# ESP32 Transpiler

[![License](https://img.shields.io/github/license/andygeiss/esp32)](https://github.com/andygeiss/esp32-transpiler/blob/master/LICENSE)
[![Releases](https://img.shields.io/github/v/release/andygeiss/esp32)](https://github.com/andygeiss/esp32-transpiler/releases)
[![Go Report Card](https://goreportcard.com/badge/github.com/andygeiss/esp32)](https://goreportcard.com/report/github.com/andygeiss/esp32)
[![Maintainability](https://api.codeclimate.com/v1/badges/90bf72e5a7b538c9e50e/maintainability)](https://codeclimate.com/github/andygeiss/esp32-transpiler/maintainability)

## Purpose

The [Arduino IDE](https://www.arduino.cc/en/Main/Software) is easy to use.
But I faced problems like maintainability and testability at more complicated IoT projects.
I needed to compile and flash the ESP32 before testing my code functionality by doing it 100% manually.

This solution transpiles Golang into Arduino code, which can be compiled to an image by using the ESP32 toolchain.
Now I am able to use a fully automated testing approach instead of doing it 100% manually.

**Important**:

The Transpiler only supports a small subset of the [Golang Language Specification](https://golang.org/ref/spec).
Look at the [mapping](https://github.com/andygeiss/esp32-transpiler/blob/master/transpile/handlers/mapping.go) and the [tests](https://github.com/andygeiss/esp32-transpiler/blob/master/transpile/service_test.go) to get the current functionality.
It is also not possible to trigger the C/C++ Garbage Collection, because Golang handles it automatically "under the hood".
Go strings will be transpiled to C constant char arrays, which could be handled on the stack.

## Installation

    go get -u github.com/andygeiss/esp32-transpiler

## Usage

    Usage of esp32-transpiler:
      -source string
            Golang source file
      -target string
            Arduino sketch file


esp32

# f4go


[![Build Status](https://travis-ci.org/Konstantin8105/f4go.svg?branch=master)](https://travis-ci.org/Konstantin8105/f4go)
[![Go Report Card](https://goreportcard.com/badge/github.com/Konstantin8105/f4go)](https://goreportcard.com/report/github.com/Konstantin8105/f4go)
[![codecov](https://codecov.io/gh/Konstantin8105/f4go/branch/master/graph/badge.svg)](https://codecov.io/gh/Konstantin8105/f4go)
[![GitHub license](https://img.shields.io/badge/license-MIT-blue.svg)](https://raw.githubusercontent.com/Konstantin8105/f4go/master/LICENSE)
[![GoDoc](https://godoc.org/github.com/Konstantin8105/f4go/fortran?status.svg)](https://godoc.org/github.com/Konstantin8105/f4go/fortran)
[![Maintainability](https://api.codeclimate.com/v1/badges/b8d0bb5533207cce5ed3/maintainability)](https://codeclimate.com/github/Konstantin8105/f4go/maintainability)

# Example of use

```cmd
> # Install golang
> # Compile f4go
> go get -u github.com/Konstantin8105/f4go
> cd $GOPATH/src/github.com/Konstantin8105/f4go
> go build
> ./f4go ./testdata/blas/caxpy.f
> # Look on Go result source
> less ./testdata/blas/caxpy.go
```

# Transpiling fortran code to golang code

Present result:
```fortran
*> \brief \b CAXPY
*
*  =========== DOCUMENTATION ===========
*
* Online html documentation available at
*            http://www.netlib.org/lapack/explore-html/
*
*  Definition:
*  ===========
*
*       SUBROUTINE CAXPY(N,CA,CX,INCX,CY,INCY)
*
*       .. Scalar Arguments ..
*       COMPLEX CA
*       INTEGER INCX,INCY,N
*       ..
*       .. Array Arguments ..
*       COMPLEX CX(*),CY(*)
*       ..
*
*
*> \par Purpose:
*  =============
*>
*> \verbatim
*>
*>    CAXPY constant times a vector plus a vector.
*> \endverbatim
*
*  Arguments:
*  ==========
*
*> \param[in] N
*> \verbatim
*>          N is INTEGER
*>         number of elements in input vector(s)
*> \endverbatim
*>
*> \param[in] CA
*> \verbatim
*>          CA is COMPLEX
*>           On entry, CA specifies the scalar alpha.
*> \endverbatim
*>
*> \param[in] CX
*> \verbatim
*>          CX is COMPLEX array, dimension ( 1 + ( N - 1 )*abs( INCX ) )
*> \endverbatim
*>
*> \param[in] INCX
*> \verbatim
*>          INCX is INTEGER
*>         storage spacing between elements of CX
*> \endverbatim
*>
*> \param[in,out] CY
*> \verbatim
*>          CY is COMPLEX array, dimension ( 1 + ( N - 1 )*abs( INCY ) )
*> \endverbatim
*>
*> \param[in] INCY
*> \verbatim
*>          INCY is INTEGER
*>         storage spacing between elements of CY
*> \endverbatim
*
*  Authors:
*  ========
*
*> \author Univ. of Tennessee
*> \author Univ. of California Berkeley
*> \author Univ. of Colorado Denver
*> \author NAG Ltd.
*
*> \date November 2017
*
*> \ingroup complex_blas_level1
*
*> \par Further Details:
*  =====================
*>
*> \verbatim
*>
*>     jack dongarra, linpack, 3/11/78.
*>     modified 12/3/93, array(1) declarations changed to array(*)
*> \endverbatim
*>
*  =====================================================================
      SUBROUTINE CAXPY(N,CA,CX,INCX,CY,INCY)
*
*  -- Reference BLAS level1 routine (version 3.8.0) --
*  -- Reference BLAS is a software package provided by Univ. of Tennessee,    --
*  -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
*     November 2017
*
*     .. Scalar Arguments ..
      COMPLEX CA
      INTEGER INCX,INCY,N
*     ..
*     .. Array Arguments ..
      COMPLEX CX(*),CY(*)
*     ..
*
*  =====================================================================
*
*     .. Local Scalars ..
      INTEGER I,IX,IY
*     ..
*     .. External Functions ..
      REAL SCABS1
      EXTERNAL SCABS1
*     ..
      IF (N.LE.0) RETURN
      IF (SCABS1(CA).EQ.0.0E+0) RETURN
      IF (INCX.EQ.1 .AND. INCY.EQ.1) THEN
*
*        code for both increments equal to 1
*
         DO I = 1,N
            CY(I) = CY(I) + CA*CX(I)
         END DO
      ELSE
*
*        code for unequal increments or equal increments
*          not equal to 1
*
         IX = 1
         IY = 1
         IF (INCX.LT.0) IX = (-N+1)*INCX + 1
         IF (INCY.LT.0) IY = (-N+1)*INCY + 1
         DO I = 1,N
            CY(IY) = CY(IY) + CA*CX(IX)
            IX = IX + INCX
            IY = IY + INCY
         END DO
      END IF
*
      RETURN
      END
```

