<?php
/************************************************************
Library to convert HTML into an approximate text equivalent
v1.0.5 update 10/06/2024 to be PHP 8 compatible
v1.0.4 update 11/10/2008 to convert HTML entities
*************************************************************
Please see http://www.howtocreate.co.uk/php/ for details
Please see http://www.howtocreate.co.uk/jslibs/termsOfUse.html for terms and
conditions of use
This version is not supported, and will no longer be updated.
Please see http://www.howtocreate.co.uk/php/ to download a supported version.
PHP 5+ is required for supported versions.
The reason this library was written was to convert HTML email contents into a text
based email content, where the rendering does not have to be as accurate as with
a text based browser. However, there must be many more uses for it.
This library attempts to deal with non-standard HTML, but may occasionally suffer
from problems with pages that are not properly written - most especially:
Tags written as <tagName attribute=somethingWithA"or'InItButNotSurroundedByQuotes>,
Closing </pre> or </textarea> tags without their corresponding opening tags,
Tags within <textarea> </textarea> tags, which will be rendered, even though they
should not be.
Conversion requires a lot of preg_replace statements, so it can be quite slow with
large HTML files.
******
To use
******
This library requires PHP 4+
To use this library, put the following line in your script before the part that needs it:
require('PATH_TO_THIS_FILE/html2text.php');
To convert HTML/PHP to text:
$textVersion = html2text( $HTMLversion );
************
Further info
************
For the technically minded, this is the process I use for converting HTML to approx text:
REMOVE php start and end tags
REMOVE <!-- -->
ensure HTML uses entities in the right places (like inside tags) so strip_tags works properly
<STYLE|SCRIPT|OPTION>
carefully remove everything between them
strip_tags except the important ones
replace all \s that are after the start or a </pre> and before <pre> or end with a single space
</TITLE|HR>
\n --------------------
<H1|H2|H3|H4|H5|H6|DIV|P|PRE>
\n\n
<SUP>
^
<UL|OL|BR|DL|DT|TABLE|CAPTION|TR->(TH|TD)>
\n
<LI>
\n·
<DD>
\n\t
<TH|TD>
\t
<A|AREA href=(!javascript:&&!#)>
[LINK:hrefWithout#]
<IMG>
[IMG:alt]
<FORM>
[FORM:action]
<INPUT|TEXTAREA|BUTTON|SELECT>
[INPUT]
strip tags again, leaving nothing this time
un-htmlspecialchars
word wrap (this will also affect pre, but as this is intended for email use, I don't care)
*/
function html2text( $badStr ) {
//remove PHP if it exists
while( substr_count( $badStr, '<'.'?' ) && substr_count( $badStr, '?'.'>' ) && strpos( $badStr, '?'.'>', strpos( $badStr, '<'.'?' ) ) > strpos( $badStr, '<'.'?' ) ) {
$badStr = substr( $badStr, 0, strpos( $badStr, '<'.'?' ) ) . substr( $badStr, strpos( $badStr, '?'.'>', strpos( $badStr, '<'.'?' ) ) + 2 ); }
//remove comments
while( substr_count( $badStr, '<!--' ) && substr_count( $badStr, '-->' ) && strpos( $badStr, '-->', strpos( $badStr, '<!--' ) ) > strpos( $badStr, '<!--' ) ) {
$badStr = substr( $badStr, 0, strpos( $badStr, '<!--' ) ) . substr( $badStr, strpos( $badStr, '-->', strpos( $badStr, '<!--' ) ) + 3 ); }
//now make sure all HTML tags are correctly written (> not in between quotes)
for( $x = 0, $goodStr = '', $is_open_tb = false, $is_open_sq = false, $is_open_dq = false; $x < strlen($badStr); $x++ ) {
$chr = $badStr[$x];
//take each letter in turn and check if that character is permitted there
switch( $chr ) {
case '<':
if( !$is_open_tb && strtolower( substr( $badStr, $x + 1, 5 ) ) == 'style' ) {
$badStr = substr( $badStr, 0, $x ) . substr( $badStr, strpos( strtolower( $badStr ), '</style>', $x ) + 7 ); $chr = '';
