INSPIRE-MIF / helpdesk-validator

Community discussion forum for INSPIRE validation issues
42 stars 23 forks source link

Local install of Validator fails test 'md common req C.1: XML Schema' for some dataset metadata #862

Closed petbod-lm closed 11 months ago

petbod-lm commented 2 years ago

We have setup a local instance of the Inspire Reference validator. We have found that certain datasets fail the test 'md common req C.1: XML Schema' on our local validator but when I test the same dataset in the EU Inspire Reference validator the dataset passes this test.

URL to dataset that fails C1: XML Schema locally but passes in EU Validator: https://www.geodata.se/geodataportalen/srv/eng/csw-inspire?request=GetRecordById&service=CSW&version=2.0.2&elementSetName=full&outputSchema=csw:IsoRecord&id=0dde1643-6323-4208-8d5b-daf0c9b718f8

Here is the Error Message from the test: The metadata record set has 1 record(s) with errors for this assertion. XML document 'csw-inspire.xml', record '0dde1643-6323-4208-8d5b-daf0c9b718f8': 'Validation failed: http://schemas.opengis.net/iso/19139/20070417/srv/1.0/serviceMetadata.xsd, 33:49: src-resolve: Cannot resolve the name 'gco:AbstractObject_Type' to a(n) 'type definition' component.'.

I have validated the metadata in XML Spy and it validates correctly.

I have updated this same dataset to use the same namespace declarations as the shown in the Dataset_metadata_2.0_example.xml (https://github.com/INSPIRE-MIF/helpdesk-validator/blob/master/examples/Dataset_metadata_2.0_example.xml) and the dataset continues to fail the XML Schema test.

I am at a loss as to why some datasets fail XML Schema validation while others pass while thier namespace declarations are the same.

Can you provide advice on how to I could find more detailed error logs from the Validator or what I should change/add to our local installation?

Thanks /Pete

petbod-lm commented 2 years ago

And here is a dataset metadata post that passes XML Schema validation in both our local validator and the EU Reference Validator https://www.geodata.se/geodataportalen/srv/eng/csw-inspire?request=GetRecordById&service=CSW&version=2.0.2&elementSetName=full&outputSchema=csw:IsoRecord&id=1f185d81-c699-48e3-9ada-c9e1ae297557

fabiovinci commented 2 years ago

Dear @petbod-lm,

could you please make a test using the official APISO schema?

You should replace the "schemaLocation" attribute in your metadata with the following value:

xsi:schemaLocation="http://www.isotc211.org/2005/gmd http://schemas.opengis.net/csw/2.0.2/profiles/apiso/1.0.0/apiso.xsd"

petbod-lm commented 2 years ago

Hi Fabio, I have modified a local copy of the dataset metadata and changed the schema location entry to use the official APISO schema xsi:schemaLocation="http://www.isotc211.org/2005/gmd http://schemas.opengis.net/csw/2.0.2/profiles/apiso/1.0.0/apiso.xsd"

I then submitted this updated metadata for validation on both the EU Inspire Reference Validator and our internal server with Inspire Valdator. The metadata continues to pass validation on the EU Inspire Reference Validator and continues to fail validation on our server. The metadata now fails validation in XML Spy as well due to the xlink namespace being already declared.

The updated metadata (with offical APISO schema) validates on the EU Inspire Reference Validator but not on our local installation . What should I look at next? Thanks /Pete

fabiovinci commented 2 years ago

Dear @petbod-lm,

I have compared the two metadata, but there are no substantial differences that can lead to different results.

Could you please download and send the report of the failed metadata?

petbod-lm commented 2 years ago

I have validated the dataset using the GetRcordById URL https://www.geodata.se/geodataportalen/srv/eng/csw-inspire?request=GetRecordById&service=CSW&version=2.0.2&elementSetName=full&outputSchema=csw:IsoRecord&id=0dde1643-6323-4208-8d5b-daf0c9b718f8 on our local installation of the Inspire Reference validator and attached are the downloaded report log file.
I can see the failure but there is no detail provided by the ETF tests about why the xml schema test fails.

Test run on 1457 - 26.10.2022 with test suite Conformance Class 2 INSPIRE data sets and data set series interoperability metadata..txt

fabiovinci commented 2 years ago

Can you send the report in html format?

petbod-lm commented 2 years ago

Test run on 1457 - 26.10.2022 with test suite Conformance Class 2 INSPIRE data sets and data set series interoperability metadata..html.zip

I have to zip the html file to attach it to the comment /Pete

fabiovinci commented 1 year ago

Dear @petbod-lm,

the different behaviour of the two instances is very strange, could you please try to install the latest version to verify if the error persists?

petbod-lm commented 1 year ago

Hej Fabio, I have updated the local instance to use 2023.1 already and continue to see the same XML Schema validation error. Can we organise a bilateral meeting together? I can show you how we have implemented our local instance and hopefully identify where we should look deeper to find the source of the problem

petbod-lm commented 1 year ago

Hello, can we origanise a bilateral meeting soon to review our setup of the ETF Validator? We continue to see the metadata failing 'md common req C1:XML Schema' regularly (but not all the time - intermittently) and I cannot see any cause in the logs on the ETF validator server.

jenriquesoriano commented 1 year ago

Dear @petbod-lm ,

based on the information shared, we cannot deduce what the issue could be. We kindly request you to contact inspire.jrc AT guadaltel.com to schedule a bilateral meeting and review your specific setup in detail.

Best regards,

petbod-lm commented 1 year ago

Thank you for the contact information.

petbod-lm commented 11 months ago

Hi, we have solved the problem! I have updated the proxy settings in the Docker file and our network team have also made some changes and now we can load XML schemas without error. Thanks for the suggestions provided in our meeting. /Pete

jenriquesoriano commented 11 months ago

These are very good news Peter, thanks a lot. In case you have further comments, please, let us know.

Best regards,