Parsing Post for BbCode on XenForo_DataWriter_DiscussionMessage_Post?

[shri]

I have a requirement to parse a post for BbCode and have access to the threadid and post id.

The goal is to extract some meta data from the text and links in the message using a text / link extraction service to figure out what the post is about (sentiment analysis / tag extraction).

Extracted tags / keywords / location information would then added to the XenForo thread tags.

Sentiments will be stored and perhaps used for user / message highlighting (most angry user - negative post / positive post etc).

The best place I've figured this could be done is in XenForo_DataWriter_DiscussionMessage_Post

I have this code which gives me the postid / threadid and the post message in $post['message']

       protected function _postSaveAfterTransaction() {
                parent::_postSaveAfterTransaction();
                $postid = $this->get('post_id');
                $threadid = $this->get('thread_id');
                $post = $this->getMergedData();
                XenForo_Error::debug("postSaveAfterTransaction: postid = $postid , threadid = $threadid");
                XenForo_Error::debug("Post Data: " . print_r($post, TRUE));
        }

Before I go further can someone help me identify if I'm on the right track.

1) Am I on the right track where this might be the best place to extract the meta data from the post?
2) How do I parse the bbcode

 [URL="http://whatever.com"]

cleanly to give me the urls in that post? I can use a regex, but want to make sure there isnt a clean XenForo way of doing this.

Thanks!

 

[Chris D]

Seems like a reasonable approach.

With regards to parsing the BB code, I probably wouldn't bother and I would just go down the regex route; there isn't really anything in XF that would help you with this, so regex is probably the best approach.

 

[shri]

Thanks Chris.

 

[shri]

Hi Chris,

So, I've managed to plug the postSaveAfterTransaction into Alchemy API and have a sanitised list of keywords. Say they're in an array $keywords.

Can you give me a pointer on how to access the tagging API so that I can add or append these keywords to the thread's tag list?

I can see references in ControllerPublic/Thread.php and DataWriter/Discussion.php - not sure where to start looking at this.

Thanks.

 

[Chris D]

Look at XenForo_ControllerPublic_Thread::actionTags()

This gets the Tagger object, sets the content, sets the tags and saves. You might be able to ignore some of the editable/uneditable and permission stuff as if you're adding this programmatically those permissions may not apply.

 

[shri]

FYI, this is part of the code that accomplished it, incase anyone else has a similar issue.

    $frontController = XenForo_Application::get('fc');
    $fakeController = new XenForo_ControllerPublic_Index($frontController->getRequest(), $frontController->getResponse(), $frontController->route());
    $ftpHelper = $fakeController->getHelper('ForumThreadPost');
    list($thread, $forum) = $ftpHelper->assertThreadValidAndViewable($threadid);
    $tagModel = $this->getModelFromCache('XenForo_Model_Tag');
    $tagger = $tagModel->getTagger('thread');
    $tagger->setContent($thread['thread_id'])->setPermissionsFromContext($thread, $forum);
    $tagger->setTags($tagModel->splitTags($tagstr));
    // XenForo_Error::debug("Tag Errors: " . print_r($tagger->getErrors(),TRUE));
    $tagger->save();