Go code:

```golang
package main

//*> \brief \b CAXPY
//*
//*  =========== DOCUMENTATION ===========
//*
//* Online html documentation available at
//*            http://www.netlib.org/lapack/explore-html/
//*
//*  Definition:
//*  ===========
//*
//*       SUBROUTINE CAXPY(N,CA,CX,INCX,CY,INCY)
//*
//*       .. Scalar Arguments ..
//*       COMPLEX CA
//*       INTEGER INCX,INCY,N
//*       ..
//*       .. Array Arguments ..
//*       COMPLEX CX(*),CY(*)
//*       ..
//*
//*
//*> \par Purpose:
//*  =============
//*>
//*> \verbatim
//*>
//*>    CAXPY constant times a vector plus a vector.
//*> \endverbatim
//*
//*  Arguments:
//*  ==========
//*
//*> \param[in] N
//*> \verbatim
//*>          N is INTEGER
//*>         number of elements in input vector(s)
//*> \endverbatim
//*>
//*> \param[in] CA
//*> \verbatim
//*>          CA is COMPLEX
//*>           On entry, CA specifies the scalar alpha.
//*> \endverbatim
//*>
//*> \param[in] CX
//*> \verbatim
//*>          CX is COMPLEX array, dimension ( 1 + ( N - 1 )*abs( INCX ) )
//*> \endverbatim
//*>
//*> \param[in] INCX
//*> \verbatim
//*>          INCX is INTEGER
//*>         storage spacing between elements of CX
//*> \endverbatim
//*>
//*> \param[in,out] CY
//*> \verbatim
//*>          CY is COMPLEX array, dimension ( 1 + ( N - 1 )*abs( INCY ) )
//*> \endverbatim
//*>
//*> \param[in] INCY
//*> \verbatim
//*>          INCY is INTEGER
//*>         storage spacing between elements of CY
//*> \endverbatim
//*
//*  Authors:
//*  ========
//*
//*> \author Univ. of Tennessee
//*> \author Univ. of California Berkeley
//*> \author Univ. of Colorado Denver
//*> \author NAG Ltd.
//*
//*> \date November 2017
//*
//*> \ingroup complex_blas_level1
//*
//*> \par Further Details:
//*  =====================
//*>
//*> \verbatim
//*>
//*>     jack dongarra, linpack, 3/11/78.
//*>     modified 12/3/93, array(1) declarations changed to array(*)
//*> \endverbatim
//*>
//*  =====================================================================
func CAXPY(N *int, CA *complex128, CX *[]complex128, INCX *int, CY *[]complex128, INCY *int) {
	I := new(int)
	IX := new(int)
	IY := new(int)
	//*
	//*  -- Reference BLAS level1 routine (version 3.8.0) --
	//*  -- Reference BLAS is a software package provided by Univ. of Tennessee,    --
	//*  -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
	//*     November 2017
	//*
	//*     .. Scalar Arguments ..
	//*     ..
	//*     .. Array Arguments ..
	//*     ..
	//*
	//*  =====================================================================
	//*
	//*     .. Local Scalars ..
	//*     ..
	//*     .. External Functions ..
	//*     ..
	if (*(N)) <= 0 {
		return
	}
	if (*SCABS1((CA))) == 0.0e+0 {
		return
	}
	if (*(INCX)) == 1 && (*(INCY)) == 1 {
		//*
		//*        code for both increments equal to 1
		//*
		for (*I) = 1; (*I) <= (*(N)); (*I)++ {
			(*(CY))[(*I)-(1)] = (*(CY))[(*I)-(1)] + (*(CA))*(*(CX))[(*I)-(1)]
		}
	} else {
		//*
		//*        code for unequal increments or equal increments
		//*          not equal to 1
		//*
		(*IX) = 1
		(*IY) = 1
		if (*(INCX)) < 0 {
			(*IX) = (-(*(N))+1)*(*(INCX)) + 1
		}
		if (*(INCY)) < 0 {
			(*IY) = (-(*(N))+1)*(*(INCY)) + 1
		}
		for (*I) = 1; (*I) <= (*(N)); (*I)++ {
			(*(CY))[(*IY)-(1)] = (*(CY))[(*IY)-(1)] + (*(CA))*(*(CX))[(*IX)-(1)]
			(*IX) = (*IX) + (*(INCX))
			(*IY) = (*IY) + (*(INCY))
		}
	}
	//*
	return
}
```


Example of simplification Go code(comments removed for short view):

```go
package main

func CAXPY(N int, CA *complex128, CX []complex128, INCX int, CY []complex128, INCY int) {
	var I int
	var IX int
	var IY int

	if N <= 0 {
		return
	}
	if (*SCABS1((CA))) == 0.0e+0 {
		return
	}
	if INCX == 1 && INCY == 1 {

		for I = 1; I <= N; I++ {
			CY[I-1] = CY[I-1] + CA*CX[I-1]
		}
	} else {

		IX = 1
		IY = 1
		if INCX < 0 {
			IX = (-N+1)*INCX + 1
		}
		if INCY < 0 {
			IY = (-N+1)*INCY + 1
		}
		for I = 1; I <= N; I++ {
			CY[IY-1] = CY[IY-1] + CA*CX[IX-1]
			IX = IX + INCX
			IY = IY + INCY
		}
	}

	return
}
```

### Notes

Fortran 77 | Golang
---------- | ------------
all arguments of function are pointers | ?
all arguments of intrisic function are pointers | ?
all internal function variables are pointers | ?


**IDENT**:
* constants
* arrays, matrixes

Operations:
* assign
* initialization
* boolean

### Fortran test sources

* [**LAPACK from lapack3.8.0**](http://netlib.org/lapack/index.html)
* [**BOSOR5**](http://shellbuckling.com/BOSOR5.php)


f4go

## GopherJS - A compiler from Go to JavaScript

[![GoDoc](https://godoc.org/github.com/gopherjs/gopherjs/js?status.svg)](https://godoc.org/github.com/gopherjs/gopherjs/js)
[![Sourcegraph](https://sourcegraph.com/github.com/gopherjs/gopherjs/-/badge.svg)](https://sourcegraph.com/github.com/gopherjs/gopherjs?badge)
[![Github Actions CI](https://github.com/gopherjs/gopherjs/actions/workflows/ci.yaml/badge.svg)](https://github.com/gopherjs/gopherjs/actions/workflows/ci.yaml)

GopherJS compiles Go code ([go.dev](https://go.dev/)) to pure JavaScript code. Its main purpose is to give you the opportunity to write front-end code in Go which will still run in all browsers.

### Help us make GopherJS better!

- ⤴️ **Help us make better decisions by filling a quick 15-question [GopherJS user survey](https://forms.gle/WEjZqZaPxTxjD9YP8)**.
- 📢 Report and discuss [issues](https://github.com/gopherjs/gopherjs/issues).
- 🎓 Share your knowledge and experience through [articles](https://github.com/gopherjs/gopherjs/wiki/Community-Tutorials-and-Blogs) and [documentation](https://github.com/gopherjs/gopherjs/tree/master/doc).
- 🛠️ Write GopherJS [bindings](https://github.com/gopherjs/gopherjs/wiki/Bindings) for other libraries or [contribute](https://github.com/gopherjs/gopherjs/wiki/Developer-Guidelines) to GopherJS itself.

### What's new?

- 2024-02-24: Go 1.19 support is [available](https://github.com/gopherjs/gopherjs/releases/tag/v1.19.0-beta1)!
- 2022-08-18: Go 1.18 support is [available](https://github.com/gopherjs/gopherjs/releases/tag/v1.18.0-beta2%2Bgo1.18.5)!
- 2021-09-19: Go 1.17 support is available!
- 2021-08-23: Go Modules are now fully supported.
- 2021-06-19: Complete `syscall/js` package implementation compatible with the upstream Go 1.16.
- 2021-04-04: **Go 1.16 is now officially supported!** 🎉 🎉 🎉

### Playground

Give GopherJS a try on the [GopherJS Playground](http://gopherjs.github.io/playground/).

### What is supported?

Nearly everything, including Goroutines ([compatibility documentation](https://github.com/gopherjs/gopherjs/blob/master/doc/compatibility.md)). Performance is quite good in most cases, see [HTML5 game engine benchmark](https://ajhager.github.io/engi/demos/botmark.html). Cgo is not supported.

