[TYPO3-Solr] index.queue.configuration for pages (tt_content)

[Johann Höchtl]

Ah sorry, 

now I see the point. Your are configuring the page indexing for the first time and didn’t know about the default behavior.
My suggestion from yesterday was based on the assumption that you don’t want to use the integrated page indexer.

First to your problem with tx_news. The field content_elements comes with the extension tx_news but is a configurable add-on. So your modification was correct if this feature is not used.

Now to page indexing. This is fairly complex because you cannot predict the content of the page without completely rendering it (there can be plugins, fe_user, partials, etc.).
That’s why the page indexing process looks like this: https://forge.typo3.org/projects/extension-solr/wiki/Indexing_Mechanism 
There is a special indexer which uses file_get_contents to call the page from the server and read the html which is then processed and indexed.
This works great and there’s even the possibility to tweak the configuration to use http-authentication or overwrite the hostname if your server isn’t able to call your domain.

Because page indexing needs to work that way there are some requirements:
- allow_url_fopen needs to be enabled, so the server can use file_get_contents to call the page
- domain-record needs to be configured properly
- server needs to be able to call himself (my tipp: add the full domain name to the /etc/hosts of the server with address 127.0.0.1, so he can use his loopback interface => performance)
- search marker in your template which surround content which should be indexed: <!--TYPO3SEARCH_begin--> und <!--TYPO3SEARCH_end—>

Best regards
Hans

Am 10.09.2014 um 09:08 schrieb Kevin von Spiczak <k.vonspiczak at q4u.de>:

> Quote: Olivier Dobberkau (oli4) wrote on Tue, 09 September 2014 19:37
> ----------------------------------------------------
>> Have you looked at the typoscript included with the ext?
>> Olivier
> ----------------------------------------------------
> HI Oliver, 
> yesterday I was working on indexing the tx_news_domain_model_news table with the included configuration from the ext. I had to alter it, to be specific, I had to remove the SOLR_RELATION for news.fields.content, because it was looking for a field "content_elements" which isn't available in the tx_news table. So my guess is, that the provided extension configuration was used in combination with some other extension for indexing news. Even the included typoscript didn't work out of the box. So there is no "best practice", or recommended way for indexing pages and their contents so far?
> 
> Kevin
> _______________________________________________
> TYPO3-project-solr mailing list
> TYPO3-project-solr at lists.typo3.org
> http://lists.typo3.org/cgi-bin/mailman/listinfo/typo3-project-solr