In the context of Spring Webflow 2.0.x......
I handle form binding "typemismatches", i.e. as a result of trying to map a String onto a Integer field, by using the following in my messages.properties
typeMismatch={0} contains invalid data.
This works fine.
The problem is that if the field that the typeMismatch error occurred on was "required" then I also receive an error for the missing required field, which is logical I guess because the value that was submitted was never bound. ("Required" being defined in a Commons Validation XML file)
So, I dont want to see the "XXX is required field" error message when the field is only missing due to the typeMismatch. How do I resolve this? I thought about overriding initBinder() on the FormAction but quickly got nowhere.....
Like Yves mentioned, among the three approaches, i have used a custom validator method and its very easy. You can use a custom validator which checks if the form field already has a xml error message of required. If the field does not have an error, then you can check for your string validation. That way it will display only one.
The other method that you could use is try a multiple xml validation, one being required and the other one being a mask which checks for a particular regular expression. In your case if your field is an integer field, then you can go and perform a mask with regex checking for only numbers. The order of mask, required or required, mask in the xml decides which message gets a higher preference.
For example:
<field property="somefield" depends="required,mask" page="2">
<arg key="somelabel"/>
<var>
<var-name>mask</var-name>
<var-value>${somepattern}</var-value>
</var>
</field>
You have many options, in order of preference:
Set selectively the message typeMismatch.target.yourFieldName or typeMismatch.int in resources files
Implement your own Validator so that you can send a dedicated message when Integer parsing will fail before the binding step
Create a BindingErrorProcessor to handle different kind of parsing issues
Related
everyone.
So, I have a SpringBoot application with a controller that has several methods, taking the following POJO as a parameter:
package com.example.dto;
import lombok.Data;
#Data
public class MyEntity {
#NotNull
private String fieldA;
private String fieldB;
}
For one of the controller endpoints I would like to apply additional validation logic, so in the validation.xml I add the following:
<constraint-mappings>
<bean class="com.example.controller.SampleController" ignore-annotations="false">
<method name="doFoo">
<parameter type="com.example.dto.MyEntity">
<valid />
<constraint annotation="com.example.validation.ValidEntity" />
</parameter>
</method>
</bean>
</constraint-mappings>
com.example.validation.ValidEntity is the constraint annotation I would like to apply.
My problem is that this additional constraint is only invoked if #NotNull checks defined in MyEntity have passed successfully. If fieldA is null, ValidEntity constraint is ignored, and the client receives an imcomplete validation result. What am I missing?
I'm not entirely sure about this because I've never worked with the validation.xml file.
However, I would say that Spring is first creating the object and then applying the validations. The #NotNull validation is performed in the creation of the instance. This means that if that validation fails the construction will throw an exception and Spring won't even try to check your constraint (which makes sense in my opinion).
I think you can "fix" it by creating an annotation with your constraint and using it in your class. If I'm right, both annotations will be checked and the thrown exception will contain all errors.
It's just a guess. Let me know if it works.
I don't know if there is an easy way to configure the validator to aggregate constraint violations from both annotation and XML configurations when first or both fails.
As demonstrated by your code Hibernate Validator can work with mixed annotation and XML configurations, but the lack of documentation for that specific case is a hint that it is at least not recommended.
When XML configuration file is used, it takes precedence over annotations by default. ignore-annotations is used to overcome this (text highlight is mine):
Setting ignore-annotations to true means that constraint
annotations placed on the configured bean are ignored. The default for
this value is true. ignore-annotations is also available for the nodes
class, fields, getter, constructor, method, parameter, cross-parameter
and return-value. If not explicitly specified on these levels the
configured bean value applies.
Using Hibernate Validator to Cover Your Validation Needs article states that:
The default for a field is ignore-annotations=”false”. This means
that by default annotations for a field are stronger (this is of
course after you indicated that that the bean itself wont ignore
annotations). If you wont that the XML will be stronger than you have
to indicate that by ignore-annotations=”true”
It seems possible to disable annotation configuration for a specific field which is configured in XML.
Another solution to switch between annotation and XML configuration is to use Grouping constraints.
