All notable changes to the Type4Py's VSCode extension will be documented in this file. The format is based on Keep a Changelog and this project adheres to Semantic Versioning.
- A setting,
workspace.localModelEnabled
, to query the local model of Type4Py on users' machines for type inference (#24).
- Making the extension compatible with the Web version of VSCode (#21).
- Adding unit tests and integration testing (#19).
- Include the name of the extension in the right side of predicted types.
- A naive support for predicting types in the case of variables redeclaration in the scope of functions (Addresses #7).
- A status bar at the bottom left of VSCode, which shows no. of in-progress and completed type prediction requests.
- An output channel for the extension to show its log messages instead of showing info windows.
- Detecting
ExtensionMode
for testing and development using the local server (ONLY FOR THE EXTENSION'S DEVELOPERS). - Submitting the canceled/rejected type predictions based on the user's consent.
- Sending the hash of files' absolute path, the extension's version, and activation ID along with the prediction request (ONLY FOR RESEARCH PURPOSE).
- Automatic type inference when a Python file is opened. A setting is created for this, which is false by default.
- Included the name of the file in the information window when type inference is invoked.
- Some revision to README and PRIVACY.
- Moved the design figure from README to a separate file,
DESIGN.md
.
- Removed the 1K LoC limitation for Python source files.
- Removed type-checking setting until we support it.
- Some improvements to README and the package file
- The list of predicted types changes when inserting an extra space after the annotation syntax, i.e,
:
or->
.
- Querying the pre-trained Type4Py model on the server.
- Handling the server's responses and convert to defined data structures for type suggestions.
- Type autocompletion for parameters and return type functions/class methods.
- Type autocompletion for variables (i.e. module-level variables, class variables, and local variables).
- Exception handlings and considering limitations.
- Adding settings for type-checking and filtering out type predictions.
- Gathering telemetry data for accepted type predictions based on the user's consent.
- Adding README