[ome-users] Substring search?

[Graham Klyne]

Josiah Johnston wrote:
> Are you still dealing with this problem? 

Not at the moment.

>From your response it seems likely I was doing something wrong.  I
didn't try very hard when this first came up, as I assumed it was a
design limitation on the way regex matching was being handled.  If it
comes up again I'll did a little deeper.

Thanks.

#g
--

I can't recreate it. Here's
> what I did:
> 
> In the image detail page (what you see when you click on an image 
> name), I entered "foo,fie,foe,fum" in the "Current Textual Annotation" 
> text box, and hit "Save". I did the same to two other images with 
> "foo,foe,fum" and "fie,fum".
> 
> I went to Search -> Other, then chose to look for "ImageAnnotation". I 
> saw all three entries, plus some old ones.
> I searched for "foo" in the Content and got 0 results.
> "*foo*" returned 1 and 2
> "foo*" also returned 1 and 2
> "fie*" returned 3
> "*fie*" returned 1 and 3
> 
> If this is not the behavior you are getting, then something unexpected 
> is going on.
> I tested this on 2.4.0 running postgresql 7.2.2
> and on a slightly old version of cvs-HEAD running 7.3.4-RH
> 
> These four queries get translated to roughly:
> select * from image_annotations where content ilike '%foo%';
> ... ilike 'foo%';
> ... ilike 'fie%';
> ... ilike '%fie%';
> 
> I get the same results in the DB as through the Web interface. e.g.
> ome=# select * from image_annotations where content ilike 'fie%';
>  attribute_id | image_id | content | the_c | the_t | 
> module_execution_id | the_z | valid
> --------------+----------+---------+-------+-------
> +---------------------+-------+-------
>        352112 |     6864 | fie,fum |       |       |               
> 37182 |       | t
> (1 row)
> 
> If you still are having this problem, can you report what the database 
> queries return?
> 
> Thanks,
> -Josiah
> 
> On Aug 6, 2005, at 6:58 AM, Graham Klyne wrote:
> 
>> I was showing OME to a new potential user yesterday, and a question 
>> came up that required searching for the occurrence of a substring 
>> *within* a string value in a new Semantic Type.
>>
>> So if we have entries with fields containing:
>>  1.  "foo,fie,foe,fum"
>>  2.  "foo,foe,fum"
>>  3.  "fie,fum"
>> a search for "fie" would return the entries 1 and 3.
>>
>> With a little experimentation, I found that I could use a wildcard to 
>> match leading substrings (e.g. searching for "foo*" would return 1
>> and  2 above) but the searching for "*fie*" didn't work.
>>
>> My question is:  is there a feature I'm overlooking that does this in 
>> the UI?  (I'm fairly sure this could be done at the database level 
>> using postgres regexp matching.)
>>
>> (This use of substring matching would be an interim "fix" until a
>> more  structured approach to organizing the data could be adopted, so
>> I'm  not suggesting here that it should be regarded as a priority
>> feature  request.)
>>
>> #g
>>
>> -- 
>> Graham Klyne
>> Image Bioinformatics Research Group (http://www.bioimage.org/)
>> Department of Zoology, University of Oxford
>> South Parks Road, Oxford OX1 3PS, UK
>> E-mail: <Graham.Klyne at zoo.ox.ac.uk>
>> Direct phone: +44-(0)1865-281991
>> Departmental fax: +44-(0)1865-310447
>>
>> _______________________________________________
>> ome-users mailing list
>> ome-users at lists.openmicroscopy.org.uk
>> http://lists.openmicroscopy.org.uk/mailman/listinfo/ome-users
>>
> 

-- 
Graham Klyne
Image Bioinformatics Research Group (http://www.bioimage.org/)
Department of Zoology, University of Oxford
South Parks Road, Oxford OX1 3PS, UK
E-mail: <Graham.Klyne at zoo.ox.ac.uk>
Direct phone: +44-(0)1865-281991
Departmental fax: +44-(0)1865-310447