Server Variables not working with Multiple Servers by Dart

[ling-qi]

Bug Report Checklist

• [ ] Have you provided a full/minimal spec to reproduce the issue?
• [ ] Have you validated the input using an OpenAPI validator (example)?
• [ ] What's the version of OpenAPI Generator used?
• [ ] Have you search for related issues/PRs?
• [ ] What's the actual output vs expected output?
• [ ] [Optional] Bounty to sponsor the fix (example)

Description

if you specify multiple servers with server variables in your yaml file, their URLs are not properly made available at template render time. Instead, it shows their URLs with the non-replaced variable names.

OpenAPI declaration file content or url

openapi: 3.0.0 info: title: Account Kit Web API description: HMS Account Kit Web API version: 1.0.0 servers:

• url: https://oauth-login.cloud.huawei.com/oauth2/v3 description: Host and Base Path of Account Kit(Obtain Access Token and Verifying ID Token)

paths: /token: post: summary: obtain access token operationId: obtainAccessToken requestBody: required: true content: application/x-www-form-urlencoded: schema: $ref: '#/components/schemas/obtainAccessToken_' responses: "200": description: "OK!" content: application/json: schema: $ref: '#/components/schemas/obtainAccessToken_response' "400": description: "INVALID_REQUEST!" content: application/json: schema: $ref: '#/components/schemas/obtainAccessToken_responsefailed' /tokeninfo: post: summary: verifying ID token operationId: verifyingIdToken requestBody: required: true content: application/x-www-form-urlencoded: schema: $ref: '#/components/schemas/verifyingIdToken' responses: "200": description: "OK!" content: application/json: schema: $ref: '#/components/schemas/verifyingIdToken_response' "400": description: "INVALID_REQUEST!" /rest.php: post: summary: get token info operationId: getTokenInfo servers:

• url: https://oauth-api.cloud.huawei.com description: Host and Base Path of Account Kit(Parsing Access Token) parameters:
• in: query name: nsp_fmt required: true schema: type: string default: JSON
• in: query name: nspsvc required: true schema: type: string default: huawei.oauth2.user.getTokenInfo requestBody: required: true content: application/x-www-form-urlencoded: schema: $ref: '#/components/schemas/getTokenInfo' responses: "200": description: "OK!" content: application/json: schema: $ref: '#/components/schemas/getTokenInfo_response' "400": description: "INVALID_REQUEST!"

components: schemas: obtainAccessToken_: type: object properties: grant_type: type: string client_id: type: string client_secret: type: string code: type: string redirect_uri: type: string refreshtoken: type: string verifyingIdToken: type: object properties: idtoken: type: string getTokenInfo: type: object properties: access_token: type: string open_id: type: string obtainAccessToken_response: type: object properties: token_type: type: string access_token: type: string scope: type: string expires_in: type: integer format: long refresh_token: type: string id_token: type: string obtainAccessToken_response_failed: type: object properties: error: type: integer sub_error: type: integer error_description: type: string verifyingIdToken_response: type: object properties: typ: type: string alg: type: string kid: type: string iss: type: string sub: type: string aud: type: string exp: type: integer format: long iat: type: integer format: long nonce: type: string at_hash: type: string azp: type: string email_verified: type: boolean email: type: string picture: type: string name: type: string locale: type: string family_name: type: string given_name: type: string display_name: type: string getTokenInfo_response: type: object properties: client_id: type: integer format: long expire_in: type: integer union_id: type: string open_id: type: string scope: type: string

Related issues/PRs

6338

Suggest a fix

One potential fix is in the pull request noted above. PHP templates deal with this by rendering all of the server information into their configuration template/file, ignoring serverVariables specified on command line to openapi-generator (--serverVariables).

Best is probably to have all templates individually handle. Command line serverVariables should be added to the bundle that is supplied to templates, and individual language templates should fill in such variables at generator run-time.

For variables specified in the yaml/json file, they can remain un-substituted at generator run-time, and instead filled in with logic in the rendered templates.

Wdyt?

[quetool]

Is there any progress on this?

[ling-qi]

Is there any progress on this? Yes，If there are local servers, replace the global servers with local servers. You need to modify the source code of openapi-generator.