lead by Brian Vinter
+# Copyright (C) 2003-2021  The MiG Project lead by Brian Vinter
 #
 # This file is part of MiG.
 #
@@ -32,6 +32,7 @@
 or email from Distinguished Name field of employee user entry. If user
 configured additional messaging protocols they can also be used.
 """
+
 from __future__ import print_function
 from __future__ import absolute_import
 
@@ -48,6 +49,7 @@
 from mig.shared.serial import load, dump
 from mig.shared.useradm import init_user_adm, search_users, default_search, \
     user_account_notify
+from mig.shared.validstring import valid_email_addresses
 
 
 def usage(name='reqacceptpeer.py'):
@@ -62,7 +64,8 @@ def usage(name='reqacceptpeer.py'):
    -c CONF_FILE        Use CONF_FILE as server configuration
    -C                  Send a copy of notifications to configured site admins
    -d DB_PATH          Use DB_PATH as user data base file path
-   -e EMAIL            Send instructions to custom email address
+   -e EMAIL            Send instructions to custom EMAIL address
+   -E EMAIL            Forward peer request to user(s) with EMAIL (AUTO to parse Comment)
    -h                  Show this help
    -I CERT_DN          Forward peer request to user(s) with ID (distinguished name)
    -s PROTOCOL         Send instructions to notification protocol from settings
@@ -92,8 +95,9 @@ def usage(name='reqacceptpeer.py'):
     # IMPORTANT: Default to nobody to avoid spam if called without -I CLIENT_ID
     search_filter['distinguished_name'] = ''
     peer_dict = {}
+    regex_keys = []
     exit_code = 0
-    opt_args = 'ac:Cd:e:hI:s:u:v'
+    opt_args = 'ac:Cd:e:E:hI:s:u:v'
     try:
         (opts, args) = getopt.getopt(args, opt_args)
     except getopt.GetoptError as err:
@@ -114,6 +118,11 @@ def usage(name='reqacceptpeer.py'):
         elif opt == '-e':
             raw_targets['email'] = raw_targets.get('email', [])
             raw_targets['email'].append(val)
+        elif opt == '-E':
+            if val != keyword_auto:
+                search_filter['email'] = val.lower()
+            else:
+                search_filter['email'] = keyword_auto
         elif opt == '-h':
             usage()
             sys.exit(0)
@@ -179,24 +188,42 @@ def usage(name='reqacceptpeer.py'):
     fill_distinguished_name(peer_dict)
     peer_id = peer_dict['distinguished_name']
 
+    if search_filter['email'] == keyword_auto:
+        peer_emails = valid_email_addresses(
+            configuration, peer_dict['comment'])
+        if peer_emails[1:]:
+            regex_keys.append('email')
+            search_filter['email'] = '(' + '|'.join(peer_emails) + ')'
+        elif peer_emails:
+            search_filter['email'] = peer_emails[0]
+        elif search_filter['distinguished_name']:
+            search_filter['email'] = '*'
+        else:
+            search_filter['email'] = ''
+
+    # If email is provided or detected DN may be almost anything
+    if search_filter['email'] and not search_filter['distinguished_name']:
+        search_filter['distinguished_name'] = '*emailAddress=*'
+
     if verbose:
-        print('Handling peer %s request to %s' %
-              (peer_id, search_filter['distinguished_name']))
+        print('Handling peer %s request to users matching %s' %
+              (peer_id, search_filter))
 
     # Lookup users to request formal acceptance from
-    (_, hits) = search_users(search_filter, conf_path, db_path, verbose)
+    (_, hits) = search_users(search_filter, conf_path,
+                             db_path, verbose, regex_match=regex_keys)
     logger = configuration.logger
     gdp_prefix = "%s=" % gdp_distinguished_field
 
     if len(hits) < 1:
         print(
             "Aborting attempt to request peer acceptance without target users")
-        print(" ... did you forget or supply too rigid -I CLIENT_ID argument?")
+        print(" ... did you forget or supply too rigid -E EMAIL or -I DN arg?")
         sys.exit(1)
     elif len(hits) > 3:
         print("Aborting attempt to request peer acceptance from %d users!" %
               len(hits))
-        print(" ... did you supply too lax -I CLIENT_ID argument?")
+        print(" ... did you supply too lax -E EMAIL or -I DN argument?")
         sys.exit(1)
     else:
         if verbose:
@@ -213,9 +240,12 @@ def usage(name='reqacceptpeer.py'):
                 print("Skip GDP project account: %s" % user_id)
             continue
 
