diff options
| author | friendica <info@friendica.com> | 2012-05-12 17:57:41 -0700 |
|---|---|---|
| committer | friendica <info@friendica.com> | 2012-07-18 20:40:31 +1000 |
| commit | 7a40f4354b32809af3d0cfd6e3af0eda02ab0e0a (patch) | |
| tree | a9c3d91209cff770bb4b613b1b95e61a7bbc5a2b /lib/htmlpurifier/library/HTMLPurifier/AttrDef | |
| parent | cd727cb26b78a1dade09d510b071446898477356 (diff) | |
| download | volse-hubzilla-7a40f4354b32809af3d0cfd6e3af0eda02ab0e0a.tar.gz volse-hubzilla-7a40f4354b32809af3d0cfd6e3af0eda02ab0e0a.tar.bz2 volse-hubzilla-7a40f4354b32809af3d0cfd6e3af0eda02ab0e0a.zip | |
some important stuff we'll need
Diffstat (limited to 'lib/htmlpurifier/library/HTMLPurifier/AttrDef')
42 files changed, 2442 insertions, 0 deletions
diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS.php new file mode 100644 index 000000000..953e70675 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS.php @@ -0,0 +1,87 @@ +<?php + +/** + * Validates the HTML attribute style, otherwise known as CSS. + * @note We don't implement the whole CSS specification, so it might be + * difficult to reuse this component in the context of validating + * actual stylesheet declarations. + * @note If we were really serious about validating the CSS, we would + * tokenize the styles and then parse the tokens. Obviously, we + * are not doing that. Doing that could seriously harm performance, + * but would make these components a lot more viable for a CSS + * filtering solution. + */ +class HTMLPurifier_AttrDef_CSS extends HTMLPurifier_AttrDef +{ + + public function validate($css, $config, $context) { + + $css = $this->parseCDATA($css); + + $definition = $config->getCSSDefinition(); + + // we're going to break the spec and explode by semicolons. + // This is because semicolon rarely appears in escaped form + // Doing this is generally flaky but fast + // IT MIGHT APPEAR IN URIs, see HTMLPurifier_AttrDef_CSSURI + // for details + + $declarations = explode(';', $css); + $propvalues = array(); + + /** + * Name of the current CSS property being validated. + */ + $property = false; + $context->register('CurrentCSSProperty', $property); + + foreach ($declarations as $declaration) { + if (!$declaration) continue; + if (!strpos($declaration, ':')) continue; + list($property, $value) = explode(':', $declaration, 2); + $property = trim($property); + $value = trim($value); + $ok = false; + do { + if (isset($definition->info[$property])) { + $ok = true; + break; + } + if (ctype_lower($property)) break; + $property = strtolower($property); + if (isset($definition->info[$property])) { + $ok = true; + break; + } + } while(0); + if (!$ok) continue; + // inefficient call, since the validator will do this again + if (strtolower(trim($value)) !== 'inherit') { + // inherit works for everything (but only on the base property) + $result = $definition->info[$property]->validate( + $value, $config, $context ); + } else { + $result = 'inherit'; + } + if ($result === false) continue; + $propvalues[$property] = $result; + } + + $context->destroy('CurrentCSSProperty'); + + // procedure does not write the new CSS simultaneously, so it's + // slightly inefficient, but it's the only way of getting rid of + // duplicates. Perhaps config to optimize it, but not now. + + $new_declarations = ''; + foreach ($propvalues as $prop => $value) { + $new_declarations .= "$prop:$value;"; + } + + return $new_declarations ? $new_declarations : false; + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/AlphaValue.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/AlphaValue.php new file mode 100644 index 000000000..292c040d4 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/AlphaValue.php @@ -0,0 +1,21 @@ +<?php + +class HTMLPurifier_AttrDef_CSS_AlphaValue extends HTMLPurifier_AttrDef_CSS_Number +{ + + public function __construct() { + parent::__construct(false); // opacity is non-negative, but we will clamp it + } + + public function validate($number, $config, $context) { + $result = parent::validate($number, $config, $context); + if ($result === false) return $result; + $float = (float) $result; + if ($float < 0.0) $result = '0'; + if ($float > 1.0) $result = '1'; + return $result; + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Background.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Background.php new file mode 100644 index 000000000..3a3d20cd6 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Background.php @@ -0,0 +1,87 @@ +<?php + +/** + * Validates shorthand CSS property background. + * @warning Does not support url tokens that have internal spaces. + */ +class HTMLPurifier_AttrDef_CSS_Background extends HTMLPurifier_AttrDef +{ + + /** + * Local copy of component validators. + * @note See HTMLPurifier_AttrDef_Font::$info for a similar impl. + */ + protected $info; + + public function __construct($config) { + $def = $config->getCSSDefinition(); + $this->info['background-color'] = $def->info['background-color']; + $this->info['background-image'] = $def->info['background-image']; + $this->info['background-repeat'] = $def->info['background-repeat']; + $this->info['background-attachment'] = $def->info['background-attachment']; + $this->info['background-position'] = $def->info['background-position']; + } + + public function validate($string, $config, $context) { + + // regular pre-processing + $string = $this->parseCDATA($string); + if ($string === '') return false; + + // munge rgb() decl if necessary + $string = $this->mungeRgb($string); + + // assumes URI doesn't have spaces in it + $bits = explode(' ', strtolower($string)); // bits to process + + $caught = array(); + $caught['color'] = false; + $caught['image'] = false; + $caught['repeat'] = false; + $caught['attachment'] = false; + $caught['position'] = false; + + $i = 0; // number of catches + $none = false; + + foreach ($bits as $bit) { + if ($bit === '') continue; + foreach ($caught as $key => $status) { + if ($key != 'position') { + if ($status !== false) continue; + $r = $this->info['background-' . $key]->validate($bit, $config, $context); + } else { + $r = $bit; + } + if ($r === false) continue; + if ($key == 'position') { + if ($caught[$key] === false) $caught[$key] = ''; + $caught[$key] .= $r . ' '; + } else { + $caught[$key] = $r; + } + $i++; + break; + } + } + + if (!$i) return false; + if ($caught['position'] !== false) { + $caught['position'] = $this->info['background-position']-> + validate($caught['position'], $config, $context); + } + + $ret = array(); + foreach ($caught as $value) { + if ($value === false) continue; + $ret[] = $value; + } + + if (empty($ret)) return false; + return implode(' ', $ret); + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/BackgroundPosition.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/BackgroundPosition.php new file mode 100644 index 000000000..fae82eaec --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/BackgroundPosition.php @@ -0,0 +1,133 @@ +<?php + +/* W3C says: + [ // adjective and number must be in correct order, even if + // you could switch them without introducing ambiguity. + // some browsers support that syntax + [ + <percentage> | <length> | left | center | right + ] + [ + <percentage> | <length> | top | center | bottom + ]? + ] | + [ // this signifies that the vertical and horizontal adjectives + // can be arbitrarily ordered, however, there can only be two, + // one of each, or none at all + [ + left | center | right + ] || + [ + top | center | bottom + ] + ] + top, left = 0% + center, (none) = 50% + bottom, right = 100% +*/ + +/* QuirksMode says: + keyword + length/percentage must be ordered correctly, as per W3C + + Internet Explorer and Opera, however, support arbitrary ordering. We + should fix it up. + + Minor issue though, not strictly necessary. +*/ + +// control freaks may appreciate the ability to convert these to +// percentages or something, but it's not necessary + +/** + * Validates the value of background-position. + */ +class HTMLPurifier_AttrDef_CSS_BackgroundPosition extends HTMLPurifier_AttrDef +{ + + protected $length; + protected $percentage; + + public function __construct() { + $this->length = new HTMLPurifier_AttrDef_CSS_Length(); + $this->percentage = new HTMLPurifier_AttrDef_CSS_Percentage(); + } + + public function validate($string, $config, $context) { + $string = $this->parseCDATA($string); + $bits = explode(' ', $string); + + $keywords = array(); + $keywords['h'] = false; // left, right + $keywords['v'] = false; // top, bottom + $keywords['ch'] = false; // center (first word) + $keywords['cv'] = false; // center (second word) + $measures = array(); + + $i = 0; + + $lookup = array( + 'top' => 'v', + 'bottom' => 'v', + 'left' => 'h', + 'right' => 'h', + 'center' => 'c' + ); + + foreach ($bits as $bit) { + if ($bit === '') continue; + + // test for keyword + $lbit = ctype_lower($bit) ? $bit : strtolower($bit); + if (isset($lookup[$lbit])) { + $status = $lookup[$lbit]; + if ($status == 'c') { + if ($i == 0) { + $status = 'ch'; + } else { + $status = 'cv'; + } + } + $keywords[$status] = $lbit; + $i++; + } + + // test for length + $r = $this->length->validate($bit, $config, $context); + if ($r !