[jira] [Comment Edited] (NUTCH-1467) nutch 1.5.1 not able to parse mutliValued metatags

["kiran (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/NUTCH-1467?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13591860#comment-13591860 ] 

kiran edited comment on NUTCH-1467 at 3/3/13 8:43 PM:
------------------------------------------------------

Sebastian,

I got what you mean. I am uploading the patch with your suggestions as v2. Thanks for your suggestions.

One odd thing, is that multipleValues for tags is not working with parse-tika but only with parse-html. For tika, only the first value is being saved. I have used the same code for HTMLMetaProcessor.java in both parse-html and parse-tika. 

Up on further investigation, the repeated tags are not sent for processing to HTMLMetaProcessor. Is it something like tika only send tags with unique names ? 

I have also used tika-app-1.2.jar and processed the URL that has multipleValues for metatag and the Tika app displayed only the first value. I have tested with http://scholar.lib.vt.edu/ejournals/ALAN/v37n1/hazlett.html URL.

Is it something I should ask in tika ? 

PS: Just to clarify, when i use only parse tika the patch is not working (Eg: parse-(tika|metatags) in plugin.includes config). The patch works just fine when i use this setting in plugin.includes config (parse-(html|tika|metatags))
                
      was (Author: kiranch):
    Sebastian,

I got what you mean. I am uploading the patch with your suggestions as v2. Thanks for your suggestions.

One odd thing, is that multipleValues for tags is not working with parse-tika but only with parse-html. For tika, only the first value is being saved. I have used the same code for HTMLMetaProcessor.java in both parse-html and parse-tika. 

Up on further investigation, the repeated tags are not sent for processing to HTMLMetaProcessor. Is it something like tika only send tags with unique names ? 

I have also used tika-app-1.2.jar and processed the URL that has multipleValues for metatag and the Tika app displayed only the first value. I have tested with http://scholar.lib.vt.edu/ejournals/ALAN/v37n1/hazlett.html URL.

Is it something I should ask in tika ? 


                  
> nutch 1.5.1 not able to parse mutliValued metatags
> --------------------------------------------------
>
>                 Key: NUTCH-1467
>                 URL: https://issues.apache.org/jira/browse/NUTCH-1467
>             Project: Nutch
>          Issue Type: Bug
>    Affects Versions: 1.5.1
>            Reporter: kiran
>            Priority: Minor
>             Fix For: 1.7
>
>         Attachments: NUTCH-1467-TEST-1.patch, NUTCH-1467-trunk.patch, NUTCH-1467-trunk_v1.patch, NUTCH-1467-trunk_v2.patch, Patch_HTMLMetaProcessor.patch, Patch_HTMLMetaTags.patch, Patch_MetadataIndexer.patch, Patch_MetaTagsParser.patch, patch.txt
>
>
> Hi,
> I have been able to parse metatags in an html page using http://wiki.apache.org/nutch/IndexMetatags. It does not work quite well when there are two metatags with same name but two different contents. 
> Does anyone encounter this kind of issue ?  
> Are there any changes that need to be made to the config files to make it work ?
> When there are two tags with same name and different content, it takes the value of the later tag and saves it rather than creating a multiValue field.
> Edit: I have attached the patch for the file and it is provided by DLA (Digital Library and Archives) http://scholar.lib.vt.edu/ of Virginia Tech. 
> Many Thanks,

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira