Using comm as an Alternative to diff
Today I discovered the comm command—an alternative way to compare files. Unlike diff, comm is particularly useful for working with sorted data.
The Problem
Every so often, I need to compare two sets of data, such as information from Service A vs B. My goal is to quickly identify what data exists in one but not the other.
What I Used to Do
I used to rely on diff with a set of specific flags:
diff --new-line-format="%L" --unchanged-line-format="" a.sorted b.sorted > missing.txt
Explanation:
--new-line-format="%L"→ Prints lines unique to the first file.--unchanged-line-format=""→ Hides common lines.
A Simpler Alternative: comm
With comm, I can achieve the same result more concisely:
comm -23 a.sorted b.sorted > missing.txt
Explanation:
-3→ Suppresses the third column (lines common to both files), leaving only unique entries from each file.
comm vs diff
While both comm and diff compare files line by line, they serve different purposes:
| Feature | comm | diff |
|---|---|---|
| Comparison Basis | Works on sorted files | Works on any files |
| Sorting Required? | Yes | No |
| Output Format | Three-column output (unique & common) | Contextual differences |
| Suppressing Output | Can hide columns (-1, -2, -3) | No direct way to suppress lines |
| Finding Unique Lines | comm -23 file1 file2 | diff --new-line-format="%L" --unchanged-line-format="" file1 file2 |
| Finding Common Lines | comm -12 file1 file2 | diff file1 file2 | grep '^ ' |
| Best Used For | Comparing sorted lists and extracting unique/common lines | Identifying line-by-line differences |
My Takeaways
For this specific task, I now prefer comm because it’s simpler and more intuitive. A key advantage is that it forces me to remember that files must be sorted—something I often forget when using diff, leading to unexpected results.