== false) { + $measures[] = $r; + $i++; + } + + // test for percentage + $r = $this->percentage->validate($bit, $config, $context); + if ($r !== false) { + $measures[] = $r; + $i++; + } + + } + + if (!$i) return false; // no valid values were caught + + $ret = array(); + + // first keyword + if ($keywords['h']) $ret[] = $keywords['h']; + elseif ($keywords['ch']) { + $ret[] = $keywords['ch']; + $keywords['cv'] = false; // prevent re-use: center = center center + } + elseif (count($measures)) $ret[] = array_shift($measures); + + if ($keywords['v']) $ret[] = $keywords['v']; + elseif ($keywords['cv']) $ret[] = $keywords['cv']; + elseif (count($measures)) $ret[] = array_shift($measures); + + if (empty($ret)) return false; + return implode(' ', $ret); + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Border.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Border.php new file mode 100644 index 000000000..42a1d1b4a --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Border.php @@ -0,0 +1,43 @@ +<?php + +/** + * Validates the border property as defined by CSS. + */ +class HTMLPurifier_AttrDef_CSS_Border extends HTMLPurifier_AttrDef +{ + + /** + * Local copy of properties this property is shorthand for. + */ + protected $info = array(); + + public function __construct($config) { + $def = $config->getCSSDefinition(); + $this->info['border-width'] = $def->info['border-width']; + $this->info['border-style'] = $def->info['border-style']; + $this->info['border-top-color'] = $def->info['border-top-color']; + } + + public function validate($string, $config, $context) { + $string = $this->parseCDATA($string); + $string = $this->mungeRgb($string); + $bits = explode(' ', $string); + $done = array(); // segments we've finished + $ret = ''; // return value + foreach ($bits as $bit) { + foreach ($this->info as $propname => $validator) { + if (isset($done[$propname])) continue; + $r = $validator->validate($bit, $config, $context); + if ($r !== false) { + $ret .= $r . ' '; + $done[$propname] = true; + break; + } + } + } + return rtrim($ret); + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Color.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Color.php new file mode 100644 index 000000000..07f95a671 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Color.php @@ -0,0 +1,78 @@ +<?php + +/** + * Validates Color as defined by CSS. + */ +class HTMLPurifier_AttrDef_CSS_Color extends HTMLPurifier_AttrDef +{ + + public function validate($color, $config, $context) { + + static $colors = null; + if ($colors === null) $colors = $config->get('Core.ColorKeywords'); + + $color = trim($color); + if ($color === '') return false; + + $lower = strtolower($color); + if (isset($colors[$lower])) return $colors[$lower]; + + if (strpos($color, 'rgb(') !== false) { + // rgb literal handling + $length = strlen($color); + if (strpos($color, ')') !== $length - 1) return false; + $triad = substr($color, 4, $length - 4 - 1); + $parts = explode(',', $triad); + if (count($parts) !== 3) return false; + $type = false; // to ensure that they're all the same type + $new_parts = array(); + foreach ($parts as $part) { + $part = trim($part); + if ($part === '') return false; + $length = strlen($part); + if ($part[$length - 1] === '%') { + // handle percents + if (!$type) { + $type = 'percentage'; + } elseif ($type !== 'percentage') { + return false; + } + $num = (float) substr($part, 0, $length - 1); + if ($num < 0) $num = 0; + if ($num > 100) $num = 100; + $new_parts[] = "$num%"; + } else { + // handle integers + if (!$type) { + $type = 'integer'; + } elseif ($type !== 'integer') { + return false; + } + $num = (int) $part; + if ($num < 0) $num = 0; + if ($num > 255) $num = 255; + $new_parts[] = (string) $num; + } + } + $new_triad = implode(',', $new_parts); + $color = "rgb($new_triad)"; + } else { + // hexadecimal handling + if ($color[0] === '#') { + $hex = substr($color, 1); + } else { + $hex = $color; + $color = '#' . $color; + } + $length = strlen($hex); + if ($length !== 3 && $length !== 6) return false; + if (!ctype_xdigit($hex)) return false; + } + + return $color; + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Composite.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Composite.php new file mode 100644 index 000000000..de1289cba --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Composite.php @@ -0,0 +1,38 @@ +<?php + +/** + * Allows multiple validators to attempt to validate attribute. + * + * Composite is just what it sounds like: a composite of many validators. + * This means that multiple HTMLPurifier_AttrDef objects will have a whack + * at the string. If one of them passes, that's what is returned. This is + * especially useful for CSS values, which often are a choice between + * an enumerated set of predefined values or a flexible data type. + */ +class HTMLPurifier_AttrDef_CSS_Composite extends HTMLPurifier_AttrDef +{ + + /** + * List of HTMLPurifier_AttrDef objects that may process strings + * @todo Make protected + */ + public $defs; + + /** + * @param $defs List of HTMLPurifier_AttrDef objects + */ + public function __construct($defs) { + $this->defs = $defs; + } + + public function validate($string, $config, $context) { + foreach ($this->defs as $i => $def) { + $result = $this->defs[$i]->validate($string, $config, $context); + if ($result !== false) return $result; + } + return false; + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/DenyElementDecorator.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/DenyElementDecorator.php new file mode 100644 index 000000000..6599c5b2d --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/DenyElementDecorator.php @@ -0,0 +1,28 @@ +<?php + +/** + * Decorator which enables CSS properties to be disabled for specific elements. + */ +class HTMLPurifier_AttrDef_CSS_DenyElementDecorator extends HTMLPurifier_AttrDef +{ + public $def, $element; + + /** + * @param $def Definition to wrap + * @param $element Element to deny + */ + public function __construct($def, $element) { + $this->def = $def; + $this->element = $element; + } + /** + * Checks if CurrentToken is set and equal to $this->element + */ + public function validate($string, $config, $context) { + $token = $context->get('CurrentToken', true); + if ($token && $token->name == $this->element) return false; + return $this->def->validate($string, $config, $context); + } +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Filter.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Filter.php new file mode 100644 index 000000000..147894b86 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Filter.php @@ -0,0 +1,54 @@ +<?php + +/** + * Microsoft's proprietary filter: CSS property + * @note Currently supports the alpha filter. In the future, this will + * probably need an extensible framework + */ +class HTMLPurifier_AttrDef_CSS_Filter extends HTMLPurifier_AttrDef +{ + + protected $intValidator; + + public function __construct() { + $this->intValidator = new HTMLPurifier_AttrDef_Integer(); + } + + public function validate($value, $config, $context) { + $value = $this->parseCDATA($value); + if ($value === 'none') return $value; + // if we looped this we could support multiple filters + $function_length = strcspn($value, '('); + $function = trim(substr($value, 0, $function_length)); + if ($function !== 'alpha' && + $function !== 'Alpha' && + $function !== 'progid:DXImageTransform.Microsoft.Alpha' + ) return false; + $cursor = $function_length + 1; + $parameters_length = strcspn($value, ')', $cursor); + $parameters = substr($value, $cursor, $parameters_length); + $params = explode(',', $parameters); + $ret_params = array(); + $lookup = array(); + foreach ($params as $param) { + list($key, $value) = explode('=', $param); + $key = trim($key); + $value = trim($value); + if (isset($lookup[$key])) continue; + if ($key !== 'opacity') continue; + $value = $this->intValidator->validate($value, $config, $context); + if ($value === false) continue; + $int = (int) $value; + if ($int > 100) $value = '100'; + if ($int < 0) $value = '0'; + $ret_params[] = "$key=$value"; + $lookup[$key] = true; + } + $ret_parameters = implode(',', $ret_params); + $ret_function = "$function($ret_parameters)"; + return $ret_function; + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Font.