Issues with DateStamp transformation

[torrin47]

We're seeing records where the datestamp isn't being accurately transformed, and are a little confused about why. Here's an example, where data.gov harvested a CSDGM record:
https://catalog.data.gov/harvest/object/9c6483ec-83d4-437c-bc16-fbfe40d5b896/original
with a //metd value of 2016-11-02, but the corresponding transformed ISO document:
https://catalog.data.gov/harvest/object/9c6483ec-83d4-437c-bc16-fbfe40d5b896
reports a //dateStamp value of 9999-01-01.

The XSLT doc looks ok, so what's causing this problem?
https://github.com/GSA/ckanext-geodatagov/blob/master/conversiontool/fgdc2iso/fgdcrse2iso19115-2.xslt

      <gmd:dateStamp>
        <xsl:choose>
          <xsl:when test="(fn:contains(fn:lower-case(fn:normalize-space(fn:string(//metd))), 'unknown'))">
            <xsl:attribute name="gco:nilReason">
              <xsl:sequence select="xs:string('unknown')"/>
            </xsl:attribute>
          </xsl:when>
          <xsl:otherwise>
            <gco:Date>
              <xsl:call-template name="fgdc2isoDate">
                <xsl:with-param name="dateField" select="normalize-space(//metd)"/>
              </xsl:call-template>
            </gco:Date>
          </xsl:otherwise>
        </xsl:choose>
      </gmd:dateStamp>

[kvuppala]

@torrin47 @FuhuXia
We are looking into the issue, the sample catalog URLs provided are not working now.

Can you provide the metadata record example link to troubleshoot the issue.

[torrin47]

I believe this is the equivalent example record in the new system:

https://catalog.data.gov/dataset/national-emissions-inventory-u-s-2014-epa-oar-oaqps-aqad

[charness39]

Here are a few more examples:

https://admin-catalog.data.gov/dataset/jobs-within-a-30-minute-transit-ride-service

https://catalog.data.gov/dataset/walkability-index

https://catalog.data.gov/dataset/smart-location-database-service

The records above also do not appear to have the links to the ISO or Original FGDC metadata.

[amilan17]

I believe that the fgdc2isoDate template is expecting a valid CSDGM date format, which in the example above would be 20161102 (without the dashes). It doesn't know how to handle ISO valid dates in the source CSDGM document...

[kvuppala]

@amilan17 thank you, that seem to be the case.

@charness39 @torrin47 I see that publication date is parsed correctly on the example https://catalog.data.gov/dataset/national-emissions-inventory-u-s-2014-epa-oar-oaqps-aqad , would you be able to update the metadata date and try reharvesting?

We will look into the other two records where the ISO / FGDC metadata links are not getting displayed.

[torrin47]

Sure, we actually have a script that updates all of those dates to reflect the date the record was last edited or updated from our catalog’s perspective, because we found if stewards made edits without updating that element, Data.gov wouldn’t recognize the edits. And we were under the impression that CKAN required ISO dates. But really quick and easy for us to adjust that globally for all our CSDGM docs. We’ve done so and reharvested, but the records don’t appear to be updated at this time – should we expect a bit of a lag? Harvest Report: https://admin-catalog.data.gov/harvest/environmental-dataset-gateway-fgdc-csdgm/job/340586a2-9c92-4025-b47e-12660ed3f2bd Data.gov record: https://catalog.data.gov/dataset/national-emissions-inventory-u-s-2014-epa-oar-oaqps-aqad Harvest Source: https://edg.epa.gov/WAFer_harvest/FGDC/epa-national-emissions-inventory-region-1-2011.xml From: Kishore K. Vuppala [mailto:[email protected]] Sent: Wednesday, April 26, 2017 11:08 AM To: GSA/ckanext-geodatagov <[email protected]> Cc: Hultgren, Torrin <[email protected]>; Mention <[email protected]> Subject: Re: [GSA/ckanext-geodatagov] Issues with DateStamp transformation (#104) @amilan17<https://github.com/amilan17> thank you, that seem to be the case. @charness39<https://github.com/charness39> @torrin47<https://github.com/torrin47> I see that publication date is parsed correctly on the example https://catalog.data.gov/dataset/national-emissions-inventory-u-s-2014-epa-oar-oaqps-aqad , would you be able to update the metadata date and try reharvesting? We will look into the other two records where the ISO / FGDC metadata links are not getting displayed. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub<#104 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/ACB0vEZwA_1SB5sfJ1t3mOb__0jsAa_Gks5rz13ggaJpZM4LBt1i>.

[FuhuXia]

WAF harvester looks for timestamp for each xml file to determine whether update is needed. The file epa-national-emissions-inventory-region-1-2011.xml has a timestamp of 6/10/2016 7:15 AM, therefore it is skipped by harvester.

https://edg.epa.gov/WAFer_harvest/FGDC/

[torrin47]

Ok. We have code that ensures that the file timestamp and the embedded last update date actually reflect the point in time when the metadata was last edited in our database. The recent fix just changed the syntax of the listed date, but continued to align it with the datestamp in the database. We’ll set all dates to today to force a full, comprehensive update of every CSDGM record, then revert to the previous approach once the records are all updated. From: Fuhu Xia [mailto:[email protected]] Sent: Wednesday, April 26, 2017 5:14 PM To: GSA/ckanext-geodatagov <[email protected]> Cc: Hultgren, Torrin <[email protected]>; Mention <[email protected]> Subject: Re: [GSA/ckanext-geodatagov] Issues with DateStamp transformation (#104) WAF harvester looks for timestamp for each xml file to determine whether update is needed. The file epa-national-emissions-inventory-region-1-2011.xml has a timestamp of 6/10/2016 7:15 AM, therefore it is skipped by harvester. https://edg.epa.gov/WAFer_harvest/FGDC/ — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub<#104 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/ACB0vE_6RTy1avoG_4rP_ACyoQbkgsjiks5rz7OsgaJpZM4LBt1i>.