Hasura CLI Installation & Notes

This post just elaborates on the existing documentation here with a few extra notes and details about installation. This can be helpful when determining how to install, deploy, or use the Hasura CLI for development, continuous integration, or continuous deployment purposes.

Installation

There are three primary operating system binary installations: Windows, MacOS, and Linux.

Linux

In the shell run the following curl command to download and install the binary.

curl -L https://github.com/hasura/graphql-engine/raw/stable/cli/get.sh | bash

This command installs the command to /usr/local/bin, but if you’d like to install it to another location you can add the follow variable INSTALL_PATH and set it to the path that you want.

curl -L https://github.com/hasura/graphql-engine/raw/stable/cli/get.sh | INSTALL_PATH=$HOME/bin bash

What this script does, in short, follows these steps. The latest version is determined, then downloaded with curl. Once that is done, it then assigns permissions to the binary for execution. It also does some checks to determine OS version, type, distro, if the CLI exists or not, and other validations. To download the binary, and source if you’d want for a particular version of the binary, check out the manual steps listed later in this post.

MacOS

For MacOS the same steps are followed as Linux, as the installation script steps through the same procedures.

curl -L https://github.com/hasura/graphql-engine/raw/stable/cli/get.sh | bash

As with the previous Linux installation, the INSTALL_PATH variable can also be set if you’d want the installation of the binary to another path.

Windows

Windows is a different installation, as one might imagine. The executable (cli-hasura-windows-amd64.exe) is available on the releases page. This leaves it up to you to determine how exactly you want this executable to be called. It’s ideal, in my opinion, to download this and then put it into a directory where you’ll map the path. You’ll also want to rename the executable itself from cli-hasura-windows-amd64.exe to hasura.exe if for any reason because it’s easier to type and then it’ll match the general examples provided in the docs.

To setup a path on Windows to point to the directory where you have the executable, you’ll want to open up ht environment variables dialog. That would be following Start > System > System Settings > Environment Variables. Scroll down until the PATH is viewable. Click the edit button to edit that path. Set it, and be sure to set it up like c:\pathWhatevsAlreadyHere;c:\newPath\directory\where\hasura\executable\is\. Save that, launch a new console and that new console should have the executable available now.

Manual Download & Installation

You can also navigate directly to the releases page and get the CLI at https://github.com/hasura/graphql-engine/releases. All of the binaries are compiled and ready for download along with source code zip file of the particular builds for those binaries.

Installation via npm

The CLI is avialable via an npm package also. It is independently maintained (the package that wraps the executable) by members in the community. If you want to provide a set Hasura CLI version to a project, using npm is a great way to do so.

For example, if you want to install the Hasura CLI, version in your project as a development dependency, use the following command to get version 1.3.0 for example.

npm install --save-dev hasura-cli@1.3.0

For version 1.3.1 it would be npm install --save-dev hasura-cli@1.3.1 for example.

The dev dependencies in the package.json file of your project would then look like the following.

"devDependencies": {
  "hasura-cli": "^1.3.0"
}

Another way you can install the CLI with npm is to just install it globally, with the same format but swap the --save-dev with --global. The following would install the latest command. You can add the @version to the end and get a specific version installed globally too, just like as shown with the dev install previously.

npm install --global hasura-cli

Notes

Using the npm option is great, if you’re installing, using, writing, or otherwise working with JavaScript, have Node.js installed on the dev and other machines, and need to have the CLI available on those particular machine instances. If not, I’d suggest installing via one of the binary options, especially if you’re creating something like a slimmed down Alpine Linux container to automate some Hasura CLI executions during a build process or something. There are a lot of variance to how you’d want to install and use the CLI, beyond just installing it to run the commands manually.

If you’re curious about any particular installation scenarios, ask me @Adron and I’ll answer there and I’ll elaborate here on this post!

Happy Hasura CLI Hacking! 🤘

References

Simple Go HTTP Server Starter in 15 Minutes

This is a quick starter, to show some of the features of Go’s http library. (sample repo here) The docs for the library are located here if you want to dig in deeper. In this post however I cover putting together a simple http server with a video on creating the http server, setting a status code, which provides the basic elements you need to further build out a fully featured server. This post is paired with a video I’ve put together, included below.

00:15 Reference to the Github Repo where I’ve put the code written in this sample.
https://github.com/Adron/coro-era-coding-go-sample
00:18 Using Goland IDE (Jetbrains) to clone the repository from Github.
00:40 Creating code files.
01:00 Pasted in some sample code, and review the code and what is being done.
02:06 First run of the web server.
02:24 First function handler to handle the request and response of an HTTP request and response cycle.
04:56 Checking out the response in the browser.
05:40 Checking out the same interaction with Postman. Also adding a header value and seeing it returned via the browser & related header information.
09:28 Starting the next function to provide further HTTP handler functionality.
10:08 Setting the status code to 200.
13:28 Changing the status code to 500 to display an error.

