* Triple quote python strings in defaults
When the model contains a string with a default value with new lines in
it, simply quoting generates invalid python code. Using triple quotes
for all string defaults makes sure the generated code will be a valid
python string.
Fixes#4862
* Wrap default string in triple quotes when it contains new lines
* Add newly generated python sample files
* Port the optional/enum template logic from async-scala to the regular scala client
* oops, accidently pulled template from akka-scala which has custom logic for enum
Just going to avoid enum for now, considering I'm not sure what scala client
will be official
* Update tests to use option
according to the help message setting dateLibrary to 'legacy' should
make the code generator use java.util.Date.
before this change, when setting dateLibrary to 'legacy' made the
generator use javax.xml.datatype.XMLGregorianCalendar.
now, the generator uses java.util.Date such that the documentation and
behavior are consistent.
* Update samples for Scalatra after 2.2.2.
* [scalatra] Revert changes to build.sbt
Changes were made in conjunction with the new finch generator. This
reverts those changes to Scalatra's previous build.sbt.
* Update scalatra samples again after revert of part of #3905.
* Improved ExampleGenerator:
- Now takes into account enum and uri/url formats for strings.
- Uses example for referenced objects if available.
- Proper examples get generated for specific numeric formats, because more specific formats now get checked before generic format.
- Honors min and max values for numerical properties, if set.
* Ran script `bin/nodejs-petstore-server.sh`.
* Renamed log to logger to conform to coding standard.
Superagent doesn't encode form parameters properly. For example it doesn't encode objects which contain array. The following is a failed test case:
var my_opts = {
'storyId': 65441,
'id': [56, 352]
};
* Update spring samples with current code.
This is done so I can see my own changes better.
* Regenerate samples after 2.2.2.
* use baseName instead of paramName for spring annotations.
* update spring petstore samples to show the effects.
While creating a static library from the generated files, compilation issues arose due to QDateTime only being forward declared. Adding the include resolves the issue.
* add boolean datatype flags to CodegenResponse and logic to set them in DefaultCodegen
* use Boolean.TRUE.equals to check for isResponseBinary or isResponseFile
CodegenOperation and CodegenProperty classes have boolean fields instead of Boolean.
This fix changes Boolean fields of CodegenParameter and CodegenResponse to boolean as well.
Boolean fields are not necessary. Other classes interacting with them have to check for null
before interacting with them which is unnecessary and leads to ugly code.
* Removed trailing spaces
* Fixed autocomplete support for bash-completion < 1.4
* Added mime type autocomplete for Zsh
* Fixed Bash url autocompletion
* Fixed Zsh space after operation arguments and added trim to descriptions
* Added generation of Dockerfile for packaging Bash rest clients
* updated test samples for Bash client generator
Before this change if a model had a property of type=string and
format=uuid, it would be generated as a Java String.
Now, it generates the property as java.util.UUID.
* jaxrs-cxf: Re-added usage of contextPath in api.mustache (basePath)
* jaxrs-cxf: Added useAnnotatedBasePath and generateNonSpringApplication flags
* useAnnotatedBasePath=true uses the @Path annotation in the generated interface to set the basePath, default is false
* generateNonSpringApplication=true generates a web.xml with CXFNonSpringJaxrsServlets. If useAnnotatedBasePath=true only one servlet will be created, otherwise one for each API
* jaxrs-cxf generateNonSpringApplication: removed creation of multiple servlets
Multiple servlets were generated if useAnnotatedBasePath was false.
This is not necessary as there is only ONE basePath per contract
* jaxrs-cxf: added sample generation for useAnnotatedBasePath and generateNonSpringApplication
* Issue #4606 (Spring) SubClass annotations are missing from the base class
* Issue #4606 (Spring) SubClass annotations are missing from the base class
* Commit updated samples from ./bin/spring-all-pestore.sh
This change removes '/' from the set of "safe" characters for the python urllib quote function.
When using PathParams containing a '/' character, the Python client does not encode the '/'. As a result, the / is interpreted as a path separator rather than a parameter value. PathSegments must encode the / character for proper transfer of information. e.g. 'my/string' should be encoded to 'my%2Fstring' for correct transmission of a PathParam argument. Path separators must only be used between path segments.
