find, locate, and xargs#

Concepts#

`find` — Search for Files by Criteria#

find walks the directory tree and tests each file against your criteria. It is powerful, flexible, and available everywhere.

find [starting-path] [tests] [actions]

Search by Name#

# Find by exact name
find /etc -name "hosts"

# Case-insensitive
find /home -iname "readme*"

# Wildcards (must be quoted to prevent shell expansion)
find . -name "*.log"
find . -name "*.txt" -o -name "*.md"     # -o = OR

Search by Type#

find /var -type f          # regular files
find /var -type d          # directories
find /var -type l          # symbolic links

Search by Size#

find /var/log -size +10M        # larger than 10 MB
find . -size -1k                # smaller than 1 KB
find . -size +100M -size -1G    # between 100 MB and 1 GB
# Units: c=bytes, k=KB, M=MB, G=GB

Search by Time#

# Modified time (days)
find /tmp -mtime -1       # modified in the last 24 hours
find /tmp -mtime +30      # modified more than 30 days ago

# Accessed time
find . -atime -7          # accessed in the last 7 days

# Modified time (minutes)
find . -mmin -60          # modified in the last 60 minutes

# Newer than a file
find . -newer reference.txt

Search by Permissions and Ownership#

# Exact permission
find /usr -perm 755

# At least these permissions set
find . -perm -644

# Owned by user
find / -user kmiguel 2>/dev/null

# Owned by group
find / -group www-data 2>/dev/null

# Files with no owner (orphaned)
find / -nouser 2>/dev/null

Search by Depth#

find . -maxdepth 1 -name "*.txt"    # current directory only
find . -maxdepth 2 -type d          # at most 2 levels deep
find . -mindepth 2 -name "*.conf"   # skip the first level

Combining Tests#

# AND (default) — both must be true
find . -name "*.log" -size +1M

# OR
find . -name "*.jpg" -o -name "*.png"

# NOT
find . ! -name "*.tmp"
find . -not -user root

# Grouping with parentheses (must be escaped)
find . \( -name "*.log" -o -name "*.tmp" \) -mtime +30

Actions#

# Default action: -print (show the path)
find . -name "*.txt"

# Delete matching files
find /tmp -name "*.tmp" -delete

# Execute a command on each result
find . -name "*.log" -exec ls -lh {} \;
# {} is replaced by the filename. \; ends the command.

# Execute with confirmation
find . -name "*.bak" -ok rm {} \;

# More efficient: pass multiple files at once (like xargs)
find . -name "*.txt" -exec grep -l "TODO" {} +
# {} + passes as many files as possible in one command invocation

`locate` — Fast Filename Search#

locate searches a pre-built database, making it much faster than find — but the database may be stale.

# Install (not always pre-installed)
sudo apt install -y plocate     # Ubuntu 22+/Debian 12+
# (older systems: sudo apt install mlocate)

# Update the database
sudo updatedb

# Search
locate hosts
locate -i readme          # case-insensitive
locate -c "*.conf"        # count matches
locate -r '\.py$'         # regex
locate --existing         # only show files that still exist

The database is usually updated automatically once a day via cron. After creating new files, run sudo updatedb to include them.

`find` vs `locate`#

	`find`	`locate`
Speed	Slower (walks filesystem)	Very fast (searches database)
Freshness	Always current	May be stale
Criteria	Name, size, time, permissions, etc.	Name/path only
Actions	Can delete, exec commands	Search only

Use locate to quickly find files by name. Use find when you need criteria beyond the name or guaranteed current results.

`xargs` — Build Commands from stdin#

xargs reads items from stdin and executes a command with those items as arguments. It is often used with find and grep.

