matoous/go-nanoid

Golang random IDs generator.

👋 Hey! I am a software engineer at SumUp. I work on Identity & Access Management, observability, and public APIs. We have a team full of wonderful and talented people, so if you are looking for a job or want to chat about any random topic, reach out to me at matous.dzivjak @ sumup.com.

I write in Go and Rust. I use Helix as my driver and also contribute to the development when I find some time.

There's more at my personal website.

go-nanoid's Issues

Haven't handle edge cases correctly.

This is got into deadlock.

package main

import (
	"fmt"

	"github.com/matoous/go-nanoid"
)

func main() {
	gonanoid.Size(0)
	gonanoid.Alphabet("abcdefghijklmnopqrstuvwxyz")
	id := gonanoid.Generate()
	fmt.Printf("ID: %s", id)
}

And this is got panic.

package main

import (
	"fmt"

	"github.com/matoous/go-nanoid"
)

func main() {
	gonanoid.Alphabet("abcdefghijklmnopqrstuvwxyz")
	gonanoid.Size(0)
	id := gonanoid.Generate()
	fmt.Printf("ID: %s", id)
}

similar error when we are using gonanoid.Alphabet("")

Dependency Dashboard

This issue lists Renovate updates and detected dependencies. Read the Dependency Dashboard docs to learn more.

Open

These updates have all been created already. Click a checkbox below to force a retry/rebase of any.

• fix(deps): update module github.com/stretchr/testify to v1.9.0
• chore(deps): update actions/checkout action to v4
• chore(deps): update actions/setup-go action to v5
• chore(deps): update golangci/golangci-lint-action action to v4
• Click on this checkbox to rebase all open PRs at once

Detected dependencies

github-actions

.github/workflows/lint.yml

• actions/checkout v2
• golangci/golangci-lint-action v2

.github/workflows/test.yml

• actions/setup-go v2
• actions/checkout v2
• actions/setup-go v2
• actions/checkout v2

gomod

go.mod

• go 1.15
• github.com/stretchr/testify v1.7.0

---

• Check this box to trigger a request for Renovate to run again on this repository

Compatibility with V3?

Just want to confirm that this package is compatible with v3 nanoid: https://github.com/ai/nanoid/releases/tag/3.0.0

Importing v2.0.0

Hi,
When I go get github.com/matoous/go-nanoid it retrieves the version 1.5.0.

go: downloading github.com/matoous/go-nanoid v1.5.0
go: github.com/matoous/go-nanoid upgrade => v1.5.0

I would like to use the most recent version (aka 2.0.0) but I'm not sure how to download/add it to my project.

The call order of the method(Size & Alphabet) affects the value of the intermediate variable(step)

step = int(math.Ceil(1.6 * float64(mask) * float64(size) / float64(len(alphabet))))	

According to to formula, step affected by size, this means:

1. If we call Size() before Alphabet(), step's value is correct;
2. If we call Size() after Alphabet(), step's value is incorrect;

Why is the default length 22?

I looked at the primary implementation as well as the implementation for other languages, and the default length for the id is 21. Is there specific reason to make it 22 in this library? In my opinion, it should be consistent with rest of them.

Broken release?

This dependency don't working

require (
	github.com/matoous/go-nanoid v1.4.0
)

go: github.com/matoous/[email protected]: reading github.com/matoous/go-nanoid/go.mod at revision v1.4.0: unknown revision v1.4.0

But working this.

require (
	github.com/matoous/go-nanoid v.1.4.0
)

Which change automatically in my go.mod to:

require (
	github.com/matoous/go-nanoid v1.3.1-0.20200518070138-513938f8e67d
)

Lack of Must / New on API

Hello!

I'm migrating from Google's UUID Package, and there is an specific detail that I find odd: the lack of Must and/or New on the API. These two functions are useful when you're quickly prototyping or when you DO want the entire world to blown up in case of errors.

These two functions have the following signature:

func Must (id string, err error) string // panics if err is not null
func New() string // returns a new id or panics

Is this feature acceptable? If so, I already made the changes locally on my machine and can send you a PR.

Non ASCII ID's

I try to use this library to generate simple human readable ID's on my native language.
Using library according to documentation with non ascii alphabet:
id, err := gonanoid.Generate("абвгдежиклмнпрстуфхшэюя0123456789-", 10)
Gives me byte garbage with wrong length

3ю�ѽ�д0
�Џ�51�б7
���Ѷ0�Ѻ�

Instead normal ID like:

2юбнг4внеэ

Original implementation has no this problem https://zelark.github.io/nano-id-cc/

Nanoid() with negative argument panics

If Nanoid() function is called with a negative argument, the function panics.

E.g.

gonanoid.Nanoid(-1)

panics with:

panic: runtime error: makeslice: len out of range [recovered]                           
        panic: runtime error: makeslice: len out of range 

The function signature already allows to return an error so I suggest to return an error in this case instead of letting it panic.

Additionally, due to the (in my opinion not very Go ideomatic) function signature, it is possible to call this function with more than one argument. I suggest to return an error in this case as well.

I can provide a PR if you wish.

func naming suggestion

wouldnt it be better to use gonanoid.New(), gonanoid.Generate() or if the package had different name generator.Nanoid() or generator.NewNanoid()?

API 2.0

Let’s update the library to be compatible with the latest Nano ID 2.0:

• Replace ~ to - in default alphabet
• Add non-secure fast generator
• Async API?

sortable ids

does it possible to get sortable nanoids ?

implement BinaryMarshaler and TextMarshaler and also sql.Scan and sql.Value

New discards 25% of the random bits

New uses bits 0-6 of each random byte (masks with 63) and the two high random bits are discarded.

I think it would be better not to discard randomness.

go-nanoid/gonanoid.go

Lines 89 to 97 in 0cc1b25

	bytes := make([]byte, size)
	_, err := rand.Read(bytes)
	if err != nil {
	return "", err
	}
	id := make([]rune, size)
	for i := 0; i < size; i++ {
	id[i] = defaultAlphabet[bytes[i]&63]
	}

It is possible to write a loop and output 4 characters per each 3 random bytes and handle the remainder separately.

Or maybe reuse the base64 encoding code from standard library.