### Installation and Usage

GopherJS [requires Go 1.19 or newer](https://github.com/gopherjs/gopherjs/blob/master/doc/compatibility.md#go-version-compatibility). If you need an older Go
version, you can use an [older GopherJS release](https://github.com/gopherjs/gopherjs/releases).

Install GopherJS with `go install`:

```
go install github.com/gopherjs/gopherjs@v1.19.0-beta1  # Or replace 'v1.19.0-beta1' with another version.
```

If your local Go distribution as reported by `go version` is newer than Go 1.19, then you need to set the `GOPHERJS_GOROOT` environment variable to a directory that contains a Go 1.19 distribution. For example:

```
go install golang.org/dl/go1.19.13@latest
go1.19.13 download
export GOPHERJS_GOROOT="$(go1.19.13 env GOROOT)"  # Also add this line to your .profile or equivalent.
```

Now you can use `gopherjs build [package]`, `gopherjs build [files]` or `gopherjs install [package]` which behave similar to the `go` tool. For `main` packages, these commands create a `.js` file and `.js.map` source map in the current directory or in `$GOPATH/bin`. The generated JavaScript file can be used as usual in a website. Use `gopherjs help [command]` to get a list of possible command line flags, e.g. for minification and automatically watching for changes.

`gopherjs` uses your platform's default `GOOS` value when generating code. Supported `GOOS` values are: `linux`, `darwin`. If you're on a different platform (e.g., Windows or FreeBSD), you'll need to set the `GOOS` environment variable to a supported value. For example, `GOOS=linux gopherjs build [package]`.

_Note: GopherJS will try to write compiled object files of the core packages to your $GOROOT/pkg directory. If that fails, it will fall back to $GOPATH/pkg._

#### gopherjs run, gopherjs test

If you want to use `gopherjs run` or `gopherjs test` to run the generated code locally, install Node.js 18 (or newer).

On supported `GOOS` platforms, it's possible to make system calls (file system access, etc.) available. See [doc/syscalls.md](https://github.com/gopherjs/gopherjs/blob/master/doc/syscalls.md) for instructions on how to do so.

#### gopherjs serve

`gopherjs serve` is a useful command you can use during development. It will start an HTTP server serving on ":8080" by default, then dynamically compile your Go packages with GopherJS and serve them.

For example, navigating to `http://localhost:8080/example.com/user/project/` should compile and run the Go package `example.com/user/project`. The generated JavaScript output will be served at `http://localhost:8080/example.com/user/project/project.js` (the .js file name will be equal to the base directory name). If the directory contains `index.html` it will be served, otherwise a minimal `index.html` that includes `<script src="project.js"></script>` will be provided, causing the JavaScript to be executed. All other static files will be served too.

Refreshing in the browser will rebuild the served files if needed. Compilation errors will be displayed in terminal, and in browser console. Additionally, it will serve $GOROOT and $GOPATH for sourcemaps.

If you include an argument, it will be the root from which everything is served. For example, if you run `gopherjs serve github.com/user/project` then the generated JavaScript for the package github.com/user/project/mypkg will be served at http://localhost:8080/mypkg/mypkg.js.

#### Environment Variables

There are some GopherJS-specific environment variables:

- `GOPHERJS_GOROOT` - if set, GopherJS uses this value as the default GOROOT
  value, instead of using the system GOROOT as the default GOROOT value
- `GOPHERJS_SKIP_VERSION_CHECK` - if set to true, GopherJS will not check
  Go version in the GOROOT for compatibility with the GopherJS release. This
  is primarily useful for testing GopherJS against unreleased versions of Go.

### Performance Tips

- Use the `-m` command line flag to generate minified code.
- Apply gzip compression (https://en.wikipedia.org/wiki/HTTP_compression).
- Use `int` instead of `(u)int8/16/32/64`.
- Use `float64` instead of `float32`.

### Community

- [#gopherjs Channel on Gophers Slack](https://gophers.slack.com/messages/gopherjs/) (invites to Gophers Slack are available [here](http://blog.gopheracademy.com/gophers-slack-community/#how-can-i-be-invited-to-join:2facdc921b2310f18cb851c36fa92369))
- [Bindings to JavaScript APIs and libraries](https://github.com/gopherjs/gopherjs/wiki/bindings)
- [GopherJS Blog](https://medium.com/gopherjs)
- [GopherJS on Twitter](https://twitter.com/GopherJS)
- [Examples, tutorials and blogs](https://github.com/gopherjs/gopherjs/wiki/Community-Tutorials-and-Blogs)

### Getting started

#### Interacting with the DOM

The package `github.com/gopherjs/gopherjs/js` (see [documentation](https://godoc.org/github.com/gopherjs/gopherjs/js)) provides functions for interacting with native JavaScript APIs. For example the line

```js
document.write("Hello world!");
```

would look like this in Go:

```go
js.Global.Get("document").Call("write", "Hello world!")
```

You may also want use the [DOM bindings](http://dominik.honnef.co/go/js/dom), the [jQuery bindings](https://github.com/gopherjs/jquery) (see [TodoMVC Example](https://github.com/gopherjs/todomvc)) or the [AngularJS bindings](https://github.com/wvell/go-angularjs). Those are some of the [bindings to JavaScript APIs and libraries](https://github.com/gopherjs/gopherjs/wiki/bindings) by community members.

#### Providing library functions for use in other JavaScript code

Set a global variable to a map that contains the functions:

```go
package main

import "github.com/gopherjs/gopherjs/js"

func main() {
	js.Global.Set("pet", map[string]interface{}{
		"New": New,
	})
}

type Pet struct {
	name string
}

func New(name string) *js.Object {
	return js.MakeWrapper(&Pet{name})
}

func (p *Pet) Name() string {
	return p.name
}

func (p *Pet) SetName(name string) {
	p.name = name
}
```

For more details see [Jason Stone's blog post](http://legacytotheedge.blogspot.de/2014/03/gopherjs-go-to-javascript-transpiler.html) about GopherJS.

### Architecture

#### General

GopherJS emulates a 32-bit environment. This means that `int`, `uint` and `uintptr` have a precision of 32 bits. However, the explicit 64-bit integer types `int64` and `uint64` are supported.

The `GOOS` value of this environment is `js`, and the `GOARCH` value is `ecmascript`. You may use these values in build constraints when [writing platform-specific code](doc/compatibility.md#how-to-write-portable-code). (GopherJS 1.17 and older used `js` as the `GOARCH` value.)

#### Application Lifecycle

The `main` function is executed as usual after all `init` functions have run. JavaScript callbacks can also invoke Go functions, even after the `main` function has exited. Therefore the end of the `main` function should not be regarded as the end of the application and does not end the execution of other goroutines.

In the browser, calling `os.Exit` (e.g. indirectly by `log.Fatal`) also does not terminate the execution of the program. For convenience, it calls `runtime.Goexit` to immediately terminate the calling goroutine.

#### Goroutines

Goroutines are fully supported by GopherJS. The only restriction is that you need to start a new goroutine if you want to use blocking code called from external JavaScript:

```go
js.Global.Get("myButton").Call("addEventListener", "click", func() {
  go func() {
    [...]
    someBlockingFunction()
    [...]
  }()
})
```

How it works:

JavaScript has no concept of concurrency (except web workers, but those are too strictly separated to be used for goroutines). Because of that, instructions in JavaScript are never blocking. A blocking call would effectively freeze the responsiveness of your web page, so calls with callback arguments are used instead.

GopherJS does some heavy lifting to work around this restriction: Whenever an instruction is blocking (e.g. communicating with a channel that isn't ready), the whole stack will unwind (= all functions return) and the goroutine will be put to sleep. Then another goroutine which is ready to resume gets picked and its stack with all local variables will be restored.

