topic Re: Site search syntax for finding files by wildcard pattern in Alfresco Archive

Site search syntax for finding files by wildcard pattern

almoehi — Mon, 18 Feb 2013 17:14:36 GMT

Hi everybody,I'd like to use the site search dashlet to find certain files in a doc lib sub folder. The are of type pdf and their names follow a defined scheme with variables. How is the syntax for that? Let's say, for example, I have a folder "Reports" in the doc lib, and in this folder are, amongs

Re: Site search syntax for finding files by wildcard pattern

almoehi — Mon, 18 Feb 2013 17:25:29 GMT

I've managed to find the files using the following search pattern:

+PATH:"//cm:Reports//*" AND cm:name:WP*rep*.pdf‍

But as soon as I add hyphens it doesn't work anymore: WP-*-rep-*.pdf
Ok, that's not so important, but why is that and how can I include them into the pattern? Escaping using backslash or _x002d_ didn't work.

Re: Site search syntax for finding files by wildcard pattern

andy — Mon, 11 Mar 2013 10:57:02 GMT

Hi

You can escape "-" using \ in query expressions.
You can also quote the expression - you are allowed to use - and * inside phrases.
Also you can prefix cm:name with = to enforce pattern matching and not token matching.

What version of Alfresco are we talking about?
*.pdf
.pdf
pdf

….should all find the docs

The lucene tokeniser does some odd stuff with 12-3-04 as it thinks it is a date.
You may be falling for this. Infact it does this for any odd number temr split by -: etc ….
If you are using lucene you probably want to consider moving to SOLR or changing the default analysis.

You can also decide to use a separator that does not mean anything in the query language.
You should also consider putting metadata on your nodes rather then encoding this in the name.
It will also much easier and more efficient to query. (you can configure the search to use these fields)

As you have to escape punctuation as the start the easiest thing is to miss it out.

Andy

Re: Site search syntax for finding files by wildcard pattern

almoehi — Mon, 11 Mar 2013 13:23:43 GMT

Hi Andy,

thanks for your answer.

<blockquote>What version of Alfresco are we talking about?</blockquote>
I'm using Alfresco Community 4.2.c (I thought it would be shown somewhere as I've selected it from the dropdown box on creating the topic).

<blockquote>You can escape "-" using \ in query expressions. …</blockquote>
Escaping the hyphens does not work, nor does quoting, but the = prefix did the trick.

<blockquote>You can also decide to use a separator that does not mean anything in the query language. …</blockquote>
Unfortunately I have to stick to the naming scheme as it is defined by the project guidelines. Meta data gets lost when the file is passed around. And yes, "*.pdf" or simply "pdf" finds all pdfs in the directory, but I really need to look for the mentioned pattern (there are other pdfs in the directory and sometimes I need to find a subset, for example the second report of all groups).

However, you helped me a lot with that, thanks again. Maybe I should spend some time learning lucene basics. The problem is that our project members definitely won't be willing to do so. They keep complaining to me about the search function. Like me they don't ever need the full power of the engine, they only need it for finding file names in the document lib and thus expect it to work like windows file explorer. It would be really nice to have some simple document-lib-only search in future releases that provides exactly that.

Alex