diff options
| author | Andreas Gohr <andi@splitbrain.org> | 2008-09-12 22:23:34 +0200 |
|---|---|---|
| committer | Andreas Gohr <andi@splitbrain.org> | 2008-09-12 22:23:34 +0200 |
| commit | 8e39ccf504d73d5d12e07bd879c5a696d5ceac3d (patch) | |
| tree | e98977e1a3b2a5f3886da350b95c4a196ea04449 /inc/geshi.php | |
| parent | 535bed707571573f6670f91094e3c8eed6d9420c (diff) | |
| download | rpg-8e39ccf504d73d5d12e07bd879c5a696d5ceac3d.tar.gz rpg-8e39ccf504d73d5d12e07bd879c5a696d5ceac3d.tar.bz2 | |
GeSHi update to 1.0.8
darcs-hash:20080912202334-7ad00-ee4e69ad5516c38729955e8919d55f3b940caf1a.gz
Diffstat (limited to 'inc/geshi.php')
| -rw-r--r-- | inc/geshi.php | 2943 |
1 files changed, 2045 insertions, 898 deletions
diff --git a/inc/geshi.php b/inc/geshi.php index 2fbf35a77..a05e1685f 100644 --- a/inc/geshi.php +++ b/inc/geshi.php @@ -41,7 +41,7 @@ // /** The version of this GeSHi file */ -define('GESHI_VERSION', '1.0.7.22'); +define('GESHI_VERSION', '1.0.8'); // Define the root directory for the GeSHi code tree if (!defined('GESHI_ROOT')) { @@ -68,6 +68,22 @@ define('GESHI_HEADER_NONE', 0); define('GESHI_HEADER_DIV', 1); /** Use a "pre" to surround the source */ define('GESHI_HEADER_PRE', 2); +/** Use a pre to wrap lines when line numbers are enabled or to wrap the whole code. */ +define('GESHI_HEADER_PRE_VALID', 3); +/** + * Use a "table" to surround the source: + * + * <table> + * <thead><tr><td colspan="2">$header</td></tr></thead> + * <tbody><tr><td><pre>$linenumbers</pre></td><td><pre>$code></pre></td></tr></tbody> + * <tfooter><tr><td colspan="2">$footer</td></tr></tfoot> + * </table> + * + * this is essentially only a workaround for Firefox, see sf#1651996 or take a look at + * https://bugzilla.mozilla.org/show_bug.cgi?id=365805 + * @note when linenumbers are disabled this is essentially the same as GESHI_HEADER_PRE + */ +define('GESHI_HEADER_PRE_TABLE', 4); // Capatalisation constants /** Lowercase keywords found */ @@ -102,7 +118,7 @@ define('GESHI_END_IMPORTANT', '<END GeSHi>'); /** Strict mode never applies (this is the most common) */ define('GESHI_NEVER', 0); /** Strict mode *might* apply, and can be enabled or - disabled by {@link GeSHi::enable_strict_mode()} */ + disabled by {@link GeSHi->enable_strict_mode()} */ define('GESHI_MAYBE', 1); /** Strict mode always applies */ define('GESHI_ALWAYS', 2); @@ -128,6 +144,75 @@ define('GESHI_CLASS', 5); /** Used in language files to mark comments */ define('GESHI_COMMENTS', 0); +/** Used to work around missing PHP features **/ +define('GESHI_PHP_PRE_433', !(version_compare(PHP_VERSION, '4.3.3') === 1)); + +/** make sure we can call stripos **/ +if (!function_exists('stripos')) { + // the offset param of preg_match is not supported below PHP 4.3.3 + if (GESHI_PHP_PRE_433) { + /** + * @ignore + */ + function stripos($haystack, $needle, $offset = null) { + if (!is_null($offset)) { + $haystack = substr($haystack, $offset); + } + if (preg_match('/'. preg_quote($needle, '/') . '/', $haystack, $match, PREG_OFFSET_CAPTURE)) { + return $match[0][1]; + } + return false; + } + } + else { + /** + * @ignore + */ + function stripos($haystack, $needle, $offset = null) { + if (preg_match('/'. preg_quote($needle, '/') . '/', $haystack, $match, PREG_OFFSET_CAPTURE, $offset)) { + return $match[0][1]; + } + return false; + } + } +} + +/** some old PHP / PCRE subpatterns only support up to xxx subpatterns in + regular expressions. Set this to false if your PCRE lib is up to date + @see GeSHi->optimize_regexp_list() + TODO: are really the subpatterns the culprit or the overall length of the pattern? + **/ +define('GESHI_MAX_PCRE_SUBPATTERNS', 500); + +//Number format specification +/** Basic number format for integers */ +define('GESHI_NUMBER_INT_BASIC', 1); //Default integers \d+ +/** Enhanced number format for integers like seen in C */ +define('GESHI_NUMBER_INT_CSTYLE', 2); //Default C-Style \d+[lL]? +/** Number format to highlight binary numbers with a suffix "b" */ +define('GESHI_NUMBER_BIN_SUFFIX', 16); //[01]+[bB] +/** Number format to highlight binary numbers with a prefix % */ +define('GESHI_NUMBER_BIN_PREFIX_PERCENT', 32); //%[01]+ +/** Number format to highlight binary numbers with a prefix 0b (C) */ +define('GESHI_NUMBER_BIN_PREFIX_0B', 64); //0b[01]+ +/** Number format to highlight octal numbers with a leading zero */ +define('GESHI_NUMBER_OCT_PREFIX', 256); //0[0-7]+ +/** Number format to highlight octal numbers with a suffix of o */ +define('GESHI_NUMBER_OCT_SUFFIX', 512); //[0-7]+[oO] +/** Number format to highlight hex numbers with a prefix 0x */ +define('GESHI_NUMBER_HEX_PREFIX', 4096); //0x[0-9a-fA-F]+ +/** Number format to highlight hex numbers with a suffix of h */ +define('GESHI_NUMBER_HEX_SUFFIX', 8192); //[0-9][0-9a-fA-F]*h +/** Number format to highlight floating-point numbers without support for scientific notation */ +define('GESHI_NUMBER_FLT_NONSCI', 65536); //\d+\.\d+ +/** Number format to highlight floating-point numbers without support for scientific notation */ +define('GESHI_NUMBER_FLT_NONSCI_F', 131072); //\d+(\.\d+)?f +/** Number format to highlight floating-point numbers with support for scientific notation (E) and optional leading zero */ +define('GESHI_NUMBER_FLT_SCI_SHORT', 262144); //\.\d+e\d+ +/** Number format to highlight floating-point numbers with support for scientific notation (E) and required leading digit */ +define('GESHI_NUMBER_FLT_SCI_ZERO', 524288); //\d+(\.\d+)?e\d+ +//Custom formats are passed by RX array + // Error detection - use these to analyse faults /** No sourcecode to highlight was specified * @deprecated @@ -137,9 +222,9 @@ define('GESHI_ERROR_NO_INPUT', 1); define('GESHI_ERROR_NO_SUCH_LANG', 2); /** GeSHi could not open a file for reading (generally a language file) */ define('GESHI_ERROR_FILE_NOT_READABLE', 3); -/** The header type passed to {@link GeSHi::set_header_type()} was invalid */ +/** The header type passed to {@link GeSHi->set_header_type()} was invalid */ define('GESHI_ERROR_INVALID_HEADER_TYPE', 4); -/** The line number type passed to {@link GeSHi::enable_line_numbers()} was invalid */ +/** The line number type passed to {@link GeSHi->enable_line_numbers()} was invalid */ define('GESHI_ERROR_INVALID_LINE_NUMBER_TYPE', 5); /**#@-*/ @@ -324,7 +409,7 @@ class GeSHi { * Styles of extra-highlighted lines * @var string */ - var $highlight_extra_lines_style = 'color: #cc0; background-color: #ffc;'; + var $highlight_extra_lines_style = 'background-color: #ffc;'; /** * The line ending @@ -344,13 +429,13 @@ class GeSHi { * The overall style for this code block * @var string */ - var $overall_style = ''; + var $overall_style = 'font-family:monospace;'; /** * The style for the actual code * @var string */ - var $code_style = 'font-family: \'Courier New\', Courier, monospace; font-weight: normal; font-style: normal;'; + var $code_style = 'font-family: monospace; font-weight: normal; font-style: normal; margin:0; padding:0; background:inherit;'; /** * The overall class for this code block @@ -377,7 +462,13 @@ class GeSHi { var $line_style2 = 'font-weight: bold;'; /** - * Flag for how line nubmers are displayed + * Style for line numbers when GESHI_HEADER_PRE_TABLE is chosen + * @var string + */ + var $table_linenumber_style = 'width:1px;font-weight: normal;text-align:right;margin:0;padding:0 2px;'; + + /** + * Flag for how line numbers are displayed * @var boolean */ var $line_numbers = GESHI_NO_LINE_NUMBERS; @@ -415,10 +506,10 @@ class GeSHi { /** * The encoding to use for entity encoding - * NOTE: no longer used + * NOTE: Used with Escape Char Sequences to fix UTF-8 handling (cf. SF#2037598) * @var string */ - var $encoding = 'ISO-8859-1'; + var $encoding = 'utf-8'; /** * Should keywords be linked? @@ -433,6 +524,44 @@ class GeSHi { */ var $loaded_language = ''; + /** + * Wether the caches needed for parsing are built or not + * + * @var bool + * @since 1.0.8 + */ + var $parse_cache_built = false; + + /** + * Work around for Suhosin Patch with disabled /e modifier + * + * Note from suhosins author in config file: + * <blockquote> + * The /e modifier inside <code>preg_replace()</code> allows code execution. + * Often it is the cause for remote code execution exploits. It is wise to + * deactivate this feature and test where in the application it is used. + * The developer using the /e modifier should be made aware that he should + * use <code>preg_replace_callback()</code> instead + * </blockquote> + * + * @var array + * @since 1.0.8 + */ + var $_kw_replace_group = 0; + var $_rx_key = 0; + + /** + * some "callback parameters" for handle_multiline_regexps + * + * @since 1.0.8 + * @access private + * @var string + */ + var $_hmr_before = ''; + var $_hmr_replace = ''; + var $_hmr_after = ''; + var $_hmr_key = 0; + /**#@-*/ /** @@ -446,13 +575,17 @@ class GeSHi { * should be automatically set correctly. If you have * renamed the language directory however, you will * still need to set the path using this parameter or - * {@link GeSHi::set_language_path()} + * {@link GeSHi->set_language_path()} * @since 1.0.0 */ - function GeSHi($source, $language, $path = '') { - $this->set_source($source); + function GeSHi($source = '', $language = '', $path = '') { + if (!empty($source)) { + $this->set_source($source); + } + if (!empty($language)) { + $this->set_language($language); + } $this->set_language_path($path); - $this->set_language($language); } /** @@ -464,15 +597,17 @@ class GeSHi { */ function error() { if ($this->error) { - $msg = $this->error_messages[$this->error]; + //Put some template variables for debugging here ... $debug_tpl_vars = array( '{LANGUAGE}' => $this->language, '{PATH}' => $this->language_path ); - foreach ($debug_tpl_vars as $tpl => $var) { - $msg = str_replace($tpl, $var, $msg); - } - return "<br /><strong>GeSHi Error:</strong> $msg (code $this->error)<br />"; + $msg = str_replace( + array_keys($debug_tpl_vars), + array_values($debug_tpl_vars), + $this->error_messages[$this->error]); + + return "<br /><strong>GeSHi Error:</strong> $msg (code {$this->error})<br />"; } return false; } @@ -505,21 +640,40 @@ class GeSHi { /** * Sets the language for this object * + * @note since 1.0.8 this function won't reset language-settings by default anymore! + * if you need this set $force_reset = true + * * @param string The name of the language to use * @since 1.0.0 */ - function set_language($language) { - $this->error = false; - $this->strict_mode = GESHI_NEVER; + function set_language($language, $force_reset = false) { + if ($force_reset) { + $this->loaded_language = false; + } + //Clean up the language name to prevent malicious code injection $language = preg_replace('#[^a-zA-Z0-9\-_]#', '', $language); - $this->language = strtolower($language); - $file_name = $this->language_path . $this->language . '.php'; + $language = strtolower($language); + + //Retreive the full filename + $file_name = $this->language_path . $language . '.php'; + if ($file_name == $this->loaded_language) { + // this language is already loaded! + return; + } + + $this->language = $language; + + $this->error = false; + $this->strict_mode = GESHI_NEVER; + + //Check if we can read the desired file if (!is_readable($file_name)) { $this->error = GESHI_ERROR_NO_SUCH_LANG; return; } + // Load the language for parsing $this->load_language($file_name); } @@ -558,15 +712,15 @@ class GeSHi { * @since 1.0.0 */ function set_header_type($type) { - if (GESHI_HEADER_DIV != $type && GESHI_HEADER_PRE != $type && GESHI_HEADER_NONE != $type) { + //Check if we got a valid header type + if (!in_array($type, array(GESHI_HEADER_NONE, GESHI_HEADER_DIV, + GESHI_HEADER_PRE, GESHI_HEADER_PRE_VALID, GESHI_HEADER_PRE_TABLE))) { $this->error = GESHI_ERROR_INVALID_HEADER_TYPE; return; } + + //Set that new header type $this->header_type = $type; - // Set a default overall style if the header is a <div> - if (GESHI_HEADER_DIV == $type && !$this->overall_style) { - $this->overall_style = 'font-family: monospace;'; - } } /** @@ -581,8 +735,7 @@ class GeSHi { function set_overall_style($style, $preserve_defaults = false) { if (!$preserve_defaults) { $this->overall_style = $style; - } - else { + } else { $this->overall_style .= $style; } } @@ -639,8 +792,7 @@ class GeSHi { function set_code_style($style, $preserve_defaults = false) { if (!$preserve_defaults) { $this->code_style = $style; - } - else { + } else { $this->code_style .= $style; } } @@ -658,15 +810,17 @@ class GeSHi { * @since 1.0.2 */ function set_line_style($style1, $style2 = '', $preserve_defaults = false) { + //Check if we got 2 or three parameters if (is_bool($style2)) { $preserve_defaults = $style2; $style2 = ''; } + + //Actually set the new styles if (!$preserve_defaults) { $this->line_style1 = $style1; $this->line_style2 = $style2; - } - else { + } else { $this->line_style1 .= $style1; $this->line_style2 .= $style2; } @@ -712,7 +866,7 @@ class GeSHi { } /** - * Get current setting for multiline spans, see GeSHi:: + * Get current setting for multiline spans, see GeSHi->enable_multiline_span(). * * @see enable_multiline_span * @return bool @@ -733,13 +887,15 @@ class GeSHi { * @since 1.0.0 */ function set_keyword_group_style($key, $style, $preserve_defaults = false) { + //Set the style for this keyword group if (!$preserve_defaults) { $this->language_data['STYLES']['KEYWORDS'][$key] = $style; - } - else { + } else { $this->language_data['STYLES']['KEYWORDS'][$key] .= $style; } - if(!isset($this->lexic_permissions['KEYWORDS'][$key])) { + + //Update the lexic permissions + if (!isset($this->lexic_permissions['KEYWORDS'][$key])) { $this->lexic_permissions['KEYWORDS'][$key] = true; } } @@ -769,8 +925,7 @@ class GeSHi { function set_comments_style($key, $style, $preserve_defaults = false) { if (!$preserve_defaults) { $this->language_data['STYLES']['COMMENTS'][$key] = $style; - } - else { + } else { $this->language_data['STYLES']['COMMENTS'][$key] .= $style; } } @@ -799,8 +954,7 @@ class GeSHi { function set_escape_characters_style($style, $preserve_defaults = false) { if (!