### GopherJS Development

If you're looking to make changes to the GopherJS compiler, see [Developer Guidelines](https://github.com/gopherjs/gopherjs/wiki/Developer-Guidelines) for additional developer information.


gopherjs

c2go - Convert C to Go
========

[![Build Status](https://github.com/goplus/c2go/actions/workflows/go.yml/badge.svg)](https://github.com/goplus/c2go/actions/workflows/go.yml)
[![Go Report Card](https://goreportcard.com/badge/github.com/goplus/c2go)](https://goreportcard.com/report/github.com/goplus/c2go)
[![GitHub release](https://img.shields.io/github/v/tag/goplus/c2go.svg?label=release)](https://github.com/goplus/c2go/releases)
[![Coverage Status](https://codecov.io/gh/goplus/c2go/branch/main/graph/badge.svg)](https://codecov.io/gh/goplus/c2go)
[![GoDoc](https://pkg.go.dev/badge/github.com/goplus/c2go.svg)](https://pkg.go.dev/mod/github.com/goplus/c2go)
[![Language](https://img.shields.io/badge/language-Go+-blue.svg)](https://github.com/goplus/gop)

This project is **DEPRECATED**! We are shifting our technical roadmap to [llgo](https://github.com/goplus/llgo).

This is a subproject of [the Go+ project](https://github.com/goplus/gop). Its goal is converting any C project into Go without any human intervention and keeping performance close to C.

* Make it passed. Stage: `Done Almost`. We have compiled [libc](https://github.com/goplus/libc) (in progress) and [sqlite3](https://github.com/goplus/sqlite-c2go) (done).
* Make it correct. Stage: `Doing`. We ported `libc-test` to test [c2go](https://github.com/goplus/c2go) and [libc](https://github.com/goplus/libc). Coverage: [![Coverage Status](https://codecov.io/gh/goplus/libc/branch/musl-go/graph/badge.svg)](https://codecov.io/gh/goplus/libc)
* Make it fast. Stage: `Planning`.

![Screen Shot1](https://user-images.githubusercontent.com/396972/160951673-30ec62ae-2981-4cdf-a1ab-bc7fcb6f7475.png)


## How to run examples?

Run an example:

- Build c2go tools: `go install -v ./...`
- Go `testdata/xxx` directory, and run `c2go .`

Run/Test multiple examples:

- Run examples: `c2go ./...`
- Test examples: `c2go -test ./...`


## How c2go is used in Go+

Here is [an example to show how Go+ interacts with C](https://github.com/goplus/gop/tree/main/testdata/helloc2go).

```go
import "C"

C.printf C"Hello, c2go!\n"
C.fprintf C.stderr, C"Hi, %7.1f\n", 3.14
```

Here we use `import "C"` to import libc. It's an abbreviation for `import "C/github.com/goplus/libc"`. It is equivalent to the following code:

```go
import "C/github.com/goplus/libc"

C.printf C"Hello, c2go!\n"
C.fprintf C.stderr, C"Hi, %7.1f\n", 3.14
```

In this example we call two C standard functions `printf` and `fprintf`, passing a C variable `stderr` and two C strings in the form of `C"xxx"` (a Go+ syntax to represent C style strings).

Run `gop run .` to see the output of this example:

```
Hello, c2go!
Hi,     3.1
```

## What's our plan?

- First, support most of the syntax of C. Stage: `Done Almost`, see [supported C syntax](#supported-c-syntax).
- Second, compile `sqlite3` to fix c2go bugs and get list of its dependent C standard libary fuctions. Stage: `Done`, see [github.com/goplus/sqlite](https://github.com/goplus/sqlite) and [its dependent fuctions](https://github.com/goplus/sqlite/blob/main/c2go_autogen.go).
- Third, support most of C standard library functions (especially that used by `sqlite3`), and can import them by Go+. Stage: `Doing`, see [detailed progress here](https://github.com/goplus/libc#whats-our-plan).
- Last, support all custom libraries, especially those well-known open source libraries. Stage: `Planning`.


## Tested Platform

- [x] MacOS: 11.x
- [x] Linux: ubuntu-20.04 (temporarily skip `testdata/qsort`)
- [ ] Windows


## Supported C syntax

### Data structures

- [x] Void: `void`
- [x] Boolean: `_Bool`, `bool`
- [x] Integer: [`signed`/`unsigned`] [`short`/`long`/`long long`] `int`
- [x] Enum: `enum`
- [x] Float: `float`, `double`, `long double`
- [x] Character: [`signed`/`unsigned`] `char`
- [ ] Wide Character: `wchar_t`
- [ ] Large Integer: [`signed`/`unsigned`] `__int128`
- [x] Complex: `_Complex` `float`/`double`/`long double`
- [x] Typedef: `typedef`
- [x] Pointer: *T, T[]
- [x] Array: T[N], T[]
- [x] Array Pointer: T(*)[N]
- [x] Function Pointer: T (*)(T1, T2, ...)
- [x] Struct: `struct`
- [x] Union: `union`
- [x] BitField: `intType :N`

### Operators

- [x] Arithmetic: a+b, a-b, a*b, a/b, a%b, -a, +a
- [x] Increment/Decrement: a++, a--, ++a, --a
- [x] Comparison: a<b, a<=b, a>b, a>=b, a==b, a!=b
- [x] Logical: a&&b, a||b, !a
- [x] Bitwise: a|b, a&b, a^b, ~a, a<<n, a>>n
- [x] Pointer Arithmetic: p+n, p-n, p-q, p++, p--
- [x] Assignment: `=`
- [x] Operator Assignment: a`<op>=`b
- [x] BitField Assignment: `=`
- [ ] BitField Operator Assignment: a`<op>=`b
- [x] Struct/Union/BitField Member: a.b
- [x] Array Member: a[n]
- [x] Pointer Member: &a, *p, p[n], n[p], p->b
- [x] Comma: `a,b`
- [x] Ternary Conditional: cond?a:b
- [x] Function Call: f(a1, a2, ...)
- [x] Conversion: (T)a
- [x] Sizeof: sizeof(T), sizeof(a)
- [x] Offsetof: __builtin_offsetof(T, member)

### Literals

- [x] Boolean, Integer
- [x] Float, Complex Imaginary
- [x] Character, String
- [x] Array: `(T[]){ expr1, expr2, ... }`
- [x] Array Pointer: `&(T[]){ expr1, expr2, ... }`
- [x] Struct: `struct T{ expr1, expr2, ... }`

### Initialization

- [x] Basic: `T a = expr`
- [x] Array: `T a[] = { expr1, expr2, ... }`, `T a[N] = { expr1, expr2, ... }`
- [x] Struct: `struct T a = { expr1, expr2, ... }`, `struct T a = { .a = expr1, .b = expr2, ... }`
- [x] Union: `union T a = { expr }, union T a = { .a = expr }`
- [x] Array in Struct: `struct { T a[N]; ... } v = { { expr1, expr2, ... }, ... }`, `struct { T a[N]; ... } v = { { [0].a = expr1, [1].a = expr2, ... }, ... }`

### Control structures

- [x] If: `if (cond) stmt1 [else stmt2]`
- [x] Switch: `switch (tag) { case expr1: stmt1 case expr2: stmt2 default: stmtN }`
- [x] For: `for (init; cond; post) stmt`
- [x] While: `while (cond) stmt`
- [x] Do While: `do stmt while (cond)`
- [x] Break/Continue: `break`, `continue`
- [x] Goto: `goto label`

### Functions

- [x] Parameters
- [x] Variadic Parameters
- [x] Variadic Parameter Access
- [x] Return


c2go

[![Go Report Card](https://goreportcard.com/badge/github.com/Konstantin8105/c4go)](https://goreportcard.com/report/github.com/Konstantin8105/c4go)
[![codecov](https://codecov.io/gh/Konstantin8105/c4go/branch/master/graph/badge.svg)](https://codecov.io/gh/Konstantin8105/c4go)
[![GitHub license](https://img.shields.io/badge/license-MIT-blue.svg)](https://raw.githubusercontent.com/Konstantin8105/c4go/master/LICENSE)
[![Go](https://github.com/Konstantin8105/c4go/actions/workflows/codecov.yml/badge.svg?branch=master)](https://github.com/Konstantin8105/c4go/actions/workflows/codecov.yml)
[![Go Reference](https://pkg.go.dev/badge/github.com/Konstantin8105/c4go.svg)](https://pkg.go.dev/github.com/Konstantin8105/c4go)


A tool for [transpiling](https://en.wikipedia.org/wiki/Source-to-source_compiler) C code to Go code.

