Merge branch 'develop'

Author: manfred-kaiser

README.md

@@ -1,9 +1,9 @@
 # Sifter3 - Sieve email filter (RFC 5228)
 
-Sifter3 is a Python 3 implementation of the Sieve email filter language (RFC 5228) and is based on the Python 2 version from <https://github.com/garyp/sifter>
+Sifter3 is a Python 3 implementation of the Sieve email filter language (RFC 5228)
 
 ![Python package](https://github.com/manfred-kaiser/sifter3/workflows/Python%20package/badge.svg)
-[![Documentation Status](https://readthedocs.org/projects/sifter3/badge/?version=master)](https://sifter3.readthedocs.io/de/master/?badge=master)
+[![Documentation Status](https://readthedocs.org/projects/sifter3/badge/?version=latest)](https://sifter3.readthedocs.io/en/latest/?badge=latest)
 [![CodeFactor](https://www.codefactor.io/repository/github/manfred-kaiser/sifter3/badge)](https://www.codefactor.io/repository/github/manfred-kaiser/sifter3)
 [![Github version](https://img.shields.io/github/v/release/manfred-kaiser/sifter3?label=github&logo=github)](https://github.com/manfred-kaiser/sifter3/releases)
 [![PyPI version](https://img.shields.io/pypi/v/sifter3.svg?logo=pypi&logoColor=FFE873)](https://pypi.org/project/sifter3/)
@@ -18,6 +18,7 @@ FEATURES
 
 -   Supports all of the base Sieve spec from RFC 5228, except for
     features still listed under TODO below
+    - multiline strings (since version 0.2.2)
 -   Extensions supported:
     -   regex (draft-ietf-sieve-regex-01)
     -   body (RFC 5173)
@@ -45,6 +46,15 @@ email message. Each action is a tuple consisting of the action name and
 action-specific arguments. It is up to the caller to manipulate the
 message and message store based on the actions returned.
 
+COMMAND LINE
+============
+
+The output of the command line tool can be parsed as json.
+
+    $ sifter tests/evaluation_1.rules tests/evaluation_1.msg
+    [['redirect', '[email protected]']]
+
+
 WARNINGS
 ========
 

docs/index.rst

@@ -9,6 +9,9 @@ FEATURES
 
 -   Supports all of the base Sieve spec from RFC 5228, except for
     features still listed under TODO below
+
+    - multiline strings (since version 0.2.2)
+
 -   Extensions supported:
 
     -   regex (draft-ietf-sieve-regex-01)
@@ -45,6 +48,17 @@ consisting of the action name and action-specific arguments. It is up to
 the caller to manipulate the message and message store based on the
 actions returned.
 
+COMMAND LINE
+------------
+
+The output of the command line tool can be parsed as json.
+
+.. code-block:: bash
+
+    $ sifter tests/evaluation_1.rules tests/evaluation_1.msg
+    [['redirect', '[email protected]']]
+
+
 WARNINGS
 --------
 
@@ -60,8 +74,8 @@ TODO
 -   An example adaptor that provides Unix LDA behavior using sieve for
     filtering
 -   Base spec features not yet implemented:
+
     -   encoded characters (section 2.4.2.4)
-    -   multi-line strings (section 2.4.2)
     -   bracketed comments (section 2.3)
     -   message uniqueness (section 2.10.3)
     -   envelope test (section 5.4)

setup.py

@@ -10,7 +10,7 @@
 
 setup(
     name="sifter3",
-    version="0.2.1",
+    version="0.2.2",
     author="Manfred Kaiser, Gary Peck",
     author_email="[email protected], [email protected]",
     url="https://sifter3.readthedocs.io/en/latest/",
@@ -60,6 +60,7 @@
             'keep = sifter.commands.keep:CommandKeep',
             'notify = sifter.commands.notify:CommandNotify',
             'redirect = sifter.commands.redirect:CommandRedirect',
+            'reject = sifter.commands.reject:CommandReject',
             'require = sifter.commands.require:CommandRequire',
             'set = sifter.commands.variables:CommandSet',
             'stop = sifter.commands.stop:CommandStop',

sifter/cli.py

@@ -3,6 +3,7 @@
 import os
 import logging
 import sys
+import json
 
 import sifter.parser
 
@@ -24,4 +25,4 @@ def main() -> None:
     rules = sifter.parser.parse_file(open(args.rulefile))
     msg = email.message_from_file(open(args.messagefile))
     msg_actions = rules.evaluate(msg)
-    print(msg_actions)
+    print(json.dumps(msg_actions, indent=4))

sifter/commands/reject.py

@@ -0,0 +1,23 @@
+from email.message import Message
+from typing import (
+    Text,
+    Optional
+)
+
+from sifter.grammar.command import Command
+from sifter.validators.stringlist import StringList
+from sifter.grammar.state import EvaluationState
+from sifter.grammar.actions import Actions
+
+
+# section 3.2
+class CommandReject(Command):
+
+    RULE_IDENTIFIER: Text = 'REJECT'
+    POSITIONAL_ARGS = [StringList()]
+
+    def evaluate(self, message: Message, state: EvaluationState) -> Optional[Actions]:
+        print(self.positional_args[0][0])
+        state.actions.append('reject', self.positional_args[0][0])
+        state.actions.cancel_implicit_keep()
+        return None

