So, I'm having trouble using an XML object definition with Swagger (Core 1.5.7). Here's my XML code:
<result status="ok">
<another>
<cards/>
<customer my_super_id="2349027834"/>
<someothers>
<someother name="who-is-this" />
</someothers>
</another>
</result>
And here's my yml code:
result:
type: object
description: Some random description
properties:
status:
type: string
xml:
attribute: true
another:
type: object
properties:
customer:
type: object
properties:
superId:
type: string
xml:
name: my_super_id
attribute: true
I can get the status with no problem, but the my_super_id is null because Swagger is generating the model class with the parameter in CamelCase, i.e., mySuperId instead of my_super_id.
In my generated model class I have this:
public ResultAnotherCustomer(#JsonProperty("mySuperId") final String mySuperId)
Is there any way to solve this?
I was facing the same issue in Java Springboot App using Swagger Open API Specification. I ended overidding a Bean to make it work. But again this is how I fixed in Springboot with OAS.
#Bean
public ModelResolver modelResolver(ObjectMapper objectMapper) {
return new ModelResolver(objectMapper.setPropertyNamingStrategy(PropertyNamingStrategy.SNAKE_CASE));
}
Stupid as it is, it seems I wasn't using the latest version of swagger-codegen
, so updating it to the latest version (2.1.5) fixed the problem.
Old topic, but I had the same issue when generating typescript-angularjs directly from swagger.io. Snake -> Camel conversion.
A modelPropertyNaming has to be configured to snake_case when generating code, by default it is Camel.
https://github.com/swagger-api/swagger-codegen/wiki/FAQ#typescript
Example using docker.
docker pull swaggerapi/swagger-codegen-cli
docker run --rm -v ${PWD}:/local swaggerapi/swagger-codegen-cli generate -i /local/swagger.yaml --additional-properties modelPropertyNaming=snake_case -l typescript-angularjs -o /local/typescript-angularjs/
If it can help some of you, fellows. It is great.
My experience working with the Swagger coder generator is that it gets me about 80% of what I want and then I have to do the rest myself. If you think this is a bug than you can log it here:
https://github.com/swagger-api/swagger-codegen/issues
Java generally favors CamelCase over snake_case so I'm guessing that it's intentional.
I also had this issue when generating a swagger client
The models were renamed from snake case 'aa_ba_ca' to camel case 'AaBaCa'
Error arose as no model was found when calling api_instance.method(swagger_client.model(parameters))
Error:
AttributeError swagger module has no attribute
Related
I am having an error when I try pushing my app to cloud foundry. The manifest.yml file is the right format and also referenced it with that given as an example on the ibm forum but I get this error which doesn't add up in my opinion. What am I doing wrong here? Anybody to help out please?
Error
yaml: unmarshal errors:
line 2: cannot unmarshal !!str https:/... into []string
I finally got an answer for this from someone else Unable to deploy to SAP Cloud foundry
My error was because I was writing buildpacks and giving it a value of a string url instead of an array. Once I enclosed it inside an array, it worked fine. This is because IBM cloudfoundry recently changed from using buildpack to buildpacks which meant using an array instead of a single url string.
I am upgrading from Spring 4.3 to Spring 5.3 and it seems that placeholders are no longer supported for the #ActiveProviles annotation.
The following code worked with the old Spring version:
#RunWith(SpringRunner.class)
#SpringBootTest
#ActiveProfiles({"${profileA}","someProfileWithoutPlaceholders"})
#ContextConfiguration(classes = MyApplication.class)
public class MyTest {...}
But it stopped working with the upgrade and now it get
Caused by: java.lang.IllegalArgumentException: Could not resolve placeholder 'nucleus.hibernate.dialect' in value "${nucleus.hibernate.dialect}"
at org.springframework.util.PropertyPlaceholderHelper.parseStringValue(PropertyPlaceholderHelper.java:178)
at org.springframework.util.PropertyPlaceholderHelper.replacePlaceholders(PropertyPlaceholderHelper.java:124)
at org.springframework.core.env.AbstractPropertyResolver.doResolvePlaceholders(AbstractPropertyResolver.java:239)
at org.springframework.core.env.AbstractPropertyResolver.resolveRequiredPlaceholders(AbstractPropertyResolver.java:210)
at org.springframework.context.support.PropertySourcesPlaceholderConfigurer.lambda$processProperties$0(PropertySourcesPlaceholderConfigurer.java:175)
at org.springframework.beans.factory.support.AbstractBeanFactory.resolveEmbeddedValue(AbstractBeanFactory.java:936)
at org.springframework.beans.factory.support.DefaultListableBeanFactory.doResolveDependency(DefaultListableBeanFactory.java:1321)
at org.springframework.beans.factory.support.DefaultListableBeanFactory.resolveDependency(DefaultListableBeanFactory.java:1300)
at org.springframework.beans.factory.annotation.AutowiredAnnotationBeanPostProcessor$AutowiredFieldElement.inject(AutowiredAnnotationBeanPostProcessor.java:640)
at org.springframework.beans.factory.annotation.InjectionMetadata.inject(InjectionMetadata.java:119)
at org.springframework.beans.factory.annotation.AutowiredAnnotationBeanPostProcessor.postProcessProperties(AutowiredAnnotationBeanPostProcessor.java:399)
... 55 more
Note, that 'nucleus.hibernate.dialect' is defined in the application properties referenced by "${profileA}".
As a workaround, I tried to specify the properties as test property source explicitly using
#TestPropertySource(locations={"classpath:/application-${profileA}.properties"}})
and that works.
I am not sure if using placeholders for selecting a Spring profile on a Spring integration test is an officially supported feature. If yes, I consider this a breaking change in the Spring test framework.