Getting a Server Running

I start off the project by pulling an empty repository that I had created before starting the video. In this I use the Jetbrains Goland IDE to pull this repository from Github.

1

Next I create two files; main.go and main_test.go. We won’t use the main_test.go file right now, but in a subsequent post I’ll put together some unit tests specifically to test out our HTTP handlers we’ll create. Once those are created I pasted in some code that just has a basic handler, using an anonymous function, provides some static file hosting, and then sets up the server and starts listening. I make a few tweaks, outlined in the video, and execute a first go with the following code.

package main

import (
    "fmt"
    "net/http"
)

func main() {
    http.HandleFunc("/", func (w http.ResponseWriter, r *http.Request) {
        fmt.Fprintf(w, "Welcome to my website!")
    })

    http.ListenAndServe(":8080", nil)
}

When that executes, opening a browser to localhost:8080 will bring up the website which then prints out “Welcome to my website!”.

2

Adding a Function as an HTTP Handler

The next thing I want to add is a function that can act as an HTTP handler for the server. To do this create a function just like we’d create any function in Go. For this example, the function I built included several print line calls to the ResponseWriter with Request properties and a string passed in.

func RootHandler(w http.ResponseWriter, r *http.Request){
    fmt.Fprintln(w, "This is my content.")
    fmt.Fprintln(w, r.Header)
    fmt.Fprintln(w, r.Body)
}

In the func main I changed out the root handler to use this newly created handler instead of the anonymous function that it currently has in place. So swap out this…

http.HandleFunc("/", func (w http.ResponseWriter, r *http.Request) {
    fmt.Fprintf(w, "Welcome to my website!")
})

with this…

http.HandleFunc("/", RootHandler)

Now the full file reads as shown.

package main

import (
    "fmt"
    "net/http"
)

func main() {
    http.HandleFunc("/", RootHandler)
    http.ListenAndServe(":8080", nil)
}

Now executing this and navigating to localhost:8080 will display the following.

3

The string is displayed first “This is my content.”, then the header, and body respectively. The body, we can see is empty. Just enclosed with two braces {}. The header is more interesting. It is returned as a map type, between the brackets []. Showing an accept, accept-encoding, accept-language, user-agent, and other header information that was passed.

This is a good thing to explore further, check out how to view or set the values associated with the header values in HTTP responses, requests, and their related metadata. To go a step further, and get into this metadata a tool like Postman comes in handy. I open this tool up, setup a GET request and add an extra header value just to test things out.

4

Printing Readable Body Contents

For the next change I wanted to get a better print out of body contents, as the previous display was actually just attempting to print out the body in an unreadable way. In this next section I used an ioutil function to get the body to print out in a readable format. The ioutil.ReadAll function takes the body, then I retrieve a body variable with the results, pending no error, the body variable is then cast as a string and print out to the ResponseWriter on the last line. The RootHandler function then reads like this with the changes.

func RootHandler(w http.ResponseWriter, r *http.Request){
    fmt.Fprintln(w, "This is my content.")
    fmt.Fprintln(w, r.Header)

    defer r.Body.Close()

    body, err := ioutil.ReadAll(r.Body)
    if err != nil {
        fmt.Fprintln(w, err)
    }
    fmt.Fprintln(w, string(body))
}

If the result is then requested using Postman again, the results now display appropriately!

5

Response Status Codes!

HTTP Status codes fit in to a set of ranges for various categories of responses. The most common code is of course the success code, which is 200 “Status OK”. Another common one is status code 500, which is a generic catch all for “Server Error”. The ranges are as follows:

  • Informational responses (100–199)
  • Successful responses (200–299)
  • Redirects (300–399)
  • Client errors (400–499)
  • and Server errors (500–599)

For the next function, to get an example working of how to set this status code, I added the following function.

func ResponseExampleHandler(w http.ResponseWriter, r *http.Request) {
    w.WriteHeader(200)
    fmt.Fprintln(w, "Testing status code. Manually added a 200 Status OK.")
    fmt.Fprintln(w, "Another line.")
}

With that, add a handler to the main function.

http.HandleFunc("/response", ResponseExampleHandler)

Now we’re ready to try that out. In the upper right of Postman, the 200 status is displayed. The other data is shown in the respective body & header details of the response.

6

Next up, let’s just write a function specifically to return an error. We’ll use the standard old default 500 status code.

func ErrorResponseHandler(w http.ResponseWriter, r *http.Request) {
    w.WriteHeader(500)
    fmt.Fprintln(w, "Server error.")
}

Then in main, as before, we’ll add the http handle for the function handler.

http.HandleFunc("/errexample", ErrorResponseHandler)

Now if the server is run again and an HTTP request is sent to the end point, the status code changes to 500 and the message “Server error.” displays on the page.