Milestones of the project:

1. Transpiling project [GNU GSL](https://www.gnu.org/software/gsl/).
2. Transpiling project [GTK+](https://www.gtk.org/).

Notes:
* This transpiler works on all linux machines. Most likely, it will work on Unix and MacOS as well. Feedback is most welcome.
* Installing the `clang` compiler is a prerequisite. See [llvm download page](http://releases.llvm.org/download.html)



# Installation

Using a terminal window, position into a working directory (E.g. $HOME/development). Then, execute the following shell commands:

```bash
# Create a local c4go repository and position into it.
git clone https://github.com/Konstantin8105/c4go
cd c4go

# Update the Go source files.
go generate ./...

# Install the c4go executable in the standard Go bin directory: $HOME/go/bin.
go install

# Make sure that $HOME/go/bin is in the path (good practice).
# Then, verify that the c4go is available for use.
c4go version
```

Sample ```c4go``` version console output:
```
Build time: May  4 20:41:31 2024 UTC
Git hash:   "c5400a76df0f5157f234547bef3ae8fb7b692685"
```

# Usage example

```bash
# Change your location to the folder with examples:
cd $GOPATH/src/github.com/Konstantin8105/c4go/examples/

# Transpile one file from the C example folder:
c4go transpile prime.c

# Look at the result
nano prime.go

# Check the result:
go run prime.go
# Enter a number
# 13
# The number is: 13
# Prime number.
```

C code of file `prime.c`:
```c
#include <stdio.h>

int main()
{
    int n, c;

    printf("Enter a number\n");
    // get value
    scanf("%d", &n);
    printf("The number is: %d\n", n);

    // -------
    if (n == 2)
        printf("Prime number.\n");
    else {
        for (c = 2; c <= n - 1; c++) {
            if (n % c == 0)
                break;
        }
        if (c != n)
            printf("Not prime.\n");
        else
            printf("Prime number.\n");
    }
    return 0;
}
```

Go code of file `prime.go`:
```golang
//
//	Package - transpiled by c4go
//
//	If you have found any issues, please raise an issue at:
//	https://github.com/Konstantin8105/c4go/
//

package main

import "unsafe"
import "github.com/Konstantin8105/c4go/noarch"
import "fmt"

// main - transpiled function from  C4GO/examples/prime.c:3
func main() {
	var n int32
	var c int32
	fmt.Printf("Enter a number\n")
	// get value
	noarch.Scanf([]byte("%d\x00"), c4goUnsafeConvert_int32(&n))
	noarch.Printf([]byte("The number is: %d\n\x00"), n)
	if n == 2 {
		// -------
		fmt.Printf("Prime number.\n")
	} else {
		for c = 2; c <= n-1; c++ {
			if n%c == 0 {
				break
			}
		}
		if c != n {
			fmt.Printf("Not prime.\n")
		} else {
			fmt.Printf("Prime number.\n")
		}
	}
	return
}

// c4goUnsafeConvert_int32 : created by c4go
func c4goUnsafeConvert_int32(c4go_name *int32) []int32 {
	return (*[1000000]int32)(unsafe.Pointer(c4go_name))[:]
}
```

# Example with binding function

C:

```c
#include <math.h>
#include <stdio.h>

int main()
{
    int n;
    double param = 8.0, result;
    result = frexp(param, &n);
    printf("result = %5.2f\n", result);
    printf("n      = %d\n", n);
    return 0;
}
```

`c4go` add automatically C binding for function without implementation:
```golang
//
//	Package - transpiled by c4go
//
//	If you have found any issues, please raise an issue at:
//	https://github.com/Konstantin8105/c4go/
//

package main

// #include </usr/include/math.h>
import "C"

import "github.com/Konstantin8105/c4go/noarch"
import "unsafe"

// main - transpiled function from  C4GO/examples/math.c:4
func main() {
	var n int32
	var param float64 = 8
	var result float64
	result = frexp(param, c4goUnsafeConvert_int32(&n))
	noarch.Printf([]byte("result = %5.2f\n\x00"), result)
	noarch.Printf([]byte("n      = %d\n\x00"), n)
	return
}

// c4goUnsafeConvert_int32 : created by c4go
func c4goUnsafeConvert_int32(c4go_name *int32) []int32 {
	return (*[1000000]int32)(unsafe.Pointer(c4go_name))[:]
}

// frexp - add c-binding for implemention function
func frexp(arg0 float64, arg1 []int32) float64 {
 	return float64(C.frexp(C.double(arg0), (*C.int)(unsafe.Pointer(&arg1[0]))))
}
```

# Example with C-pointers and C-arrays

```c
#include <stdio.h>

// input argument - C-pointer
void a(int* v1) { printf("a: %d\n", *v1); }

// input argument - C-array
void b(int v1[], int size)
{
    for (size--; size >= 0; size--) {
        printf("b: %d %d\n", size, v1[size]);
    }
}

int main()
{
    // value
    int i1 = 42;
    a(&i1);
    b(&i1, 1);

    // C-array
    int i2[] = { 11, 22 };
    a(i2);
    b(i2, 2);

    // C-pointer from value
    int* i3 = &i1;
    a(i3);
    b(i3, 1);

    // C-pointer from array
    int* i4 = i2;
    a(i4);
    b(i4, 2);

    // C-pointer from array
    int* i5 = i2[1];
    a(i5);
    b(i5, 1);

    return 0;
}
```

```go
//
//	Package - transpiled by c4go
//
//	If you have found any issues, please raise an issue at:
//	https://github.com/Konstantin8105/c4go/
//

package main

import "unsafe"
import "github.com/Konstantin8105/c4go/noarch"

// a - transpiled function from  C4GO/examples/ap.c:4
func a(v1 []int32) {
	// input argument - C-pointer
	noarch.Printf([]byte("a: %d\n\x00"), v1[0])
}

// b - transpiled function from  C4GO/examples/ap.c:7
func b(v1 []int32, size int32) {
	{
		// input argument - C-array
		for size -= 1; size >= 0; size-- {
			noarch.Printf([]byte("b: %d %d\n\x00"), size, v1[size])
		}
	}
}

// main - transpiled function from  C4GO/examples/ap.c:14
func main() {
	var i1 int32 = 42
	// value
	a(c4goUnsafeConvert_int32(&i1))
	b(c4goUnsafeConvert_int32(&i1), 1)
	var i2 []int32 = []int32{11, 22}
	// C-array
	a(i2)
	b(i2, 2)
	var i3 []int32 = c4goUnsafeConvert_int32(&i1)
	// C-pointer from value
	a(i3)
	b(i3, 1)
	var i4 []int32 = i2
	// C-pointer from array
	a(i4)
	b(i4, 2)
	var i5 []int32 = i2[1:]
	// C-pointer from array
	a(i5)
	b(i5, 1)

	return
}

// c4goUnsafeConvert_int32 : created by c4go
func c4goUnsafeConvert_int32(c4go_name *int32) []int32 {
	return (*[1000000]int32)(unsafe.Pointer(c4go_name))[:]
}
```