After debugging around the test case I got a bit more insight on this issue.
It seems that in
org.springframework.core.env.PropertySourcesPropertyResolver#getProperty(java.lang.String, java.lang.Class<T>, boolean)
the test property source 'test' splits up the profiles and has two entries:
spring.profiles.active[0] -> "${profileA}
spring.profiles.active[1] -> "someProfileWithoutPlaceholders"
So when looking up with key "spring.profiles.active", it does not find any value, and org.springframework.core.env.AbstractPropertyResolver#resolveNestedPlaceholders is not called.
This is done in org.springframework.core.env.AbstractEnvironment#doGetActiveProfiles when building the test application context..
In the old Spring version, there was only one map entry:
spring.profiles.active -> "${profileA},someProfileWithoutPlaceholders"
so it found the entry and resolved the placeholder.
I'm trying to create a Sitelet with SiteletBuilder in C#:
return WebSharper.Sitelets.Content.Page(...)
However, the class Websharper.Sitelet contains Content both as Struct and Class.
So, this does not compile.
Versions of Zafir-Libraries are
Zafir 4.0.152.29-beta5
Zafir.CSharp 4.0.152.29-beta5
Zafir.Html 4.0.56.95-beta5
Zafir.UI.Next 4.0.102.33-beta5
How to reference WebSharper.Sitelets.Content proberly?
Or is this indeed a bug?
Thanks for the report, created ticket: https://github.com/intellifactory/websharper/issues/645
I have been testing with having using WebSharper.Sitelets; and then using with shorter form Content.Page(...). C# can resolve this for some reason, although the name conflict indeed exists in WebSharper.Sitelets.dll
when deploying the following webservice
#WebService(serviceName = "TestService")
#SOAPBinding(use = Use.LITERAL, style = Style.DOCUMENT, parameterStyle= SOAPBinding.ParameterStyle.WRAPPED)
public class KekeDummyWebservice implements kekeService {...
on one of my servers I do get the following error:
javax.wsdl.WSDLException: WSDLException:faultCode=CONFIGURATION_ERROR: Unsupported Java encoding for writing wsdl file: 'ISO8859_15'.
I don't know where the 'ISO8859_15' encoding comes from. The wildfly prints out
-Dfile.encoding=ISO-8859-15
while starting. Another point is that during wildfly start the warning
[jacorb.codeset] (MSC service thread 1-7) Warning - unknown codeset (ISO8859_15) - defaulting to ISO-8859-1
can be seen.
Thanks
I had similar issue on our system, with following error in the log during deployment on Wildlfy10:
Caused by: javax.wsdl.WSDLException: WSDLException: faultCode=CONFIGURATION_ERROR: Unsupported Java encoding for writing wsdl file: 'Cp1252'
Finally it was solved by resaving of xsd/wsld in utf-8 and setting targetNamespace in WS implementaion accordingly to what was defined in xsd/wsld. This aproach is not fully applicable on your case, but maybe it helps.
Try to run command "locale". It will give "LANG=en_US.ISO-8859-15".
Update this LANG to use "en_US.UTF-8" using export.
This solved the issue.
I decided to return to Dropwizard after a very long affair with Spring. I quickly got the absolute barebones REST service built, and it runs without any problems.
Using Dropwizard 0.7.1 and Java 1.8, only POM entries are the dropwizard-core dependency and the maven compiler plugin to enforce Java 1.8, as recommended by the Dropwizard user manual
However, as soon as I try to add an Optional QueryParam to the basic controller, the application fails to start with the following error (cut for brevity):
INFO [2015-01-03 17:44:58,059] io.dropwizard.jersey.DropwizardResourceConfig: The following paths were found for the configured resources:
GET / (edge.dw.sample.controllers.IndexController)
ERROR [2015-01-03 17:44:58,158] com.sun.jersey.spi.inject.Errors: The following errors and warnings have been detected with resource and/or provider classes:
SEVERE: Missing dependency for method public java.lang.String edge.dw.sample.controllers.IndexController.index(java.util.Optional) at parameter at index 0
Exception in thread "main" javax.servlet.ServletException: com.sun.jersey.spi.container.servlet.ServletContainer-6c2ed0cd#330103b7==com.sun.jersey.spi.container.servlet.ServletContainer,1,false
The code for the controller is as follows:
#Path("/")
public class IndexController {
#GET
#Timed
public String index(#QueryParam("name") Optional<String> name) {
String saying = "Hi";
if(name != null && name.isPresent()) {
saying += " " + name.get();
}
return saying;
}
}
If I remove Optional from the mix, the application runs just fine. I replace the Optional-specific code with null checks and it works perfectly.
Am I missing something fundamental here? Both Google Guava Optional and java.util.Optional fail with the same error. (And yes, I did narrow it down to the Optional object)
A quick Google/SO search yielded nothing useful, but feel free to point me to a resource I may have missed
Thanks in advance!
Moments after posting this, I found that the issue was my use of Java 1.8. If using Java 1.8, I have to add the Java8Bundle to my app:
POM Entry:
<dependency>
<groupId>io.dropwizard.modules</groupId>
<artifactId>dropwizard-java8</artifactId>
<version>0.7.0-1</version>
</dependency>
And code in the Application class:
#Override
public void initialize(Bootstrap<SampleConfiguration> bootstrap) {
bootstrap.addBundle(new Java8Bundle());
}
See: https://github.com/dropwizard/dropwizard-java8
This enables both Google Guava Optional and java.util.Optional to work just fine.
If I revert to Java 1.7 and use the Google Guava Optional, it works just fine as well and I don't have to include the Java8Bundle. I'll opt for the Java8Bundle for now, though, as using Java8 features is lucrative for me :)
Cheers!