Help creating schema for indexable document
Hi Guys. I am struggling to create a schema with a determinist content model for a set of documents I want to index. My indexable documents will look something like: 1 code1 code2 mycategory My service will be mission critical and will accept batch imports from a potentially unreliable source. Are there any xml schema guru's who can help me with creating xn xsd which will work with my sample document? Thanks in advance for your help, -- Ross -- View this message in context: http://www.nabble.com/Help-creating-schema-for-indexable-document-tp24862700p24862700.html Sent from the Solr - User mailing list archive at Nabble.com.
Additional metadata when using Solr Cell
Hi.
I am indexing a PDF document with the ExtractingRequestHandler. My curl
post has a URL like:
../solr/update/extract?ext.idx.attr=true&ext.def.fl=text&ext.literal.id=123&ext.literal.author=Somebody
Sure enough I see in the server logs:
params={ext.def.fl=text&ext.literal.id=123&ext.idx.attr=true&ext.literal.author=Somebody}
I am trying to get my field back in the results from a query:
../solr/select?indent=on&version=2.2&q=hello&start=0&rows=10&fl=author%2Cscore&qt=standard&wt=standard&explainOther=&hl.fl=
I see the score in the results 'doc' but no reference to author.
Can anyone advise on what I am forgetting to do, to get hold of this field?
Thanks in advance for your help,
-- Ross
--
View this message in context:
http://www.nabble.com/Additional-metadata-when-using-Solr-Cell-tp23541256p23541256.html
Sent from the Solr - User mailing list archive at Nabble.com.
Re: Additional metadata when using Solr Cell
There is no reference to the author field I am trying to set.. I am using the
latest nightly download.
-- Ross
Grant Ingersoll-6 wrote:
>
> what does /admin/luke show for fields and terms in the fields?
>
> On May 14, 2009, at 10:03 AM, rossputin wrote:
>
>>
>> Hi.
>>
>> I am indexing a PDF document with the ExtractingRequestHandler. My
>> curl
>> post has a URL like:
>>
>> ../solr/update/extract?
>> ext
>> .idx
>> .attr
>> =true&ext.def.fl=text&ext.literal.id=123&ext.literal.author=Somebody
>>
>> Sure enough I see in the server logs:
>>
>> params
>> =
>> {ext
>> .def
>> .fl
>> =
>> text&ext.literal.id=123&ext.idx.attr=true&ext.literal.author=Somebody}
>>
>> I am trying to get my field back in the results from a query:
>>
>> ../solr/select?
>> indent=on&version=2.2&q=hello&start=0&rows=10&fl=author
>> %2Cscore&qt=standard&wt=standard&explainOther=&hl.fl=
>>
>> I see the score in the results 'doc' but no reference to author.
>>
>> Can anyone advise on what I am forgetting to do, to get hold of this
>> field?
>>
>> Thanks in advance for your help,
>>
>> -- Ross
>> --
>> View this message in context:
>> http://www.nabble.com/Additional-metadata-when-using-Solr-Cell-tp23541256p23541256.html
>> Sent from the Solr - User mailing list archive at Nabble.com.
>>
>
> --
> Grant Ingersoll
> http://www.lucidimagination.com/
>
> Search the Lucene ecosystem (Lucene/Solr/Nutch/Mahout/Tika/Droids)
> using Solr/Lucene:
> http://www.lucidimagination.com/search
>
>
>
--
View this message in context:
http://www.nabble.com/Additional-metadata-when-using-Solr-Cell-tp23541256p23541857.html
Sent from the Solr - User mailing list archive at Nabble.com.
Re: Additional metadata when using Solr Cell
There is now, thanks for your help. On the same topic.. is there a best
practice for modifying schema, in a future-proof way ?
-- Ross
Grant Ingersoll-6 wrote:
>
> Do you have an author field in your schema?
>
> On May 14, 2009, at 10:31 AM, rossputin wrote:
>
>>
>> There is no reference to the author field I am trying to set.. I am
>> using the
>> latest nightly download.
>>
>> -- Ross
>>
>>
>> Grant Ingersoll-6 wrote:
>>>
>>> what does /admin/luke show for fields and terms in the fields?
>>>
>>> On May 14, 2009, at 10:03 AM, rossputin wrote:
>>>
>>>>
>>>> Hi.