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Font.php new file mode 100644 index 000000000..699ee0b70 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Font.php @@ -0,0 +1,149 @@ +<?php + +/** + * Validates shorthand CSS property font. + */ +class HTMLPurifier_AttrDef_CSS_Font extends HTMLPurifier_AttrDef +{ + + /** + * Local copy of component validators. + * + * @note If we moved specific CSS property definitions to their own + * classes instead of having them be assembled at run time by + * CSSDefinition, this wouldn't be necessary. We'd instantiate + * our own copies. + */ + protected $info = array(); + + public function __construct($config) { + $def = $config->getCSSDefinition(); + $this->info['font-style'] = $def->info['font-style']; + $this->info['font-variant'] = $def->info['font-variant']; + $this->info['font-weight'] = $def->info['font-weight']; + $this->info['font-size'] = $def->info['font-size']; + $this->info['line-height'] = $def->info['line-height']; + $this->info['font-family'] = $def->info['font-family']; + } + + public function validate($string, $config, $context) { + + static $system_fonts = array( + 'caption' => true, + 'icon' => true, + 'menu' => true, + 'message-box' => true, + 'small-caption' => true, + 'status-bar' => true + ); + + // regular pre-processing + $string = $this->parseCDATA($string); + if ($string === '') return false; + + // check if it's one of the keywords + $lowercase_string = strtolower($string); + if (isset($system_fonts[$lowercase_string])) { + return $lowercase_string; + } + + $bits = explode(' ', $string); // bits to process + $stage = 0; // this indicates what we're looking for + $caught = array(); // which stage 0 properties have we caught? + $stage_1 = array('font-style', 'font-variant', 'font-weight'); + $final = ''; // output + + for ($i = 0, $size = count($bits); $i < $size; $i++) { + if ($bits[$i] === '') continue; + switch ($stage) { + + // attempting to catch font-style, font-variant or font-weight + case 0: + foreach ($stage_1 as $validator_name) { + if (isset($caught[$validator_name])) continue; + $r = $this->info[$validator_name]->validate( + $bits[$i], $config, $context); + if ($r !== false) { + $final .= $r . ' '; + $caught[$validator_name] = true; + break; + } + } + // all three caught, continue on + if (count($caught) >= 3) $stage = 1; + if ($r !== false) break; + + // attempting to catch font-size and perhaps line-height + case 1: + $found_slash = false; + if (strpos($bits[$i], '/') !== false) { + list($font_size, $line_height) = + explode('/', $bits[$i]); + if ($line_height === '') { + // ooh, there's a space after the slash! + $line_height = false; + $found_slash = true; + } + } else { + $font_size = $bits[$i]; + $line_height = false; + } + $r = $this->info['font-size']->validate( + $font_size, $config, $context); + if ($r !== false) { + $final .= $r; + // attempt to catch line-height + if ($line_height === false) { + // we need to scroll forward + for ($j = $i + 1; $j < $size; $j++) { + if ($bits[$j] === '') continue; + if ($bits[$j] === '/') { + if ($found_slash) { + return false; + } else { + $found_slash = true; + continue; + } + } + $line_height = $bits[$j]; + break; + } + } else { + // slash already found + $found_slash = true; + $j = $i; + } + if ($found_slash) { + $i = $j; + $r = $this->info['line-height']->validate( + $line_height, $config, $context); + if ($r !== false) { + $final .= '/' . $r; + } + } + $final .= ' '; + $stage = 2; + break; + } + return false; + + // attempting to catch font-family + case 2: + $font_family = + implode(' ', array_slice($bits, $i, $size - $i)); + $r = $this->info['font-family']->validate( + $font_family, $config, $context); + if ($r !== false) { + $final .= $r . ' '; + // processing completed successfully + return rtrim($final); + } + return false; + } + } + return false; + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/FontFamily.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/FontFamily.php new file mode 100644 index 000000000..0d9a4e12c --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/FontFamily.php @@ -0,0 +1,197 @@ +<?php + +/** + * Validates a font family list according to CSS spec + */ +class HTMLPurifier_AttrDef_CSS_FontFamily extends HTMLPurifier_AttrDef +{ + + protected $mask = null; + + public function __construct() { + $this->mask = '- '; + for ($c = 'a'; $c <= 'z'; $c++) $this->mask .= $c; + for ($c = 'A'; $c <= 'Z'; $c++) $this->mask .= $c; + for ($c = '0'; $c <= '9'; $c++) $this->mask .= $c; // cast-y, but should be fine + // special bytes used by UTF-8 + for ($i = 0x80; $i <= 0xFF; $i++) { + // We don't bother excluding invalid bytes in this range, + // because the our restriction of well-formed UTF-8 will + // prevent these from ever occurring. + $this->mask .= chr($i); + } + + /* + PHP's internal strcspn implementation is + O(length of string * length of mask), making it inefficient + for large masks. However, it's still faster than + preg_match 8) + for (p = s1;;) { + spanp = s2; + do { + if (*spanp == c || p == s1_end) { + return p - s1; + } + } while (spanp++ < (s2_end - 1)); + c = *++p; + } + */ + // possible optimization: invert the mask. + } + + public function validate($string, $config, $context) { + static $generic_names = array( + 'serif' => true, + 'sans-serif' => true, + 'monospace' => true, + 'fantasy' => true, + 'cursive' => true + ); + $allowed_fonts = $config->get('CSS.AllowedFonts'); + + // assume that no font names contain commas in them + $fonts = explode(',', $string); + $final = ''; + foreach($fonts as $font) { + $font = trim($font); + if ($font === '') continue; + // match a generic name + if (isset($generic_names[$font])) { + if ($allowed_fonts === null || isset($allowed_fonts[$font])) { + $final .= $font . ', '; + } + continue; + } + // match a quoted name + if ($font[0] === '"' || $font[0] === "'") { + $length = strlen($font); + if ($length <= 2) continue; + $quote = $font[0]; + if ($font[$length - 1] !== $quote) continue; + $font = substr($font, 1, $length - 2); + } + + $font = $this->expandCSSEscape($font); + + // $font is a pure representation of the font name + + if ($allowed_fonts !== null && !isset($allowed_fonts[$font])) { + continue; + } + + if (ctype_alnum($font) && $font !== '') { + // very simple font, allow it in unharmed + $final .= $font . ', '; + continue; + } + + // bugger out on whitespace. form feed (0C) really + // shouldn't show up regardless + $font = str_replace(array("\n", "\t", "\r", "\x0C"), ' ', $font); + + // Here, there are various classes of characters which need + // to be treated differently: + // - Alphanumeric characters are essentially safe. We + // handled these above. + // - Spaces require quoting, though most parsers will do + // the right thing if there aren't any characters that + // can be misinterpreted + // - Dashes rarely occur, but they fairly unproblematic + // for parsing/rendering purposes. + // The above characters cover the majority of Western font + // names. + // - Arbitrary Unicode characters not in ASCII. Because + // most parsers give little thought to Unicode, treatment + // of these codepoints is basically uniform, even for + // punctuation-like codepoints. These characters can + // show up in non-Western pages and are supported by most + // major browsers, for example: "ＭＳ 明朝" is a + // legitimate font-name + // <http://ja.wikipedia.org/wiki/MS_明朝>. See + // the CSS3 spec for more examples: + // <http://www.w3.org/TR/2011/WD-css3-fonts-20110324/localizedfamilynames.png> + // You can see live samples of these on the Internet: + // <http://www.google.co.jp/search?q=font-family+ＭＳ+明朝|ゴシック> + // However, most of these fonts have ASCII equivalents: + // for example, 'MS Mincho', and it's considered + // professional to use ASCII font names instead of + // Unicode font names. Thanks Takeshi Terada for + // providing this information. + // The following characters, to my knowledge, have not been + // used to name font names. + // - Single quote. While theoretically you might find a + // font name that has a single quote in its name (serving + // as an apostrophe, e.g. Dave's Scribble), I haven't + // been able to find any actual examples of this. + // Internet Explorer's cssText translation (which I + // believe is invoked by innerHTML) normalizes any + // quoting to single quotes, and fails to escape single + // quotes. (Note that this is not IE's behavior for all + // CSS properties, just some sort of special casing for + // font-family). So a single quote *cannot* be used + // safely in the font-family context if there will be an + // innerHTML/cssText translation. Note that Firefox 3.x + // does this too. + // - Double quote. In IE, these get normalized to + // single-quotes, no matter what the encoding. (Fun + // fact, in IE8, the 'content' CSS property gained + // support, where they special cased to preserve encoded + // double quotes, but still translate unadorned double + // quotes into single quotes.) So, because their + // fixpoint behavior is identical to single quotes, they + // cannot be allowed either. Firefox 3.x displays + // single-quote style behavior. + // - Backslashes are reduced by one (so \\ -> \) every + // iteration, so they cannot be used safely. This shows + // up in IE7, IE8 and FF3 + // - Semicolons, commas and backticks are handled properly. + // - The rest of the ASCII punctuation is handled properly. + // We haven't checked what browsers do to unadorned + // versions, but this is not important as long as the + // browser doesn't /remove/ surrounding quotes (as IE does + // for HTML). + // + // With these results in hand, we conclude that there are + // various levels of safety: + // - Paranoid: alphanumeric, spaces and dashes(?) + // - International: Paranoid + non-ASCII Unicode + // - Edgy: Everything except quotes, backslashes + // - NoJS: Standards compliance, e.g. sod IE. Note that + // with some judicious character escaping (since certain + // types of escaping doesn't work) this is theoretically + // OK as long as innerHTML/cssText is not called. + // We believe that international is a reasonable default + // (that we will implement now), and once we do more + // extensive research, we may feel comfortable with dropping + // it down to edgy. + + // Edgy: alphanumeric, spaces, dashes and Unicode. Use of + // str(c)spn assumes that the string was already well formed + // Unicode (which of course it is). + if (strspn($font, $this->mask) !