sort

Sort a file alphabetically

$ sort [file]

$ sort -r [file]

$ sort -n [file]

$ sort -k2 [file]

$ sort -k1,1 -k2,2n [file]

$ sort -u [file]

$ sort -f [file]

$ sort [file] -o [output_file]

$ sort -h [file]

sort [options] [file...]

sort sorts lines of text files according to specified criteria. By default, it performs lexicographic (dictionary) sorting using the current locale.
Multiple input files are merged and sorted together. Output goes to stdout by default; use -o to write to a file (can be same as input file safely).
The -k option is powerful for structured data. Combined with -t to set the delimiter, it can sort CSV, TSV, and other columnar data by specific fields.
For large files, sort automatically uses temporary files and can utilize multiple CPU cores with --parallel.

-r, --reverse

Reverse the sort order

Compare according to string numerical value

Compare human-readable numbers (2K, 1G)

Fold lowercase to uppercase (case-insensitive)

Output only unique lines

Sort by specified key/column

Use SEP as field separator

Write result to FILE instead of stdout

Check if input is sorted; exit with status

Stabilize sort by disabling last-resort comparison

Merge already sorted files

Ignore leading blanks in keys

Natural sort of version numbers

Use N parallel threads

-k POS1[,POS2]: Sort by field from POS1 to POS2
Format: F[.C][OPTS]
- F = field number (1-based)
- C = character position within field
- OPTS = ordering options (n, r, b, f, etc.)
Examples:
- -k2: Sort by field 2 to end
- -k2,2: Sort by field 2 only
- -k2n: Sort by field 2 numerically
- -k1,1 -k2,2n: Primary alpha, secondary numeric

Locale affects sort order. Use LC_ALL=C for byte-value sorting, which is faster and more predictable for machine processing.
Numeric sort (-n) handles integers and decimals. For version numbers like "1.10" vs "1.9", use -V (version sort).
Memory usage can be significant for large files. Use -S to limit buffer size or -T to specify temp directory.

uniq(1), join(1), comm(1), cut(1), tsort(1)