Previous

Next

Write a function that completes tasks

Completed

The complete_task function is responsible for trying to remove a task from the to-do list, which is stored in a file. The function needs to complete these actions:

• Read the file.
• Collect existing tasks, if there are any.
• Remove the task that's at the indicated position, if there is one.
• Write the updated vector of tasks back into the file.

A first implementation of the complete_task function could look like the following code, but we can already see signs of code duplication, so we'll need to do some refactoring.

use std::io::{Error, ErrorKind, Result, Seek, SeekFrom};  // Include the `Error` type.


pub fn complete_task(journal_path: PathBuf, task_position: usize) -> Result<()> {
    // Open the file.
    let file = OpenOptions::new()
        .read(true)
        .write(true)
        .open(journal_path)?;

    // Consume the file's contents as a vector of tasks.
    let tasks = match serde_json::from_reader(file) {
        Ok(tasks) => tasks,
        Err(e) if e.is_eof() => Vec::new(),
        Err(e) => Err(e)?,
    };

    // Remove the task.
    if task_position == 0 || task_position > tasks.len() {
        return Err(Error::new(ErrorKind::InvalidInput, "Invalid Task ID"));
    }
    tasks.remove(task_position - 1);

    // Rewind and truncate the file.
    file.seek(SeekFrom::Start(0))?;
    file.set_len(0)?;

    // Write the modified task list back into the file.
    serde_json::to_writer(file, &tasks)?;
    Ok(())
}

Before we start writing this function, we can see that the same code for reading the file that we used in the add_task function is required here. It will also be required again when we implement the list_tasks function. This need for duplication indicates that we should refactor our code and encapsulate that behavior in a dedicated function. We can then reuse the code in the logic for all of our three actions.

Refactor the Task collection

We can create a collect_tasks function that handles file parsing:

fn collect_tasks(mut file: &File) -> Result<Vec<Task>> {
    file.seek(SeekFrom::Start(0))?; // Rewind the file before.
    let tasks = match serde_json::from_reader(file) {
        Ok(tasks) => tasks,
        Err(e) if e.is_eof() => Vec::new(),
        Err(e) => Err(e)?,
    };
    file.seek(SeekFrom::Start(0))?; // Rewind the file after.
    Ok(tasks)
}

The function receives a reference to a File and returns a std::io::Result<Vec<Task>>. That means it's expecting an io::Error to happen. As a bonus, it also rewinds the file, before reading its contents and before returning it back to the caller.

We can now refactor the add_task function to use the new function:

pub fn add_task(journal_path: PathBuf, task: Task) -> Result<()> {
    let file = OpenOptions::new()
        .read(true)
        .write(true)
        .create(true)
        .open(journal_path)?;
    let mut tasks = collect_tasks(&file)?;
    tasks.push(task);
    serde_json::to_writer(file, &tasks)?;
    Ok(())
}

After the refactor, add-task looks much better and is easier to read!

The final version of complete_task

We can now finally use the refactored code in the complete_task function:

pub fn complete_task(journal_path: PathBuf, task_position: usize) -> Result<()> {
    // Open the file.
    let file = OpenOptions::new()
        .read(true)
        .write(true)
        .open(journal_path)?;

    // Consume file's contents as a vector of tasks.
    let mut tasks = collect_tasks(&file)?;

    // Try to remove the task.
    if task_position == 0 || task_position > tasks.len() {
        return Err(Error::new(ErrorKind::InvalidInput, "Invalid Task ID"));
    }
    tasks.remove(task_position - 1);

    // Write the modified task list back into the file.
    file.set_len(0)?;
    serde_json::to_writer(file, &tasks)?;
    Ok(())
}

The first and second parts, and some of the fourth part, are doing the same thing we did in our add_task function, with a couple of exceptions:

• We're not creating the journal file. It doesn't exist.
• We're truncating the file before writing to it because we're performing a removal operation. So the file will be smaller than the original. If we ignored this step, the rewound cursor would stop behind the previously written bytes of the file, resulting in a malformed JSON file. When we truncate the file by using the file.set_len(0) operation, we ensure that we're writing the bytes in a blank page.

In the third section, we remove a task from the vector by pointing out its position. If the position isn't valid, we return early with a custom-made io::Error that explains the problem.

Continue