Dozer partially case sensitive in 5.5.1

[alangibson]

Field mappings are apparently now partially case sensitive in Dozer 5.5.1. There appears to be no option to switch back to the 5.4.0 case-insensitive mappings.

Example from discussion at https://groups.google.com/forum/#!topic/dozer-mapper/k6i_Bf6ZIMc.

import org.dozer.DozerBeanMapper;
public class App {
public static void main(String[] args) {
DozerBeanMapper mapper = new DozerBeanMapper();
UpperCaseBean src = new UpperCaseBean();
src.setMYSTRING("This is from the upper case bean");
LowerCaseBean dst = new LowerCaseBean();
mapper.map(src, dst);
System.out.println(dst.getmystring());
}
public static class UpperCaseBean {
private String MYSTRING;
public String getMYSTRING() {
return MYSTRING;
}
public void setMYSTRING(String MYSTRING) {
this.MYSTRING = MYSTRING;
}
}
public static class LowerCaseBean {
private String mystring;
public void setmystring(String mystring) {
this.mystring = mystring;
}
public String getmystring() {
return this.mystring;
}
}
}

With the following mapping, the "mystring" field will not be mapped:

UpperCaseBean LowerCaseBean

However, with the following mapping, the field "mystring" will be mapped:

UpperCaseBean LowerCaseBean mystring mystring

This seems like very counter-intuitive behavior. It would be nice to simply add a configuration option to control this behavior.

[brian428]

This seems like kind of a HUGE problem, since it totally breaks automatic mapping for any properties with different cases in the getters or setters.

[madhurimav]

I am also facing the same problem with version 5.5.1 .please share the fix for this case sensitive issue.
Is there a configuration flag available to map the case sensitive fields?

[brian428]

Does anyone know if this has been fixed? The only workaround we've found is to manually list out every single property name in the mapping files, which is obviously a really undesirable option.

[obozek]

+1

[RaphC]

I'm also facing the same issue

[brian428]

I getting the impression that Dozer development is stopped, unfortunately. I see there's only been 4 commits in the last year. :-( It might be fork time...

[garethahealy]

@brian428 ; I am trying to get Dozer active again. Happy for any pull requests / help.

[garethahealy]

Closing due to inactivity.

[grubeninspekteur]

I'd like to maintain backwards compatibility by optionally supporting case insensitive mapping. I can think of two ways to implement this:

• add a configuration flag "wildcard-case-insensitive" to global and per-mapping configurations.
• use a wildcard mapping strategy interface that takes the lists of unmapped fields, src and dest classes and determines pairs of src/dest fields to map. The implementation could be configured at global or class-pair-mapping level. A case sensitive strategy would be provided as default and the case insensitive one as an alternative.

@garethahealy Which would you prefer? Also, I'm unsure what should happen if a class defines two parameters that are equal except for their case. This affects both the filter step for already mapped fields as well as the field alignment.

[garethahealy]

@grubeninspekteur ; Configuration options sounds best.