# Example of transpilation [neatvi](https://github.com/aligrudi/neatvi).

```bash
# clone git repository
git clone https://github.com/aligrudi/neatvi.git

# move in folder
cd neatvi

# look in Makefile
cat Makefile
```
Makefile:
```
CC = cc
CFLAGS = -Wall -O2
LDFLAGS =

OBJS = vi.o ex.o lbuf.o mot.o sbuf.o ren.o dir.o syn.o reg.o led.o \
	uc.o term.o rset.o regex.o cmd.o conf.o

all: vi

conf.o: conf.h

%.o: %.c
	$(CC) -c $(CFLAGS) $<
vi: $(OBJS)
	$(CC) -o $@ $(OBJS) $(LDFLAGS)
clean:
rm -f *.o vi
```
As we see not need any specific, for example `-clang-flag`.

Let's try transpile:
```bash
c4go transpile *.c
```
Result:
```
panic: clang failed: exit status 1:

/temp/neatvi/reg.c:6:14: error: redefinition of 'bufs' with a different type: 'char *[256]' vs 'struct buf [8]'
static char *bufs[256];
             ^
/temp/neatvi/ex.c:35:3: note: previous definition is here
} bufs[8];
  ^
/temp/neatvi/reg.c:12:9: error: returning 'struct buf' from a function with incompatible result type 'char *'
 return bufs[c];
        ^~~~~~~
/temp/neatvi/reg.c:17:81: error: invalid operands to binary expression ('int' and 'struct buf')
 char *pre = ((*__ctype_b_loc ())[(int) ((c))] & (unsigned short int) _ISupper) && bufs[tolower(c)] ? bufs[tolower(c)] : "";
             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^  ~~~~~~~~~~~~~~~~
/temp/neatvi/reg.c:21:7: error: passing 'struct buf' to parameter of incompatible type 'void *'
 free(bufs[tolower(c)]);
      ^~~~~~~~~~~~~~~~
/usr/include/stdlib.h:483:25: note: passing argument to parameter '__ptr' here
extern void free (void *__ptr) __attribute__ ((__nothrow__ ));
                        ^
/temp/neatvi/reg.c:22:19: error: assigning to 'struct buf' from incompatible type 'char *'
 bufs[tolower(c)] = buf;
                  ^ ~~~
/temp/neatvi/reg.c:43:8: error: passing 'struct buf' to parameter of incompatible type 'void *'
  free(bufs[i]);
       ^~~~~~~
/usr/include/stdlib.h:483:25: note: passing argument to parameter '__ptr' here
extern void free (void *__ptr) __attribute__ ((__nothrow__ ));
                        ^
/temp/neatvi/regex.c:98:12: error: static declaration of 'uc_len' follows non-static declaration
static int uc_len(char *s)
           ^
/temp/neatvi/vi.h:83:5: note: previous declaration is here
int uc_len(char *s);
    ^
/temp/neatvi/regex.c:200:14: error: static declaration of 'uc_beg' follows non-static declaration
static char *uc_beg(char *beg, char *s)
             ^
/temp/neatvi/vi.h:101:7: note: previous declaration is here
char *uc_beg(char *beg, char *s);
      ^
/temp/neatvi/vi.c:635:12: error: redefinition of 'linecount'
static int linecount(char *s)
           ^
/temp/neatvi/lbuf.c:124:12: note: previous definition is here
static int linecount(char *s)
           ^
9 errors generated.


goroutine 1 [running]:
main.generateAstLines(0x1, 0x0, 0xc000010140, 0x10, 0x10, 0x0, 0x0, 0x0, 0x0, 0x0, ...)
	/go/src/github.com/Konstantin8105/c4go/main.go:438 +0xd40
main.Start(0x1, 0x0, 0xc000010140, 0x10, 0x10, 0x0, 0x0, 0x0, 0x0, 0x0, ...)
	/go/src/github.com/Konstantin8105/c4go/main.go:356 +0xa9
main.runCommand(0x0)
	/go/src/github.com/Konstantin8105/c4go/main.go:723 +0xa45
main.main()
	/go/src/github.com/Konstantin8105/c4go/main.go:556 +0x22
```
That error is good and enought information for next step of transpilation.
Let's clarify one of them - we see 2 function with same function name and
if you open files and compare.
```
/temp/neatvi/vi.c:635:12: error: redefinition of 'linecount'
static int linecount(char *s)

/temp/neatvi/lbuf.c:124:12: note: previous definition is here
static int linecount(char *s)
```
This is 2 absolute identical function, so we can remove one of them.
I choose remove function `linecount` from file `vi.c`, because in
according to error - it is duplicate.

Also remove next function in according to errors:
* `vi.c` function `linecount`
* `regex.c` function `uc_len`
* `regex.c` function `uc_bed`

Next error is about duplicate of variable names:
```
/temp/neatvi/reg.c:6:14: error: redefinition of 'bufs' with a different type: 'char *[256]' vs 'struct buf [8]'
static char *bufs[256];
             ^
/temp/neatvi/ex.c:35:3: note: previous definition is here
} bufs[8];
```
Let's replace variable name in file `reg.c` and run:
```
sed -i.bak 's/bufs/bufs_postfix/g' reg.c
```

Now, let's try again and output in file `vi.go`:
```bash
c4go transpile -o vi.go *.c
```
Now transpiled without `clang` error. Look the result:
```bash
less vi.go
```
We see warning of lose some types:
```
// Warning (*ast.BinaryOperator):  /temp/neatvi/cmd.c:20 :Cannot transpile BinaryOperator with type 'int' : result type = {}. Error: operator is `=`. Cannot casting {__pid_t -> int}. err = Cannot resolve type '__pid_t' : I couldn't find an appropriate Go type for the C type '__pid_t'.
```

For fix that add flag `-s` for adding all type from C standard library:
```bash
c4go transpile -o vi.go -s *.c
```

Now, all is Ok. Look the result:
```bash
less vi.go
```


# C standard library implementation

```
            assert.h	       1/1	         100%
             ctype.h	     13/13	         100%
             errno.h	       0/1	           0%
             float.h	          	    undefined
            iso646.h	          	    undefined
            limits.h	          	    undefined
            locale.h	       0/3	           0%
              math.h	     22/22	         100%
            setjmp.h	       0/3	           0%
            signal.h	       3/3	         100%
            stdarg.h	       4/4	         100%
            stddef.h	       4/4	         100%
             stdio.h	     37/41	        90.2%
            stdlib.h	     31/37	        83.8%
            string.h	     21/24	        87.5%
              time.h	      1/15	        6.67%
             wchar.h	      3/60	           5%
            wctype.h	      0/21	           0%
```

# Contributing

Feel free to submit PRs or open issues.
Main information from: [en.cppreference.com](http://en.cppreference.com/w/c)

## Testing

By default, only unit tests are run with `go test`. You can also include
all tests:

```bash
go test ./...
```

For tests work like this:

1. `clang` compiles the C to a binary as normal.
2. `c4go` converts the C file to Go.
3. Both binaries are executed and the output is compared. All C files will
contain some output so the results can be verified.