== strlen($font)) { + continue; + } + + // Historical: + // In the absence of innerHTML/cssText, these ugly + // transforms don't pose a security risk (as \\ and \" + // might--these escapes are not supported by most browsers). + // We could try to be clever and use single-quote wrapping + // when there is a double quote present, but I have choosen + // not to implement that. (NOTE: you can reduce the amount + // of escapes by one depending on what quoting style you use) + // $font = str_replace('\\', '\\5C ', $font); + // $font = str_replace('"', '\\22 ', $font); + // $font = str_replace("'", '\\27 ', $font); + + // font possibly with spaces, requires quoting + $final .= "'$font', "; + } + $final = rtrim($final, ', '); + if ($final === '') return false; + return $final; + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Ident.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Ident.php new file mode 100644 index 000000000..779794a0b --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Ident.php @@ -0,0 +1,24 @@ +<?php + +/** + * Validates based on {ident} CSS grammar production + */ +class HTMLPurifier_AttrDef_CSS_Ident extends HTMLPurifier_AttrDef +{ + + public function validate($string, $config, $context) { + + $string = trim($string); + + // early abort: '' and '0' (strings that convert to false) are invalid + if (!$string) return false; + + $pattern = '/^(-?[A-Za-z_][A-Za-z_\-0-9]*)$/'; + if (!preg_match($pattern, $string)) return false; + return $string; + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/ImportantDecorator.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/ImportantDecorator.php new file mode 100644 index 000000000..4e6b35e5a --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/ImportantDecorator.php @@ -0,0 +1,40 @@ +<?php + +/** + * Decorator which enables !important to be used in CSS values. + */ +class HTMLPurifier_AttrDef_CSS_ImportantDecorator extends HTMLPurifier_AttrDef +{ + public $def, $allow; + + /** + * @param $def Definition to wrap + * @param $allow Whether or not to allow !important + */ + public function __construct($def, $allow = false) { + $this->def = $def; + $this->allow = $allow; + } + /** + * Intercepts and removes !important if necessary + */ + public function validate($string, $config, $context) { + // test for ! and important tokens + $string = trim($string); + $is_important = false; + // :TODO: optimization: test directly for !important and ! important + if (strlen($string) >= 9 && substr($string, -9) === 'important') { + $temp = rtrim(substr($string, 0, -9)); + // use a temp, because we might want to restore important + if (strlen($temp) >= 1 && substr($temp, -1) === '!') { + $string = rtrim(substr($temp, 0, -1)); + $is_important = true; + } + } + $string = $this->def->validate($string, $config, $context); + if ($this->allow && $is_important) $string .= ' !important'; + return $string; + } +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Length.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Length.php new file mode 100644 index 000000000..a07ec5813 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Length.php @@ -0,0 +1,47 @@ +<?php + +/** + * Represents a Length as defined by CSS. + */ +class HTMLPurifier_AttrDef_CSS_Length extends HTMLPurifier_AttrDef +{ + + protected $min, $max; + + /** + * @param HTMLPurifier_Length $max Minimum length, or null for no bound. String is also acceptable. + * @param HTMLPurifier_Length $max Maximum length, or null for no bound. String is also acceptable. + */ + public function __construct($min = null, $max = null) { + $this->min = $min !== null ? HTMLPurifier_Length::make($min) : null; + $this->max = $max !== null ? HTMLPurifier_Length::make($max) : null; + } + + public function validate($string, $config, $context) { + $string = $this->parseCDATA($string); + + // Optimizations + if ($string === '') return false; + if ($string === '0') return '0'; + if (strlen($string) === 1) return false; + + $length = HTMLPurifier_Length::make($string); + if (!$length->isValid()) return false; + + if ($this->min) { + $c = $length->compareTo($this->min); + if ($c === false) return false; + if ($c < 0) return false; + } + if ($this->max) { + $c = $length->compareTo($this->max); + if ($c === false) return false; + if ($c > 0) return false; + } + + return $length->toString(); + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/ListStyle.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/ListStyle.php new file mode 100644 index 000000000..4406868c0 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/ListStyle.php @@ -0,0 +1,78 @@ +<?php + +/** + * Validates shorthand CSS property list-style. + * @warning Does not support url tokens that have internal spaces. + */ +class HTMLPurifier_AttrDef_CSS_ListStyle extends HTMLPurifier_AttrDef +{ + + /** + * Local copy of component validators. + * @note See HTMLPurifier_AttrDef_CSS_Font::$info for a similar impl. + */ + protected $info; + + public function __construct($config) { + $def = $config->getCSSDefinition(); + $this->info['list-style-type'] = $def->info['list-style-type']; + $this->info['list-style-position'] = $def->info['list-style-position']; + $this->info['list-style-image'] = $def->info['list-style-image']; + } + + public function validate($string, $config, $context) { + + // regular pre-processing + $string = $this->parseCDATA($string); + if ($string === '') return false; + + // assumes URI doesn't have spaces in it + $bits = explode(' ', strtolower($string)); // bits to process + + $caught = array(); + $caught['type'] = false; + $caught['position'] = false; + $caught['image'] = false; + + $i = 0; // number of catches + $none = false; + + foreach ($bits as $bit) { + if ($i >= 3) return; // optimization bit + if ($bit === '') continue; + foreach ($caught as $key => $status) { + if ($status !== false) continue; + $r = $this->info['list-style-' . $key]->validate($bit, $config, $context); + if ($r === false) continue; + if ($r === 'none') { + if ($none) continue; + else $none = true; + if ($key == 'image') continue; + } + $caught[$key] = $r; + $i++; + break; + } + } + + if (!$i) return false; + + $ret = array(); + + // construct type + if ($caught['type']) $ret[] = $caught['type']; + + // construct image + if ($caught['image']) $ret[] = $caught['image']; + + // construct position + if ($caught['position']) $ret[] = $caught['position']; + + if (empty($ret)) return false; + return implode(' ', $ret); + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Multiple.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Multiple.php new file mode 100644 index 000000000..4d62a40d7 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Multiple.php @@ -0,0 +1,58 @@ +<?php + +/** + * Framework class for strings that involve multiple values. + * + * Certain CSS properties such as border-width and margin allow multiple + * lengths to be specified. This class can take a vanilla border-width + * definition and multiply it, usually into a max of four. + * + * @note Even though the CSS specification isn't clear about it, inherit + * can only be used alone: it will never manifest as part of a multi + * shorthand declaration. Thus, this class does not allow inherit. + */ +class HTMLPurifier_AttrDef_CSS_Multiple extends HTMLPurifier_AttrDef +{ + + /** + * Instance of component definition to defer validation to. + * @todo Make protected + */ + public $single; + + /** + * Max number of values allowed. + * @todo Make protected + */ + public $max; + + /** + * @param $single HTMLPurifier_AttrDef to multiply + * @param $max Max number of values allowed (usually four) + */ + public function __construct($single, $max = 4) { + $this->single = $single; + $this->max = $max; + } + + public function validate($string, $config, $context) { + $string = $this->parseCDATA($string); + if ($string === '') return false; + $parts = explode(' ', $string); // parseCDATA replaced \r, \t and \n + $length = count($parts); + $final = ''; + for ($i = 0, $num = 0; $i < $length && $num < $this->max; $i++) { + if (ctype_space($parts[$i])) continue; + $result = $this->single->validate($parts[$i], $config, $context); + if ($result !== false) { + $final .= $result . ' '; + $num++; + } + } + if ($final === '') return false; + return rtrim($final); + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Number.