Page MenuHomestyx hydra
Files F233288

No OneTemporary
Actions

View Options

diff --git a/src/applications/metamta/parser/PhabricatorMetaMTAEmailBodyParser.php b/src/applications/metamta/parser/PhabricatorMetaMTAEmailBodyParser.php
index a53d9bf089..7ee2e02485 100644
--- a/src/applications/metamta/parser/PhabricatorMetaMTAEmailBodyParser.php
+++ b/src/applications/metamta/parser/PhabricatorMetaMTAEmailBodyParser.php
@@ -1,121 +1,128 @@
<?php
final class PhabricatorMetaMTAEmailBodyParser {
/**
* Mails can have bodies such as
*
* !claim
*
* taking this task
*
* Or
*
* !assign epriestley
*
* please, take this task I took; its hard
*
* This function parses such an email body and returns a dictionary
* containing a clean body text (e.g. "taking this task"), a $command
* (e.g. !claim, !assign) and a $command_value (e.g. "epriestley" in the
* !assign example.)
*
* @return dict
*/
public function parseBody($body) {
$body = $this->stripTextBody($body);
$lines = explode("\n", trim($body));
$first_line = head($lines);
$command = null;
$command_value = null;
$matches = null;
if (preg_match('/^!(\w+)\s*(\S+)?/', $first_line, $matches)) {
$lines = array_slice($lines, 1);
$body = implode("\n", $lines);
$body = trim($body);
$command = $matches[1];
$command_value = idx($matches, 2);
}
return array(
'body' => $body,
'command' => $command,
'command_value' => $command_value);
}
public function stripTextBody($body) {
return trim($this->stripSignature($this->stripQuotedText($body)));
}
private function stripQuotedText($body) {
// Look for "On <date>, <user> wrote:". This may be split across multiple
// lines. We need to be careful not to remove all of a message like this:
//
// On which day do you want to meet?
//
// On <date>, <user> wrote:
// > Let's set up a meeting.
$start = null;
$lines = phutil_split_lines($body);
foreach ($lines as $key => $line) {
if (preg_match('/^\s*>?\s*On\b/', $line)) {
$start = $key;
}
if ($start !== null) {
if (preg_match('/\bwrote:/', $line)) {
$lines = array_slice($lines, 0, $start);
$body = implode('', $lines);
break;
}
}
}
// Outlook english
$body = preg_replace(
'/^\s*(> )?-----Original Message-----.*?/imsU',
'',
$body);
// Outlook danish
$body = preg_replace(
'/^\s*(> )?-----Oprindelig Meddelelse-----.*?/imsU',
'',
$body);
// See example in T3217.
$body = preg_replace(
'/^________________________________________\s+From:.*?/imsU',
'',
$body);
return rtrim($body);
}
private function stripSignature($body) {
// Quasi-"standard" delimiter, for lols see:
// https://bugzilla.mozilla.org/show_bug.cgi?id=58406
$body = preg_replace(
'/^-- +$.*/sm',
'',
$body);
+ // Mailbox seems to make an attempt to comply with the "standard" but
+ // omits the leading newline and uses an em dash?
+ $body = preg_replace(
+ "/\s*\xE2\x80\x94 \nSent from Mailbox\s*\z/su",
+ '',
+ $body);
+
// HTC Mail application (mobile)
$body = preg_replace(
'/^\s*^Sent from my HTC smartphone.*/sm',
'',
$body);
// Apple iPhone
$body = preg_replace(
'/^\s*^Sent from my iPhone\s*$.*/sm',
'',
$body);
return rtrim($body);
}
}
diff --git a/src/applications/metamta/parser/__tests__/PhabricatorMetaMTAEmailBodyParserTestCase.php b/src/applications/metamta/parser/__tests__/PhabricatorMetaMTAEmailBodyParserTestCase.php
index 4a9c2cc845..94473f498c 100644
--- a/src/applications/metamta/parser/__tests__/PhabricatorMetaMTAEmailBodyParserTestCase.php
+++ b/src/applications/metamta/parser/__tests__/PhabricatorMetaMTAEmailBodyParserTestCase.php
@@ -1,183 +1,189 @@
<?php
final class PhabricatorMetaMTAEmailBodyParserTestCase
extends PhabricatorTestCase {
public function testQuotedTextStripping() {
$bodies = $this->getEmailBodies();
foreach ($bodies as $body) {
