Lectures

Let’s break down why a path in Go is not “just a string”, how to choose correctly between the path and path/filepath packages, how to build and normalize paths via filepath.Join and filepath.Clean, when to convert relative paths to absolute ones via filepath.Abs, and how to diagnose path issues (the %q format) and errors (wrapping via %w).

We break down file I/O in Go as working with resources: why it’s important to close * os . File via defer, how to place defer correctly after checking err, how os . Open, os . Create, and os . OpenFile differ, and how to use the flags os . O_APPEND, os . O_CREATE, os . O_EXCL. Separately, we discuss the normal io . EOF, common errors, and build a mini example: a “todo in a file”.

We’ll break down when it makes sense in Go to read and write a file “all at once” using os.ReadFile and os.WriteFile: why reading returns []byte, how to explicitly convert bytes to text via string and back, how this approach differs from working through os.Open/ Read, and how to build it into a simple TODO app with the functions loadText, parseTasks, saveText.

We break down why it is important to prepare the directory structure before writing files, how os.MkdirAll works, what the octal modes 0755/ 0644 and the os.FileMode type mean, and which “safe defaults” to choose for public vs private data. At the end: the “directory → file” protocol, error wrapping via fmt.Errorf, and a walkthrough of common mistakes.

We break down why in Go it’s useful to classify file errors by “classes” ( not exist, permission, invalid path), why you can’t reliably rely on err.Error(), and how to correctly check causes via os.IsNotExist(err), os.IsPermission(err), errors.Is/ errors.As and wrapping with %w. At the end: a mini “todo” example and a set of common mistakes.