Lucene searching question

[hsjawanda]

Given a node's id, how can I get its Lucene-path (i.e., a string that I can use in a Lucene "PATH:x" query)?

I am having trouble with a lucene query which includes TYPE and PATH fields that used to work fine (or at least as I expected it to work 🙂 ) before upgrading from PR6 to v1.0.  For example:

+TYPE:"{http://www.alacre.com/cms/content-model-1.0}project" AND PATH:"app:company_home//*/cms:lost"‍

works, but the following doesn't:

+TYPE:"{http://www.alacre.com/cms/content-model-1.0}project" AND PATH:"app:company_home//*/cms:get lost"‍

I thought maybe it was because of the space in "get lost", which I replaced with "_", but it still doesn't work.

I am confused: in bootstrapped data, which properties determine the name of a node as used in a Lucene path query?

[andy]

Hi

There has been a change that will affect this query.

The PATH field now matches what would be required in XPATH.
Your query would not be a valid XPATH as it contains a space.

Characters not allowed in XML element names are encoded using ISO9075.

So you would need:

+TYPE:"{http://www.alacre.com/cms/content-model-1.0}project" AND PATH:"app:company_home//*/cms:get_x0020_lost"
‍‍‍

A restricted set of characters were all mapped to _ which was wrong.

The nodeService can be used to get Path objects which give the correct path via toString()

Cheers

Andy

[hsjawanda]

Hi,

Is searching in the TEXT field the same as searching inside a content property?  Basically, if I am already searching on

TEXT:"access*"‍

do I need to search on:

@\{http\://www.alfresco.org/model/content/1.0\}content:'access*"‍

A comment in SearchContext.java refers to using the "TEXT" field as doing a full text search. What does this full text include and where is this configured?

Thanks,

[andy]

Hi

TEXT is a combined index of all content properties.
So you just need to search TEXT in the case you describe.

The individual properties are there for when you do not want to search all content attributes.

Regards

Andy

[cyman]

Sorry for refreshing such an old topic, but I'm chasing down an issue with the search functionality and the last entry in this topic troubles me somehow:

TEXT is a combined index of all content properties.

Can somebody confirm whether this still applies to the current version (3.2)?

I have a custom metadata property (called "additionalComments") which shall also be searchable through the fulltext search. I could solve this for the simple search (upper-right-corner) by adding the following configuration:

   <config>
      <client>
         <simple-search-additional-attributes>
            <qname>{http://www.alfresco.org/model/content/1.0}description</qname>
            <qname>{http://www.alfresco.org/model/content/1.0}title</qname>
            <qname>{my.model}additionalComments</qname>
         </simple-search-additional-attributes>
      </client>
   </config>
‍‍‍‍‍‍‍‍‍‍‍

However this does not affect the TEXT (Look for field on the Advanced Search page. If the information above was still correct, I would assume that no additional configuration should be needed.