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Number.php new file mode 100644 index 000000000..3f99e12ec --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Number.php @@ -0,0 +1,69 @@ +<?php + +/** + * Validates a number as defined by the CSS spec. + */ +class HTMLPurifier_AttrDef_CSS_Number extends HTMLPurifier_AttrDef +{ + + /** + * Bool indicating whether or not only positive values allowed. + */ + protected $non_negative = false; + + /** + * @param $non_negative Bool indicating whether negatives are forbidden + */ + public function __construct($non_negative = false) { + $this->non_negative = $non_negative; + } + + /** + * @warning Some contexts do not pass $config, $context. These + * variables should not be used without checking HTMLPurifier_Length + */ + public function validate($number, $config, $context) { + + $number = $this->parseCDATA($number); + + if ($number === '') return false; + if ($number === '0') return '0'; + + $sign = ''; + switch ($number[0]) { + case '-': + if ($this->non_negative) return false; + $sign = '-'; + case '+': + $number = substr($number, 1); + } + + if (ctype_digit($number)) { + $number = ltrim($number, '0'); + return $number ? $sign . $number : '0'; + } + + // Period is the only non-numeric character allowed + if (strpos($number, '.') === false) return false; + + list($left, $right) = explode('.', $number, 2); + + if ($left === '' && $right === '') return false; + if ($left !== '' && !ctype_digit($left)) return false; + + $left = ltrim($left, '0'); + $right = rtrim($right, '0'); + + if ($right === '') { + return $left ? $sign . $left : '0'; + } elseif (!ctype_digit($right)) { + return false; + } + + return $sign . $left . '.' . $right; + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Percentage.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Percentage.php new file mode 100644 index 000000000..c34b8fc3c --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/Percentage.php @@ -0,0 +1,40 @@ +<?php + +/** + * Validates a Percentage as defined by the CSS spec. + */ +class HTMLPurifier_AttrDef_CSS_Percentage extends HTMLPurifier_AttrDef +{ + + /** + * Instance of HTMLPurifier_AttrDef_CSS_Number to defer number validation + */ + protected $number_def; + + /** + * @param Bool indicating whether to forbid negative values + */ + public function __construct($non_negative = false) { + $this->number_def = new HTMLPurifier_AttrDef_CSS_Number($non_negative); + } + + public function validate($string, $config, $context) { + + $string = $this->parseCDATA($string); + + if ($string === '') return false; + $length = strlen($string); + if ($length === 1) return false; + if ($string[$length - 1] !== '%') return false; + + $number = substr($string, 0, $length - 1); + $number = $this->number_def->validate($number, $config, $context); + + if ($number === false) return false; + return "$number%"; + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/TextDecoration.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/TextDecoration.php new file mode 100644 index 000000000..772c922d8 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/TextDecoration.php @@ -0,0 +1,38 @@ +<?php + +/** + * Validates the value for the CSS property text-decoration + * @note This class could be generalized into a version that acts sort of + * like Enum except you can compound the allowed values. + */ +class HTMLPurifier_AttrDef_CSS_TextDecoration extends HTMLPurifier_AttrDef +{ + + public function validate($string, $config, $context) { + + static $allowed_values = array( + 'line-through' => true, + 'overline' => true, + 'underline' => true, + ); + + $string = strtolower($this->parseCDATA($string)); + + if ($string === 'none') return $string; + + $parts = explode(' ', $string); + $final = ''; + foreach ($parts as $part) { + if (isset($allowed_values[$part])) { + $final .= $part . ' '; + } + } + $final = rtrim($final); + if ($final === '') return false; + return $final; + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/URI.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/URI.php new file mode 100644 index 000000000..c2f767e57 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/CSS/URI.php @@ -0,0 +1,61 @@ +<?php + +/** + * Validates a URI in CSS syntax, which uses url('http://example.com') + * @note While theoretically speaking a URI in a CSS document could + * be non-embedded, as of CSS2 there is no such usage so we're + * generalizing it. This may need to be changed in the future. + * @warning Since HTMLPurifier_AttrDef_CSS blindly uses semicolons as + * the separator, you cannot put a literal semicolon in + * in the URI. Try percent encoding it, in that case. + */ +class HTMLPurifier_AttrDef_CSS_URI extends HTMLPurifier_AttrDef_URI +{ + + public function __construct() { + parent::__construct(true); // always embedded + } + + public function validate($uri_string, $config, $context) { + // parse the URI out of the string and then pass it onto + // the parent object + + $uri_string = $this->parseCDATA($uri_string); + if (strpos($uri_string, 'url(') !== 0) return false; + $uri_string = substr($uri_string, 4); + $new_length = strlen($uri_string) - 1; + if ($uri_string[$new_length] != ')') return false; + $uri = trim(substr($uri_string, 0, $new_length)); + + if (!empty($uri) && ($uri[0] == "'" || $uri[0] == '"')) { + $quote = $uri[0]; + $new_length = strlen($uri) - 1; + if ($uri[$new_length] !== $quote) return false; + $uri = substr($uri, 1, $new_length - 1); + } + + $uri = $this->expandCSSEscape($uri); + + $result = parent::validate($uri, $config, $context); + + if ($result === false) return false; + + // extra sanity check; should have been done by URI + $result = str_replace(array('"', "\\", "\n", "\x0c", "\r"), "", $result); + + // suspicious characters are ()'; we're going to percent encode + // them for safety. + $result = str_replace(array('(', ')', "'"), array('%28', '%29', '%27'), $result); + + // there's an extra bug where ampersands lose their escaping on + // an innerHTML cycle, so a very unlucky query parameter could + // then change the meaning of the URL. Unfortunately, there's + // not much we can do about that... + + return "url(\"$result\")"; + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Clone.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Clone.php new file mode 100644 index 000000000..ce68dbd54 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Clone.php @@ -0,0 +1,28 @@ +<?php + +/** + * Dummy AttrDef that mimics another AttrDef, BUT it generates clones + * with make. + */ +class HTMLPurifier_AttrDef_Clone extends HTMLPurifier_AttrDef +{ + /** + * What we're cloning + */ + protected $clone; + + public function __construct($clone) { + $this->clone = $clone; + } + + public function validate($v, $config, $context) { + return $this->clone->validate($v, $config, $context); + } + + public function make($string) { + return clone $this->clone; + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Enum.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Enum.php new file mode 100644 index 000000000..5d603ebcc --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Enum.php @@ -0,0 +1,65 @@ +<?php + +// Enum = Enumerated +/** + * Validates a keyword against a list of valid values. + * @warning The case-insensitive compare of this function uses PHP's + * built-in strtolower and ctype_lower functions, which may + * cause problems with international comparisons + */ +class HTMLPurifier_AttrDef_Enum extends HTMLPurifier_AttrDef +{ + + /** + * Lookup table of valid values. + * @todo Make protected + */ + public $valid_values = array(); + + /** + * Bool indicating whether or not enumeration is case sensitive. + * @note In general this is always case insensitive. + */ + protected $case_sensitive = false; // values according to W3C spec + + /** + * @param $valid_values List of valid values + * @param $case_sensitive Bool indicating whether or not case sensitive + */ + public function __construct( + $valid_values = array(), $case_sensitive = false + ) { + $this->valid_values = array_flip($valid_values); + $this->case_sensitive = $case_sensitive; + } + + public function validate($string, $config, $context) { + $string = trim($string); + if (!$this->case_sensitive) { + // we may want to do full case-insensitive libraries + $string = ctype_lower($string) ? $string : strtolower($string); + } + $result = isset($this->valid_values[$string]); + + return $result ? $string : false; + } + + /** + * @param $string In form of comma-delimited list of case-insensitive + * valid values. Example: "foo,bar,baz". Prepend "s:" to make + * case sensitive + */ + public function make($string) { + if (strlen($string) > 2 && $string[0] == 's' && $string[1] == ':') { + $string = substr($string, 2); + $sensitive = true; + } else { + $sensitive = false; + } + $values = explode(',', $string); + return new HTMLPurifier_AttrDef_Enum($values, $sensitive); + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Bool.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Bool.php new file mode 100644 index 000000000..e06987eb8 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Bool.php @@ -0,0 +1,28 @@ +<?php + +/** + * Validates a boolean attribute + */ +class HTMLPurifier_AttrDef_HTML_Bool extends HTMLPurifier_AttrDef +{ + + protected $name; + public $minimized = true; + + public function __construct($name = false) {$this->name = $name;} + + public function validate($string, $config, $context) { + if (empty($string)) return false; + return $this->name; + } + + /** + * @param $string Name of attribute + */ + public function make($string) { + return new HTMLPurifier_AttrDef_HTML_Bool($string); + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Class.