} elseif( !$is_open_tb && strtolower( substr( $badStr, $x + 1, 6 ) ) == 'script' ) {
$badStr = substr( $badStr, 0, $x ) . substr( $badStr, strpos( strtolower( $badStr ), '</script>', $x ) + 8 ); $chr = '';
} elseif( !$is_open_tb ) { $is_open_tb = true; } else { $chr = '<'; }
break;
case '>':
if( !$is_open_tb || $is_open_dq || $is_open_sq ) { $chr = '>'; } else { $is_open_tb = false; }
break;
case '"':
if( $is_open_tb && !$is_open_dq && !$is_open_sq ) { $is_open_dq = true; }
elseif( $is_open_tb && $is_open_dq && !$is_open_sq ) { $is_open_dq = false; }
else { $chr = '"'; }
break;
case "'":
if( $is_open_tb && !$is_open_dq && !$is_open_sq ) { $is_open_sq = true; }
elseif( $is_open_tb && !$is_open_dq && $is_open_sq ) { $is_open_sq = false; }
} $goodStr .= $chr;
}
//now that the page is valid (I hope) for strip_tags, strip all unwanted tags
$goodStr = strip_tags( $goodStr, '<title><hr><h1><h2><h3><h4><h5><h6><div><p><pre><sup><ul><ol><br><dl><dt><table><caption><tr><li><dd><th><td><a><area><img><form><input><textarea><button><select><option>' );
//strip extra whitespace except between <pre> and <textarea> tags
$badStr = preg_split( "/<\/?pre[^>]*>/i", $goodStr );
for( $x = 0; isset($badStr[$x]) && is_string( $badStr[$x] ); $x++ ) {
if( $x % 2 ) { $badStr[$x] = '<pre>'.$badStr[$x].'</pre>'; } else {
$goodStr = preg_split( "/<\/?textarea[^>]*>/i", $badStr[$x] );
for( $z = 0; isset($goodStr[$z]) && is_string( $goodStr[$z] ); $z++ ) {
if( $z % 2 ) { $goodStr[$z] = '<textarea>'.$goodStr[$z].'</textarea>'; } else {
$goodStr[$z] = preg_replace( "/\s+/", ' ', $goodStr[$z] );
} }
$badStr[$x] = implode('',$goodStr);
} }
$goodStr = implode('',$badStr);
//remove all options from select inputs
$goodStr = preg_replace( "/<option[^>]*>[^<]*/i", '', $goodStr );
//replace all tags with their text equivalents
$goodStr = preg_replace( "/<(\/title|hr)[^>]*>/i", "\n --------------------\n", $goodStr );
$goodStr = preg_replace( "/<(h|div|p)[^>]*>/i", "\n\n", $goodStr );
$goodStr = preg_replace( "/<sup[^>]*>/i", '^', $goodStr );
$goodStr = preg_replace( "/<(ul|ol|br|dl|dt|table|caption|\/textarea|tr[^>]*>\s*<(td|th))[^>]*>/i", "\n", $goodStr );
$goodStr = preg_replace( "/<li[^>]*>/i", "\n· ", $goodStr );
$goodStr = preg_replace( "/<dd[^>]*>/i", "\n\t", $goodStr );
$goodStr = preg_replace( "/<(th|td)[^>]*>/i", "\t", $goodStr );
$goodStr = preg_replace( "/<a[^>]* href=(\"((?!\"|#|javascript:)[^\"#]*)(\"|#)|'((?!'|#|javascript:)[^'#]*)('|#)|((?!'|\"|>|#|javascript:)[^#\"'> ]*))[^>]*>/i", "[LINK: $2$4$6] ", $goodStr );
$goodStr = preg_replace( "/<img[^>]* alt=(\"([^\"]+)\"|'([^']+)'|([^\"'> ]+))[^>]*>/i", "[IMAGE: $2$3$4] ", $goodStr );
$goodStr = preg_replace( "/<form[^>]* action=(\"([^\"]+)\"|'([^']+)'|([^\"'> ]+))[^>]*>/i", "\n[FORM: $2$3$4] ", $goodStr );
$goodStr = preg_replace( "/<(input|textarea|button|select)[^>]*>/i", "[INPUT] ", $goodStr );
//strip all remaining tags (mostly closing tags)
$goodStr = strip_tags( $goodStr );
//convert HTML entities
$goodStr = strtr( $goodStr, array_flip( get_html_translation_table( HTML_ENTITIES ) ) );
if( function_exists('preg_replace_callback') ) {
//needed for PHP 7 support (*sigh*, yes the script got updated even though it is unsupported)
$goodStr = preg_replace_callback( "/&#(\d+);/m", "html2text_getchr", $goodStr );
} else {
//early 4.0 releases
$goodStr = preg_replace( "/&#(\d+);/me", "chr('$1')", $goodStr );
}
//wordwrap
$goodStr = wordwrap( $goodStr );
//make sure there are no more than 3 linebreaks in a row and trim whitespace
return preg_replace( "/^\n*|\n*$/", '', preg_replace( "/[ \t]+(\n|$)/", "$1", preg_replace( "/\n(\s*\n){2}/", "\n\n\n", preg_replace( "/\r\n?|\f/", "\n", str_replace( chr(160), ' ', $goodStr ) ) ) ) );
}
function html2text_getchr($matches) { return chr($matches[1]); }
?>