scc

Count lines in current directory

$ scc

$ scc [path/to/dir]

$ scc --include-ext [py]

$ scc --exclude-dir [vendor,node_modules]

$ scc --sort lines

$ scc --format json

$ scc -w

$ scc --by-file

scc [--include-ext ext] [--exclude-dir dirs] [--format fmt] [options] [paths]

scc (Succinct Code Counter) counts lines of code in many languages. It's extremely fast, using parallelism and optimized parsing.
Output shows files, lines, blanks, comments, and code per language. Wide mode adds complexity estimates based on branching.
Language detection uses file extensions and content inspection. Hundreds of languages are recognized, from common to obscure.
Generated and minified file detection prevents counting boilerplate. Duplicates can be ignored to avoid counting vendored code multiple times.
COCOMO estimation provides rough effort metrics. These are theoretical but useful for comparisons.
Output formats support integration with other tools. JSON and CSV enable processing by scripts and dashboards.

--include-ext EXT

Include only extensions.

Exclude directories.

Sort by field (files, lines, blanks, code, comments).

Output format (default, json, csv, html, etc.).

Show complexity and weighted lines.

Show per-file breakdown.

Ignore duplicate files.

Ignore generated files.

Ignore minified files.

Show COCOMO estimates.

Match filenames.

Max file size.

Complexity estimates are heuristic. Language detection may misidentify some files. Very large codebases may need memory.

scc was created by Ben Boyter around 2018 as a faster alternative to cloc and other line counters. Written in Go, it emphasizes speed through parallelism and efficient algorithms.

cloc(1), tokei(1), loc(1), wc(1)