+        if peer_id == user_id:
+            print("Skip same user account %s as own peer" % user_id)
+            continue
+
         if not peers_permit_allowed(configuration, user_dict):
-            if verbose:
-                print("Skip account %s without vouching permission" % user_id)
+            print("Skip account %s without vouching permission" % user_id)
             continue
 
         if not manage_pending_peers(configuration, user_id, "add",

mig/shared/base.py

@@ -26,6 +26,7 @@
 #
 
 """Base helper functions"""
+
 from __future__ import print_function
 from __future__ import absolute_import
 

mig/shared/useradm.py

@@ -4,7 +4,7 @@
 # --- BEGIN_HEADER ---
 #
 # useradm - user administration functions
-# Copyright (C) 2003-2020  The MiG Project lead by Brian Vinter
+# Copyright (C) 2003-2021  The MiG Project lead by Brian Vinter
 #
 # This file is part of MiG.
 #
@@ -26,6 +26,7 @@
 #
 
 """User administration functions"""
+
 from __future__ import print_function
 from __future__ import absolute_import
 
@@ -65,7 +66,7 @@
 from mig.shared.settings import update_settings, update_profile, update_widgets
 from mig.shared.sharelinks import load_share_links, update_share_link, \
     get_share_link, mode_chars_map
-from mig.shared.validstring import possible_user_id
+from mig.shared.validstring import possible_user_id, valid_email_addresses
 from mig.shared.vgrid import vgrid_add_owners, vgrid_remove_owners, \
     vgrid_add_members, vgrid_remove_members, in_vgrid_share, \
     vgrid_sharelinks, vgrid_add_sharelinks
@@ -226,8 +227,7 @@ def create_user(
         peer_email_list = []
         # extract email of vouchee from comment if possible
         comment = user.get('comment', '')
-        # NOTE: address must end in letter(s) to avoid trailing period, etc.
-        all_matches = re.findall(r'[\w\.-]+@[\w\.-]+[\w]+', comment)
+        all_matches = valid_email_addresses(configuration, comment)
         for i in all_matches:
             peer_email = "%s" % i
             if not possible_user_id(configuration, peer_email):
@@ -257,26 +257,45 @@ def create_user(
         peer_notes = []
         if not hits:
             peer_notes.append("no match for peers")
-        for (user_id, _) in hits:
-            _logger.debug("check %s in peers for %s" % (client_id, user_id))
-            accepted_peers = get_accepted_peers(configuration, user_id)
+        for (sponsor_id, sponsor_dict) in hits:
+            _logger.debug("check %s in peers for %s" % (client_id, sponsor_id))
+            if client_id == sponsor_id:
+                warn_msg = "users cannot vouch for themselves: %s for %s" % \
+                           (client_id, sponsor_id)
+                _logger.warning(warn_msg)
+                continue
+            sponsor_expire = sponsor_dict.get('expire', -1)
+            if sponsor_expire >= 0 and time.time() > sponsor_expire:
+                warn_msg = "expire %s prevents %s as peer for %s" % \
+                           (sponsor_expire, sponsor_id, client_id)
+                _logger.warning(warn_msg)
+                peer_notes.append(warn_msg)
+                continue
+            sponsor_status = sponsor_dict.get('status', 'active')
+            if sponsor_status not in ['active', 'temporal']:
+                warn_msg = "status %s prevents %s as peer for %s" % \
+                           (sponsor_status, sponsor_id, client_id)
+                _logger.warning(warn_msg)
+                peer_notes.append(warn_msg)
+                continue
+            accepted_peers = get_accepted_peers(configuration, sponsor_id)
             peer_entry = accepted_peers.get(client_id, None)
             if not peer_entry:
                 _logger.warning("could not validate %s as peer for %s" %
-                                (user_id, client_id))
+                                (sponsor_id, client_id))
                 continue
             peer_expire = datetime.datetime.strptime(
                 peer_entry.get('expire', 0), '%Y-%m-%d')
-            user_expire = datetime.datetime.fromtimestamp(user['expire'])
-            if peer_expire < user_expire:
+            client_expire = datetime.datetime.fromtimestamp(user['expire'])
+            if peer_expire < client_expire:
                 warn_msg = "expire %s vs %s prevents %s as peer for %s" % \
-                           (peer_expire, user_expire, user_id, client_id)
+                           (peer_expire, client_expire, sponsor_id, client_id)
                 _logger.warning(warn_msg)
                 peer_notes.append(warn_msg)
                 continue
-            _logger.debug("validated %s accepts %s as peer" % (user_id,
+            _logger.debug("validated %s accepts %s as peer" % (sponsor_id,
                                                                client_id))
-            accepted_peer_list.append(user_id)
+            accepted_peer_list.append(sponsor_id)
         if not accepted_peer_list:
             _logger.error("requested peer validation with %r for %s failed" %
                           (verify_pattern, client_id))