## Note

### Use lastest version of clang.

If you use `Ubuntu`, then use command like next for choose `clang` version:

```bash
sudo update-alternatives --install /usr/bin/clang clang /usr/bin/clang-6.0 1000
```

### Performance

Main time of transpilation takes `clang`, for example run:
```bash
go test -run=Benchmark -bench=. -benchmem
```
Result looks for example:
```
goos: linux
goarch: amd64
pkg: github.com/Konstantin8105/c4go
BenchmarkTranspile/Full-6         	       5	 274922964 ns/op	43046865 B/op	  379676 allocs/op
BenchmarkTranspile/GoCode-6       	      20	  86806808 ns/op	36577533 B/op	  308060 allocs/op
PASS
```
So, transpilation time is just 30% of full time. In my point of view
no need of performance optimization, see [Amdahl's law](https://en.wikipedia.org/wiki/Amdahl%27s_law).

### Example of performance analyse

Please run:

```bash
# Run cpuprofiling for sqlite transpilation example
time ./scripts/sqlite.sh 

# Example of output:
#
# % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
#                                 Dload  Upload   Total   Spent    Left  Speed
#100 2217k  100 2217k    0     0   235k      0  0:00:09  0:00:09 --:--:--  357k
#Archive:  /tmp/SQLITE/sqlite-amalgamation-3250200.zip
#   creating: /tmp/SQLITE/sqlite-amalgamation-3250200/
#  inflating: /tmp/SQLITE/sqlite-amalgamation-3250200/sqlite3ext.h  
#  inflating: /tmp/SQLITE/sqlite-amalgamation-3250200/sqlite3.c  
#  inflating: /tmp/SQLITE/sqlite-amalgamation-3250200/sqlite3.h  
#  inflating: /tmp/SQLITE/sqlite-amalgamation-3250200/shell.c  
#After transpiling shell.c and sqlite3.c together, have summary: 695 warnings.
#In file sqlite.go summary : 3 warnings in go build.
#Amount unsafe package using: 2902
#
#real	0m18.434s
#user	0m14.212s
#sys	0m1.434s

# Run profiler
go tool pprof ./testdata/cpu.out
```

For more information, see [Profiling Go Programs](https://blog.golang.org/profiling-go-programs).



c4go

Golang to Haxe to CPP/CSharp/Java/JavaScript transpiler.

# TARDIS Go -> Haxe transpiler

#### Haxe -> C++ / C# / Java / JavaScript 

[![Build Status](https://travis-ci.org/tardisgo/tardisgo.png?branch=master)](https://travis-ci.org/tardisgo/tardisgo)
[![GoDoc](https://godoc.org/github.com/tardisgo/tardisgo?status.png)](https://godoc.org/github.com/tardisgo/tardisgo)
[![status](https://sourcegraph.com/api/repos/github.com/tardisgo/tardisgo/badges/status.png)](https://sourcegraph.com/github.com/tardisgo/tardisgo)

## Project status: a non-working curiosity, development currently on-ice

### The advent of Go 1.5, with support for both iOS and Android, together with [GopherJS](http://www.gopherjs.org/), mean that the cross-platform UI objectives of this project are likely to soon be redundant, as libraries to achieve this can now be written directly in Go. [The author](https://github.com/elliott5) has therefore paused development of the Haxe aspect of this project, freezing the runtime at Go 1.4. With the advent of Go 1.6, the code compiles but the tests no-longer pass.

All of the core [Go language specification](http://golang.org/ref/spec) is implemented, including single-threaded goroutines and channels. However the package "reflect", which is mentioned in the core specification, is not yet fully supported. 

Goroutines are implemented as co-operatively scheduled co-routines. Other goroutines are automatically scheduled every time there is a channel operation or goroutine creation (or call to a function which uses channels or goroutines through any called function). So loops without channel operations may never give up control. The function runtime.Gosched() provides a convenient way to allow other goroutines to run.  

[Well over half of the standard packages pass their tests for all targets](https://github.com/tardisgo/tardisgo/blob/master/STDPKGSTATUS.md). 

The GOOS for TARDISgo is ['nacl'](https://github.com/golang/go/wiki/NativeClient), complete with an in-memory file system. However please note that NaCl provides no access to traditional networking. Go programs written for TARDISgo must use Haxe APIs to access host OS functionality.

A start has been made on the automated integration with Haxe libraries, but this is incomplete and the API unstable, see the haxe/hx directory and gohaxelib repository for the story so far. 

The code is developed and tested on OS X 10.10.2, using Go 1.5rc1 and Haxe 3.2.0. The short CI test runs on 64-bit Ubuntu. No other platforms are currently regression tested. 

Please note that the Haxe compiler may require in excess of 2Gb of memory to compile a non-trivial program to Java or C#.

To see a handful of live visual examples of code generated by this project in operation, please see the top and right-hand-side of [tardisgo.github.io](http://tardisgo.github.io).

## Installation and use:
 
```
go get -u github.com/tardisgo/tardisgo
```

If tardisgo is not installing and there is a green "build:passing" icon at the top of this page, please e-mail [Elliott](https://github.com/elliott5)!

To translate Go to Haxe, from the directory containing your .go files type the command line: 
```
tardisgo yourfilename.go 
``` 
A large number of .hx files will be created in the tardis subdirectory, of which Go.hx contains the entry-point. The use of a file per Haxe class makes second and subsequent compilations using C++ much faster, as only the altered classes are recompiled.

To run your transpiled code you will first need to install [Haxe](http://haxe.org), version 3.2.0.

Then to run the tardis/Go.hx file generated above, for example in JavaScript, type the command lines: 
```
haxe -main tardis.Go -cp tardis -dce full -D uselocalfunctions -js tardis/go.js
node < tardis/go.js
```
... or whatever [Haxe compilation options](http://haxe.org/documentation/introduction/compiler-usage.html) you want to use. 
See the [tgoall.sh](https://github.com/tardisgo/tardisgo-samples/blob/master/scripts/tgoall.sh) script for simple examples. Note that in this example "-dce full" causes Haxe to do dead code elimination and that "-D uselocalfunctions" is a tardisgo haxe flag to generate JS code that is more likely to be optimized by V8.

The default memory model is fast, but requires more memory than you might expect (an int per byte) and only allows some unsafe pointer usages. If your code uses unsafe pointers to re-use memory as different types (say writing a float64 but reading back a uint64), there is a Haxe compilation flag for "fullunsafe" mode (this is slower, but has a smaller memory footprint and allows most unsafe pointers to be modeled accurately). In JS fullunsafe uses the dataview method of object access, for other targets it simulates memory access. Fullunsafe is little-endian only at present and pointer arithmetic (via uintptr) will panic. A command line example: 
```
tardisgo mycode.go
haxe -main tardis.Go -cp tardis -D fullunsafe -js tardis/go-fu.js
node < tardis/go-fu.js
```

While on the subject of JS, the closure compiler seems to work, but only using the default "SIMPLE_OPTIMIZATIONS" option. It currently generates a large number of warnings.

The in-memory filesystem used by the nacl target is implemented, it can be pre-loaded with files by using the haxe command line flag "-resource" with the name "local/file/path/a.txt@/nacl/file/path/a.txt" thus (for example in JS):
```
tardisgo your_code_using_package_os.go
haxe -main tardis.Go -cp tardis -js tardis/go.js -resource testdata/config.xml@/myapp/static/config.xml
node < tardis/go.js
```
To add more than one file, use multiple -resource flags (the haxe ".hxml" compiler parameter file format can be helpful here). The files are stored as part of the executable code, in a target-specific way. The only resources that will be loaded are those named with a leading "/". A log file of the load process can be found at "/fsinit.log" in the in-memory file-system.