I'm not sure if anything of the above is of any use for you, but if it is possible I would probably switch to a single configuration (XML, assuming that annotation config comes from external library you cannot modify) and enforce it everywhere instead of relying on undocumented features.
My requirement is as below:
I have a properties file which has values as:
/message/header/messagetype ='DATA'
So the XPaths are the keys and the values are the dates I want to check in Camel at runtime
I read an ActiveMQ queue which gets me an XML message. Now at runtime I need to check the respective XPaths from the properties file and check there respective values in the XML message I get from the queue. There could be any number of XPath checks defined in the properties file. This is defined by the business needs and the code should take care of it.
There is no standard way to read a properties file and "transform" it to XPath checks. However, you can always use a plain Java Bean and call it from your Camel route. For example like this:
from()
...
.bean(YourBean.class)
...
With this you can use Java code as a "component" in your Camel route.
For more details see the Camel docs about Bean binding and the Bean component.
Thank you for the help. I figured out how to do this by using Predicate functionality in Camel.So what I am doing now is, creating multiple Predicate objects based on my property file entries and adding them to 'PredicateBuilder' as below ::
PredicateBuilder.and(p1,p2....pn);
This returns me an object of Predicate which I can use in when() of Camel.
for eg. choice().
when(predicateBuilder()).
I am trying to make my #webparam mandatory using (#XmlElement(required=true) but the generated XSD still shows this as minOccurs="0". Also tried setting nillable=false but still not working.
Here is my web method :
#WebMethod
#WebResult(name = "Biller")
public Biller getBiller(#XmlElement(required=true) #WebParam(name = "billerId") Integer billerId){}
Please suggest.
As specified by me in comments - either you need to wrap your Integer variable into some Java POJO and apply rules for specific fields in that POJO or change to primitive type because - reference types are always optional but constituent fields of wrapper types can be made required or optional.
Primitive types are always required.
Refer this answer
Then comes the question about default values - and for that refer Answers to this question and summary is - if nothing is specified -
minOccurs and maxOccurs default to 1
Now why your call succeeds with SOAP UI - Your xsd is correct and that is acknowledged by SOAP UI so my guess is client might be attaching some default values when its missing. I haven't used SOAP UI. Try examining your request ans see if value is really missing. If value is indeed missing in request then try to examine as why validation is not kicking in.
Is there any way of getting the metadata for a solr core ?
For instance I know the core name, and can obtain a SolServer from that and I also know the field name.
Is there any way to determine the metadata though. Specifically I would like to know whether the field type is an int or a double.
Thanks
You can make a request to the luke request handler:
http://localhost:8983/solr/corename/admin/luke?show=schema&wt=json&_=1453816769771
The output will include the schema for the core, along with the defined fields, their settings and their types:
{"fields":{"xyz":{"type":"string","flags":"I-S-M---OF-----l","copyDests":[],"copySources":[]}, .... }
A neat trick to find these endpoints is to watch the 'network' tab when browsing the admin interface to Solr, as the admin interface is just a static HTML / Javascript frontend that makes all the requests for actual content from the Solr server behind the scenes.
I am working on a project which made use of an old (but nice) framework Struts 1.3, and I am trying to accomplish a simple task.
I have a bean containing a property, which is a key corresponding to a property file. I would like to use it for recall the appropriate translation.
Variable in property file
props.myprop.sample=This is my sample property
The property is in a bean passed to the jsp called for convenience AllProps which has a getter for the property, and this should be a pseudo code:
<bean:define id="sample" name="AllProps" property="sample" type="java.lang.String"/> // should result in sample = props.myprop.sample
<div><bean:message key="sample"/></div>
Which should output:
<div>This is my sample property</div>
But obviously result in a property not found, can you give me help on how to deal with this ?
I would like to stick as much as possible to Struts tag, then Jsp tag, and scriptlet as last resource.
Thanks
Straight from the documentation:
<bean:message>
Render an internationalized message string to the response.
Retrieves an internationalized message for the specified locale, using
the specified message key, and write it to the output stream. Up to
five parametric replacements (such as "{0}") may be specified.
The message key may be specified directly, using the key attribute, or
indirectly, using the name and property attributes to obtain it from a
bean.
(emphasis mine)
So, all you need is
<bean:message name="AllProps" property="sample"/>