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Class.php new file mode 100644 index 000000000..370068d97 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Class.php @@ -0,0 +1,34 @@ +<?php + +/** + * Implements special behavior for class attribute (normally NMTOKENS) + */ +class HTMLPurifier_AttrDef_HTML_Class extends HTMLPurifier_AttrDef_HTML_Nmtokens +{ + protected function split($string, $config, $context) { + // really, this twiddle should be lazy loaded + $name = $config->getDefinition('HTML')->doctype->name; + if ($name == "XHTML 1.1" || $name == "XHTML 2.0") { + return parent::split($string, $config, $context); + } else { + return preg_split('/\s+/', $string); + } + } + protected function filter($tokens, $config, $context) { + $allowed = $config->get('Attr.AllowedClasses'); + $forbidden = $config->get('Attr.ForbiddenClasses'); + $ret = array(); + foreach ($tokens as $token) { + if ( + ($allowed === null || isset($allowed[$token])) && + !isset($forbidden[$token]) && + // We need this O(n) check because of PHP's array + // implementation that casts -0 to 0. + !in_array($token, $ret, true) + ) { + $ret[] = $token; + } + } + return $ret; + } +} diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Color.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Color.php new file mode 100644 index 000000000..00d865723 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Color.php @@ -0,0 +1,32 @@ +<?php + +/** + * Validates a color according to the HTML spec. + */ +class HTMLPurifier_AttrDef_HTML_Color extends HTMLPurifier_AttrDef +{ + + public function validate($string, $config, $context) { + + static $colors = null; + if ($colors === null) $colors = $config->get('Core.ColorKeywords'); + + $string = trim($string); + + if (empty($string)) return false; + if (isset($colors[strtolower($string)])) return $colors[$string]; + if ($string[0] === '#') $hex = substr($string, 1); + else $hex = $string; + + $length = strlen($hex); + if ($length !== 3 && $length !== 6) return false; + if (!ctype_xdigit($hex)) return false; + if ($length === 3) $hex = $hex[0].$hex[0].$hex[1].$hex[1].$hex[2].$hex[2]; + + return "#$hex"; + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/FrameTarget.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/FrameTarget.php new file mode 100644 index 000000000..ae6ea7c01 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/FrameTarget.php @@ -0,0 +1,21 @@ +<?php + +/** + * Special-case enum attribute definition that lazy loads allowed frame targets + */ +class HTMLPurifier_AttrDef_HTML_FrameTarget extends HTMLPurifier_AttrDef_Enum +{ + + public $valid_values = false; // uninitialized value + protected $case_sensitive = false; + + public function __construct() {} + + public function validate($string, $config, $context) { + if ($this->valid_values === false) $this->valid_values = $config->get('Attr.AllowedFrameTargets'); + return parent::validate($string, $config, $context); + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/ID.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/ID.php new file mode 100644 index 000000000..0015fa1eb --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/ID.php @@ -0,0 +1,80 @@ +<?php + +/** + * Validates the HTML attribute ID. + * @warning Even though this is the id processor, it + * will ignore the directive Attr:IDBlacklist, since it will only + * go according to the ID accumulator. Since the accumulator is + * automatically generated, it will have already absorbed the + * blacklist. If you're hacking around, make sure you use load()! + */ + +class HTMLPurifier_AttrDef_HTML_ID extends HTMLPurifier_AttrDef +{ + + // selector is NOT a valid thing to use for IDREFs, because IDREFs + // *must* target IDs that exist, whereas selector #ids do not. + + /** + * Determines whether or not we're validating an ID in a CSS + * selector context. + */ + protected $selector; + + public function __construct($selector = false) { + $this->selector = $selector; + } + + public function validate($id, $config, $context) { + + if (!$this->selector && !$config->get('Attr.EnableID')) return false; + + $id = trim($id); // trim it first + + if ($id === '') return false; + + $prefix = $config->get('Attr.IDPrefix'); + if ($prefix !== '') { + $prefix .= $config->get('Attr.IDPrefixLocal'); + // prevent re-appending the prefix + if (strpos($id, $prefix) !== 0) $id = $prefix . $id; + } elseif ($config->get('Attr.IDPrefixLocal') !== '') { + trigger_error('%Attr.IDPrefixLocal cannot be used unless '. + '%Attr.IDPrefix is set', E_USER_WARNING); + } + + if (!$this->selector) { + $id_accumulator =& $context->get('IDAccumulator'); + if (isset($id_accumulator->ids[$id])) return false; + } + + // we purposely avoid using regex, hopefully this is faster + + if (ctype_alpha($id)) { + $result = true; + } else { + if (!ctype_alpha(@$id[0])) return false; + $trim = trim( // primitive style of regexps, I suppose + $id, + 'A..Za..z0..9:-._' + ); + $result = ($trim === ''); + } + + $regexp = $config->get('Attr.IDBlacklistRegexp'); + if ($regexp && preg_match($regexp, $id)) { + return false; + } + + if (!$this->selector && $result) $id_accumulator->add($id); + + // if no change was made to the ID, return the result + // else, return the new id if stripping whitespace made it + // valid, or return false. + return $result ? $id : false; + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Length.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Length.php new file mode 100644 index 000000000..a242f9c23 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Length.php @@ -0,0 +1,41 @@ +<?php + +/** + * Validates the HTML type length (not to be confused with CSS's length). + * + * This accepts integer pixels or percentages as lengths for certain + * HTML attributes. + */ + +class HTMLPurifier_AttrDef_HTML_Length extends HTMLPurifier_AttrDef_HTML_Pixels +{ + + public function validate($string, $config, $context) { + + $string = trim($string); + if ($string === '') return false; + + $parent_result = parent::validate($string, $config, $context); + if ($parent_result !== false) return $parent_result; + + $length = strlen($string); + $last_char = $string[$length - 1]; + + if ($last_char !== '%') return false; + + $points = substr($string, 0, $length - 1); + + if (!is_numeric($points)) return false; + + $points = (int) $points; + + if ($points < 0) return '0%'; + if ($points > 100) return '100%'; + + return ((string) $points) . '%'; + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/LinkTypes.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/LinkTypes.php new file mode 100644 index 000000000..76d25ed08 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/LinkTypes.php @@ -0,0 +1,53 @@ +<?php + +/** + * Validates a rel/rev link attribute against a directive of allowed values + * @note We cannot use Enum because link types allow multiple + * values. + * @note Assumes link types are ASCII text + */ +class HTMLPurifier_AttrDef_HTML_LinkTypes extends HTMLPurifier_AttrDef +{ + + /** Name config attribute to pull. */ + protected $name; + + public function __construct($name) { + $configLookup = array( + 'rel' => 'AllowedRel', + 'rev' => 'AllowedRev' + ); + if (!isset($configLookup[$name])) { + trigger_error('Unrecognized attribute name for link '. + 'relationship.', E_USER_ERROR); + return; + } + $this->name = $configLookup[$name]; + } + + public function validate($string, $config, $context) { + + $allowed = $config->get('Attr.' . $this->name); + if (empty($allowed)) return false; + + $string = $this->parseCDATA($string); + $parts = explode(' ', $string); + + // lookup to prevent duplicates + $ret_lookup = array(); + foreach ($parts as $part) { + $part = strtolower(trim($part)); + if (!isset($allowed[$part])) continue; + $ret_lookup[$part] = true; + } + + if (empty($ret_lookup)) return false; + $string = implode(' ', array_keys($ret_lookup)); + + return $string; + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/MultiLength.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/MultiLength.php new file mode 100644 index 000000000..c72fc76e4 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/MultiLength.php @@ -0,0 +1,41 @@ +<?php + +/** + * Validates a MultiLength as defined by the HTML spec. + * + * A multilength is either a integer (pixel count), a percentage, or + * a relative number. + */ +class HTMLPurifier_AttrDef_HTML_MultiLength extends HTMLPurifier_AttrDef_HTML_Length +{ + + public function validate($string, $config, $context) { + + $string = trim($string); + if ($string === '') return false; + + $parent_result = parent::validate($string, $config, $context); + if ($parent_result !== false) return $parent_result; + + $length = strlen($string); + $last_char = $string[$length - 1]; + + if ($last_char !== '*') return false; + + $int = substr($string, 0, $length - 1); + + if ($int == '') return '*'; + if (!is_numeric($int)) return false; + + $int = (int) $int; + + if ($int < 0) return false; + if ($int == 0) return '0'; + if ($int == 1) return '*'; + return ((string) $int) . '*'; + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Nmtokens.