8.8 KiB
8.8 KiB
Update Doc String
Systematically manage multilingual docstrings/comments and maintain high-quality documentation.
Usage
# Run with automatic language detection
"Please add docstrings to classes and functions without them, and update comments that don't meet standards"
# Run with specified language
/update-doc-string --lang python
"Please update docstrings in Python files to comply with PEP 257"
# Maintain documentation for specific directories
"Please add JSDoc to functions under src/components/"
Options
--lang <language>: Documentation language (default: en)--style <style>: Specify documentation style (has language-specific defaults)--marker <true|false>: Whether to add Claude markers (default: true)
Basic Examples
# 1. Analyze target files (programming language is auto-detected)
find . -type f \( -name "*.py" -o -name "*.js" -o -name "*.ts" -o -name "*.dart" -o -name "*.go" -o -name "*.rs" \) | grep -v test
"Please identify elements with insufficient docstrings (0 comment lines or fewer than 30 characters)"
# 2. Add documentation (uses English by default)
"Please add docstrings containing language-specific required elements to the identified elements"
# → Uses English for all documentation
# 3. Add documentation (explicitly specify English)
/update-doc-string --lang en
"Add docstrings with required elements to the identified elements"
# 4. Check markers
"Please confirm that all added/updated docstrings have Claude markers"
Execution Steps
1. Priority of Target Elements
- 🔴 Highest Priority: Elements without docstrings/comments (0 comment lines)
- 🟡 Next Priority: Elements not meeting standards (fewer than 30 characters or missing required elements)
- 🟢 Verification Target: Existing comments without Claude markers
Target Elements (Common Across Languages):
- Class definitions
- Functions/methods
- Modules (Python, Go)
- Enums
- Interfaces (TypeScript, Go)
2. Language-Specific Documentation Rules
Python (PEP 257):
# English version (default)
def calculate_total(items: List[Item]) -> float:
"""Calculate the total amount for a list of items. (30-60 characters)
Multiplies the price and quantity of each item and returns
the total with tax. Returns 0.0 for empty lists. (50-200 characters)
Args:
items: List of items to calculate
Returns:
Total amount with tax
Generated by Claude 🤖
"""
# English version (--lang en)
def calculate_total(items: List[Item]) -> float:
"""Calculate the total amount for a list of items. (30-60 chars)
Multiplies the price and quantity of each item and returns
the total with tax. Returns 0.0 for empty lists. (50-200 chars)
Args:
items: List of items to calculate
Returns:
Total amount with tax
Generated by Claude 🤖
"""
JavaScript/TypeScript (JSDoc):
/**
* Component that displays a user profile. (30-60 characters)
*
* Displays avatar image, username, and status information,
* and navigates to the profile detail screen when clicked. (50-200 characters)
*
* @param {Object} props - Component properties
* @param {string} props.userId - User ID
* @param {boolean} [props.showStatus=true] - Status display flag
* @returns {JSX.Element} Rendered component
*
* @generated by Claude 🤖
*/
const UserProfile = ({ userId, showStatus = true }) => {
Go:
// CalculateTotal calculates the total amount for a list of items. (30-60 characters)
//
// Multiplies the price and quantity of each item and returns
// the total with tax. Returns 0.0 for empty slices. (50-200 characters)
//
// Generated by Claude 🤖
func CalculateTotal(items []Item) float64 {
Rust:
/// Calculate the total amount for a list of items. (30-60 characters)
///
/// Multiplies the price and quantity of each item and returns
/// the total with tax. Returns 0.0 for empty vectors. (50-200 characters)
///
/// Generated by Claude 🤖
pub fn calculate_total(items: &[Item]) -> f64 {
Dart (DartDoc):
/// Widget that displays a user profile. (30-60 characters)
///
/// Vertically arranges avatar image, username, and status information,
/// and navigates to the profile detail screen when tapped. (50-200 characters)
///
/// Generated by Claude 🤖
class UserProfileWidget extends StatelessWidget {
3. Existing Content Retention Rules
