From 1cb0d41367b7a79f0fdaa1be20e348ce5bd8b628 Mon Sep 17 00:00:00 2001 From: epriestley Date: Wed, 22 Nov 2017 07:35:22 -0800 Subject: [PATCH] Fix inbound mail handling for messages with no HTML body part Summary: See D18776. See . The change in D18776 to improve handling of non-utf8 HTML parts broke handling of mail with //no// HTML parts. Partly, this is because MimeMailParser has a "traditional" PHP-style API where the return type is an exciting surprise. Test Plan: - Sent a text-only message in `Mail.app`. - Used "Show Raw" to copy it to `mail.txt`, verifying that the raw message contains ONLY a text body. - Ran `cat mail.txt | ./scripts/mail/mail_handler.php --trace --process-duplicates`. - Before patch: error about bad `idx()` on a non-array. - After patch: clean mail processing. - Did the same with a message with both HTML and text bodies to make sure I didn't break anything. Ideally we'd probably get test coverage on this, but it's been touched roughly once a year since 2013 so it'll probably hold. Reviewers: amckinley, alexmv Reviewed By: amckinley, alexmv Differential Revision: https://secure.phabricator.com/D18778 --- scripts/mail/mail_handler.php | 20 ++++++++++++-------- 1 file changed, 12 insertions(+), 8 deletions(-) diff --git a/scripts/mail/mail_handler.php b/scripts/mail/mail_handler.php index b76b3910df..1c3c71f305 100755 --- a/scripts/mail/mail_handler.php +++ b/scripts/mail/mail_handler.php @@ -38,15 +38,19 @@ $parser->setText(file_get_contents('php://stdin')); $content = array(); foreach (array('text', 'html') as $part) { $part_body = $parser->getMessageBody($part); - $part_headers = $parser->getMessageBodyHeaders($part); - $content_type = idx($part_headers, 'content-type'); - if ( - !phutil_is_utf8($part_body) && - (preg_match('/charset="(.*?)"/', $content_type, $matches) || - preg_match('/charset=(\S+)/', $content_type, $matches)) - ) { - $part_body = phutil_utf8_convert($part_body, 'UTF-8', $matches[1]); + + if (strlen($part_body) && !phutil_is_utf8($part_body)) { + $part_headers = $parser->getMessageBodyHeaders($part); + if (!is_array($part_headers)) { + $part_headers = array(); + } + $content_type = idx($part_headers, 'content-type'); + if (preg_match('/charset="(.*?)"/', $content_type, $matches) || + preg_match('/charset=(\S+)/', $content_type, $matches)) { + $part_body = phutil_utf8_convert($part_body, 'UTF-8', $matches[1]); + } } + $content[$part] = $part_body; }