>>>>
>>>> I am indexing a PDF document with the ExtractingRequestHandler. My
>>>> curl
>>>> post has a URL like:
>>>>
>>>> ../solr/update/extract?
>>>> ext
>>>> .idx
>>>> .attr
>>>> =true&ext.def.fl=text&ext.literal.id=123&ext.literal.author=Somebody
>>>>
>>>> Sure enough I see in the server logs:
>>>>
>>>> params
>>>> =
>>>> {ext
>>>> .def
>>>> .fl
>>>> =
>>>> text
>>>> &ext.literal.id=123&ext.idx.attr=true&ext.literal.author=Somebody}
>>>>
>>>> I am trying to get my field back in the results from a query:
>>>>
>>>> ../solr/select?
>>>> indent=on&version=2.2&q=hello&start=0&rows=10&fl=author
>>>> %2Cscore&qt=standard&wt=standard&explainOther=&hl.fl=
>>>>
>>>> I see the score in the results 'doc' but no reference to author.
>>>>
>>>> Can anyone advise on what I am forgetting to do, to get hold of this
>>>> field?
>>>>
>>>> Thanks in advance for your help,
>>>>
>>>> -- Ross
>>>> --
>>>> View this message in context:
>>>> http://www.nabble.com/Additional-metadata-when-using-Solr-Cell-tp23541256p23541256.html
>>>> Sent from the Solr - User mailing list archive at Nabble.com.
>>>>
>>>
>>> --
>>> Grant Ingersoll
>>> http://www.lucidimagination.com/
>>>
>>> Search the Lucene ecosystem (Lucene/Solr/Nutch/Mahout/Tika/Droids)
>>> using Solr/Lucene:
>>> http://www.lucidimagination.com/search
>>>
>>>
>>>
>>
>> --
>> View this message in context:
>> http://www.nabble.com/Additional-metadata-when-using-Solr-Cell-tp23541256p23541857.html
>> Sent from the Solr - User mailing list archive at Nabble.com.
>>
>
> --
> Grant Ingersoll
> http://www.lucidimagination.com/
>
> Search the Lucene ecosystem (Lucene/Solr/Nutch/Mahout/Tika/Droids)
> using Solr/Lucene:
> http://www.lucidimagination.com/search
>
>
>
--
View this message in context:
http://www.nabble.com/Additional-metadata-when-using-Solr-Cell-tp23541256p23542620.html
Sent from the Solr - User mailing list archive at Nabble.com.
highlight results from pdf search
Hi. I have some PDF documents indexed through solr cell. My highlighting queries work fine on standard xml doc types, eg the samples. I would now like to highlight some queries on a PDF document. Currently for my simple examples I am just indexing a PDF, providing an id, and an arbitrary ext.literal. I would like to be able to get highlighted snippets back from the extracted content of the PDF. Is this possible? Thanks in advance for your help, - Ross -- View this message in context: http://www.nabble.com/highlight-results-from-pdf-search-tp23791905p23791905.html Sent from the Solr - User mailing list archive at Nabble.com.
Solr document structure for preserving version information
Hi Guys. This is a schema design question I suppose. I would like to store a series of version elements comprising of two attributes, 'updated' (a date) and 'reason' (just a simple string). I aim to produce xml based on a search which would look something like: So I realise I could use multiValued fields, but I want to avoid doing something like: 01/04/2009 10:30:00|changes made (using | or some other separator) As I would need to split the field in my code. This approach does not seem the best. Has anyone got an approach they could share ? Thanks in advance for your help, - Ross -- View this message in context: http://www.nabble.com/Solr-document-structure-for-preserving-version-information-tp23885262p23885262.html Sent from the Solr - User mailing list archive at Nabble.com.
posting binary file and metadata in two separate documents
Hi. I am currently using Solr Cell to extract content from binary files, and I am passing along some additional metadata with ext.literal params. Sample below: curl "http://localhost:8983/solr/update/extract?ext.literal.id=2&ext.literal.some_code1=code1&ext.literal.some_code2=code2&ext.idx.attr=true\&ext.def.fl=text"; -F "myfi...@myfile.pdf" Where I have large numbers of ext.literal params this becomes a bit of a chore.. and it would be the same case in an html form with many params... can I pass both files to '/update/extract' as documents, (files) linked together? Or are there any other options like this? Perhaps something I can do with Solrj. Thanks in advance for your help, regards, Ross. -- View this message in context: http://www.nabble.com/posting-binary-file-and-metadata-in-two-separate-documents-tp24375649p24375649.html Sent from the Solr - User mailing list archive at Nabble.com.
