Name

utf8trans — Transliterate UTF-8 characters according to a table

Synopsis

utf8trans charmap [file...]

Description

utf8trans transliterates characters in the specified files (or standard input, if they are not specified) and writes the output to standard output. All input and output is in the UTF-8 encoding.

This program is usually used to render characters in Unicode text files as some markup escapes or ASCII transliterations. (It is not intended for general charset conversions.) It provides functionality similar to the character maps in XSLT 2.0 (XML Stylesheet Language – Transformations, version 2.0).

Options

-m, --modify

Modifies the given files in-place with their transliterated output, instead of sending it to standard output.

This option is useful for efficient transliteration of many files at once.

--help

Show brief usage information and exit.

--version

Show version and exit.

Usage

The translation is done according to the rules in the “character map”, named in the file charmap. It has the following format:

  1. Each line represents a translation entry, except for blank lines and comment lines, which are ignored.

  2. Any amount of whitespace (space or tab) may precede the start of an entry.

  3. Comment lines begin with #. Everything on the same line is ignored.

  4. Each entry consists of the Unicode codepoint of the character to translate, in hexadecimal, followed one space or tab, followed by the translation string, up to the end of the line.

  5. The translation string is taken literally, including any leading and trailing spaces (except the delimeter between the codepoint and the translation string), and all types of characters. The newline at the end is not included.

The above format is intended to be restrictive, to keep utf8trans simple. But if a XML-based format is desired, there is a xmlcharmap2utf8trans script that comes with the docbook2X distribution, that converts character maps in XSLT 2.0 format to the utf8trans format.

Limitations

  • utf8trans does not work with binary files, because malformed UTF-8 sequences in the input are substituted with U+FFFD characters. However, null characters in the input are handled correctly. This limitation may be removed in the future.

  • There is no way to include a newline or null in the substitution string.