Unix Programs

When programs conclude they return an exit status.

• 0 is success, else is failure
• positive integers represent an error state, usually 1 is used
• in Bash, $? can be used to access last commands exit status

A helpful resources outside of man pages:

• tldr.sh
  • web client tldr.inbrowser.app

Where Should Programs Go?

• /usr - all system-wide, read-only files installed by (or provided by) the OS
• /usr/local - system-wide, read-only files installed by the local administrator (usually, you). And that's why most directory names from /usr are duplicated here.
• /opt - an atrocity meant for system-wide, read-only and self-contained software. That is, software that does not split their files over bin, lib, share, include like well-behaved software should.
• ~/.local - the per-user counterpart of /usr/local, that is: software installed by (and for) each user. Like /usr, it has its own ~/.local/share, ~/.local/bin, ~/.local/lib.
• ~/.local/opt - the per-user counterpart of /opt

MestreLion on StackOverflow

Useful Programs

• apropos <search>
  • searches man page titles
• cut [-d <delim>] -f <fields to remain> <files>
  • takes ASCII structured file separated by a delimiter like TAB and cut returns the fields indicated
  • cut -d "," -f 1,5-7 datafile.csv
• paste [-d <sep>] <files>
  • steps through multiple files in parallel
  • delimiter is what separates each line from each file e.g.
• tr <options> <string1> <string>
  • converts characters in first string to characters in the second string
  • takes from stdin
  • -c complement the first string
  • -s squeeze character repetitions of string1 characters into a single occurrence ("####" -> "#")
  • tr '[A-Z]' '[a-z]' < file.txt
• comm [<output options>] <file1> <file2>
  • two sorted files are compared, 3 columns are produced
  • col 1 contains only lines unique to file1
  • col 2 contains only lines unique to file2
  • col 3 contains common lines
  • -1 suppress col 1
  • -2 duh
  • -3 duh
• uniq <file>
  • compresses adjacent repeated lines into a single line
  • -c outputs the counts of the number of copies
• sort [options] <file>
  • outputs to stdout
  • -u removes duplicate entries
  • -t <char> used as field separator, TAB is default
  • -k can sort on a particular column of structured file
• basename /path/file.txt
  • strips the directory and file extensions from a path
• grep [options] <pattern string> <file>
  • searches for patterns in one or more files
  • returns matches to stdout
  • returns 0 on successful find, else 1
• test <expression>
  • does conditional checks for a number of common operations
  • some examples
    • -e FileName FileName exists
    • -b Filename Returns a True exit value if the specified FileName exists and is a block special file
    • -c FileName - FileName is a character special file
    • -d FileName FileName is a directory
    • -f FileName FileName is a regular file
    • -g FileName FileName's Set Group ID bit is set
    • -h FileName FileName is a symbolic link
    • -k FileName FileName's sticky bit is set
    • -L FileName FileName is a symbolic link
    • -p FileName FileName is a named pipe (FIFO)
    • -r FileName FileName is readable by the current process
    • -s FileName FileName has a size greater than 0
    • -t FileDescriptor FileDescriptor is open and associated with a terminal
    • -u FileName FileName's Set User ID bit is set
• find [options] <path> [<expression>]
  • recursively searches for file patterns
  • -name <pattern> test file name matches pattern
  • -type <c> type of file is specified by d/f/l
  • -newer <file> test file has been accessed more recently than <file> was modified
  • -print prints the full path name of file
  • -exec <command> execute command on all files found (like JavaScript .map())
    • command terminated with \;
    • {} refers to file found
  • -ok <command> same as command except user is prompted first
  • -perm <mode> finds only files with certain permissions
    • mode can be octal or symbolic
• diff <file1> <file2>
  • shows the differences between to files
• grep [<options>] <regex> <file>
  • -i case insensitive
  • -n prepend numbers of matching lines
  • -v invert match so only non-matching lines reported
• sed
• head and tail
  • display n lines from top or bottom of
  • -n <number|-number> show n lines, if negative show all but final n lines

Interesting Links

• Unix Program Design
  • Rob Pike and Brian W. Kernighan on the cat program and how it is a good example of Unix program philosphy.