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Nmtokens.php new file mode 100644 index 000000000..aa34120bd --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Nmtokens.php @@ -0,0 +1,52 @@ +<?php + +/** + * Validates contents based on NMTOKENS attribute type. + */ +class HTMLPurifier_AttrDef_HTML_Nmtokens extends HTMLPurifier_AttrDef +{ + + public function validate($string, $config, $context) { + + $string = trim($string); + + // early abort: '' and '0' (strings that convert to false) are invalid + if (!$string) return false; + + $tokens = $this->split($string, $config, $context); + $tokens = $this->filter($tokens, $config, $context); + if (empty($tokens)) return false; + return implode(' ', $tokens); + + } + + /** + * Splits a space separated list of tokens into its constituent parts. + */ + protected function split($string, $config, $context) { + // OPTIMIZABLE! + // do the preg_match, capture all subpatterns for reformulation + + // we don't support U+00A1 and up codepoints or + // escaping because I don't know how to do that with regexps + // and plus it would complicate optimization efforts (you never + // see that anyway). + $pattern = '/(?:(?<=\s)|\A)'. // look behind for space or string start + '((?:--|-?[A-Za-z_])[A-Za-z_\-0-9]*)'. + '(?:(?=\s)|\z)/'; // look ahead for space or string end + preg_match_all($pattern, $string, $matches); + return $matches[1]; + } + + /** + * Template method for removing certain tokens based on arbitrary criteria. + * @note If we wanted to be really functional, we'd do an array_filter + * with a callback. But... we're not. + */ + protected function filter($tokens, $config, $context) { + return $tokens; + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Pixels.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Pixels.php new file mode 100644 index 000000000..4cb2c1b85 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/HTML/Pixels.php @@ -0,0 +1,48 @@ +<?php + +/** + * Validates an integer representation of pixels according to the HTML spec. + */ +class HTMLPurifier_AttrDef_HTML_Pixels extends HTMLPurifier_AttrDef +{ + + protected $max; + + public function __construct($max = null) { + $this->max = $max; + } + + public function validate($string, $config, $context) { + + $string = trim($string); + if ($string === '0') return $string; + if ($string === '') return false; + $length = strlen($string); + if (substr($string, $length - 2) == 'px') { + $string = substr($string, 0, $length - 2); + } + if (!is_numeric($string)) return false; + $int = (int) $string; + + if ($int < 0) return '0'; + + // upper-bound value, extremely high values can + // crash operating systems, see <http://ha.ckers.org/imagecrash.html> + // WARNING, above link WILL crash you if you're using Windows + + if ($this->max !== null && $int > $this->max) return (string) $this->max; + + return (string) $int; + + } + + public function make($string) { + if ($string === '') $max = null; + else $max = (int) $string; + $class = get_class($this); + return new $class($max); + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Integer.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Integer.php new file mode 100644 index 000000000..d59738d2a --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Integer.php @@ -0,0 +1,73 @@ +<?php + +/** + * Validates an integer. + * @note While this class was modeled off the CSS definition, no currently + * allowed CSS uses this type. The properties that do are: widows, + * orphans, z-index, counter-increment, counter-reset. Some of the + * HTML attributes, however, find use for a non-negative version of this. + */ +class HTMLPurifier_AttrDef_Integer extends HTMLPurifier_AttrDef +{ + + /** + * Bool indicating whether or not negative values are allowed + */ + protected $negative = true; + + /** + * Bool indicating whether or not zero is allowed + */ + protected $zero = true; + + /** + * Bool indicating whether or not positive values are allowed + */ + protected $positive = true; + + /** + * @param $negative Bool indicating whether or not negative values are allowed + * @param $zero Bool indicating whether or not zero is allowed + * @param $positive Bool indicating whether or not positive values are allowed + */ + public function __construct( + $negative = true, $zero = true, $positive = true + ) { + $this->negative = $negative; + $this->zero = $zero; + $this->positive = $positive; + } + + public function validate($integer, $config, $context) { + + $integer = $this->parseCDATA($integer); + if ($integer === '') return false; + + // we could possibly simply typecast it to integer, but there are + // certain fringe cases that must not return an integer. + + // clip leading sign + if ( $this->negative && $integer[0] === '-' ) { + $digits = substr($integer, 1); + if ($digits === '0') $integer = '0'; // rm minus sign for zero + } elseif( $this->positive && $integer[0] === '+' ) { + $digits = $integer = substr($integer, 1); // rm unnecessary plus + } else { + $digits = $integer; + } + + // test if it's numeric + if (!ctype_digit($digits)) return false; + + // perform scope tests + if (!$this->zero && $integer == 0) return false; + if (!$this->positive && $integer > 0) return false; + if (!$this->negative && $integer < 0) return false; + + return $integer; + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Lang.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Lang.php new file mode 100644 index 000000000..10e6da56d --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Lang.php @@ -0,0 +1,73 @@ +<?php + +/** + * Validates the HTML attribute lang, effectively a language code. + * @note Built according to RFC 3066, which obsoleted RFC 1766 + */ +class HTMLPurifier_AttrDef_Lang extends HTMLPurifier_AttrDef +{ + + public function validate($string, $config, $context) { + + $string = trim($string); + if (!$string) return false; + + $subtags = explode('-', $string); + $num_subtags = count($subtags); + + if ($num_subtags == 0) return false; // sanity check + + // process primary subtag : $subtags[0] + $length = strlen($subtags[0]); + switch ($length) { + case 0: + return false; + case 1: + if (! ($subtags[0] == 'x' || $subtags[0] == 'i') ) { + return false; + } + break; + case 2: + case 3: + if (! ctype_alpha($subtags[0]) ) { + return false; + } elseif (! ctype_lower($subtags[0]) ) { + $subtags[0] = strtolower($subtags[0]); + } + break; + default: + return false; + } + + $new_string = $subtags[0]; + if ($num_subtags == 1) return $new_string; + + // process second subtag : $subtags[1] + $length = strlen($subtags[1]); + if ($length == 0 || ($length == 1 && $subtags[1] != 'x') || $length > 8 || !ctype_alnum($subtags[1])) { + return $new_string; + } + if (!ctype_lower($subtags[1])) $subtags[1] = strtolower($subtags[1]); + + $new_string .= '-' . $subtags[1]; + if ($num_subtags == 2) return $new_string; + + // process all other subtags, index 2 and up + for ($i = 2; $i < $num_subtags; $i++) { + $length = strlen($subtags[$i]); + if ($length == 0 || $length > 8 || !ctype_alnum($subtags[$i])) { + return $new_string; + } + if (!ctype_lower($subtags[$i])) { + $subtags[$i] = strtolower($subtags[$i]); + } + $new_string .= '-' . $subtags[$i]; + } + + return $new_string; + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Switch.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Switch.php new file mode 100644 index 000000000..c9e3ed193 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Switch.php @@ -0,0 +1,34 @@ +<?php + +/** + * Decorator that, depending on a token, switches between two definitions. + */ +class HTMLPurifier_AttrDef_Switch +{ + + protected $tag; + protected $withTag, $withoutTag; + + /** + * @param string $tag Tag name to switch upon + * @param HTMLPurifier_AttrDef $with_tag Call if token matches tag + * @param HTMLPurifier_AttrDef $without_tag Call if token doesn't match, or there is no token + */ + public function __construct($tag, $with_tag, $without_tag) { + $this->tag = $tag; + $this->withTag = $with_tag; + $this->withoutTag = $without_tag; + } + + public function validate($string, $config, $context) { + $token = $context->get('CurrentToken', true); + if (!$token || $token->name !== $this->tag) { + return $this->withoutTag->validate($string, $config, $context); + } else { + return $this->withTag->validate($string, $config, $context); + } + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Text.