find & sed

The find and sed commands are powerful tools in Linux for searching and manipulating text.

Here’s an overview related to how I used them both to rename specific words in multiple files. Basically, I wanted to change all instances of “loose” to “relaxed” in text files within a project.

find + sed Pattern

find [path] [filters] -exec sed -i '' 's/old/new/g' {} +

Components

find: Locate files matching criteria

• -name "*.ext": Filter by filename pattern
• -type f: Only regular files (not directories)
• -exec cmd {} +: Run command on all found files

sed: Stream editor for text replacement

• -i '': Edit in-place (macOS syntax)
  • On Linux, use -i without quotes
• s/pattern/replacement/g: substitute globally
• \b: Word boundary (prevents partial matches)

{}: Placeholder for filenames found by find +: Batch multiple files (efficient) vs ; (one at a time)

Patterns

Word Boundaries

# WITHOUT \b: "loosely" → "relaxedly" (wrong!)
sed 's/loose/relaxed/g'

# WITH \b: Only "loose" → "relaxed" (correct!)
sed 's/\bloose\b/relaxed/g'

Case Sensitivity

# Separate patterns for different cases
sed 's/\bLoose mode\b/Relaxed mode/g'  # Capital L
sed 's/\bloose mode\b/relaxed mode/g'  # Lowercase l

Delimiters

# When pattern contains /, use different delimiter
sed 's|path/to/loose|path/to/relaxed|g'  # Using | instead of /