Replace bad quotes (#3439)

specification/botservice/resource-manager/Microsoft.BotService/preview/2017-12-01/botservice.json

@@ -245,7 +245,7 @@
         ],
         "responses": {
           "200": {
-            "description": "The resource provider should return 200 (OK) to indicate that the operation completed successfully. For other errors (e.g. internal errors) use the appropriate HTTP error code.\nThe nextLink field is expected to point to the URL the client should use to fetch the next page (per server side paging). This matches the OData guidelines for paged responses here. If a resource provider does not support paging, it should return the same body (JSON object with “value” property) but omit nextLink entirely (or set to null, *not* empty string) for future compatibility.\nThe nextLink should be implemented using following query parameters:\n· skipToken: opaque token that allows the resource provider to skip resources already enumerated. This value is defined and returned by the RP after first request via nextLink.\n· top: the optional client query parameter which defines the maximum number of records to be returned by the server.\nImplementation details:\n· NextLink may include all the query parameters (specifically OData $filter) used by the client in the first query. \n· Server may return less records than requested with nextLink. Returning zero records with NextLink is an acceptable response. \nClients must fetch records until the nextLink is not returned back / null. Clients should never rely on number of returned records to determinate if pagination is completed.",
+            "description": "The resource provider should return 200 (OK) to indicate that the operation completed successfully. For other errors (e.g. internal errors) use the appropriate HTTP error code.\nThe nextLink field is expected to point to the URL the client should use to fetch the next page (per server side paging). This matches the OData guidelines for paged responses here. If a resource provider does not support paging, it should return the same body (JSON object with \"value\" property) but omit nextLink entirely (or set to null, *not* empty string) for future compatibility.\nThe nextLink should be implemented using following query parameters:\n· skipToken: opaque token that allows the resource provider to skip resources already enumerated. This value is defined and returned by the RP after first request via nextLink.\n· top: the optional client query parameter which defines the maximum number of records to be returned by the server.\nImplementation details:\n· NextLink may include all the query parameters (specifically OData $filter) used by the client in the first query. \n· Server may return less records than requested with nextLink. Returning zero records with NextLink is an acceptable response. \nClients must fetch records until the nextLink is not returned back / null. Clients should never rely on number of returned records to determinate if pagination is completed.",
             "schema": {
               "$ref": "#/definitions/BotResponseList"
             }
@@ -593,7 +593,7 @@
         ],
         "responses": {
           "200": {
-            "description": "The resource provider should return 200 (OK) to indicate that the operation completed successfully. For other errors (e.g. internal errors) use the appropriate HTTP error code.\nThe nextLink field is expected to point to the URL the client should use to fetch the next page (per server side paging). This matches the OData guidelines for paged responses here. If a resource provider does not support paging, it should return the same body (JSON object with “value” property) but omit nextLink entirely (or set to null, *not* empty string) for future compatibility.\nThe nextLink should be implemented using following query parameters:\n· skipToken: opaque token that allows the resource provider to skip resources already enumerated. This value is defined and returned by the RP after first request via nextLink.\n· top: the optional client query parameter which defines the maximum number of records to be returned by the server.\nImplementation details:\n· NextLink may include all the query parameters (specifically OData $filter) used by the client in the first query. \n· Server may return less records than requested with nextLink. Returning zero records with NextLink is an acceptable response. \nClients must fetch records until the nextLink is not returned back / null. Clients should never rely on number of returned records to determinate if pagination is completed.",
+            "description": "The resource provider should return 200 (OK) to indicate that the operation completed successfully. For other errors (e.g. internal errors) use the appropriate HTTP error code.\nThe nextLink field is expected to point to the URL the client should use to fetch the next page (per server side paging). This matches the OData guidelines for paged responses here. If a resource provider does not support paging, it should return the same body (JSON object with \"value\" property) but omit nextLink entirely (or set to null, *not* empty string) for future compatibility.\nThe nextLink should be implemented using following query parameters:\n· skipToken: opaque token that allows the resource provider to skip resources already enumerated. This value is defined and returned by the RP after first request via nextLink.\n· top: the optional client query parameter which defines the maximum number of records to be returned by the server.\nImplementation details:\n· NextLink may include all the query parameters (specifically OData $filter) used by the client in the first query. \n· Server may return less records than requested with nextLink. Returning zero records with NextLink is an acceptable response. \nClients must fetch records until the nextLink is not returned back / null. Clients should never rely on number of returned records to determinate if pagination is completed.",
             "schema": {
               "$ref": "#/definitions/ChannelResponseList"
             }
@@ -996,7 +996,7 @@
         ],
         "responses": {
           "200": {
-            "description": "The resource provider should return 200 (OK) to indicate that the operation completed successfully. For other errors (e.g. internal errors) use the appropriate HTTP error code.\nThe nextLink field is expected to point to the URL the client should use to fetch the next page (per server side paging). This matches the OData guidelines for paged responses here. If a resource provider does not support paging, it should return the same body (JSON object with “value” property) but omit nextLink entirely (or set to null, *not* empty string) for future compatibility.\nThe nextLink should be implemented using following query parameters:\n· skipToken: opaque token that allows the resource provider to skip resources already enumerated. This value is defined and returned by the RP after first request via nextLink.\n· top: the optional client query parameter which defines the maximum number of records to be returned by the server.\nImplementation details:\n· NextLink may include all the query parameters (specifically OData $filter) used by the client in the first query. \n· Server may return less records than requested with nextLink. Returning zero records with NextLink is an acceptable response. \nClients must fetch records until the nextLink is not returned back / null. Clients should never rely on number of returned records to determinate if pagination is completed.",
+            "description": "The resource provider should return 200 (OK) to indicate that the operation completed successfully. For other errors (e.g. internal errors) use the appropriate HTTP error code.\nThe nextLink field is expected to point to the URL the client should use to fetch the next page (per server side paging). This matches the OData guidelines for paged responses here. If a resource provider does not support paging, it should return the same body (JSON object with \"value\" property) but omit nextLink entirely (or set to null, *not* empty string) for future compatibility.\nThe nextLink should be implemented using following query parameters:\n· skipToken: opaque token that allows the resource provider to skip resources already enumerated. This value is defined and returned by the RP after first request via nextLink.\n· top: the optional client query parameter which defines the maximum number of records to be returned by the server.\nImplementation details:\n· NextLink may include all the query parameters (specifically OData $filter) used by the client in the first query. \n· Server may return less records than requested with nextLink. Returning zero records with NextLink is an acceptable response. \nClients must fetch records until the nextLink is not returned back / null. Clients should never rely on number of returned records to determinate if pagination is completed.",
             "schema": {
               "$ref": "#/definitions/ConnectionSettingResponseList"
             }

