codeclass | Golang语言社区 -- 腾讯课堂、网易云课堂、字节教育课程PPT及代码

It is better to use doOnResponse or doAfterResponseSuccess it depends on the use case which one is more appropriate.

However waiting on RELEASED should not be a problem

If the connection is released after an error status code, and the observer is closing that connection, can a new request acquire the same connection in parallel? Does the framework internally handle things gracefully or is this a race condition that invalidates the above approach?

The connection pool runs with FIFO leasing strategy by default, so if there are idle connections in the pool you will not obtain the same connection, which is not the case if you switch the connection pool to LIFO leasing strategy. When acquiring, every connection is checked whether it is active or not and only an active connection will be provided for use.

UPDATE:

You can try also the approach below which uses ONLY WebClient API and not Reactor Netty API:

The main issue with your query is that you are missing join conditions between the tables: this results in a cartesian product between the 3 tables. Note that, as far as concerns, you don't need to bring in the eleve table to generate the expected resultset.

Also, I think that you need aggregation to gather together notes of each (eleve, annee) across all codeMats.

I think that the following query does what you want:

The best way to extend a schema depends on the way the schema is written, and in particular whether it is designed with extensibility in mind. This schema doesn't explicitly permit elements to be added in a third party namespace (that is, the schema appears to contain no wildcard elements), but at least all the types like AirspaceAssociationType are global, which means you can (if you choose) define new types that are derived from these by extension. If you go down this route, then your instance documents will have to name the extended type in an xsi:type attribute, which is pretty ugly.

Given what you say about the project and given the way this particular schema is written, I think my preferred approach would be to write an XSLT transformation that creates a variant of AIXM-Features.xsd that either explicitly includes your extensions, or that adds wildcards (xs:any elements) making your extensions valid.

If you're prepared to use XML Schema 1.1 (which would restrict your choice of schema processors) then you could consider defining all these types to have an open content model which allows elements in a third-party namespace to appear after the standard schema-defined elements. That can be done simply by adding an xs:defaultOpenContent element to the schema - you could define a new schema document that contains this declaration and then does an xs:include on the standard AIXM-defined schema.

UPDATE

I made a mistake here. Reading the spec again, it seems the defaultOpenContent element is scoped to a schema document (module): it only affects xs:complexType definitions within the same module, not in included modules. So it looks as if a change to AIXM-Features.xsd is needed after all.

If you do use defaultOpenContent, the xs:any element can specify a namespace, and can specify processContents="strict", so you can ensure that the elements you add to the model are validated. Of course you'll also need to ensure that the aggregate schema includes your own element declarations as well as the AIXM-defined ones.