https://github.com/borud/hdlc

HDLC'ish unframer
https://github.com/borud/hdlc

Last synced: about 2 months ago
JSON representation

HDLC'ish unframer

• Host: GitHub
• URL: https://github.com/borud/hdlc
• Owner: borud
• License: apache-2.0
• Created: 2022-04-01T17:46:13.000Z (about 4 years ago)
• Default Branch: main
• Last Pushed: 2022-05-30T18:54:29.000Z (about 4 years ago)
• Last Synced: 2025-02-24T10:43:35.516Z (over 1 year ago)
• Language: Go
• Size: 13.7 KB
• Stars: 1
• Watchers: 2
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# HDLC

[![Godoc](http://img.shields.io/badge/godoc-reference-blue.svg?style=flat)](https://godoc.org/github.com/borud/hdlc)

HDLC is a library for unframing HDLC-like frames. Instead of implementing this as an io.Reader,

we read whole frames from a channel.  This avoids misunderstandings that might arise if you use the io.Reader interface.

Assuming you have an `io.Reader` named `myReader` this is how you unframe things.

```go

import "github.com/borud/hdlc"

...

 unframer := NewUnframer(myReader)

...

 for range unframer.Frames() {

     // do something with the frames

 }

```

## Framing format

The framing format defines three special values:

```go

const (

    FlagEscape = byte(0x7d)

    FlagSep    = byte(0x7e)

    FlagAbort  = byte(0x7f)

    XORMask    = byte(0x20)

)

```

- Each frame starts and ends with a `FlagSep` byte

- If we encounter a `FlagAbort` value it means we abort the current frame and wait for the next frame beginning before accumulating data.

- If values equal to `FlagEscape`, `FlagSep` or `FlagAbort` occur in the payload it needs to be escaped. Escaping happens by replacing the value with two bytes: FlagEscape followed by the value we are escaping XOR'ed by XORMask (we flip the 5th bit).

Unframing is just applying this process in reverse.

### API choice

Rather than implement the unframer as an `io.Reader` we explicitly chose an API that reflects that this is not an `io.Reader`.  We are not reading a stream, we are reading discrete frames. This means that if you were to wrap it in something that assumes a stream, you would run into problems.