bpo-35805: Add parser for Message-ID email header. (GH-13397)

* bpo-35805: Add parser for Message-ID header.

This parser is based on the definition of Identification Fields from RFC 5322
Sec 3.6.4.

This should also prevent folding of Message-ID header using RFC 2047 encoded
words and hence fix bpo-35805.

* Prevent folding of non-ascii message-id headers.
* Add fold method to MsgID token to prevent folding.

2 files changed, 135 insertions, 15 deletions

diff --git a/Lib/email/_header_value_parser.py b/Lib/email/_header_value_parser.py
index 14cc00c..34969ab 100644
--- a/Lib/email/_header_value_parser.py
+++ b/Lib/email/_header_value_parser.py
@@ -179,37 +179,30 @@ class WhiteSpaceTokenList(TokenList):
 
 
 class UnstructuredTokenList(TokenList):
-
     token_type = 'unstructured'
 
 
 class Phrase(TokenList):
-
     token_type = 'phrase'
 
 class Word(TokenList):
-
     token_type = 'word'
 
 
 class CFWSList(WhiteSpaceTokenList):
-
     token_type = 'cfws'
 
 
 class Atom(TokenList):
-
     token_type = 'atom'
 
 
 class Token(TokenList):
-
     token_type = 'token'
     encode_as_ew = False
 
 
 class EncodedWord(TokenList):
-
     token_type = 'encoded-word'
     cte = None
     charset = None
@@ -496,16 +489,19 @@ class Domain(TokenList):
 
 
 class DotAtom(TokenList):
-
     token_type = 'dot-atom'
 
 
 class DotAtomText(TokenList):
-
     token_type = 'dot-atom-text'
     as_ew_allowed = True
 
 
+class NoFoldLiteral(TokenList):
+    token_type = 'no-fold-literal'
+    as_ew_allowed = False
+
+
 class AddrSpec(TokenList):
 
     token_type = 'addr-spec'
@@ -809,7 +805,6 @@ class ParameterizedHeaderValue(TokenList):
 
 
 class ContentType(ParameterizedHeaderValue):
-
     token_type = 'content-type'
     as_ew_allowed = False
     maintype = 'text'
@@ -817,27 +812,35 @@ class ContentType(ParameterizedHeaderValue):
 
 
 class ContentDisposition(ParameterizedHeaderValue):
-
     token_type = 'content-disposition'
     as_ew_allowed = False
     content_disposition = None
 
 
 class ContentTransferEncoding(TokenList):
-
     token_type = 'content-transfer-encoding'
     as_ew_allowed = False
     cte = '7bit'
 
 
 class HeaderLabel(TokenList):
-
     token_type = 'header-label'
     as_ew_allowed = False
 
 
-class Header(TokenList):
+class MsgID(TokenList):
+    token_type = 'msg-id'
+    as_ew_allowed = False
+
+    def fold(self, policy):
+        # message-id tokens may not be folded.
+        return str(self) + policy.linesep
+
+class MessageID(MsgID):
+    token_type = 'message-id'
 
+
+class Header(TokenList):
     token_type = 'header'
 
 
@@ -1583,7 +1586,7 @@ def get_addr_spec(value):
     addr_spec.append(token)
     if not value or value[0] != '@':
         addr_spec.defects.append(errors.InvalidHeaderDefect(
-            "add-spec local part with no domain"))
+            "addr-spec local part with no domain"))
         return addr_spec, value
     addr_spec.append(ValueTerminal('@', 'address-at-symbol'))
     token, value = get_domain(value[1:])
@@ -1968,6 +1971,110 @@ def get_address_list(value):
             value = value[1:]
     return address_list, value
 
