The tr Command

The tr command is used to translate specified characters into other characters or to delete them.

In contrast to many command line programs, tr does not accept file names as arguments (i.e., input data). Instead, it only accepts inputs via standard input, (i.e., from the keyboard) or from the output of other programs via redirection.

The general syntax of tr is

tr [options] set1 [set2]

The items in the square brackets are optional. tr requires at least one argument and accepts a maximum of two. The first, designated set1, lists the characters in the text to be replaced or removed. The second, set2, lists the characters that are to be substituted for the characters listed in the first argument.

When used without any options, tr will replace the characters provided in set1 with those provided in set1. Thus, for example, every instance of the letter a in text typed in at the keyboard can be replaced with the letter b by entering the following command, pressing the ENTER key to start a new line, typing the text, and then pressing the ENTER key again:

tr a b

This substitution can repeated for additional text by typing it in and then pressing the ENTER key again. It can be terminated by simultaneously pressing the CONTROL and c keys.

Although tr cannot accept the names of files as arguments, it can nevertheless be used to modify copies of their contents. All that is necessary is to use the input redirection operator, which is represented by a leftward pointing angular bracket, before the name of the file whose text is to be modified. For example, the following would redirect a copy of a file named file1 to tr, which would display its contents on the monitor screen with every instance of an a replaced with a b:

tr a b < file1

It is typically more convenient to have the modified output written to a file rather than being displayed on the screen, and this can be easily accomplished by using the output redirection operator, which is represented by a rightward pointing angular bracket. Thus, for example, the following would replace the letter c with the letter d in the text from file1 and write it to a file named file2:

tr c d < file1 > file2

If file2 does not already exist, it will be created automatically by the output redirection operator. If it already exists, it will be overwritten.

Care should be taken to avoid attempting to use the output redirection operator to write to the same file from which the text is being read, as this will erase all of the text in that file, including that outputted by tr. Thus, for example, the following should not be used:

tr c d < file1 > file1

What can be done, however, is to use the append redirection operator, which is represented by two successive rightward pointing angular brackets, to add the modified text to the end of the existing text in the input file. Thus, the above example could be modified to append the revised text to file1 as follows:

tr c d < file1 >> file1

An alternative way to obtain a copy of the content of a file is to use a command such as cat to read it and then send its output to tr via a pipe, which is represented by the vertical bar character. For example, the following sends a copy of the text in a file named file3 to tr for translation before writing the result to another file named file4:

cat file3 | tr c d > file4

tr can do much more than just translate a single letter. It can also translate any number of specified characters into other characters. In such case, each of the two sets needs to be enclosed in square brackets, and these brackets, in turn, need to be enclosed in quotation marks. Either single of double quotation marks can be used.

One way of performing such translation of multiple characters is by listing the characters that are to be changed in the first argument and and listing their respective replacements in order in the second argument. For example, the following would replace all instances of the letters e, f and g in a file named file5 with the letters x, y and z, respectively, and write the output to another file named file6:

cat file5 | tr '[efg]' '[xyz]' > file6

tr can also replace characters in a specified range by their counterparts in another specified range, which can be more convenient when numerous characters are involved. A range is indicated by inserting a hyphen between the first and last characters, placing all of this in square brackets and then enclosing the brackets in quotation marks. Thus, the above example could be rewritten as follows:

cat file5 | tr '[e-g]' '[x-z]' > file6

Likewise, the following would replace every upper case letter in a file named file7 by its lower case counterpart and write the result to a file called file8:

cat file7 | tr '[A-Z]' '[a-z]' > file8

tr can also be used with control characters. Also referred to as non-printing characters, these are bit sequences that represent basic formatting instructions, etc., and each begins with a backslash. They are \b for backspace, \f for form feed, \n for new line, \r for return, \t for horizontal tab, \v for vertical tab, \a for buzzer sound, \\ for backslash and \NNN for octal value NNN (consisting of one to three digits).

tr also has a number of standardized arguments that are often more convenient to use instead of making lists or ranges of characters to be translated. Each consists of a word or abbreviation surrounded by colons and then enclosed in a set of square brackets. For example, all of the letters of the alphabet are represented by [:alpha:], all numerals are represented by [:digit:], all alphanumeric characters are represented by [:alnum:], all horizontal white spaces are represented by [:blank:], all horizontal or vertical white spaces are represented by [:space:], all printable characters exclusive of spaces are represented by [:graph:], all control characters are represented by [:cntrl:], all lower case letters are represented by [:lower:], all upper case letters are represented by [:upper:], all punctuation characters are represented by [:punct:], all printable characters including spaces are represented by [:print:], and all hexadecimal digits are represented by [:xdigit:].

Thus, for example, the previous example could be rewritten as

cat file7 | tr [:upper:] [:lower:] > file8

Probably the most useful of tr's several options is -s, which replaces each instance of a sequence of repeated characters listed in set1 with a single instance of the character specified in set2. This squeeze option is commonly used to replace each sequence of multiple blank spaces in text with a single blank space. For example, the following will remove all instances of successive blank spaces from a copy of the text in a file named file9 and write its output to a new file named file10:

tr -s ' ' ' ' < file9 > file10

The -d option is used to delete every instance of the string (i.e., sequence of characters) specified in set1. Thus, for example, the following would remove every instance of the word soft from a copy of the text in a file named file11 and write the modified text to a file named file12:

cat file11 | tr -d 'soft' > file12

The quotation marks are necessary for tr to treat the argument as a string. If they are not used, everything in the argument is instead treated as individual characters. Thus, if the above example were rewritten without the quotation marks, it would remove every instance of the letters s, o, f and t. Interestingly, the quotation marks cannot be used to treat arguments as strings when not using the -d option.

Among the few remaining options is -c, which causes tr to work on the complement of the specified characters, that is, on the characters that are not in the given set.

tr contains much of most basic functionality of the command line program sed, which is used to perform basic editing on streams of text supplied by a pipe. However, it often advantageous to use tr instead of sed because the former is simpler and requires less typing and because it is easier to incorporate into scripts.