# Basic usage: pass stdin as arguments
echo "file1.txt file2.txt file3.txt" | xargs rm

# With find (alternative to -exec)
find . -name "*.log" | xargs ls -lh

# Handle filenames with spaces (null-delimited)
find . -name "*.txt" -print0 | xargs -0 rm
# -print0: find outputs null-separated names
# -0: xargs reads null-separated input

Controlling Argument Placement#

# -I {} — replace {} with each input item (one at a time)
find . -name "*.conf" | xargs -I {} cp {} /backup/

# Process one item at a time
echo "a b c" | xargs -n 1 echo
# a
# b
# c

# Process two items at a time
echo "a b c d" | xargs -n 2 echo
# a b
# c d

Parallel Execution#

# Run up to 4 processes in parallel
find . -name "*.gz" -print0 | xargs -0 -P 4 gunzip

# Combine with -n for parallel, one-item-at-a-time processing
cat urls.txt | xargs -n 1 -P 8 curl -O

Practical Examples#

# Delete all .tmp files (safe with spaces in names)
find /tmp -name "*.tmp" -print0 | xargs -0 rm -f

# Grep across all Python files
find . -name "*.py" -print0 | xargs -0 grep -l "import os"

# Count lines in all shell scripts
find . -name "*.sh" -print0 | xargs -0 wc -l

# Change permissions on all directories
find . -type d -print0 | xargs -0 chmod 755

# Change permissions on all files
find . -type f -print0 | xargs -0 chmod 644

# Show sizes of all found files
find /var/log -name "*.log" -print0 | xargs -0 du -sh

# Rename files (using -I for placeholder)
ls *.txt | xargs -I {} mv {} {}.bak

Why Use `xargs` Instead of `find -exec`?#

xargs can run commands in parallel (-P)
xargs batches arguments efficiently (fewer command invocations)
xargs works with any input, not just find

For simple cases, find -exec ... {} + is equivalent and simpler.

Lab#

Exercise 1: find Basics#

# Find all .conf files in /etc
find /etc -name "*.conf" 2>/dev/null | head -10

# Find directories in your home
find ~ -maxdepth 2 -type d | head -20

# Find files modified in the last hour
find /tmp -mmin -60 -type f 2>/dev/null

# Find files larger than 10MB
find /var -size +10M -type f 2>/dev/null

Exercise 2: find with Actions#

# Create test files
mkdir -p /tmp/findlab
touch /tmp/findlab/{a,b,c}.txt /tmp/findlab/{x,y}.log

# Find .txt files and list them in detail
find /tmp/findlab -name "*.txt" -exec ls -l {} \;

# Find .log files and show their names
find /tmp/findlab -name "*.log" -exec basename {} \;

# Delete .log files
find /tmp/findlab -name "*.log" -delete
ls /tmp/findlab

# Clean up
rm -rf /tmp/findlab

Exercise 3: locate#

# Update the database
sudo updatedb

# Find all files named "bashrc"
locate bashrc

# Find all .service files (systemd units)
locate -c ".service"

# Case-insensitive search
locate -i readme | head -10

Exercise 4: xargs#

# Create test files
mkdir -p /tmp/xargslab
for i in {1..5}; do echo "line $i" > /tmp/xargslab/file$i.txt; done

# Count lines in all files using xargs
find /tmp/xargslab -name "*.txt" | xargs wc -l

# Safe version with spaces
find /tmp/xargslab -name "*.txt" -print0 | xargs -0 wc -l

# Use xargs to create backup copies
find /tmp/xargslab -name "*.txt" | xargs -I {} cp {} {}.bak
ls /tmp/xargslab

# Clean up
rm -rf /tmp/xargslab

Review#

1. How do you find all .log files larger than 10MB?

find / -name "*.log" -size +10M. Combine -name for the pattern and -size +10M for the size filter.

2. What is the difference between find and locate?

find walks the filesystem in real-time — always current but slower. locate searches a pre-built database — very fast but may be stale. locate can only search by name/path; find supports size, time, permissions, and more.

3. How do you handle filenames with spaces when using find and xargs?

Use find ... -print0 | xargs -0 .... The -print0 flag outputs null-delimited filenames, and -0 tells xargs to read null-delimited input.

4. What does find's -exec do?

Runs a command on each matching file. {} is replaced by the filename, and the command ends with \; (one invocation per file) or + (batch multiple files into one invocation).

5. How do you run xargs in parallel?

Use the -P flag: xargs -P 4 runs up to 4 processes simultaneously. Combine with -n 1 to process one item per invocation.

Previous: Shell Customization | Next: Scheduling with cron and at

find, locate, and xargs