Mastering Git Internals: What Happens When You Commit
An exploration into how Git handles files under the hood, breaking down blobs, trees, and commits.
မြန်မာဘာသာဖြင့် ဖတ်ရန်Mastering Git Internals
Today I decided to look past the basic git add and git commit surface commands to understand how Git actually tracks our code changes under the hood.
Git isn't just a simple file tracker—it's a content-addressable filesystem.
The Three Core Objects
When you commit a file, Git generates compressed snapshots and saves them inside the hidden .git/objects directory using SHA-1 hashes. There are three essential object types:
- Blobs: Stored compressed binary data representing raw file contents. A blob doesn't store file names or directory structures—just the file data.
- Trees: The equivalent of folders. A tree object lists file names, permissions, and links them to their corresponding file blobs or other nested trees.
- Commits: A tiny text file containing a link to the top-level tree snapshot, author credentials, timestamps, and a pointer to the parent commit hash. မြန်မာလိုရောအဆင်ပြေလား
Simulating Object Generation
I ran some low-level Git plumbing commands in my terminal to see exactly how Git hashes content.
# Hash a piece of text and write it directly into the Git object store
echo "Hello, Git Internals!" | git hash-object -w --stdin