This content originally appeared on Level Up Coding - Medium and was authored by Adam Berg
In the first part of this series, we created a basic rust CLI program that allows us to create notes and save them in an sqlite database. If you have not already read that one, you should start there as this builds off of where that post left off.
This next part will cover the rest of CRUD: reading, updating, and deleting.
The Rust application described in this series has landed in the open-source repo for engram. If you would like to follow development more closely, please star on Github so you can remember to check back in.
Read
The next logical step is adding the ability to read whatever has just been created. In a command line application, the options here are much more limited which allows us to skip discussing all kinds of Graphical User Interface (GUI) related topics. This doesn’t necessarily make it simple though.
There are some considerations for how this might work in our application. Namely how do we want to trigger and display historical notes to the user.
Triggers or Commands
The notes application will continue running after started until an empty string is submitted. In order to query for some kind of information, we need to invent some functionality that allows the user to differentiate between a simple note and a more advanced command.
To accomplish this, we will implement “slash commands”. These have been popularized by the likes of Slack and Notion and should work well for what we need. If a note is started with a “/”, the text that follows will be interpreted as a command. More spcecifically for the reading of notes, we will add a “/list” command. To keep things simple, it will dump all notes that are in the database. We will follow up on this in a later post to improve how this works. Every new piece of functionality has many more considerations than you first realize, so delaying certain things prevents you from getting hung up on details that don’t matter yet.
/list
let mut running = true;
while running == true {
let mut buffer = String::new();
io::stdin().read_line(&mut buffer)?;
let trimmed_body = buffer.trim();
if trimmed_body == "" {
running = false;
}
else if trimmed_body == "/list" {
let mut stmt = conn.prepare("SELECT id, body from notes")?;
let mut rows = stmt.query(rusqlite::params![])?;
while let Some(row) = rows.next()? {
let id: i32 = row.get(0)?;
let body: String = row.get(1)?;
println!("{} {}", id, body.to_string());
}
}
else {
conn.execute("INSERT INTO notes (body) values (?1)", [trimmed_body])?;
}
}
Whenever a note is submitted, we check to see if “/list” was entered. If it was, we proceed into that else if block to print out existing notes.
let mut stmt = conn.prepare("SELECT id, body from notes")?;
This prepares an SQL statement using the connection we initialized at the start of our main function. SELECT id, body from notes specifies that we want to return the id and body columns from the notes table.
let mut rows = stmt.query(rusqlite::params![])?;
We then issue the query method on the prepared statement. Since we are selecting all notes there are no params which is why we pass empty params using: rusqlite::params![].
while let Some(row) = rows.next()? {
let id: i32 = row.get(0)?;
let body: String = row.get(1)?;
println!("{} {}", id, body.to_string());
}The above code iterated through all of the returned rows. Each row will be a note that has previously been entered in to our database. row.get(0) and row.get(1) fetch the value of the associated column index. In our case, our query specified: SELECT id, body from notes which means id will be at index 0 and body will be at index 1.
Rust cannot infer the types of these properties, which is why they must be specified as let id: i32 = row.get(0)?; and let body: String = row.get(1)?;. With i32 identifying that id is a 32 bit integer and String identifying that body is a string.
Finally we pass both of these along to the println function in order to output them back to the terminal.
Once again, you can now run the app with cargo run and if you issue a /list command, you should now see all of the notes you have submitted printed back at you.
Delete
I almost always implement deleting before updating. It’s a simple operation which makes it very quick to implement. It can also act as a makeshift way to edit items. In our notes example, if I want to edit a note, I could first delete it and then create a new one with the correct body. As mentioned above, this is less relevant in a small example like this, but in a larger application, the edit GUI can be surprisingly complex.
