Searches

find

find [<path>] [-maxdepth n] [-type [dfl]] [-name <file-name>, –regex “<regex>”]
     [-mmin -<max-minutes>] [-mtime -<max-days>]
     [-exec <command> [options] {} \;] [-print,-print0]

Options:
    <path>    : Path to search from (default: . ) (required on Solaris)
    -maxdepth : Maximum depth to search (default: no limit)
    -type  (by default, searches for all types)
        d: directory
        f: file
        l: link
    -name   : File name pattern (default: * )
    -regex  : Regular-expression file name pattern

    -mmin   : Files changed within the last n minutes
    -mtime  : Files changed within the last n days

    -exec   : Run the specified command for each file that's found.
              "{}" is replaced with the name of the current file.
              Only a single command can be run.
    -print  : Print the file name followed by a newline (this option is used by default)
    -print0 : Print the file name followed by a null character
              (for supporting file names with blanks spaces or newlines)

• Locates a file

# Search “data” directories for .txt files containing the text “hello”
find -type d -name data -exec find {} -type f -name "*.txt" \; | xargs -l1 grep -li 'hello'

# Replace "hi" with "bye" for all .html and .txt files in "data" directories
for ext in html txt; do
  find -type d -name data -exec \
    find {} -type f -name "*.$ext" \; | xargs -l1 \
      perl -pi -e "s|/hi/bye|ig;"
done

# The following lines are equivalent (recursively list all files in the current directory)
find -type f -name "*"
find . -type f
find -type f

-regex

• Has to match the entire relative-path file name, starting with ./
• Regex characters that have to be escaped: ( ) |
  
• Regex characters not supported: { }

-exec

# Search .js files
find -name "*.js" -exec grep -il "text" {} \;
 
# Search a subset of .js files (excluding *_data.js)
find \( -name "*.js" -a ! -name "*_data.js" \) -exec grep -il "text" {} \;
 
# Search .php and .js files
find \( -name "*.php" -o -name "*.js" \) -exec grep -il "text" {} \;
 
# Search .php and a subset of .js files
find \( -name "*.php" -o \( -name "*.js" -a ! -name "*_data.js" \) \) -exec grep -il "text" {} \;

• Only one command can be run for each file
  
• {} can be used more than once
  

for ... in

# Search .js files
for i in `find -name "*.js"`; do grep -il "text" $i; done
 
# Search .php and .js files
for i in `find \( -name "*.php" -o -name "*.js" \)`; do grep -il "text" $i; done

• Using find in combination with a for loop allows multiple commands to be run for each file

vi $(...)

# Edit a set of .php files
vi $(find -name "*.php" -exec grep -il "text" {} \;)

grep

grep [-<options>] <pattern> <file-name>

Options:
    l: Print the file name (and nothing else) for each match
    L: Print the file name (and nothing else) for each unsuccessful match
    H: Print the file name (and the full line of text) for each match
    n: Print the line number for each match

    i: Case insensitive
    w: Match whole words only
    E: Extended regular expression
    P: Perl regular expression

    R, r: Recursive
    I: Skip binary files
    z: Treat the entire file as a single line

• Searches for a string within a file
• Matches patterns on a line-by-line basis
  
• Prints the entire line for any matches; no obvious way to only display the matching portion of a line
  

# recursively search for "text" in all files
grep -r "text" *

# Identify where paths are added to the PATH variable
cd /etc
find -type f -exec grep -q PATH {} \; -print

# For each file containing the text "hello"
grep -lr 'hello' * | xargs -L1 ls

# For each *.txt file containing the text "hello"
find -type f -name "*.txt" -exec grep -l 'hello' {} \; | xargs -L1 ls