The files you may edit to adapt the validation are as follows:
resources\v1.2 contains the files for version 1.2.resources\config.properties contains configuration that you may want to adapt.To publish changes commit and push your updates. In 1-2 minutes the online service will be updated.
The online service is accessible as follows:
Compliance of a dataset to a data standard is often implemented as a data validation check. Data validation is an process step happening in the course of an data exchange. Every data exhange happens within its context. It is that context that determines the executed checks.
The context and the executed checks determine a level of trust in the compliance of a dataset to the data standard. If the validation checks are weak and do not cover all cases, the trust will be low. But naively executing all checks in a specification might trigger errors/warnings on data that is not present in the exchange. Blocking so unnecessary the data exchange or forcing the exchanged data to contain dummy data. An example of the latter case happens when the data standard covers a wide range of entities and the to-be-validated dataset contains only information about 1 entity. Errors relating to the not-supplied entities are not usefull for the dataset owner.
So validation is not a unique single process fitting all data exchange processes. One cannot set from the data standard a unique set of validation rules perfectly matching all data exchange processes.
Expressing the expectations in a data exchange process is thus key. To support this communication for DCAT-AP several configurations of data validations are proposed. The dimensions of the configurations are the kind of validation checks and the incorporation of background knowledge (vocabularies). These configurations can initiate discussions on the definition the validation checks to be performed for your data exchange.
DCAT-AP 2.1.1 Base Zero (no background knowledge)
All constraints that must be satisfied to be technical coherent, excluding range class membership constraints and usage of controlled vocabularies.
DCAT-AP 2.1.1 Ranges Zero (no background knowledge)
All range class membership constraints. These are separated from the previous set because this information almost never leads to a validation error. The declaration of a sender that a value V is a member of the class R cannot be contradicted by an independent external source. Therefore to minimise the data volume between sender and receiver, the reciever can rely on that (trust that) the received value is a member of the range class. And therefore these rules should not be executed because they are always valid. Observe that validating or not these constraints is an implementation choice for the exchange between sender and reciever.
DCAT-AP 2.1.1 Base (with background knowledge)
Case 1 extended with background knowledge. The background knowledge is compiled of all the vocabularies that are used in DCAT-AP.
DCAT-AP 2.1.1 Ranges (with background knowledge)
Case 2 extended with background knowledge.
DCAT-AP 2.1.1 Recommendations (with background knowledge)
All constraints concerning recommended properties.
DCAT-AP 2.1.1 Controlled Vocabularies
All constraints concerning controlled vocabularies. This option will load also the controlled vocabulary in order to validate if a value is a member of the controlled vocabulary
DCAT-AP 2.1.1 Full (with background knowledge)
The union of cases 3, 4, 5 and 6.