7

Summary

That’s a quick intro to writing an HTTP server with Go. From here, we can take many next steps such as writing tests to verify the function handlers, or setup a Docker image in which to deploy the server itself. In subsequent blog entries I’ll write up just those and many other step by step procedures. For now, a great next step is to expand on trying out the different functions and features of the http library.

That’s it for now. However, if you’re interested in joing me to write some JavaScript, Go, Python, Terraform, and more infrastructure, web dev, and coding in general I stream regularly on Twitch at https://twitch.tv/adronhall, post the VOD’s to YouTube along with entirely new tech and metal content at https://youtube.com/c/ThrashingCode. Feel free to check out a coding session, ask questions, interject, or just come and enjoy the tunes!

Creating a Go Module & Writing Tests in Less Than 3 Minutes

In this short (less than 4 minutes) video I put together a Go module library, then setup some first initial tests calling against functions in the module.

 

00:12 – Writing unit tests.

00:24 – Create the first go file.

00:40 – Creating the Go go.mod file. When creating the go.mode file, note the command is go mod init [repopath] where repopath is the actual URI to the repo. The go.mod file that is generated would look like this.

module github.com/adron/awesomeLib

go 1.13

00:56 Enabling Go mod integration for Goland IDE. This dialog has an option to turn off the proxy or go direct to repo bypassing the proxy. For more details on the Go proxy, check out goproxy.io and the Go docs for more details.

01:09Instead of TDD, first I cover a simple function implementation. Note the casing is very important in setting up a test in Go. The test function needs to be public, thus capitalized, and the function needs accessible, thus also capitalized. The function in the awesomeLib.go file, just to have a function that would return some value, looks like this.

package awesomeLib

func GetKnownResult() string {
    return "known"
}

01:30Creating the test file, awesomeLib_test.go, then creating the test. In this I also show some of the features of the Goland IDE that extracts and offers the function names of a module prefaced with Test and other naming that provides they perform a test, performance, or related testing functionality.

package awesomeLib

import "testing"

func TestGetKnownResult(t *testing.T) {
    got := GetKnownResult()
    if got != "known" {
        t.Error("Known result not received, test failed.")
    }
}

02:26Running the standard go test to execute the test. To run tests for all files within this module, such as if we’ve added multiple directories and other files, would be go test ./....

02:36Using Goland to run the tests with singular or multiple tests being executed. With Goland there are other capabilities to show test coverage, what percentage of functionality is covered by tests, and other various code metrics around testing, performance, and other telemetry.

That’s it, now the project is ready for elaboration and is setup for a TDD, BDD, or implement and test style approach to development.

For JavaScript, Go, Python, Terraform, and more infrastructure, web dev, and coding in general I stream regularly on Twitch at https://twitch.tv/adronhall, post the VOD’s to YouTube along with entirely new tech and metal content at https://youtube.com/c/ThrashingCode.

5 Ways to Learn & Improve Your Programming Language Use

Posed the question,

What to write when starting to learn and programming language?

I started an answer to the question in this video. But read on for more extensive list and details!

I’ve written up this article. It started last week when I wrote up “Top 5 Ways to get the coding basics down!” This week I wanted to delve deeper into a specific item on that list and provide what I came up with, which is this List of 10 things to write in order to effectively learn a programming language.

1. Find a Known Domain

What I mean by this is to find a domain that you personally know. It seemed the 80’s generation of programmer dorks wrote role playing game character generators for Dungeons & Dragons. I did myself but for Robotech. But there’s a million other domains that you would know, and those domains are what you should pick and find data and actions in that domain you could implement with code.

Two examples I’ve used recently includes the made up e-commerce store of Better Botz, a store that works with the domain of robots and mecha, and Railroads such as Union Pacific, Norfolk Southern, Amtrak, and the respective domain around that. Many of the other things you could use might be transportation, automotive, health care, or more specific the universe of Harry Potter, Star Wars, Tumble Leaf, Sesame Street, style and fashion trends,

2. Model a Domain According to the Language

Let’s say you’re learning JavaScript. Whatever the domain you’ll want to find actions and data to implement and use specificly to how the language is best used. With JavaScript that would be to use the language features and design to build a working application, like; perform mathematical calculations, such as with bigint, using dynamic import, chaining, promises, async and await, global variables, arrow functions, inheritence through prototyping, and how the properties and methods, or functions, for starters work best for the application domain.

In C#, Java, Rust, Go, Erlang, F#, and other languages these features and characteristics would be different. Whatever the language, make a list of the features and characteristics that are widely used and work with those to best implement the application domain.