To load a zipped file system (very slow to un-zip, but useful for testing) use go code
`syscall.UnzipFS("myfs.zip")` 
and include 
`-resource myfs.zip` on the haxe command line.

To add Go build tags, use the "-tags 'name1 name2'" tardisgo compilation flag. Note that particular Go build tags are required when compiling for OpenFL using the [pre-built Haxe API definitions](https://github.com/tardisgo/gohaxelib). 

Use the "-debug" tardisgo compilation flag to instrument the code and add automated comments to the Haxe. When you experience a panic in this mode the latest Go source code line information and local variables appears in the stack dump. For the C++ & Neko (--interp) targets, a very simple debugger is also available by using the "-D godebug" Haxe flag, for example to use it in C++ type:
```
tardisgo -debug myprogram.go
haxe -main tardis.Go -cp tardis -dce full -D godebug -cpp tardis/cpp
./tardis/cpp/Go
``` 
To get a list of commands type "?" followed by carriage return, after the 1st break location is printed (there is no prompt character). 

To run cross-target command-line tests as quickly as possible, the "-haxe X" flag concurrently runs the Haxe compiler and executes the resulting code as follows:
- "-haxe all" - all supported targets (C++, C#, Java, JavaScript)
- "-haxe bench" - all supported targets (C++, C#, Java, JavaScript) but using benchmark settings
- "-haxe js" - only compiles and runs nodeJS (for automated testing, exits with an error if one occurs)
- "-haxe jsfu" - only compiles (-D fullunsafe) and runs nodeJS (for automated testing, exits with an error if one occurs)
- "-haxe cpp" - only compiles and runs C++ (for automated testing, exits with an error if one occurs)
- "-haxe cs" - only compiles and runs C# (for automated testing, exits with an error if one occurs)
- "-haxe java" - only compiles and runs Java (for automated testing, exits with an error if one occurs)
- "-haxe math" - only runs C++ and JS with the -D fullunsafe haxe flag (using JS dataview)
- "-haxe interp" - only runs the haxe interpreter (for automated testing, exits with an error if one occurs)

Compiler output is suppressed and results appear in the order they complete, with an execution time, for example:
```
tardisgo -haxe all myprogram.go
```

When using the -haxe flag with the -test flag, if the file "tgotestfs.zip" exists in the current directory, it will be added as a haxe resource and its contents auto-loaded into the in-memory file system. 

If you can't work-out what is going on prior to a panic, you can add the "-trace" tardisgo compilation flag to instrument the code even further, printing out every part of the code visited. But be warned, the output can be huge.

Please note that strings in Go are held as Haxe strings, but encoded as UTF-8 even when strings for that host are encoded as UTF-16. The system should automatically do the translation to/from the correct format at the Go/Haxe boundary, but there are certain to be some occasions when a translation has to be done explicitly (see Force.toHaxeString/Force.fromHaxeString in haxe/haxeruntime.go).

## Benchmarks

Tabulating the very simple indicative [benchmarking](https://github.com/tardisgo/tardisgo-samples/blob/master/benchmarks) results, looking only at elapsed (rather than cpu) time in seconds, as a multiple of the Go time: 

| Test - of what functionality         | C++      | C#       | Java     | JS       |Closure/JS| [GopherJS](http://www.gopherjs.org/) |
| ------------------------------------ | -------- | -------- | -------- | -------- | -------- | -------- |
| mandel.go - floating point           | 1.07x    | 2.73x    | 1.25x    | 1.24x    | 1.11x    | 0.99x    |
| fannkuch.go - slice & array indexing | 2.09x    | 4.22x    | 3.35x    | 5.31x    | 5.23x    | 3.54x    |
| binarytree.go - garbage collection   | 16.92x   | 11.62x   | 1.61x    | 8.56x    | 6.58x    | 0.32x (!)|

Figures above are the latest results as at 2nd July 2015, including performace figures after running the [Google Closure Compiler](https://developers.google.com/closure/compiler/) on the JS and from the parallel project GopherJS (un-minified). The Haxe compilation flag "-D inlinepointers" was used for all targets, the additional flag "-D useloacalfunctions" was used for the JS target.

Execution speed significantly improved after:
* re-writing code generation for non-goroutine functions to reconstruct Haxe "while" and "if" control structures from the SSA form where possible; and
* optimising away local pointers that are only used in a local sub-block and creating temporary local variables to speed execution and reduce code size; however
* since garbage collection currently relies on the target language runtime, the same Haxe code produces very different results.

Expect further gradual speed improvements. 

## Unsupported Haxe targets: ActionScript, PHP, Python and Neko

The nature of ActionScript/Flash means that it is not possible to run automated tests, although it seems to be a reliable target. 

The PHP, Python and Neko targets are not currently reliable enough to permit automated testing. 

PHP specific issues:
* to compile for PHP you currently need to add the haxe compilation option "--php-prefix tgo" to avoid name conflicts
* very long PHP class/file names may cause name resolution problems on some platforms

## Next steps:
Please go to http://github.com/tardisgo/tardisgo-samples for example Go code modified to work with tardisgo. Including some very simple [example code](http://github.com/tardisgo/tardisgo-samples). 

For a small technical FAQ, please see the [Wiki page](https://github.com/tardisgo/tardisgo/wiki). 

For public help or discussion please go to the [Google Group](https://groups.google.com/d/forum/tardisgo); or feel free to e-mail [Elliott](https://github.com/elliott5) direct to discuss any issues if you prefer.

The documentation is sparse at present, if there is some aspect of the system that you want to know more about, please let [Elliott](https://github.com/elliott5) know and he will prioritise that area to add to the wiki.

If you transpile your own code using TARDIS Go, please report the bugs that you find here, so that they can be fixed.

## Why do it at all?

The internal objective of the project was for [Elliott](https://github.com/elliott5) to learn Go and understand the Go runtime. For both the Go language as a whole, and for the parts of the runtime implemented in TARDISgo, this objective been achieved. However some of the early code that was written would benefit from being refactored.

The external objective of this project was to enable the same [Go](http://golang.org) code to be re-deployed in  as many different execution environments as possible, thus saving development time and effort. 

However the advent of Go 1.5, with support for both iOS and Android, together with the excellent GopherJS project, mean that the cross-platform UI objectives of this project are now largely redundant.

But it does work. The example code demonstrates both simple command-line examples and multi-platform client-side applications in Go using the APIs available in the Haxe ecosystem:
- The standard [Haxe APIs](http://api.haxe.org/) for JavaScript and Flash.
- [OpenFL](http://openfl.org) "Open Flash" to target HTML5, Windows, Mac, Linux, iOS, Android, BlackBerry, Firefox OS, Tizen and Flash, using compiled auto-generated C++ code where possible (star users are TiVo and Prezzi).

Target long-term use-cases (if the generated code and runtime environment were developed further to make them more efficient) would be:
- For the Haxe community: provide access to the portable elements of Go's extensive libraries and open-source code base.
- For the Go community: write code in Go and call to-and-from existing Haxe, JavaScript, Java or C#  applications (in C++ you would probably just link as normal through CGo). 

## Future plans:

The project is currently on ice, pending inspiration for developing it further. As at August 2015, it is the only way to target the JVM or CLR from Go.

The Author retains an interest in collaborating in the development of related project ideas, like:
- cross-platform development for Go;
- generating code from Go's SSA form; or
- making Go a target for Haxe.

If you would like to discuss this or any of the above project ideas, that would be wonderful, please contact [Elliott](https://github.com/elliott5). 

## License:

MIT license, please see the license file.


The Best Go Libraries For Go Compilers (6)

esp32

f4go

gopherjs

c2go

c4go

tardisgo