diff options
Diffstat (limited to 'test/mailbox.lm')
-rw-r--r-- | test/mailbox.lm | 106 |
1 files changed, 106 insertions, 0 deletions
diff --git a/test/mailbox.lm b/test/mailbox.lm new file mode 100644 index 0000000..ca3b9a9 --- /dev/null +++ b/test/mailbox.lm @@ -0,0 +1,106 @@ +##### LM ##### + +# lines, and fromlines +lex + rl day /[A-Z][a-z][a-z]/ + rl month /[A-Z][a-z][a-z]/ + rl year /[0-9][0-9][0-9][0-9]/ + rl time /[0-9][0-9] ':' [0-9][0-9] ( ':' [0-9][0-9] )? / + rl letterZone /[A-Z][A-Z][A-Z]/ + rl numZone /[+\-][0-9][0-9][0-9][0-9]/ + rl zone / letterZone | numZone/ + rl dayNum /[0-9 ][0-9]/ + + # These are the different formats of the date minus an obscure + # type that has a funny string 'remote from xxx' on the end. Taken + # from c-client in the imap-2000 distribution. + rl date / day ' ' month ' ' dayNum ' ' time ' ' + ( year | year ' ' zone | zone ' ' year ) / + + # From lines separate messages. We will exclude from_line from a message + # body line. This will cause us to stay in message line up until an + # entirely correct from line is matched. + token from_line / 'From ' (any-'\n')* ' ' date '\n' / + token simple_line / [^\n]* '\n' / +end + +rl hchar /print - [ :]/ +token header_name /hchar+/ + +token colon /':' ' '*/ +token header_content / ([^\n] | '\n' [ \t])* '\n'/ +token blank_line / '\n' / + +def header + [header_name colon header_content] + +def message + [from_line header* blank_line simple_line*] + +def start + [message*] + +parse S: start[ stdin ] +print_xml( S ) +print( '\n' ) +##### IN ##### +From thurston Tue Jan 2 21:16:50 2007 +Return-Path: <unknown> +X-Spam-Level: * +Received: from [109.111.71.111] (helo=twfmtr) + by zifreax with smtp (Exim 4.43) + id 1H1vfs-0005LN-HW; Tue, 2 Jan 2007 21:16:16 -0500 +Message-ID: <459B113F.8050903@immoarthabitatge.com> +X-Keywords: +X-UID: 1 + +Content-Type: text/html; charset=ISO-8859-1 +</body> +</html> + +From thurston Wed Jan 3 02:35:48 2007 +Return-Path: <unknown> +X-Spam-Checker-Version: SpamAssassin 3.1.1 (2006-03-10) on mambo.cs.queensu.ca +X-Spam-Level: ** +X-Spam-Status: No, score=2.9 required=5.0 tests=BAYES_20,EXTRA_MPART_TYPE, + HTML_40_50,HTML_IMAGE_ONLY_16,HTML_MESSAGE,RCVD_IN_BL_SPAMCOP_NET + autolearn=no version=3.1.1 +X-Bogosity: Unsure, tests=bogofilter, spamicity=0.971708, version=1.0.2 +Status: RO +X-UID: 2 + +------=_NextPart_000_0010_01C72F11.F137BD60 + charset="windows-1252" +Content-Transfer-Encoding: quoted-printable + +##### EXP ##### +<start><_repeat_message><message><from_line>From thurston Tue Jan 2 21:16:50 2007 +</from_line><_repeat_header><header><header_name>Return-Path</header_name><colon>: </colon><header_content><unknown> +</header_content></header><header><header_name>X-Spam-Level</header_name><colon>: </colon><header_content>* +</header_content></header><header><header_name>Received</header_name><colon>: </colon><header_content>from [109.111.71.111] (helo=twfmtr) + by zifreax with smtp (Exim 4.43) + id 1H1vfs-0005LN-HW; Tue, 2 Jan 2007 21:16:16 -0500 +</header_content></header><header><header_name>Message-ID</header_name><colon>: </colon><header_content><459B113F.8050903@immoarthabitatge.com> +</header_content></header><header><header_name>X-Keywords</header_name><colon>: </colon><header_content> +</header_content></header><header><header_name>X-UID</header_name><colon>: </colon><header_content>1 +</header_content></header></_repeat_header><blank_line> +</blank_line><_repeat_simple_line><simple_line>Content-Type: text/html; charset=ISO-8859-1 +</simple_line><simple_line></body> +</simple_line><simple_line></html> +</simple_line><simple_line> +</simple_line></_repeat_simple_line></message><message><from_line>From thurston Wed Jan 3 02:35:48 2007 +</from_line><_repeat_header><header><header_name>Return-Path</header_name><colon>: </colon><header_content><unknown> +</header_content></header><header><header_name>X-Spam-Checker-Version</header_name><colon>: </colon><header_content>SpamAssassin 3.1.1 (2006-03-10) on mambo.cs.queensu.ca +</header_content></header><header><header_name>X-Spam-Level</header_name><colon>: </colon><header_content>** +</header_content></header><header><header_name>X-Spam-Status</header_name><colon>: </colon><header_content>No, score=2.9 required=5.0 tests=BAYES_20,EXTRA_MPART_TYPE, + HTML_40_50,HTML_IMAGE_ONLY_16,HTML_MESSAGE,RCVD_IN_BL_SPAMCOP_NET + autolearn=no version=3.1.1 +</header_content></header><header><header_name>X-Bogosity</header_name><colon>: </colon><header_content>Unsure, tests=bogofilter, spamicity=0.971708, version=1.0.2 +</header_content></header><header><header_name>Status</header_name><colon>: </colon><header_content>RO +</header_content></header><header><header_name>X-UID</header_name><colon>: </colon><header_content>2 +</header_content></header></_repeat_header><blank_line> +</blank_line><_repeat_simple_line><simple_line>------=_NextPart_000_0010_01C72F11.F137BD60 +</simple_line><simple_line> charset="windows-1252" +</simple_line><simple_line>Content-Transfer-Encoding: quoted-printable +</simple_line><simple_line> +</simple_line></_repeat_simple_line></message></_repeat_message></start> |