...
let trimmed_body = buffer.trim();
let cmd_split = trimmed_body.split_once(" ");
let mut cmd = trimmed_body;
let mut msg = "";
if cmd_split != None {
cmd = cmd_split.unwrap().0;
msg = cmd_split.unwrap().1;
}
if cmd == "/del" {
let id = msg;
conn.execute("delete from notes where id = (?1)", [id])?;
}
...The /del command has something the /list command doesn’t - an additional parameter. We need to specify which note we want to delete. After thinking about it for a moment, I decided I’d like to pass /del 1 to delete the note with id 1.
In order to differentiate between the “command” and the “parameter”, I decided to use the split_once method.
let cmd_split = trimmed_body.split_once(" ");
The split_once method splits a string based on the passed delimiter. In our example “/del 1” would return as Some(("/del", "1")). I then proceed to unwrap those values and store them in cmd and msg variables.
if cmd_split != None {
This equality check covers the case where there are no spaces. In this condition the split_once method returns None to identify that the “ ” delimiter doesn’t exist.
I’m still new to Rust and found this a bit clunky. I suspect there’s probably a better way to write this, but for now it does the job. I’ve learned many times to not get hung up on small details as this one alone could lead to a 30 minute rabbit hole down Rust documentation. If you have any recommendations, feel free to share them in the comments section.
if cmd == "/del" {
let id = msg;
conn.execute("delete from notes where id = (?1)", [id])?;
}We now check that the first part of the entered text was /del and if so, we know that we can get the id to delete from the msg variable.
"delete from notes where id = (?1)", [id]
This is the SQL command to delete a row from the notes table that matches the id specified.
Once again, you can run cargo run and now try /del 1 which should delete the first message you created. You can confirm it worked by running /list and you should no longer see a note with index 1.
Update
There are a few options for how the update can work. To continue to keep things simple, I decided that the edit should be issued all as one command. /edit 1 the new body I want to have. Similar to the delete, an id is passed to identify which note to edit. Everything following the id is then treated as the new body to overwrite the existing one.
else if cmd == "/edit" {
let msg_split = msg.split_once(" ").unwrap();
let id = msg_split.0;
let body = msg_split.1; conn.execute("update notes set body = (?1) where id = (?2)", [body, id])?;
}The /edit command starts similar to the /del. The main difference is we need to again split the msg by a white space. Using the split_once only splits on the first white space, which allows the body to remain intact.
"update notes set body = (?1) where id = (?2)", [body, id]
This update command specifies that we will set the body column to what we’ve parsed from the input on any rows that have an id that matches the id specified.
(?1) and (?2)
These denote positional parameters. All of our previous SQL statements had only one, but in this case there are two. (?1) will be replaced by the first entry in the supplied parameters [body, id] (or body in this case). (?2) will be replaced by the id variable.
Fire this up one more time and try editing one of your existing notes. You can see previous ones with the /list command, then issue and /edit command, and finally issue another /list command to confirm the note was correctly modified.
Installing Your Notes Application
cargo install --path .
Running the above will compile the rust applciation and add it to your system path. If you used the cargo new notes command at the start, you should now have access to the notes command from your terminal. If you would like to update the name of your executable, you can modify the name property in the Cargo.toml file to match whatever you prefer.
I have been working on a notes application for a while called engram and to keep my executable name short, I’ve shorted this to eg. Now anytime I am in the terminal I can type eg and have immediate access to my notes.
Wrap-up
If you followed along through this whole tutorial, you should now have a functional notes app written in Rust accessible from a terminal. I will continue to write posts that build upon this foundation, but they will likely be about more specific topics as I build new functionality for engram’s CLI application.
How to Build a Note Taking Command Line Application With Rust: Part 2 was originally published in Level Up Coding on Medium, where people are continuing the conversation by highlighting and responding to this story.
This content originally appeared on Level Up Coding - Medium and was authored by Adam Berg
Adam Berg | Sciencx (2021-12-10T05:54:25+00:00) How to Build a Note Taking Command Line Application With Rust: Part 2. Retrieved from https://www.scien.cx/2021/12/10/how-to-build-a-note-taking-command-line-application-with-rust-part-2/
Please log in to upload a file.
There are no updates yet.
Click the Upload button above to add an update.