Open APIs

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?

Michael, you closing tag YOGO (You Only Generate Once), is an interesting point.

In the proof of concept I am building, because of the ease with which the Java code can be generated (excluding the @ issue), I've added a Maven project that just builds the API & model classes, and it deliberately does not commit the generated source. So, if another developer checks out the code from git, they too would end up generating the source again. So it is a no go for yogo.

It would be a simple enough matter to change this, so that we only generate once, but I consider this to be somewhat sub-optimal. You should always be able to regenerate the code and use it without modifying it.

If we can address this issue through updating the TM Forum API design guidelines, I do think it would be the best outcome. I really don't like seeing UML models where a special character denotes some kind of additional trait for an entity or attribute, so if we can avoid this, I think it would be the best outcome.

I am also considering forking the swagger.io code generator, to see if we can get the Maven plugin to address this through prefix-replacement, as I am sure that the TM Forum model is not the only model to be affected by these sort of design decisions.

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?

Michael, you closing tag YOGO (You Only Generate Once), is an interesting point.

In the proof of concept I am building, because of the ease with which the Java code can be generated (excluding the @ issue), I've added a Maven project that just builds the API & model classes, and it deliberately does not commit the generated source. So, if another developer checks out the code from git, they too would end up generating the source again. So it is a no go for yogo.

It would be a simple enough matter to change this, so that we only generate once, but I consider this to be somewhat sub-optimal. You should always be able to regenerate the code and use it without modifying it.

If we can address this issue through updating the TM Forum API design guidelines, I do think it would be the best outcome. I really don't like seeing UML models where a special character denotes some kind of additional trait for an entity or attribute, so if we can avoid this, I think it would be the best outcome.

I am also considering forking the swagger.io code generator, to see if we can get the Maven plugin to address this through prefix-replacement, as I am sure that the TM Forum model is not the only model to be affected by these sort of design decisions.

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?

Michael, you closing tag YOGO (You Only Generate Once), is an interesting point.

In the proof of concept I am building, because of the ease with which the Java code can be generated (excluding the @ issue), I've added a Maven project that just builds the API & model classes, and it deliberately does not commit the generated source. So, if another developer checks out the code from git, they too would end up generating the source again. So it is a no go for yogo.

It would be a simple enough matter to change this, so that we only generate once, but I consider this to be somewhat sub-optimal. You should always be able to regenerate the code and use it without modifying it.

If we can address this issue through updating the TM Forum API design guidelines, I do think it would be the best outcome. I really don't like seeing UML models where a special character denotes some kind of additional trait for an entity or attribute, so if we can avoid this, I think it would be the best outcome.

I am also considering forking the swagger.io code generator, to see if we can get the Maven plugin to address this through prefix-replacement, as I am sure that the TM Forum model is not the only model to be affected by these sort of design decisions.

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?

Michael, you closing tag YOGO (You Only Generate Once), is an interesting point.

In the proof of concept I am building, because of the ease with which the Java code can be generated (excluding the @ issue), I've added a Maven project that just builds the API & model classes, and it deliberately does not commit the generated source. So, if another developer checks out the code from git, they too would end up generating the source again. So it is a no go for yogo.

It would be a simple enough matter to change this, so that we only generate once, but I consider this to be somewhat sub-optimal. You should always be able to regenerate the code and use it without modifying it.

If we can address this issue through updating the TM Forum API design guidelines, I do think it would be the best outcome. I really don't like seeing UML models where a special character denotes some kind of additional trait for an entity or attribute, so if we can avoid this, I think it would be the best outcome.

I am also considering forking the swagger.io code generator, to see if we can get the Maven plugin to address this through prefix-replacement, as I am sure that the TM Forum model is not the only model to be affected by these sort of design decisions.

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?

Michael, you closing tag YOGO (You Only Generate Once), is an interesting point.

In the proof of concept I am building, because of the ease with which the Java code can be generated (excluding the @ issue), I've added a Maven project that just builds the API & model classes, and it deliberately does not commit the generated source. So, if another developer checks out the code from git, they too would end up generating the source again. So it is a no go for yogo.

It would be a simple enough matter to change this, so that we only generate once, but I consider this to be somewhat sub-optimal. You should always be able to regenerate the code and use it without modifying it.

If we can address this issue through updating the TM Forum API design guidelines, I do think it would be the best outcome. I really don't like seeing UML models where a special character denotes some kind of additional trait for an entity or attribute, so if we can avoid this, I think it would be the best outcome.

I am also considering forking the swagger.io code generator, to see if we can get the Maven plugin to address this through prefix-replacement, as I am sure that the TM Forum model is not the only model to be affected by these sort of design decisions.

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?

Hi, if touching the swagger codegen is an option, take a look at the function  sanitizeName in DefaultCodegen.java (package io.swagger.codegen).

It should be possible to do stuff like

Br,
Knut Johannessen
Telenor

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?

I would suggest avoiding symbol@ in element names to avoid additional effort and complexity.

Hi, if touching the swagger codegen is an option, take a look at the function  sanitizeName in DefaultCodegen.java (package io.swagger.codegen).

It should be possible to do stuff like

Br,
Knut Johannessen
Telenor

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?

I would suggest avoiding symbol@ in element names to avoid additional effort and complexity.

Hi, if touching the swagger codegen is an option, take a look at the function  sanitizeName in DefaultCodegen.java (package io.swagger.codegen).

It should be possible to do stuff like

Br,
Knut Johannessen
Telenor

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?

I would suggest avoiding symbol@ in element names to avoid additional effort and complexity.

Hi, if touching the swagger codegen is an option, take a look at the function  sanitizeName in DefaultCodegen.java (package io.swagger.codegen).

It should be possible to do stuff like

Br,
Knut Johannessen
Telenor

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?

I would suggest avoiding symbol@ in element names to avoid additional effort and complexity.

Hi, if touching the swagger codegen is an option, take a look at the function  sanitizeName in DefaultCodegen.java (package io.swagger.codegen).

It should be possible to do stuff like

Br,
Knut Johannessen
Telenor

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?

I would suggest avoiding symbol@ in element names to avoid additional effort and complexity.

Hi, if touching the swagger codegen is an option, take a look at the function  sanitizeName in DefaultCodegen.java (package io.swagger.codegen).

It should be possible to do stuff like

Br,
Knut Johannessen
Telenor

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?

I've posted this on the GitHub for the TMF657_ServiceQualityManagement repo on GitHub, but thought it worthy of posting here, as it would impact on the API Design Guidelines.

When generating models using the Swagger definition using the swagger-codegen-maven-plugin (v2.3.1), there are some entities where the use of the @ to prefix attribute names is causing issues where there is another attribute named the same. eg. on ServiceLevelSpecParameter @type & type.

By default, when the model code is generated, in our case into Java, the preceding @ symbol is stripped, resulting in duplicate methods being generated in the code.

So, the use of @ to proceed the attributes is problematic. Would it be possible to review the use of @ to prefix attribute names? Or, should we raise this as an issue with swagger.io to modify the code generator?