CVS: tmda/TMDA Util.py,1.98,1.99 FilterParser.py,1.57,1.58 Defaults.py,1.182,1.183
Timothy Legant <[email protected]>
| Newsgroups | gmane.mail.spam.tmda.cvs |
|---|---|
| Message-ID | <[email protected]> |
Update of /cvsroot/tmda/tmda/TMDA
In directory sc8-pr-cvs1:/tmp/cvs-serv21143/TMDA
Modified Files:
Util.py FilterParser.py Defaults.py
Log Message:
Added support for from-sql/to-sql rules. These rules use Python DB
API 2.0 modules to support various databases with the same code.
Also, fixed two regular expressions in FilterParser that were too
"greedy". The old behavior made it impossible to have two custom
headers with quoted values. The text between the first quote of the
first header and the last quote of the last header was concatenated
into one string.
Index: Util.py
===================================================================
RCS file: /cvsroot/tmda/tmda/TMDA/Util.py,v
retrieving revision 1.98
retrieving revision 1.99
diff -u -r1.98 -r1.99
--- Util.py 9 Jun 2003 02:20:51 -0000 1.98
+++ Util.py 1 Jul 2003 06:29:51 -0000 1.99
@@ -939,8 +939,9 @@
def filter_match(filename, recip, sender=None):
"""Check if the give e-mail addresses match lines in filename."""
+ import Defaults
import FilterParser
- filter = FilterParser.FilterParser()
+ filter = FilterParser.FilterParser(Defaults.DB_CONNECTION)
filter.read(filename)
(actions, matchline) = filter.firstmatch(recip, [sender])
# print the results
Index: FilterParser.py
===================================================================
RCS file: /cvsroot/tmda/tmda/TMDA/FilterParser.py,v
retrieving revision 1.57
retrieving revision 1.58
diff -u -r1.57 -r1.58
--- FilterParser.py 9 Jun 2003 02:20:51 -0000 1.57
+++ FilterParser.py 1 Jul 2003 06:29:51 -0000 1.58
@@ -37,6 +37,7 @@
import time
import types
+import Defaults
import Util
@@ -221,7 +222,7 @@
bol_comment = re.compile(r'\s*#')
most_sources = re.compile(r"""
- ( (?:to|from)-(?:file|cdb|dbm|ezmlm|mailman|mysql)
+ ( (?:to|from)-(?:file|cdb|dbm|ezmlm|mailman|mysql|sql)
| size | pipe
| (?:to|from) (?!-) )
""", re.VERBOSE | re.IGNORECASE)
@@ -232,14 +233,14 @@
""", re.VERBOSE | re.IGNORECASE)
matches = re.compile(r"""
- (?: ([\'\"]) ( (?: \\\1 | [^\1] )+ ) \1
+ (?: ([\'\"]) ( (?: \\\1 | . )+? ) \1
| ( \S+ ) )
""", re.VERBOSE)
tag_action = re.compile(r"""
( [A-Za-z][-\w]+ )
\s+
- (?: ([\'\"]) ( (?: \\\2 | [^\2] )+ ) \2
+ (?: ([\'\"]) ( (?: \\\2 | . )+? ) \2
| ( \S+ ) )
""", re.VERBOSE)
@@ -273,6 +274,10 @@
'to-mailman' : ('attr', 'optional' ),
'from-mysql' : ('like', 'rlike'),
'to-mysql' : ('like', 'rlike'),
+ 'from-sql' : ('action_column', 'addr_column',
+ 'domains', 'wildcards'),
+ 'to-sql' : ('action_column', 'addr_column',
+ 'domains', 'wildcards'),
'body' : ('case',),
'headers' : ('case',),
'body-file' : ('case', 'optional'),
@@ -285,7 +290,8 @@
MySQL = None
- def __init__(self):
+ def __init__(self, db_instance=None):
+ self.db_instance = db_instance
self.macros = []
self.files = []
self.filterlist = []
@@ -496,7 +502,6 @@
def __findvarsub(self, var):
"""Look up 'var' in the Defaults namespace and the environment."""
# First, check the Defaults namespace.
- import Defaults
sub = Defaults.__dict__.get(var, None)
# Then, try the environment.
if not sub:
@@ -686,22 +691,18 @@
return actions
- def __search_file(self, pathname, keys, actions, source):
- """
- Search a text file for match in first column.
- """
- found_match = 0
- match_list = Util.file_to_list(pathname)
+ def __search_list(self, addrlist, keys, actions, source):
+ """Search addrlist for match in field 1, optional action in 2."""
# This list comprehension splits each line in the list into
# two columns, lowercases the first column and stuffs the
# columns back together. The result is the original list of
# lines from the file with the first field in each line
# lowercased.
- match_list = [' '.join(
+ addrlist = [' '.join(
apply(lambda f1, f2=None: f2 and [f1.lower(), f2]
or [f1.lower()],
- line.split(None, 1))) for line in match_list]
- found_match = Util.findmatch(match_list, keys)
+ line.split(None, 1))) for line in addrlist]
+ found_match = Util.findmatch(addrlist, keys)
if found_match:
# The second column of the line may contain an
# overriding action specification.
@@ -713,11 +714,21 @@
return found_match
+ def __search_file(self, pathname, keys, actions, source):
+ """
+ Search a text file for match in first column.