List Interupt: Refactoring

I’ve written this list in order of what I personally have found the easiest path in gaining a high level of profiecience with a programming language. You could however mix the list to your preferred style. However at this point between the first two item efforts and the next a new skill will be learned both out of necessity and intent: refactoring.

Code Refactoring is the process of restructuring existing computer code—changing the factoring—without changing its external behavior. Refactoring is intended to improve the design, structure, and/or implementation of the software (its non-functional attributes), while preserving the functionality of the software. Potential advantages of refactoring may include improved code readability and reduced complexity; these can improve source code maintainability and create a simpler, cleaner, or more expressive internal architecture or object model to improve extensibility.”

3. Learn to Test

Learning to test an application will help you tremendously in writing better code. If you’ve committed the actions of the first two items in this list, this is an opportunity to significantly refactor the application you’ve built around your known domain. You’ll need to break apart existing functions so that they’re bit size enough to test. Sometimes you’ll need to differentiate between making some testable at a small unit size versus across an application’s hard boundaries, like an integration test between a database and the application’s data access code.

Pushing for testing the core domain functionality of the application is the best thing to focus on. Don’t worry about the specifics of testing if data is written to the database, or data is retrieved from the database, as much as testing that a function returns the specific answer you expected based on passing it the specific parameters you’ve assumed it would be given. This isn’t to say sometimes you may want to test what you’ve written to the database, so sure, that is something that you can test. But focusing on the core application domain functionality will ensure the best return on your investment of time!

4. Clean Coding Practices

Once you’ve gained familiarity with the language by building an application or two using the features, rooted in a known domain that you understand well, and have some test coverage of the application, refactor that application using clean coding practices. Clean code is often described using the SOLID principles:

  • Single Responsibility Principle
  • Open / Closed Principle
  • Liskov Substitution Principle
  • Interface Segregation Principle
  • Dependency Inversion Principle

See also: KISS (Keep It Simple Stupid), DRY (Don’t Repeat Yourself), YAGNI (You Aint Gonna Need It), and others.

The idea behind these, is to keep the application code regardless of languages as simple as it can possibly be. With a secondary effect that once software becomes painful to modify, it should be changed as to make modification simpler. Thus the idea of refactoring to the SOLID principles listed, and doing so continuously will keep software in a usable, maintainable, and one could arguable say even a performant state.

5. The Hardest Suggestion: Naming & Thinking

It is often stated, “The hardest problem in computer science is naming things.” to which I, not in jest, whole heartedly agree. Elements in programming, programming design, programming implementation, and every permeable space around programming is littered with poorly named things. I don’t blame anybody because it is damned hard to get a good name for features and functionality, and have truth in the naming of those features and functionality remain throughout the course of its lifetime.

Naming is hard, which I’d combine to say stop programming regularly and think about your naming. Think about refactoring to names that are more accurate to the intent. Stop and ponder what a good name would be, and if the function or feature will replication what that name implies. Add in some metaphysical thought to the whole effort, talk it through with someone you know (or a stranger, it’s a good skill for debuggin!), but think, think, ponder, analyze, and think some more.

The short of the long part of this suggetion can simply be summarized as: Practice naming things, think a bunch, then rename things, and repeat this cycle.

In Summary

To really get a programming language down by heart, be able to readily implement pretty much anything, this is my go to list of actions and practices to follow. Hopefully, it’ll provide you with some ideas to explore how best you would also like to go about learning a programming language. I’m always open to suggestions, comments, thoughts, and discussions on the topic, so feel free to ping me via the Twitters @Adron.

Search Words

Type in any of the following, in pretty much any search engine, like duckduckgo to dig up the latest and greatest on this topic.

  • Single Responsibility Principle
  • Open / Closed Principle
  • Liskov Substitution Principle
  • Interface Segregation Principle
  • Dependency Inversion Principle
  • How to Model a Domain
  • Clean Coding Practices
  • Unit Testing
  • Unit Testing Practices
  • SOLID Principles
  • Package Principles (REP, CRP, CCP, ADP, SDP, SAP)
  • Don’t Repeat Yourself (DRY)
  • Keep It Simple Stupid (KISS)
  • You Aint Gonna Need It (YAGNI)

Finally Solidified, Twitch Scheduling

twitch-logo-sizedIt’s been a year plus now that I’ve been streaming on Twitch (FOLLOW on the channel, it’s free not a paid subscription or anything). At points it has been weekly, sometimes every other week, sometimes every day of the week, or in some cases even more intermittently or frequently. The schedule, considering, has been kind of ridiculous. But that has now changed.

Continue reading “Finally Solidified, Twitch Scheduling”