Patent application number | Description | Published |
20110276950 | NAME BINDING EXTENSIBILITY FOR TYPED PROGRAMMING LANGUAGE - Architecture that provides a way to plug into a compiler (e.g., F#) and the associated language service. The language components define a standard plug-in mechanism that allows any component to plug into the services and, provide name binding and type resolution services. The compiler and language service utilize the protocol defined (e.g., in the shape of interfaces and classes that must be implemented in order to plug-in to the system) to provide compile time and design time extensions. The architecture provides the ability to define a customized protocol to interact with the compiler's name binding mechanism. The use of this protocol allows not only compilation to succeed, but to also provide rich IDE services, such as name lookup, and value lookup for quick information. Additionally, the capability to provide static type information to satisfy the requirements of a typed programming language is supported. | 11-10-2011 |
20120266141 | API DESCRIPTIONS - API description techniques are described for consumption by dynamically-typed languages. In one or more implementations, machine-readable data is parsed to locate descriptions of one or more application programming interfaces (APIs). The descriptions of the one or more application programming interfaces are projected into an alternate form that is different than a form of the machine-readable data. | 10-18-2012 |
20130007702 | STATIC METADATA IN DYNAMIC PROGRAMS - A dynamic programming environment includes a dynamic runtime infrastructure configured to receive static metadata as source text in a source code written in a dynamic programming language. The dynamic runtime infrastructure determines an object representation based on the static metadata from the source text in the source code. | 01-03-2013 |
20130067433 | HIDDEN REPRESENTATIONS FOR MARSHALING THROUGH HOSTED SCRIPTING ENVIRONMENTS - Systems and methods are described for passing a data item to and obtaining the data item from a scripting environment hosted by a host environment, wherein each environment uses a different representation of the data item. The host environment passes the data item to the scripting environment by passing a host environment representation of the data item that is hidden from developers of scripts to be executed by the scripting environment. This enables the original host environment representation of the data item to be preserved even after the data item has been passed to and from the scripting environment in the case where the scripting environment does not modify the data item, and also facilitates lazy conversion of the data item or avoidance of conversion entirely in the case where the scripting environment does not modify the data item. | 03-14-2013 |
20140282441 | STATIC TYPE CHECKING ACROSS MODULE UNIVERSES - Static type checking can be performed on types and values defined in modules in a system that dynamically composes programs from modules. The types and values do not share a global namespace. Each module defines its own module universe, disjoint from other modules. A language mechanism can establish a local name binding to one module within the content of another module. When type checking at compile time an environment can be established that corresponds to a runtime instance of the program. The static type system can be arranged to align with the runtime values, such that the names used to refer to objects at runtime are the same as the names used to refer to the types of those objects in the static type system. Aliases of a particular type are resolved to a known compile time description of the type. | 09-18-2014 |
20140282442 | STATICALLY EXTENSIBLE TYPES - Statically extensible types allow a static type system to model the behavior of dynamic object model extension in dynamic runtime systems. Static types that model dynamically extended types can be extended using additional declarations in the same compilation. Declarations for a particular type can come from multiple independent sources which can have been defined at multiple points in time. Extension declarations can use the same syntax as the initial type declaration. Hence presence of one or more declarations for the same type in a compilation can indicate that the type has been extended. These features allow static type checking of dynamic plug-ins to be supported using statically extensible types. Declarations and extension declarations for a type can be merged together to create an extended type that enables different processing paths and dependencies. | 09-18-2014 |
20140282443 | CONTEXTUAL TYPING - Type inference is used to provide the benefits of static type checking without explicitly declaring type in the source code. Type inference that proceeds in a bottom up direction is augmented by contextual typing that proceeds in a top down direction. Contextual typing types enclosed expressions by associating the type of a directly enclosing expression with the enclosed expression. Object literals, function literals and array literals enclosed in assignment and call expressions, can be contextually typed. | 09-18-2014 |
20140372993 | OVERLOADING ON CONSTANTS - A function in a type system can be overloaded using specified constants. The constant can be can be the result of evaluating an expression to: a string, a number, a Boolean, a pattern or any type of constant. The return type of the function can depend on the specified constant that is passed into the function. The return type of the function can depend on the type of the specified constant that is passed into the function. The type of the parameter that is passed into the function can depend on the value of the constant. The function overloads can be validated to ensure that the constant-based overload is a subtype of a more general overload. A constant can be an expression used at compile time during type checking. | 12-18-2014 |