tokei

tokei é um programa extremamente rápido para contar linhas de código, comentários e linhas em branco. Escrito em Rust, suporta mais de 200 linguagens de programação, fornece múltiplos formatos de saída e respeita padrões gitignore.

Instalação

# Using cargo (Rust)
cargo install tokei

# macOS
brew install tokei

# Ubuntu/Debian (via snap)
sudo snap install tokei

# Arch Linux
sudo pacman -S tokei

# Nix
nix-env -i tokei

# Download prebuilt binary
curl -sSfL https://github.com/XAMPPRocky/tokei/releases/latest/download/tokei-x86_64-unknown-linux-gnu.tar.gz | \
    tar xz -C /usr/local/bin/

# Verify installation
tokei --version

Uso Básico

# Count lines in current directory
tokei

# Count lines in a specific directory
tokei /path/to/project

# Count lines in multiple directories
tokei src/ tests/ lib/

# Count a specific file
tokei main.rs

Reading Saída

# Example tokei output:
# ===============================================================================
#  Language            Files        Lines         Code     Comments       Blanks
# ===============================================================================
#  Rust                   45        12345         9876          987         1482
#  TOML                    3          120           95           10           15
#  Markdown                5          450            0          350          100
#  YAML                    2           80           65            5           10
# ===============================================================================
#  Total                  55        12995        10036         1352         1607
# ===============================================================================

# Columns:
# Files    - Number of files for that language
# Lines    - Total lines (Code + Comments + Blanks)
# Code     - Lines of actual code
# Comments - Comment lines
# Blanks   - Empty lines

Language Breakdown

# Show detailed file-level breakdown
tokei --files

# Output includes per-file statistics:
# ===============================================================================
#  Language            Files        Lines         Code     Comments       Blanks
# ===============================================================================
#  Rust                   45        12345         9876          987         1482
# -------------------------------------------------------------------------------
#  src/main.rs                        250          200           20           30
#  src/lib.rs                         180          150           10           20
#  ...

# Show only specific languages
tokei -t Rust,Python,JavaScript

# List all supported languages
tokei --languages

Exclude Patterns

# Exclude specific directories
tokei -e node_modules -e target -e .git

# Exclude multiple patterns
tokei -e "*.min.js" -e vendor -e dist

# Exclude test files
tokei -e tests -e "*_test.go" -e "test_*.py"

# tokei respects .gitignore by default
# Files in .gitignore are automatically excluded

Formatos de Saída

JSON Saída

# JSON output for scripting
tokei -o json > stats.json

# Pretty-print JSON
tokei -o json | python3 -m json.tool

# Parse with jq
tokei -o json | python3 -c "
import sys, json
data = json.load(sys.stdin)
for lang, info in sorted(data.items()):
    if isinstance(info, dict) and 'code' in info:
        print(f'{lang}: {info[\"code\"]} lines of code')
"

YAML Saída

# YAML output
tokei -o yaml > stats.yaml

CBOR Saída

# CBOR (Concise Binary Object Representation) output
tokei -o cbor > stats.cbor

Sorting

# Sort by lines of code (default)
tokei -s code

# Sort by number of files
tokei -s files

# Sort by total lines
tokei -s lines

# Sort by comments
tokei -s comments

# Sort by blank lines
tokei -s blanks

.tokeignore

# Create a .tokeignore file (same syntax as .gitignore)
# This excludes paths from tokei's count

# .tokeignore example:
# node_modules/
# dist/
# *.min.js
# *.min.css
# vendor/
# __pycache__/
# *.generated.go
# coverage/

# .tokeignore
node_modules/
dist/
build/
*.min.js
*.min.css
vendor/
__pycache__/
*.generated.*
coverage/
.next/

Badges for READMEs

# Generate a badge showing lines of code
# Use with shields.io or similar badge services

# Get total lines of code
LINES=$(tokei -o json | python3 -c "
import sys, json
data = json.load(sys.stdin)
total = sum(v.get('code', 0) for v in data.values() if isinstance(v, dict))
print(total)
")
echo "Total lines of code: $LINES"

# Markdown badge (using shields.io)
echo "![Lines of Code](https://img.shields.io/badge/lines%20of%20code-${LINES}-blue)"

Comparing Projects

# Compare code statistics between directories
echo "=== Project A ==="
tokei /path/to/project-a
echo ""
echo "=== Project B ==="
tokei /path/to/project-b

# Track growth over time with JSON
tokei -o json > stats-$(date +%Y%m%d).json

# Compare two snapshots
python3 -c "
import json
with open('stats-20260101.json') as f:
    old = json.load(f)
with open('stats-20260601.json') as f:
    new = json.load(f)

for lang in new:
    if isinstance(new[lang], dict) and 'code' in new[lang]:
        old_code = old.get(lang, {}).get('code', 0)
        new_code = new[lang]['code']
        diff = new_code - old_code
        if diff != 0:
            print(f'{lang}: {old_code} -> {new_code} ({diff:+d})')
"

CI Integração

# Use tokei in CI to track code metrics
# GitHub Actions example:
# - name: Count lines of code
#   run: |
#     cargo install tokei
#     tokei -o json > code-stats.json
#     tokei

# Fail if code exceeds a threshold
MAX_LINES=100000
CURRENT=$(tokei -o json | python3 -c "
import sys, json
data = json.load(sys.stdin)
print(sum(v.get('code', 0) for v in data.values() if isinstance(v, dict)))
")
if [ "$CURRENT" -gt "$MAX_LINES" ]; then
    echo "Code exceeds $MAX_LINES lines ($CURRENT)"
    exit 1
fi

Hidden Files and Configuração

# Count hidden files too
tokei --hidden

# Specify number of columns for output
tokei --columns 100

# Show processing speed
tokei --verbose
# Shows how many files were processed and how fast