$parser = new PhabricatorMetaMTAEmailBodyParser();
$stripped = $parser->stripTextBody($body);
$this->assertEqual('OKAY', $stripped);
}
}
public function testEmailBodyCommandParsing() {
$bodies = $this->getEmailBodiesWithFullCommands();
foreach ($bodies as $body) {
$parser = new PhabricatorMetaMTAEmailBodyParser();
$body_data = $parser->parseBody($body);
$this->assertEqual('OKAY', $body_data['body']);
$this->assertEqual('whatevs', $body_data['command']);
$this->assertEqual('dude', $body_data['command_value']);
}
$bodies = $this->getEmailBodiesWithPartialCommands();
foreach ($bodies as $body) {
$parser = new PhabricatorMetaMTAEmailBodyParser();
$body_data = $parser->parseBody($body);
$this->assertEqual('OKAY', $body_data['body']);
$this->assertEqual('whatevs', $body_data['command']);
$this->assertEqual(null, $body_data['command_value']);
}
}
public function testFalsePositiveForOnWrote() {
$body = <<<EOEMAIL
On which horse shall you ride?
On Sep 23, alincoln wrote:
> Hey bro do you want to go ride horses tomorrow?
EOEMAIL;
$parser = new PhabricatorMetaMTAEmailBodyParser();
$stripped = $parser->stripTextBody($body);
$this->assertEqual('On which horse shall you ride?', $stripped);
}
private function getEmailBodiesWithFullCommands() {
$bodies = $this->getEmailBodies();
$with_commands = array();
foreach ($bodies as $body) {
$with_commands[] = "!whatevs dude\n".$body;
}
return $with_commands;
}
private function getEmailBodiesWithPartialCommands() {
$bodies = $this->getEmailBodies();
$with_commands = array();
foreach ($bodies as $body) {
$with_commands[] = "!whatevs\n".$body;
}
return $with_commands;
}
private function getEmailBodies() {
$trailing_space = ' ';
+ $emdash = "\xE2\x80\x94";
return array(
<<<EOEMAIL
OKAY
On May 30, 2011, at 8:36 PM, Someone wrote:
> ...
EOEMAIL
,
<<<EOEMAIL
OKAY
On Fri, May 27, 2011 at 9:39 AM, Someone <
somebody@somewhere.com> wrote:
> ...
EOEMAIL
,
<<<EOEMAIL
OKAY
On Fri, May 27, 2011 at 9:39 AM, Someone
<somebody@somewhere.com> wrote:
> ...
EOEMAIL
,
<<<EOEMAIL
OKAY
-----Oprindelig Meddelelse-----
> ...
EOEMAIL
,
<<<EOEMAIL
OKAY
-----Original Message-----
> ...
EOEMAIL
,
<<<EOEMAIL
OKAY
-----oprindelig meddelelse-----
> ...
EOEMAIL
,
<<<EOEMAIL
OKAY
-----original message-----
> ...
EOEMAIL
,
<<<EOEMAIL
OKAY
Sent from my HTC smartphone on the Now Network from Sprint!
-Reply message ----- From: "somebody (someone)" <
somebody@somewhere.com>
To: <somebody@somewhere.com>
Subject: Some Text Date: Mon, Apr 2, 2012 1:42 pm
> ...
EOEMAIL
,
<<<EOEMAIL
OKAY
--{$trailing_space}
Abraham Lincoln
Supreme Galactic Emperor
EOEMAIL
,
<<<EOEMAIL
OKAY
Sent from my iPhone
EOEMAIL
,
<<<EOMAIL
OKAY
________________________________________
From: Abraham Lincoln <alincoln@logcab.in>
Subject: Core World Tariffs
EOMAIL
,
<<<EOMAIL
OKAY
> On 17 Oct 2013, at 17:47, "Someone" <somebody@somewhere> wrote:
> ...
EOMAIL
,
<<<EOMAIL
OKAY
> -----Original Message-----
>
> ...
+EOMAIL
+,
+<<<EOMAIL
+OKAY {$emdash}{$trailing_space}
+Sent from Mailbox
EOMAIL
);
}
}

File Metadata

Mime Type
text/x-diff
Expires
Mon, Jul 28, 1:34 AM (1 w, 20 h ago)
Storage Engine
blob
Storage Format
Raw Data
Storage Handle
186325
Default Alt Text
(7 KB)

Event Timeline

styx copyright 2020 sirocyl / ellie (kymkdd) / ncommander / styx contributors
Use of other names and brands is not intended to reflect a claim of, or right to, use the name or brand.
styx is an operating system using the Linux kernel and components, and "styx Linux" as a phrase does not reflect the inclusion of the Linux brand or trademark in the styx project.
Linux® is the registered trademark of Linus Torvalds in the U.S. and other countries.
Terms and Conditions (TODO: fix this!)