+ """
+ return self.__search_list(Util.file_to_list(pathname),
+ keys,
+ actions,
+ source)
+
+
def __search_mysql(self, Table, Args, Keys, Actions, Source):
"Search MySQL table."
if not self.MySQL:
# Connect to the database if we have not yet connected
- import Defaults, _mysql
+ import _mysql
self.MySQL = _mysql.connect \
(
host = Defaults.MYSQL_HOST,
@@ -838,7 +849,75 @@
except IndexError:
pass
return domains
-
+
+
+ def __create_sql_params(self, dbkeys):
+ """Return dictionary of parameters for sql statement."""
+ params = { 'recipient': Defaults.USERNAME+'@'+Defaults.HOSTNAME,
+ 'username' : Defaults.USERNAME,
+ 'hostname' : Defaults.HOSTNAME
+ }
+ for i in range(len(dbkeys)):
+ params['criterion'+str(i)] = dbkeys[i]
+ return params
+
+
+ def __create_sql_criteria(self, dbkeys, addresscolumn):
+ """Return condition string for insertion into SQL statement."""
+ if not dbkeys:
+ return ''
+ criteria = "("
+ for i in range(len(dbkeys)):
+ if i > 0:
+ criteria += " OR "
+ criteria += "%s = %%(criterion%d)s" % (addresscolumn, i)
+ criteria += ")"
+ return criteria
+
+
+ def __get_column_index(self, colname, cursor):
+ """Return index of column named 'colname'."""
+ for i in range(len(cursor.description)):
+ if colname == cursor.description[i][0]:
+ return i
+ return -1
+
+
+ def __search_sql(self, selectstmt, args, keys, actions, source, lineno):
+ """Search SQL DB (Python DB API 2.0)."""
+ found_match = 0
+ dbkeys = keys
+ if args.has_key('wildcards'):
+ dbkeys = []
+ params = self.__create_sql_params(dbkeys)
+ cursor = self.db_instance.cursor()
+ cursor.execute(selectstmt, params)
+ rows = cursor.fetchall()
+ if cursor.rowcount <= 0:
+ return 0
+ if args.has_key('wildcards'):
+ if len(cursor.description) > 1:
+ dblist = [' '.join([row[0], row[1] or '']) for row in rows]
+ else:
+ dblist = [row[0] for row in rows]
+ found_match = self.__search_list(dblist, keys, actions, source)
+ else:
+ action_column = args.get('action_column')
+ if action_column:
+ actcolidx = self.__get_column_index(action_column, cursor)
+ if actcolidx == -1:
+ actcolidx = self.__get_column_index(action_column.lower(),
+ cursor)
+ if actcolidx == -1:
+ err = "no action column (%s)" % (action_column,)
+ raise MatchError(lineno, err)
+ action = rows[0][actcolidx]
+ if action:
+ actions.clear()
+ actions.update(self.__buildactions(action, source))
+ found_match = 1
+ return found_match
+
def firstmatch(self, recipient, senders=None,
msg_body=None, msg_headers=None, msg_size=None):
@@ -988,6 +1067,29 @@
)
if found_match:
break
+ # Generic SQL. Expects a SELECT statement as the 'match' field.
+ # There are two "modes", depending on the presence of TMDA-style
+ # wildcards in the database. See the filter source documentation
+ # for more information.
+ if source in ('from-sql', 'to-sql'):
+ selectstmt = match
+ if args.has_key('domains'):
+ keys += self.__extract_domains(keys)
+ addr_column = args.get('addr_column')
+ if args.has_key('wildcards'):
+ if addr_column:
+ raise MatchError(lineno,
+ "-addr_column and -wildcards " +
+ "cannot be used together")
+ elif not addr_column:
+ raise MatchError(lineno, "-addr_column must be specified")
+ else:
+ criteria = self.__create_sql_criteria(keys, addr_column)
+ selectstmt = selectstmt.replace('%(criteria)s', criteria)
+ found_match = self.__search_sql(
+ selectstmt, args, keys, actions, source, lineno)
+ if found_match:
+ break
# A match is found if the command exits with a zero exit
# status.
if source == 'pipe' and msg_body and msg_headers:
Index: Defaults.py
===================================================================
RCS file: /cvsroot/tmda/tmda/TMDA/Defaults.py,v
retrieving revision 1.182
retrieving revision 1.183
diff -u -r1.182 -r1.183
--- Defaults.py 5 Jun 2003 14:04:29 -0000 1.182
+++ Defaults.py 1 Jul 2003 06:29:51 -0000 1.183
@@ -1088,6 +1088,20 @@
if MYSQL_ENABLED and not vars().has_key('MYSQL_PASSWORD'):
MYSQL_PASSWORD = ""
+# DB_CONNECTION
+# Python DB API Connection object. This is specific to the database
+# and should be created in the ~/.tmda/config file. This is typically
+# created by importing the appropriate module and calling the connect()
+# function.
+#
+# Example:
+# import MySQLdb
+# DB_CONNECTION = MySQLdb.connect("...")
+#
+# Default is None
+if not vars().has_key('DB_CONNECTION'):
+ DB_CONNECTION = None
+
# PENDING_DIR
# Full path to the directory containing messages pending confirmation
# (aka, the "pending queue"). If this directory doesn't exist, it
_______________________________________
tmda-cvs mailing list
http://tmda.net/lists/listinfo/tmda-cvs