+
+def get_no_fold_literal(value):
+    """ no-fold-literal = "[" *dtext "]"
+    """
+    no_fold_literal = NoFoldLiteral()
+    if not value:
+        raise errors.HeaderParseError(
+            "expected no-fold-literal but found '{}'".format(value))
+    if value[0] != '[':
+        raise errors.HeaderParseError(
+            "expected '[' at the start of no-fold-literal "
+            "but found '{}'".format(value))
+    no_fold_literal.append(ValueTerminal('[', 'no-fold-literal-start'))
+    value = value[1:]
+    token, value = get_dtext(value)
+    no_fold_literal.append(token)
+    if not value or value[0] != ']':
+        raise errors.HeaderParseError(
+            "expected ']' at the end of no-fold-literal "
+            "but found '{}'".format(value))
+    no_fold_literal.append(ValueTerminal(']', 'no-fold-literal-end'))
+    return no_fold_literal, value[1:]
+
+def get_msg_id(value):
+    """msg-id = [CFWS] "<" id-left '@' id-right  ">" [CFWS]
+       id-left = dot-atom-text / obs-id-left
+       id-right = dot-atom-text / no-fold-literal / obs-id-right
+       no-fold-literal = "[" *dtext "]"
+    """
+    msg_id = MsgID()
+    if value[0] in CFWS_LEADER:
+        token, value = get_cfws(value)
+        msg_id.append(token)
+    if not value or value[0] != '<':
+        raise errors.HeaderParseError(
+            "expected msg-id but found '{}'".format(value))
+    msg_id.append(ValueTerminal('<', 'msg-id-start'))
+    value = value[1:]
+    # Parse id-left.
+    try:
+        token, value = get_dot_atom_text(value)
+    except errors.HeaderParseError:
+        try:
+            # obs-id-left is same as local-part of add-spec.
+            token, value = get_obs_local_part(value)
+            msg_id.defects.append(errors.ObsoleteHeaderDefect(
+                "obsolete id-left in msg-id"))
+        except errors.HeaderParseError:
+            raise errors.HeaderParseError(
+                "expected dot-atom-text or obs-id-left"
+                " but found '{}'".format(value))
+    msg_id.append(token)
+    if not value or value[0] != '@':
+        msg_id.defects.append(errors.InvalidHeaderDefect(
+            "msg-id with no id-right"))
+        # Even though there is no id-right, if the local part
+        # ends with `>` let's just parse it too and return
+        # along with the defect.
+        if value and value[0] == '>':
+            msg_id.append(ValueTerminal('>', 'msg-id-end'))
+            value = value[1:]
+        return msg_id, value
+    msg_id.append(ValueTerminal('@', 'address-at-symbol'))
+    value = value[1:]
+    # Parse id-right.
+    try:
+        token, value = get_dot_atom_text(value)
+    except errors.HeaderParseError:
+        try:
+            token, value = get_no_fold_literal(value)
+        except errors.HeaderParseError as e:
+            try:
+                token, value = get_domain(value)
+                msg_id.defects.append(errors.ObsoleteHeaderDefect(
+                    "obsolete id-right in msg-id"))
+            except errors.HeaderParseError:
+                raise errors.HeaderParseError(
+                    "expected dot-atom-text, no-fold-literal or obs-id-right"
+                    " but found '{}'".format(value))
+    msg_id.append(token)
+    if value and value[0] == '>':
+        value = value[1:]
+    else:
+        msg_id.defects.append(errors.InvalidHeaderDefect(
+            "missing trailing '>' on msg-id"))
+    msg_id.append(ValueTerminal('>', 'msg-id-end'))
+    if value and value[0] in CFWS_LEADER:
+        token, value = get_cfws(value)
+        msg_id.append(token)
+    return msg_id, value
+
+
+def parse_message_id(value):
+    """message-id      =   "Message-ID:" msg-id CRLF
+    """
+    message_id = MessageID()
+    try:
+        token, value = get_msg_id(value)
+    except errors.HeaderParseError:
+        message_id.defects.append(errors.InvalidHeaderDefect(
+            "Expected msg-id but found {!r}".format(value)))
+    message_id.append(token)
+    return message_id
+
 #
 # XXX: As I begin to add additional header parsers, I'm realizing we probably
 # have two level of parser routines: the get_XXX methods that get a token in
diff --git a/Lib/email/headerregistry.py b/Lib/email/headerregistry.py
index 0065204..452c6ad 100644
--- a/Lib/email/headerregistry.py
+++ b/Lib/email/headerregistry.py
@@ -520,6 +520,18 @@ class ContentTransferEncodingHeader:
         return self._cte
 
 
+class MessageIDHeader:
+
+    max_count = 1
+    value_parser = staticmethod(parser.parse_message_id)
+
+    @classmethod
+    def parse(cls, value, kwds):
+        kwds['parse_tree'] = parse_tree = cls.value_parser(value)
+        kwds['decoded'] = str(parse_tree)
+        kwds['defects'].extend(parse_tree.all_defects)
+
+
 # The header factory #
 
 _default_header_map = {
@@ -542,6 +554,7 @@ _default_header_map = {
     'content-type':                 ContentTypeHeader,
     'content-disposition':          ContentDispositionHeader,
     'content-transfer-encoding':    ContentTransferEncodingHeader,
+    'message-id':                   MessageIDHeader,
     }
 
 class HeaderRegistry: