Use Kubo (go-ipfs) as a library to spawn a node and add a file

Note: if you are trying to customize or extend Kubo, you should read the Customizing Kubo doc

By the end of this tutorial, you will learn how to:

Spawn an IPFS node that runs in process (no separate daemon process)
Create an IPFS repo
Add files and directories to IPFS
Retrieve those files and directories using cat and get
Connect to other nodes in the network
Retrieve a file that only exists on the network
The difference between a node in DHT client mode and full DHT mode

All of this using only golang!

In order to complete this tutorial, you will need:

golang installed on your machine. See how at https://golang.org/doc/install
git installed on your machine (so that go can download the repo and the necessary dependencies). See how at https://git-scm.com/downloads
IPFS Desktop (for convenience) installed and running on your machine. See how at https://github.com/ipfs-shipyard/ipfs-desktop#ipfs-desktop

Disclaimer: The example code is quite large (more than 300 lines of code) and it has been a great way to understand the scope of the go-ipfs Core API, and how it can be improved to further the user experience. You can expect to be able to come back to this example in the future and see how the number of lines of code have decreased and how the example have become simpler, making other go-ipfs programs simpler as well.

Getting started

Note: Make sure you have installed.

Download Kubo and jump into the example folder:

console

$ git clone https://github.com/ipfs/kubo.git
$ cd kubo/docs/examples/kubo-as-a-library

Running the example as-is

To run the example, simply do:

console

$ go run main.go

You should see the following as output:

-- Getting an IPFS node running --
Spawning Kubo node on a temporary repo
IPFS node is running

-- Adding and getting back files & directories --
Added file to IPFS with CID /ipfs/QmV9tSDx9UiPeWExXEeH6aoDvmihvx6jD5eLb4jbTaKGps
Added directory to IPFS with CID /ipfs/QmdQdu1fkaAUokmkfpWrmPHK78F9Eo9K2nnuWuizUjmhyn
Got file back from IPFS (IPFS path: /ipfs/QmV9tSDx9UiPeWExXEeH6aoDvmihvx6jD5eLb4jbTaKGps) and wrote it to ./example-folder/QmV9tSDx9UiPeWExXEeH6aoDvmihvx6jD5eLb4jbTaKGps
Got directory back from IPFS (IPFS path: /ipfs/QmdQdu1fkaAUokmkfpWrmPHK78F9Eo9K2nnuWuizUjmhyn) and wrote it to ./example-folder/QmdQdu1fkaAUokmkfpWrmPHK78F9Eo9K2nnuWuizUjmhyn

-- Going to connect to a few nodes in the Network as bootstrappers --
Fetching a file from the network with CID QmUaoioqU7bxezBQZkUcgcSyokatMY71sxsALxQmRRrHrj
Wrote the file to ./example-folder/QmUaoioqU7bxezBQZkUcgcSyokatMY71sxsALxQmRRrHrj

All done! You just finalized your first tutorial on how to use Kubo as a library

Understanding the example

In this example, we add a file and a directory with files; we get them back from IPFS; and then we use the IPFS network to fetch a file that we didn't have in our machines before.

Each section below has links to lines of code in the file main.go. The code itself will have comments explaining what is happening for you.

The `func main() {}`

The main function is where the magic starts, and it is the best place to follow the path of what is happening in the tutorial.

Part 1: Getting an IPFS node running

To get get a node running as an ephemeral node (that will cease to exist when the run ends), you will need to:

As soon as you construct the IPFS node instance, the node will be running.

Part 2: Adding a file and a directory to IPFS

Part 3: Getting the file and directory you added back

Part 4: Getting a file from the IPFS network

Bonus: Spawn a daemon on your existing IPFS repo (on the default path ~/.ipfs)

As a bonus, you can also find lines that show you how to spawn a node over your default path (~/.ipfs) in case you had already started a node there before. To try it:

Voilá! You are now a Kubo hacker

You've learned how to spawn a Kubo node using the Core API. There are many more methods to experiment next. Happy hacking!

README