* Use {{unescapedDescription}} in html, html2
This addresses https://github.com/swagger-api/swagger-codegen/issues/4611
Replace {{description}} with {{unescapedDescription}}
This is an improvement but because of issue #1203 and/or #3436 (fix not
yet impleented), Markdown is not being processed and converted to HTML.
* Add generated petstore samples index.html
* Copy unescapedDescription
Outputs warnings about ASP.NET Core's src/ requirement, and adjusts
sourceFolder if the input doesn't start with src/.
As a consequence of this change, apiFileFolder and modelFileFolder may
possibly return different values if users are setting both sourceFolder
and packageName. This could be considered a breaking change in those
sceanrios.
* issue 3914, removed logic that remove header parameters. update template files to handle header parameters in the same manner as other parameters
* update swift and swift3 samples
* Fixed issue #4665 [go] required object parameters are marked omitempty.
* Run ./bin/go-petstore.sh as suggested by @wing328.
Please note, that there are quite few changes, that doesn’t come from the diff I’ve applied. Seems the script wasn’t run on the previous commit.
* [JAXRS-CXF] [issue #4386] add builder-style methods to model classes
before this change, model classes had only a default constructor and
setter methods, resulting in code like this:
myModel = MyModel();
myModel.setFirstField(firstField);
myModel.setSecondField(secondField);
return myModel;
this change adds builder style methods, such that the above code can be
written more compactly:
return MyModel().firstField(firstField).secondField(secondField);
this is consistent with other JAVA generators in swagger-codegen.
* update jaxrs-cxf sample code
* Add elixir client
* Add test for elixir client
* Add shell script for generating sample codes for elixir client
It just copied from bin/bash-petstore.sh
* Make elixir-petstore.sh to generate sample codes for elixir client
* Add sample codes for elixir client
error TS2454: Variable 'contentTypeHeader' is used before being assigned.
error TS2345: Argument of type 'FetchAPI | undefined' is not assignable to parameter of type 'FetchAPI'.
* Feature/objc tasks 2.3.0 (#3522)
* change api and ApiClient to return cancellable NSURLSessionTasks instead of NSNumber
* define a configuration protocol for custom configurations, which can be passed to api clients instead of a global configuration, provide a default implementation with a singleton option
* integrate a workaround for a current JSONModel concurrency bug
* update to new ISO8601 pod
* add missing call to super
* integrate new templates into codegen
* updates documentation templates
* updates petstore objc generated code
* fixes objc client tests
* [ObjC] Add version define and share default headers of each client
* add finch generator and resource
* update license, add errros
* Fix problem with multitheard api client
* fix some errors for finch
* [finch] Remove license header
* [finch] Remove finatra stuff, fix a few issues
* WIP: Finch server generator
* [finch] WIP: server generator impl
This puts parameters (input/output) in the right format. Currently, this
is done in the generator class using vendorExtensions, but should be
refactored to imported templates to clean up.
Previous commits of the server generator output to appropriate
models/api directories. I've made no changes to this logic, but code
currently generates to the root scala package directory. This will need
to be fixed.
There's also an issue with circe's and Option[Date] in the Order type.
This issue will need to be resolved. As well, there's some unused
imports to clean up.
Initial implementation lacks support for custom imports, type mappings,
etc.
* [finch] Update api/model package and imports
* [finch] Explicit import/type mappings
* [finch] Regenerate example
* [Python] Improve model equality implementation
The model equality implementation template blindly tried accessing the
`__dict__` attribute of the variable to compare equality against. This
attribute does not exist for a lot of Python built-in types (`None`,
`str` etc.) and the equality check would simply crash.
This adds a simple guard to only continue with equality check if the
variable to compare is an instance of the model being compared against.
* Remove wrong auto-update of Python requirements
* two fixes: 1) extra ?'s at end of some url's 2) enums not being called out w/ rawValue to get the proper string name
* update samples
* one step closer
* closer implementation
