'Is there any chance where the name of documentum metadata gets changed
while reaching the output connector'
Hi Tamizh,
If you have a Metadata Adjuster transformation connector in your pipeline,
then you might be mapping the metadata name to something else. But that is
the only way.
The metadata names that are available from Documentum for a document type
we find using this DQL query:
String strDQL = "select distinct attr_name FROM dmi_dd_attr_info
where type_name = '" + docType + "' order by attr_name asc";
If there's any way that fetching the metadata value must be done using some
private or internal attribute name instead on your setup, then this code
might be incorrect. You might want to inspect your dmi_dd_attr_info table
to see if there are any columns that seem like they might be the internal
name of the attribute. If so, please let me know, and we can try an
experiment to see whether that's what the trouble is.
Karl
On Fri, Nov 18, 2016 at 12:43 AM, Tamizh Kumaran Thamizharasan <
[email protected]> wrote:
> Hi Karl,
>
> DFC which we are using is compatible with the documentum.
>
> Is there any chance where the name of documentum metadata gets changed
> while reaching the output connector?
>
> Thanks,
> Tamizh
>
> On Nov 17, 2016 4:52 PM, "Karl Wright" <[email protected]> wrote:
>
>> Hi Tamizh,
>>
>> The code which fetches attribute values uses straight access to object
>> attributes via the DFC IDfPersistentObject API.
>>
>> Here's the actual code. "object" is an object of type
>> IDfPersistentObject.
>>
>> >>>>>>
>> int valueCount = object.getValueCount(attribute);
>> String[] values = new String[valueCount];
>> int y = 0;
>> while (y < valueCount)
>> {
>> // Fetch the attribute.
>> // It's supposed to work for all attribute types...
>> String value = object.getRepeatingString(attribute,y);
>> values[y++] = value;
>> }
>> return values;
>> <<<<<<
>>
>> Is there any chance that your version of DFC is not compatible with the
>> version of Documentum you are trying to crawl?
>>
>> Karl
>>
>>
>> On Thu, Nov 17, 2016 at 6:13 AM, Karl Wright <[email protected]> wrote:
>>
>>> Hi Tamizh,
>>>
>>> I do not have a documentum instance here. All I can do is help you
>>> debug what is happening.
>>>
>>> The code which puts the metadata into the RepositoryDocument object is
>>> this:
>>>
>>> >>>>>>
>>> rval = new RepositoryDocument();
>>>
>>> if (contentType != null)
>>> rval.setMimeType(contentType);
>>>
>>> List<String> attributeDescriptions =
>>> sDesc.getMetadataFields(typeName);
>>> if (attributeDescriptions != null)
>>> {
>>> for (String attrName : attributeDescriptions)
>>> {
>>> // Fetch the attributes from the object
>>> String[] values = object.getAttributeValues(attrName);
>>> // Add the attribute to the rd
>>> rval.addField(attrName,values);
>>> }
>>> }
>>> <<<<<<
>>>
>>> There's unfortunately no log output here that would help us see whether
>>> or not the attributeDescriptions variable gets set with the proper list of
>>> attributes, but since you see it display correctly in the UI I bet that is
>>> fine. I will have to look at what object.getAttributeValues() does in the
>>> server process to see what DSQL query it runs. Stay tuned.
>>>
>>> Karl
>>>
>>>
>>> On Thu, Nov 17, 2016 at 5:58 AM, Tamizh Kumaran Thamizharasan <
>>> [email protected]> wrote:
>>>
>>>> Hi Karl,
>>>>
>>>> I can see the metadata under view page.
>>>> Even after selecting all metadata check box, I am facing the same issue.
>>>>
>>>> Can you please share me steps to index documents from documentum?
>>>> So that I can cross check the same.
>>>>
>>>> Thanks,
>>>> Tamizh
>>>>
>>>> On Nov 17, 2016 4:20 PM, "Karl Wright" <[email protected]> wrote:
>>>>
>>>>> Hi Tamizh,
>>>>>
>>>>> Can you view the job and make sure when you do that the metadata
>>>>> fields you have selected shows up on the view page? The UI for this is a
>>>>> bit tricky and I have seen users not use it properly before.
>>>>>
>>>>> Thanks,
>>>>> Karl
>>>>>
>>>>>
>>>>> On Thu, Nov 17, 2016 at 3:27 AM, Tamizh Kumaran Thamizharasan <
>>>>> [email protected]> wrote:
>>>>>
>>>>>> Hi Karl,
>>>>>>
>>>>>> Thanks for your reply on short time.
>>>>>>
>>>>>> I can index content and some metadata from Documentum, but problem
>>>>>> arises when I try to index the metadata listed under job>document
>>>>>> types>specific metadata or all metadata.
>>>>>>
>>>>>> I cannot see any errors in both Solr and Manifoldcf.
>>>>>>
>>>>>> I am using the binary distribution of Apache Manifoldcf 2.5
>>>>>>
>>>>>> Thanks,
>>>>>> Tamizh
>>>>>>
>>>>>> On Nov 17, 2016 12:27 PM, "Karl Wright" <[email protected]> wrote:
>>>>>>
>>>>>>> Hi Tamizh,
>>>>>>>
>>>>>>> I know of a number of people who are indexing documents and metadata
>>>>>>> from documentum. I also know of many people who are indexing documents
>>>>>>> and
>>>>>>> metadata into Solr. So I know this works. But there are a lot of
>>>>>>> places
>>>>>>> you will need to configure correctly for this to work the way you
>>>>>>> expect.
>>>>>>>
>>>>>>> For a start, I would verify that Solr is getting sent the
>>>>>>> information you expect. Please have a look at the Solr logs,
>>>>>>> specifically
>>>>>>> the INFO statements that describe the posted URLs that come from
>>>>>>> ManifoldCF. There should be one for each document that gets sent.
>>>>>>> These
>>>>>>> URLs contain a lot of "parameter=value" fields. Please check that they
>>>>>>> contain the metadata you want to index.
>>>>>>>
>>>>>>> If the Solr logs have the metadata but Solr isn't indexing it, then
>>>>>>> you will need to look at your Solr configuration, specifically your solr
>>>>>>> schema. If the Solr logs don't have the metadata, then let us know and
>>>>>>> we'll have to look at your Documentum job specification.
>>>>>>>
>>>>>>> Thanks,
>>>>>>> Karl
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On Wed, Nov 16, 2016 at 11:48 PM, Tamizh Kumaran Thamizharasan <
>>>>>>> [email protected]> wrote:
>>>>>>>
>>>>>>>> Hi,
>>>>>>>>
>>>>>>>>
>>>>>>>> I am trying to index documentum content to Apache Solr through
>>>>>>>> manifoldcf.
>>>>>>>>
>>>>>>>>
>>>>>>>> I can able to index documents but couldn't index all the metadata
>>>>>>>> from documentum to solr.
>>>>>>>>
>>>>>>>>
>>>>>>>> Any help on the above mentioned issue will be highly appreciated.
>>>>>>>>
>>>>>>>>
>>>>>>>> Thanks,
>>>>>>>>
>>>>>>>> Tamizh
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>
>>>
>>