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Text.php new file mode 100644 index 000000000..c6216cc53 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/Text.php @@ -0,0 +1,15 @@ +<?php + +/** + * Validates arbitrary text according to the HTML spec. + */ +class HTMLPurifier_AttrDef_Text extends HTMLPurifier_AttrDef +{ + + public function validate($string, $config, $context) { + return $this->parseCDATA($string); + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI.php new file mode 100644 index 000000000..c2b684671 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI.php @@ -0,0 +1,77 @@ +<?php + +/** + * Validates a URI as defined by RFC 3986. + * @note Scheme-specific mechanics deferred to HTMLPurifier_URIScheme + */ +class HTMLPurifier_AttrDef_URI extends HTMLPurifier_AttrDef +{ + + protected $parser; + protected $embedsResource; + + /** + * @param $embeds_resource_resource Does the URI here result in an extra HTTP request? + */ + public function __construct($embeds_resource = false) { + $this->parser = new HTMLPurifier_URIParser(); + $this->embedsResource = (bool) $embeds_resource; + } + + public function make($string) { + $embeds = ($string === 'embedded'); + return new HTMLPurifier_AttrDef_URI($embeds); + } + + public function validate($uri, $config, $context) { + + if ($config->get('URI.Disable')) return false; + + $uri = $this->parseCDATA($uri); + + // parse the URI + $uri = $this->parser->parse($uri); + if ($uri === false) return false; + + // add embedded flag to context for validators + $context->register('EmbeddedURI', $this->embedsResource); + + $ok = false; + do { + + // generic validation + $result = $uri->validate($config, $context); + if (!$result) break; + + // chained filtering + $uri_def = $config->getDefinition('URI'); + $result = $uri_def->filter($uri, $config, $context); + if (!$result) break; + + // scheme-specific validation + $scheme_obj = $uri->getSchemeObj($config, $context); + if (!$scheme_obj) break; + if ($this->embedsResource && !$scheme_obj->browsable) break; + $result = $scheme_obj->validate($uri, $config, $context); + if (!$result) break; + + // Post chained filtering + $result = $uri_def->postFilter($uri, $config, $context); + if (!$result) break; + + // survived gauntlet + $ok = true; + + } while (false); + + $context->destroy('EmbeddedURI'); + if (!$ok) return false; + + // back to string + return $uri->toString(); + + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/Email.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/Email.php new file mode 100644 index 000000000..bfee9d166 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/Email.php @@ -0,0 +1,17 @@ +<?php + +abstract class HTMLPurifier_AttrDef_URI_Email extends HTMLPurifier_AttrDef +{ + + /** + * Unpacks a mailbox into its display-name and address + */ + function unpack($string) { + // needs to be implemented + } + +} + +// sub-implementations + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/Email/SimpleCheck.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/Email/SimpleCheck.php new file mode 100644 index 000000000..94c715ab4 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/Email/SimpleCheck.php @@ -0,0 +1,21 @@ +<?php + +/** + * Primitive email validation class based on the regexp found at + * http://www.regular-expressions.info/email.html + */ +class HTMLPurifier_AttrDef_URI_Email_SimpleCheck extends HTMLPurifier_AttrDef_URI_Email +{ + + public function validate($string, $config, $context) { + // no support for named mailboxes i.e. "Bob <bob@example.com>" + // that needs more percent encoding to be done + if ($string == '') return false; + $string = trim($string); + $result = preg_match('/^[A-Z0-9._%-]+@[A-Z0-9.-]+\.[A-Z]{2,4}$/i', $string); + return $result ? $string : false; + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/Host.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/Host.php new file mode 100644 index 000000000..125decb2d --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/Host.php @@ -0,0 +1,101 @@ +<?php + +/** + * Validates a host according to the IPv4, IPv6 and DNS (future) specifications. + */ +class HTMLPurifier_AttrDef_URI_Host extends HTMLPurifier_AttrDef +{ + + /** + * Instance of HTMLPurifier_AttrDef_URI_IPv4 sub-validator + */ + protected $ipv4; + + /** + * Instance of HTMLPurifier_AttrDef_URI_IPv6 sub-validator + */ + protected $ipv6; + + public function __construct() { + $this->ipv4 = new HTMLPurifier_AttrDef_URI_IPv4(); + $this->ipv6 = new HTMLPurifier_AttrDef_URI_IPv6(); + } + + public function validate($string, $config, $context) { + $length = strlen($string); + // empty hostname is OK; it's usually semantically equivalent: + // the default host as defined by a URI scheme is used: + // + // If the URI scheme defines a default for host, then that + // default applies when the host subcomponent is undefined + // or when the registered name is empty (zero length). + if ($string === '') return ''; + if ($length > 1 && $string[0] === '[' && $string[$length-1] === ']') { + //IPv6 + $ip = substr($string, 1, $length - 2); + $valid = $this->ipv6->validate($ip, $config, $context); + if ($valid === false) return false; + return '['. $valid . ']'; + } + + // need to do checks on unusual encodings too + $ipv4 = $this->ipv4->validate($string, $config, $context); + if ($ipv4 !== false) return $ipv4; + + // A regular domain name. + + // This doesn't match I18N domain names, but we don't have proper IRI support, + // so force users to insert Punycode. + + // The productions describing this are: + $a = '[a-z]'; // alpha + $an = '[a-z0-9]'; // alphanum + $and = '[a-z0-9-]'; // alphanum | "-" + // domainlabel = alphanum | alphanum *( alphanum | "-" ) alphanum + $domainlabel = "$an($and*$an)?"; + // toplabel = alpha | alpha *( alphanum | "-" ) alphanum + $toplabel = "$a($and*$an)?"; + // hostname = *( domainlabel "." ) toplabel [ "." ] + if (preg_match("/^($domainlabel\.)*$toplabel\.?$/i", $string)) { + return $string; + } + + // If we have Net_IDNA2 support, we can support IRIs by + // punycoding them. (This is the most portable thing to do, + // since otherwise we have to assume browsers support + + if ($config->get('Core.EnableIDNA')) { + $idna = new Net_IDNA2(array('encoding' => 'utf8', 'overlong' => false, 'strict' => true)); + // we need to encode each period separately + $parts = explode('.', $string); + try { + $new_parts = array(); + foreach ($parts as $part) { + $encodable = false; + for ($i = 0, $c = strlen($part); $i < $c; $i++) { + if (ord($part[$i]) > 0x7a) { + $encodable = true; + break; + } + } + if (!$encodable) { + $new_parts[] = $part; + } else { + $new_parts[] = $idna->encode($part); + } + } + $string = implode('.', $new_parts); + if (preg_match("/^($domainlabel\.)*$toplabel\.?$/i", $string)) { + return $string; + } + } catch (Exception $e) { + // XXX error reporting + } + } + + return false; + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/IPv4.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/IPv4.php new file mode 100644 index 000000000..ec4cf591b --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/IPv4.php @@ -0,0 +1,39 @@ +<?php + +/** + * Validates an IPv4 address + * @author Feyd @ forums.devnetwork.net (public domain) + */ +class HTMLPurifier_AttrDef_URI_IPv4 extends HTMLPurifier_AttrDef +{ + + /** + * IPv4 regex, protected so that IPv6 can reuse it + */ + protected $ip4; + + public function validate($aIP, $config, $context) { + + if (!$this->ip4) $this->_loadRegex(); + + if (preg_match('#^' . $this->ip4 . '$#s', $aIP)) + { + return $aIP; + } + + return false; + + } + + /** + * Lazy load function to prevent regex from being stuffed in + * cache. + */ + protected function _loadRegex() { + $oct = '(?:25[0-5]|2[0-4][0-9]|1[0-9]{2}|[1-9][0-9]|[0-9])'; // 0-255 + $this->ip4 = "(?:{$oct}\\.{$oct}\\.{$oct}\\.{$oct})"; + } + +} + +// vim: et sw=4 sts=4 diff --git a/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/IPv6.php b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/IPv6.php new file mode 100644 index 000000000..9454e9be5 --- /dev/null +++ b/lib/htmlpurifier/library/HTMLPurifier/AttrDef/URI/IPv6.php @@ -0,0 +1,99 @@ +<?php + +/** + * Validates an IPv6 address. + * @author Feyd @ forums.devnetwork.net (public domain) + * @note This function requires brackets to have been removed from address + * in URI. + */ +class HTMLPurifier_AttrDef_URI_IPv6 extends HTMLPurifier_AttrDef_URI_IPv4 +{ + + public function validate($aIP, $config, $context) { + + if (!$this->ip4) $this->_loadRegex(); + + $original = $aIP; + + $hex = '[0-9a-fA-F]'; + $blk = '(?:' . $hex . '{1,4})'; + $pre = '(?:/(?:12[0-8]|1[0-1][0-9]|[1-9][0-9]|[0-9]))'; // /0 - /128 + + // prefix check + if (strpos($aIP, '/') !== false) + { + if (preg_match('#' . $pre . '$#s', $aIP, $find)) + { + $aIP = substr($aIP, 0, 0-strlen($find[0])); + unset($find); + } + else + { + return false; + } + } + + // IPv4-compatiblity check + if (preg_match('#(?<=:'.')' . $this->ip4 . '$#s', $aIP, $find)) + { + $aIP = substr($aIP, 0, 0-strlen($find[0])); + $ip = explode('.', $find[0]); + $ip = array_map('dechex', $ip); + $aIP .= $ip[0] . $ip[1] . ':' . $ip[2] . $ip[3]; + unset($find, $ip); + } + + // compression check + $aIP = explode('::', $aIP); + $c = count($aIP); + if ($c > 2) + { + return false; + } + elseif ($c == 2) + { + list($first, $second) = $aIP; + $first = explode(':', $first); + $second = explode(':', $second); + + if (count($first) + count($second) > 8) + { + return false; + } + + while(count($first) < 8) + { + array_push($first, '0'); + } + + array_splice($first, 8 - count($second), 8, $second); + $aIP = $first; + unset($first,$second); + } + else + { + $aIP = explode(':', $aIP[0]); + } + $c = count($aIP); + + if ($c != 8) + { + return false; + } + + // All the pieces should be 16-bit hex strings. Are they? + foreach ($aIP as $piece) + { + if (!preg_match('#^[0-9a-fA-F]{4}$#s', sprintf('%04s', $piece))) + { + return false; + } + } + + return $original; + + } + +} + +// vim: et sw=4 sts=4 |