sifter/extensions/__init__.py

@@ -34,14 +34,15 @@ class ExtensionRegistry():
         ]
     ] = {}
     DEFAULT_EXTENSION: List[Text] = [
-        'regex',
+        'body',
         'comparator-i;ascii-casemap',
         'comparator-i;octet',
+        'enotify',
         'fileinto',
-        'body'
+        'imap4flags',
+        'regex',
+        'reject',
         'variables',
-        'enotify',
-        'imap4flags'
     ]
 
     def __init__(self) -> None:

sifter/grammar/grammar.py

@@ -32,7 +32,7 @@ def __init__(self, debug: bool = False) -> None:
     def make_parser(mod: Any, debug: bool = False) -> 'LRParser':
         return yacc(
             module=mod,
-            debug=debug,
+            debug=True,
             write_tables=False,
             errorlog=NullLogger() if not debug else None
         )
@@ -167,6 +167,10 @@ def p_argument_tag(self, p: 'YaccProduction') -> None:
         """argument : TAG"""
         p[0] = Tag(p[1])
 
+    def p_argument_multiline_string(self, p: 'YaccProduction') -> None:
+        """argument : MULTILINE_STRING"""
+        p[0] = [p[1]]
+
     def p_stringlist_error(self, p: 'YaccProduction') -> None:
         """argument : '[' error ']'"""
         print("Syntax error in string list that starts on line %d" % p.lineno(1))
@@ -186,4 +190,9 @@ def p_string(self, p: 'YaccProduction') -> None:
         p[0] = String(p[1])
 
     def p_error(self, p: 'YaccProduction') -> None:
-        pass
+        if p is None:
+            token = "end of file"
+        else:
+            token = f"{p.type}({p.value}) on line {p.lineno}"
+
+        print(f"Syntax error: Unexpected {token}")

sifter/grammar/lexer.py

@@ -54,7 +54,7 @@ def t_BRACKET_COMMENT(self, t: 'LexToken') -> Optional['LexToken']:
 
     # section 2.4.2
     def t_MULTILINE_STRING(self, t: 'LexToken') -> Optional['LexToken']:
-        r'"@@@@@@@@@@@@@@@"'
+        r'text:\s?(?:\#.*)\r?\n(?P<multilinetext>[\r\n\S\s.]*?\r?\n)\.\r?\n'
         # TODO: For entering larger amounts of text, such as an email message,
         # a multi-line form is allowed.  It starts with the keyword "text:",
         # followed by a CRLF, and ends with the sequence of a CRLF, a single
@@ -68,7 +68,8 @@ def t_MULTILINE_STRING(self, t: 'LexToken') -> Optional['LexToken']:
         # that is, ".foo" is interpreted as ".foo".  However, because this is
         # potentially ambiguous, scripts SHOULD be properly dot-stuffed so such
         # lines do not appear.
-        return None
+        t.value = t.lexer.lexmatch.group('multilinetext')
+        return t
 
     # section 2.4.2
     def t_QUOTED_STRING(self, t: 'LexToken') -> Optional['LexToken']:

sifter/tests/body.py

@@ -65,26 +65,26 @@ def evaluate(self, message: Message, state: EvaluationState) -> Optional[bool]:
             #     if there are no headers, we match ^\n, which is guaranteed to be there
             (_, bodystr) = re.split(r'^\r?\n|\r?\n\r?\n', message.as_string(False), 1)
             return self.evaluate_part(bodystr, state)
-        else:
-            for msgpart in message.walk():
-                if msgpart.is_multipart():
-                    # TODO: If "multipart/*" extract prologue and epilogue and make that searcheable
-                    # TODO: If "message/rfc822" extract headers and make that searchable
-                    # Insetad we skip multipart objects and descend into its children
-                    continue
-                msgtxt = msgpart.get_payload()
-                for mimetype in self.body_transform:
-                    if not mimetype:  # empty body_transform matches all
-                        if self.evaluate_part(msgtxt, state):
-                            return True
-                    match = re.match(r'^([^/]+)(?:/([^/]+))?$', mimetype)
-                    if not match:
-                        continue  # malformed body_transform is skipped
-                    (maintype, subtype) = match.groups()
-                    if maintype == msgpart.get_content_maintype() and (
-                            not subtype or subtype == msgpart.get_content_subtype()):
-                        if self.evaluate_part(msgtxt, state):
-                            return True
+
+        for msgpart in message.walk():
+            if msgpart.is_multipart():
+                # TODO: If "multipart/*" extract prologue and epilogue and make that searcheable
+                # TODO: If "message/rfc822" extract headers and make that searchable
+                # Insetad we skip multipart objects and descend into its children
+                continue
+            msgtxt = msgpart.get_payload()
+            for mimetype in self.body_transform:
+                if not mimetype:  # empty body_transform matches all
+                    if self.evaluate_part(msgtxt, state):
+                        return True
+                match = re.match(r'^([^/]+)(?:/([^/]+))?$', mimetype)
+                if not match:
+                    continue  # malformed body_transform is skipped
+                (maintype, subtype) = match.groups()
+                if maintype == msgpart.get_content_maintype() and (
+                        not subtype or subtype == msgpart.get_content_subtype()):
+                    if self.evaluate_part(msgtxt, state):
+                        return True
         return False
 
     def evaluate_part(self, part_str: Text, state: EvaluationState) -> bool:

sifter/validators/tag.py

@@ -125,6 +125,5 @@ def validate(
                 raise ValueError('unexpected return value from StringList.validate')
             if content_args > 0:
                 return validated_args + content_args
-            else:
-                raise RuleSyntaxError("body :content requires argument")
+            raise RuleSyntaxError("body :content requires argument")
         return validated_args