- If existing comments meet standards: Keep as-is (do not add new ones)
- Standards: At least 30 characters and includes required elements (summary, details, marker)
- If existing comments do not meet standards: Completely replace (no duplication)
- If no existing comments: Add new comments
Important Information to Retain:
- URLs and links:
See also:,@see,References:etc. - TODO comments:
TODO:,FIXME:,XXX:format - Notes:
Note:,Warning:,Important:etc. - Examples:
Example:,Usage:,# Examplesetc. - Existing parameter and return value descriptions
Language-Specific Settings
# Language-specific default settings
languages:
python:
style: "google" # google, numpy, sphinx
indent: 4
quotes: '"""
javascript:
style: "jsdoc"
indent: 2
prefix: "/**"
suffix: "*/"
typescript:
style: "tsdoc"
indent: 2
prefix: "/**"
suffix: "*/"
go:
style: "godoc"
indent: 0
prefix: "//"
rust:
style: "rustdoc"
indent: 0
prefix: "///"
dart:
style: "dartdoc"
indent: 0
prefix: "///"
Quality Checklist
- ✅ Character Count: Strictly adhere to 30-60 characters for summary, 50-200 for details
- ✅ Required Elements: Always include summary, detailed description, and Claude marker
- ✅ Completeness: Describe role, usage context, and notes
- ✅ Language Conventions: Comply with official style guides for each language
- ✅ Exceptions: Explain errors and exceptions (when applicable)
- ✅ Accuracy: Analyze implementation and only include fact-based descriptions
Notes
🔴 Strict Prohibitions:
- ❌ Code changes other than documentation comments
- ❌ Speculation about implementation details (only facts)
- ❌ Formats that violate language conventions
- ❌ Deletion or modification of existing type annotations
- ❌ Duplication with existing comments
- ❌ Comments below character count standards in test files
Execution and Verification:
# Record execution results
ADDED_COMMENTS=0
UPDATED_COMMENTS=0
ERRORS=0
# Auto-detect language from existing comments
# If Japanese characters (hiragana, katakana, kanji) are detected, use ja; otherwise, use en
DOC_LANGUAGE="en" # Default
if grep -E -r '[\u3040-\u309F\u30A0-\u30FF\u4E00-\u9FAF]' --include="*.py" --include="*.js" --include="*.ts" --include="*.dart" --include="*.go" --include="*.rs" . 2>/dev/null | head -n 1; then
DOC_LANGUAGE="ja"
fi
# Auto-detect programming language and perform static analysis
if [ -f "*.py" ]; then
pylint --disable=all --enable=missing-docstring .
elif [ -f "*.js" ] || [ -f "*.ts" ]; then
eslint . --rule 'jsdoc/require-jsdoc: error'
elif [ -f "*.go" ]; then
golint ./...
elif [ -f "*.rs" ]; then
cargo doc --no-deps
elif [ -f "*.dart" ]; then
melos analyze
fi
if [ $? -ne 0 ]; then
echo "🔴 Error: Static analysis failed"
exit 1
fi
# Output execution summary
echo "📊 Execution Results:"
echo "- Documentation Language: $DOC_LANGUAGE"
echo "- Added Comments: $ADDED_COMMENTS"
echo "- Updated Comments: $UPDATED_COMMENTS"
echo "- Errors: $ERRORS"
Success Criteria
-
Completion Criteria: Success when all of the following are met:
- Language-specific static analysis PASSED
- Error count is 0
- All added/updated comments meet standards
-
Partial Success: In the following cases:
- Error count is less than 5
- More than 90% meet standards
-
Failure: In the following cases:
- Static analysis FAILED
- Error count is 5 or more
Integration with Claude
# Analyze entire project (auto language detection)
find . -type f \( -name "*.py" -o -name "*.js" -o -name "*.ts" \)
/update-doc-string
"Update this project's docstrings following language-specific best practices"
# → If existing comments contain Japanese, run with ja; otherwise, run with en
# Explicitly run with English documentation
/update-doc-string --lang en
"Update docstrings following language-specific best practices"
# Explicitly run with Japanese documentation
/update-doc-string --lang ja
"Update this project's docstrings following language-specific best practices"
# Run without markers (auto language detection)
/update-doc-string --marker false
"Improve existing docstrings without adding Claude markers"
# English documentation, no markers
/update-doc-string --lang en --marker false
"Improve existing docstrings without adding Claude markers"