rjnienaber / amazonite Goto Github PK

Add API implementation for CloudWatch Logs.

• Generate and add cloudwatch_logs_v1 into src
• Add integration tests
• Add examples

Add API implementation for Systems Manager (SSM).

• Generate and add ssm_v1 into src
• Add integration tests
• Add examples

Add API implementation for Kinesis.

• Generate and add kinesis_v1 into src
• Add integration tests
• Add examples

The properties of inputs sometimes specify constraints for the values that are being assigned. For example, a DynamoDB TableName property specifies a min/max length and a regex pattern that should be followed.

These constraints should be turned into validations that:

• Can be called by users on the input object
• Are called by the client before sending the request by default
• Can be turned off on the client before the request is sent

The http timeouts used by Amazonite are current the default for the HTTP::Client library (which is the default for the socket library of the underlying platform?).

The following timeouts should be made configurable by users:

• dns_timeout
• connect_timeout
• read_timeout
• write_timeout

Each of their overloads should also be accepted i.e. Number and Time::Span

All input and output objects should override the == and hash(hasher) functions. Ostensibly this looks like it can be done fairly easily with the def_equals and def_hash macros. Failing that, they should be manually implemented.

At the moment, code generation will parse all the APIs it knows how to parse and output the generated code to tmp. Some command line flags should be added that will allow a developer to:

• Point to an aws-sdk-js dir that is in some arbitrary location, not just the one in the root of the repo.
• Specify an arbitrary output directory
• Add varying degrees of log statements that can be turned on or off
• Filter APIs generated by type or name