mig/shared/validstring.py

@@ -4,7 +4,7 @@
 # --- BEGIN_HEADER ---
 #
 # validstring - string validators
-# Copyright (C) 2003-2018  The MiG Project lead by Brian Vinter
+# Copyright (C) 2003-2021  The MiG Project lead by Brian Vinter
 #
 # This file is part of MiG.
 #
@@ -26,9 +26,11 @@
 #
 
 """String validation"""
+
 from __future__ import absolute_import
 
 import os.path
+import re
 
 from mig.shared.base import invisible_path
 from mig.shared.conf import get_configuration_object
@@ -100,10 +102,28 @@ def is_valid_email_address(addr, logger):
         if addr[c] in rfc822_specials:
             return False
         c += 1
-    logger.debug('%s is a valid email address' % addr)
+    logger.debug('%s is a valid email address: %s' % (addr, count >= 1))
     return count >= 1
 
 
+def valid_email_addresses(configuration, text, lowercase=True):
+    """Extract list of all valid email addresses found in free-form text"""
+    _logger = configuration.logger
+    # NOTE: address must end in letter(s) to avoid trailing period, etc.
+    email_list = []
+    all_matches = re.findall(r'[\w\._-]+@[\w\.-]+[\w]+', text)
+    for i in all_matches:
+        email = "%s" % i
+        if lowercase:
+            email = email.lower()
+        if not is_valid_email_address(email, _logger):
+            _logger.warning('skip invalid email: %s' % email)
+            continue
+        _logger.debug('found valid email: %s' % email)
+        email_list.append(email)
+    return email_list
+
+
 def possible_user_id(configuration, user_id):
     """Check if user_id is a possible user ID based on knowledge about
     contents. We always use email or hexlified version of cert DN.
@@ -115,6 +135,7 @@ def possible_user_id(configuration, user_id):
             return False
     return True
 
+
 def possible_gdp_user_id(configuration, gdp_user_id):
     """Check if gdp_user_id is a possible user ID based on knowledge about
     contents. We always use the format email@project
@@ -129,6 +150,7 @@ def possible_gdp_user_id(configuration, gdp_user_id):
         return False
     return True
 
+
 def possible_job_id(configuration, job_id):
     """Check if job_id is a possible job ID based on knowledge about contents
     and length. We use hexlify on a random string of session_id_bytes, which
@@ -299,3 +321,13 @@ def valid_user_path(configuration, path, home_dir, allow_equal=False,
 
             same = False
         return inside or same
+
+
+if __name__ == "__main__":
+    from mig.shared.conf import get_configuration_object
+    conf = get_configuration_object()
+    comment = """Testing email extract with [email protected] and
+nosuchemail@abc.%!?.com and [email protected]
+with whatever text trailing."""
+    print("Extract email addresses from:\n%r" % comment)
+    print(', '.join(valid_email_addresses(conf, comment)))