specification/cognitiveservices/data-plane/LUIS/Authoring/stable/v2.0/LUIS-Authoring.json

@@ -3075,7 +3075,7 @@
         ],
         "responses": {
           "200": {
-            "description": "A successful response provides an array of training status details for a LUIS app that you submitted for training. Each element in the response array provides training status for a model (intent or entity) in the LUIS app. Note: Training status is not provided for prebuilt entities because they are pretrained. The \"modelId\" property identifies which intent or entity the training status corresponds to. To get the name and type of each model, use the models API which has a request URL in this format: https://westus.api.cognitive.microsoft.com/luis/api/v2.0/apps/{appId}/versions/{versionId}/models The details property for each model can contain the following fields: \"statusId\": An integer from 0 to 3 that corresponds to the value of the status field. \"status\": A string with one of the following values: “Success”, “UpToDate”, “InProgress”, “Fail”. If the status is “Fail”, the \"failureReason\" property provides the reason for failure. \"exampleCount\": The number of examples used for training. In the case of the None intent or prebuilt domain intents and entities, this number includes example provided internally by the system as well as examples you added to your LUIS app. \"failureReason\": A string that indicates the reason training failed.  The value \"FewLabels\" means that no labeled examples were provided for training. \"trainingDateTime\": A string indicating the time the model was last trained. The value uses the ISO 8601 format for representing times in UTC (Coordinated Universal Time) with a UTC designator (“Z”), for example: \"2017-08-10T01:08:34Z\".",
+            "description": "A successful response provides an array of training status details for a LUIS app that you submitted for training. Each element in the response array provides training status for a model (intent or entity) in the LUIS app. Note: Training status is not provided for prebuilt entities because they are pretrained. The \"modelId\" property identifies which intent or entity the training status corresponds to. To get the name and type of each model, use the models API which has a request URL in this format: https://westus.api.cognitive.microsoft.com/luis/api/v2.0/apps/{appId}/versions/{versionId}/models The details property for each model can contain the following fields: \"statusId\": An integer from 0 to 3 that corresponds to the value of the status field. \"status\": A string with one of the following values: \"Success\", \"UpToDate\", \"InProgress\", \"Fail\". If the status is \"Fail\", the \"failureReason\" property provides the reason for failure. \"exampleCount\": The number of examples used for training. In the case of the None intent or prebuilt domain intents and entities, this number includes example provided internally by the system as well as examples you added to your LUIS app. \"failureReason\": A string that indicates the reason training failed.  The value \"FewLabels\" means that no labeled examples were provided for training. \"trainingDateTime\": A string indicating the time the model was last trained. The value uses the ISO 8601 format for representing times in UTC (Coordinated Universal Time) with a UTC designator (\"Z\"), for example: \"2017-08-10T01:08:34Z\".",
             "schema": {
               "$ref": "#/definitions/ModelTrainingInfoList"
             }