Overview

Some tools for code editing.

Features

• Unify file's line feed
• Unify file's character encoding
• Editing command with LLM
  • Support OpenAI API
  • Support local API endpoint
• Others
  • Load settings from a file. Able to use an individual setting for each projects.

AI Features

Just right click on the target code, and select command from Handy Tools submenu.

Command	Place	Description
Add Multi Line Complete	Cursor Position	Fill in the middle at the cursor position.
Explain Function	Selection or one line	Explain a funciton which the selection or one line is overlapping.
Translate	Selection	Translate the selected text on the chat view.
Add Doxygen	Selection or one line	Add Doxygen style document for a function,
		which the selection or one line is overlapping.
Preview Doxygen	Selection of one line	Preview Doxygen style document for a function on the chat view.
Add One Line Completion	Cursor Position	Fill in the middle at the cursor position only one line

• Add Doxygen comment

• One line completion

Notice: now supports only C/C++

Settings

General

Category	Item	Description	Default
General
	Load Setting Files	Load settings from a file	true
Unify Encoding
	Encoding	Character encoding for unifying	UTF8
Unify Line Feed
	C/C++	Line feed for C/C++	LF
	CSharp	Line feed for C#	LF
	Others	Line feed for other text format	LF

AI

AI	Item	Description	Default
API	API Endpoint	Endpoint Address	empty
	API Key	API Key	XXX
	Completion Endpoint	Endpoint Address for completion	empty
Model	Model for General	Model for general tasks	llama2
	Model for Generation	Model for generation or code FIM tasks	llama2
	Model for Translation	Model for translation	llama2
	Format After Generation	Whether format text after generation	false
	Max Text Length	Max text length for passing to LLM (in chars, not context size)	llama2
	Temperature	Temperature in generation parameters	0.1
	RealTime Completion	Automatic suggest completions on every key strokes	False
	Completion Interval	Interval time in milliseconds for suggestion	1000
	MaxCompletionInputSize	Max text length around cursor point for completion	4000
	MaxCompletionOutputSize	Max output text length for line completion	64
Prompt	Completion	Prompt for completion tasks
	Documentation	Prompt for documentation tasks
	Explanation	Prompt for explanation tasks
	Translation	Prompt for translation tasks

Setting File

Try to find a setting file,

1. From the directory of the focused file
2. Search _handytools.xml
3. If the current directory don't have ".git" or ".svn", move to upper directory and repeat from 2.

General Settings

An example general setting file "_handytools.xml" is below,

<?xml version="1.0" encoding="utf-8"?>
<HandyTools>
    <UnifyLineFeed>
        <Code lang="C/C++">
            LF
        </Code>
        <Code lang="CSharp">
            CRLF
        </Code>
        <Code lang="Others">
            LF
        </Code>
    </UnifyLineFeed>
    <UnifyEncoding>
        <Encoding>UTF8</Encoding>
    </UnifyEncoding>
</HandyTools>

AI Settings

An example settings for OpenAI API endpoint,

<?xml version="1.0" encoding="utf-8"?>
<HandyTools>
	<AI>
		<ModelGeneral>llama2</ModelGeneral>
		<ModelGeneration>llama2</ModelGeneration>
		<ModelTranslation>llama2</ModelTranslation>
		<ApiKey>XXX</ApiKey>
		<ApiEndpoint>http://localhost:9090</ApiEndpoint>
		<CompletionEndpoint>http://localhost:9090</CompletionEndpoint>
		<FormatResponse>False</FormatResponse>
		<Temperature>0.1</Temperature>
		<Timeout>30</Timeout>
		<RealTimeCompletion>False</RealTimeCompletion>
		<MaxTextLength>3000</MaxTextLength>
		<CompletionIntervalInMilliseconds>1000</CompletionIntervalInMilliseconds>
		<MaxCompletionInputSize>4000</MaxCompletionInputSize>
		<MaxCompletionOutputSize>64</MaxCompletionOutputSize>
		<PromptCompletion>&lt;|fim_prefix|&gt;{prefix}&lt;|fim_suffix|&gt;{suffix}&lt;|fim_middle|&gt;</PromptCompletion>
		<PromptExplanation>Explain the next {filetype} code.\ncode:{content}</PromptExplanation>
		<PromptTranslation>Translate in English\n\n{content}</PromptTranslation>
		<PromptDocumentation>Create a doxygen comment for the following C++ Function. doxygen comment only\n\n{content}</PromptDocumentation>
	</AI>
</HandyTools>

ToDo

License

This software is distributed under two licenses 'The MIT License' or 'Public Domain', choose whichever you like.

Release

• v.0.5.9
  • Work around to fix exceptions while accessing CodeElements.GetEnumerator
  • Change algorithm of searching the definition
• v.0.5.8
  • Change logging method of one line completion
• v.0.5.7
  • Change progress display
  • Output logs for debugging
• v.0.5.6
  • Fix null exception
• v.0.5.5
  • Delete Ollama API support
  • Add AI one line completion
• v.0.5.4
  • Fix failing to detect function's declarations
• v.0.5.3
  • Remove supports of Unreal
• v.0.5.2
  • Add template and Unreal's FUNCTION macro
• v.0.5.1
  • Fix some bugs about async
  • Apply short cut keybindings
• v.0.5.0
  • Remove full-text search, because Visual Studio itself has integrated full-text search
  • Add AI commands
• v.0.4.0
  • Add a search window menu
• v.0.3.0
  • Fix failures on loading settings
• v.0.2.0
  • Add full-text search
• v.0.1.0
  • Initial release