Re: posting binary file and metadata in two separate documents
Hi. Apologies for bumping this one, but another question occurred to me... is there a limit to the number of &ext.literal components I can put in my curl command... if so, i will definitely need to find another way to get this data in, as I am building up relationships between documents, and there will be many of them. Thanks in advance for your help, regards, Ross rossputin wrote: > > Hi. > > I am currently using Solr Cell to extract content from binary files, and I > am passing along some additional metadata with ext.literal params. Sample > below: > > curl > "http://localhost:8983/solr/update/extract?ext.literal.id=2&ext.literal.some_code1=code1&ext.literal.some_code2=code2&ext.idx.attr=true\&ext.def.fl=text"; > -F "myfi...@myfile.pdf" > > Where I have large numbers of ext.literal params this becomes a bit of a > chore.. and it would be the same case in an html form with many params... > can I pass both files to '/update/extract' as documents, (files) linked > together? Or are there any other options like this? Perhaps something I > can do with Solrj. > > Thanks in advance for your help, > > regards, > > Ross. > > > -- View this message in context: http://www.nabble.com/posting-binary-file-and-metadata-in-two-separate-documents-tp24375649p24423267.html Sent from the Solr - User mailing list archive at Nabble.com.
Re: posting binary file and metadata in two separate documents
Hi. Thanks for your reply, shame nobody has already implemented the multiple 'ContentStreams' idea :-) With regards to posting in a form, I had considered that, but unfortunately there can be an arbitrary number of 'ext.literals', so it would be difficult to build a form which would handle all cases. Regards, -- Ross hossman wrote: > > > : Subject: posting binary file and metadata in two separate documents > > there was some discussion a while back about that fact that you can push > multiple "ContentStreams" to SOlr in a single request, and while the > existing handelrs all just iterate over and process them seperately, it > would be *possible* for a variant of ExtractingRequest handler to use the > first stream to get document metadat, and have that metdata refrence the > other streams in some way for large chunks of text) > > But no one has attempted to implement that as far as i know. > > : > "http://localhost:8983/solr/update/extract?ext.literal.id=2&ext.literal.some_code1=code1&ext.literal.some_code2=code2&ext.idx.attr=true\&ext.def.fl=text"; > : -F "myfi...@myfile.pdf" > : > : Where I have large numbers of ext.literal params this becomes a bit of a > : chore.. and it would be the same case in an html form with many > params... > : can I pass both files to '/update/extract' as documents, (files) linked > : together? Or are there any other options like this? Perhaps something > I > : can do with Solrj. > > there's no reason those params have ot be in the URL. you can do a > multipart POST with application/x-www-form-urlencoded in one part and your > pdf file in another part (just like doing a POST from a massive HTML form > with an '' option) > > > -Hoss > > > -- View this message in context: http://www.nabble.com/posting-binary-file-and-metadata-in-two-separate-documents-tp24375649p24530051.html Sent from the Solr - User mailing list archive at Nabble.com.
post error - ERROR:unknown field 'title'
Hi guys. I have two different solr versions as I am evaluating nightly builds. On a more recent one.. I think 15th July I am getting the following error : ERROR:unknown field 'title' I am posting to 'solr/update/extract' with the following: curl "http://localhost:8983/solr/update/extract?ext.literal.id=1&ext.literal.code=somecode&ext.literal.url=someurl/file.pdf&ext.literal.category=somecat&ext.literal.updated=2009-06-01T09:10:30.000Z&ext.idx.attr=true\&ext.def.fl=text"; -F "myfi...@1411_9.pdf" My schema does not, and is not intended to contain a 'title' field. Thanks in advance for your help, -- Ross -- View this message in context: http://www.nabble.com/post-error---ERROR%3Aunknown-field-%27title%27-tp24567235p24567235.html Sent from the Solr - User mailing list archive at Nabble.com.
