html2text

Section: User Commands (1)
Updated: 2004-01-14
Page Index

NAME

html2text - an advanced HTML-to-text converter

SYNOPSIS

html2text -help
html2text -version
html2text [ -unparse | -check ] [ -debug-scanner ] [ -debug-parser ] [ -rcfile path ] [ -style ( compact | pretty ) ] [ -width width ] [ -o output-file ] [ -nobs ] [ -ascii ] [ input-url ... ]

DESCRIPTION

html2text reads HTML documents from the input-urls, formats each of them into a stream of plain text characters, and writes the result to standard output (or into output-file, if the -o command line option is used).

Documents that are specified by a URL (RFC 1738) that begins with "http:" are retrieved with the Hypertext Transfer Protocol (RFC 1945). URLs that begin with "file:" and URLs that do not contain a colon specify local files. All other URLs are invalid.

If no input-urls are specified on the command line, html2text reads from standard input. A dash as the input-url is an alternate way to specify standard input.

html2text understands all HTML 3.2 constructs, but can render only part of them due to the limitations of the text output format. However, the program attempts to provide good substitutes for the elements it cannot render. html2text parses HTML 4 input, too, but not always as successful as other HTML processors. It also accepts syntactically incorrect input, and attempts to interpret it "reasonably".

The way html2text formats the HTML documents is controlled by formatting properties read from an RC file. html2text attempts to read $HOME/.html2textrc (or the file specified by the -rcfile command line option); if that file cannot be read, html2text attempts to read /etc/html2textrc. If no RC file can be read (or if the RC file does not override all formatting properties), then "reasonable" defaults are assumed. The RC file format is described in the html2textrc(5) manual page.

OPTIONS

-ascii: By default, html2text uses ISO 8859-1 for the output. Specifying this option, plain ASCII is used instead. To find out how non-ASCII characters are rendered, refer to the file "ascii.substitutes".
-check: This option is for diagnostic purposes: The HTML document is only parsed and not processed otherwise. In this mode of operation, html2text will report on parse errors and scan errors, which it does not in other modes of operation. Note that parse and scan errors are not fatal for html2text, but may cause mis-interpretation of the HTML code and/or portions of the document being swallowed.
-debug-parser: Let html2text report on the tokens being shifted, rules being applied, etc., while scanning the HTML document. This option is for diagnostic purposes.
-debug-scanner: Let html2text report on each lexical token scanned, while scanning the HTML document. This option is for diagnostic purposes.
-help: Print command line summary and exit.
-nobs: By default, html2text renders underlined letters with sequences like "underscore-backspace-character" and boldface letters like "character-backspace-character", which works fine when the output is piped into more(1), less(1), or similar. For other applications, or when redirecting the output into a file, it may be desirable not to render character attributes with such backspace sequences, which can be accomplished with this command line option.
-o output-file: Write the output to output-file instead of standard output. A dash as the output-file is an alternate way to specify the standard output.
-rcfile path: Attempt to read the file specified in path as RC file.
-style ( compact | pretty ): Style pretty changes some of the default values of the formatting parameters documented in html2textrc(5). To find out which and how the formatting parameter defaults are changed, check the file "pretty.style". If this option is omitted, style compact is assumed as default.
-unparse: This option is for diagnostic purposes: Instead of formatting the parsed document, generate HTML code, that is guaranteed to be syntactically correct. If html2text has problems parsing a syntactically incorrect HTML document, this option may help you to understand what html2text thinks that the original HTML code means.
-version: Print program version and exit.
-width width: By default, html2text formats the HTML documents for a screen width of 79 characters. If redirecting the output into a file, or if your terminal has a width other than 80 characters, or if you just want to get an idea how html2text deals with large tables and different terminal widths, you may want to specify a different width.

FILES

/etc/html2textrc: System wide parser configuration file.
$HOME/.html2textrc: Personal parser configuration file, overrides the system wide values.

CONFORMING TO

HTML 3.2 (HTML 3.2 Reference Specification - http://www.w3.org/TR/REC-html32),
RFC 1945 (Hypertext Transfer Protocol - HTTP).

RESTRICTIONS

html2text provides only a basic implementation of the Hypertext Transfer Protocol (HTTP). It requires the complete and exactly matching URL to be given as argument and will not follow redirections (HTTP 301/ 307).

html2text was written to convert HTML 3.2 documents. When using it with HTML 4 or even XHTML 1 documents, some constructs present only in these HTML versions might not be rendered.

AUTHOR

html2text was written up to version 1.2.2 by Arno Unkrig <arno@unkrig.de> for GMRS Software GmbH, Unterschleissheim.

Current maintainer and primary download location is:
Martin Bayer <mbayer@zedat.fu-berlin.de>
http://userpage.fu-berlin.de/~mbayer/tools/html2text.html