$preserve_defaults) { $this->language_data['STYLES']['ESCAPE_CHAR'][0] = $style; - } - else { + } else { $this->language_data['STYLES']['ESCAPE_CHAR'][0] .= $style; } } @@ -832,8 +986,7 @@ class GeSHi { function set_brackets_style($style, $preserve_defaults = false) { if (!$preserve_defaults) { $this->language_data['STYLES']['BRACKETS'][0] = $style; - } - else { + } else { $this->language_data['STYLES']['BRACKETS'][0] .= $style; } } @@ -864,14 +1017,15 @@ class GeSHi { * @since 1.0.1 */ function set_symbols_style($style, $preserve_defaults = false, $group = 0) { + // Update the style of symbols if (!$preserve_defaults) { $this->language_data['STYLES']['SYMBOLS'][$group] = $style; - } - else { + } else { $this->language_data['STYLES']['SYMBOLS'][$group] .= $style; } + // For backward compatibility - if(0 == $group) { + if (0 == $group) { $this->set_brackets_style ($style, $preserve_defaults); } } @@ -883,7 +1037,9 @@ class GeSHi { * @since 1.0.0 */ function set_symbols_highlighting($flag) { + // Update lexic permissions for this symbol group $this->lexic_permissions['SYMBOLS'] = ($flag) ? true : false; + // For backward compatibility $this->set_brackets_highlighting ($flag); } @@ -901,8 +1057,7 @@ class GeSHi { function set_strings_style($style, $preserve_defaults = false) { if (!$preserve_defaults) { $this->language_data['STYLES']['STRINGS'][0] = $style; - } - else { + } else { $this->language_data['STYLES']['STRINGS'][0] .= $style; } } @@ -930,8 +1085,7 @@ class GeSHi { function set_numbers_style($style, $preserve_defaults = false) { if (!$preserve_defaults) { $this->language_data['STYLES']['NUMBERS'][0] = $style; - } - else { + } else { $this->language_data['STYLES']['NUMBERS'][0] .= $style; } } @@ -962,8 +1116,7 @@ class GeSHi { function set_methods_style($key, $style, $preserve_defaults = false) { if (!$preserve_defaults) { $this->language_data['STYLES']['METHODS'][$key] = $style; - } - else { + } else { $this->language_data['STYLES']['METHODS'][$key] .= $style; } } @@ -991,8 +1144,7 @@ class GeSHi { function set_regexps_style($key, $style, $preserve_defaults = false) { if (!$preserve_defaults) { $this->language_data['STYLES']['REGEXPS'][$key] = $style; - } - else { + } else { $this->language_data['STYLES']['REGEXPS'][$key] .= $style; } } @@ -1030,7 +1182,7 @@ class GeSHi { * @since 1.0.1 */ function set_case_keywords($case) { - if(in_array($case, array( + if (in_array($case, array( GESHI_CAPS_NO_CHANGE, GESHI_CAPS_UPPER, GESHI_CAPS_LOWER))) { $this->language_data['CASE_KEYWORDS'] = $case; } @@ -1046,8 +1198,9 @@ class GeSHi { */ function set_tab_width($width) { $this->tab_width = intval($width); + //Check if it fit's the constraints: - if($this->tab_width < 1) { + if ($this->tab_width < 1) { //Return it to the default $this->tab_width = 8; } @@ -1089,7 +1242,7 @@ class GeSHi { */ function enable_strict_mode($mode = true) { if (GESHI_MAYBE == $this->language_data['STRICT_MODE_APPLIES']) { - $this->strict_mode = ($mode) ? true : false; + $this->strict_mode = ($mode) ? GESHI_ALWAYS : GESHI_NEVER; } } @@ -1121,11 +1274,11 @@ class GeSHi { foreach ($value as $k => $v) { $this->lexic_permissions[$key][$k] = $flag; } - } - else { + } else { $this->lexic_permissions[$key] = $flag; } } + // Context blocks $this->enable_important_blocks = $flag; } @@ -1155,6 +1308,7 @@ class GeSHi { 'caddcl' => array(), 'cadlisp' => array(), 'cdfg' => array('cdfg'), + 'cobol' => array('cbl'), 'cpp' => array('cpp', 'h', 'hpp'), 'csharp' => array(), 'css' => array('css'), @@ -1164,6 +1318,8 @@ class GeSHi { 'html4strict' => array('html', 'htm'), 'java' => array('java'), 'javascript' => array('js'), + 'klonec' => array('kl1'), + 'klonecpp' => array('klx'), 'lisp' => array('lisp'), 'lua' => array('lua'), 'mpasm' => array(), @@ -1214,8 +1370,7 @@ class GeSHi { if (is_readable($file_name)) { $this->set_source(file_get_contents($file_name)); $this->set_language($this->get_language_name_from_extension(substr(strrchr($file_name, '.'), 1), $lookup)); - } - else { + } else { $this->error = GESHI_ERROR_FILE_NOT_READABLE; } } @@ -1228,7 +1383,15 @@ class GeSHi { * @since 1.0.0 */ function add_keyword($key, $word) { - $this->language_data['KEYWORDS'][$key][] = $word; + if (!in_array($word, $this->language_data['KEYWORDS'][$key])) { + $this->language_data['KEYWORDS'][$key][] = $word; + + //NEW in 1.0.8 don't recompile the whole optimized regexp, simply append it + if ($this->parse_cache_built) { + $subkey = count($this->language_data['CACHED_KEYWORD_LISTS'][$key]) - 1; + $this->language_data['CACHED_KEYWORD_LISTS'][$key][$subkey] .= '|' . preg_quote($word, '/'); + } + } } /** @@ -1236,11 +1399,24 @@ class GeSHi { * * @param int The key of the keyword group to remove the keyword from * @param string The word to remove from the keyword group + * @param bool Wether to automatically recompile the optimized regexp list or not. + * Note: if you set this to false and @see GeSHi->parse_code() was already called once, + * for the current language, you have to manually call @see GeSHi->optimize_keyword_group() + * or the removed keyword will stay in cache and still be highlighted! On the other hand + * it might be too expensive to recompile the regexp list for every removal if you want to + * remove a lot of keywords. * @since 1.0.0 */ - function remove_keyword($key, $word) { - $this->language_data['KEYWORDS'][$key] = - array_diff($this->language_data['KEYWORDS'][$key], array($word)); + function remove_keyword($key, $word, $recompile = true) { + $key_to_remove = array_search($word, $this->language_data['KEYWORDS'][$key]); + if ($key_to_remove !== false) { + unset($this->language_data['KEYWORDS'][$key][$key_to_remove]); + + //NEW in 1.0.8, optionally recompile keyword group + if ($recompile && $this->parse_cache_built) { + $this->optimize_keyword_group($key); + } + } } /** @@ -1254,10 +1430,21 @@ class GeSHi { */ function add_keyword_group($key, $styles, $case_sensitive = true, $words = array()) { $words = (array) $words; + if (empty($words)) { + // empty word lists mess up highlighting + return false; + } + + //Add the new keyword group internally $this->language_data['KEYWORDS'][$key] = $words; $this->lexic_permissions['KEYWORDS'][$key] = true; $this->language_data['CASE_SENSITIVE'][$key] = $case_sensitive; $this->language_data['STYLES']['KEYWORDS'][$key] = $styles; + + //NEW in 1.0.8, cache keyword regexp + if ($this->parse_cache_built) { + $this->optimize_keyword_group($key); + } } /** @@ -1267,10 +1454,25 @@ class GeSHi { * @since 1.0.0 */ function remove_keyword_group ($key) { + //Remove the keyword group internally unset($this->language_data['KEYWORDS'][$key]); unset($this->lexic_permissions['KEYWORDS'][$key]); unset($this->language_data['CASE_SENSITIVE'][$key]); unset($this->language_data['STYLES']['KEYWORDS'][$key]); + + //NEW in 1.0.8 + unset($this->language_data['CACHED_KEYWORD_LISTS'][$key]); + } + + /** + * compile optimized regexp list for keyword group + * + * @param int The key of the keyword group to compile & optimize + * @since 1.0.8 + */ + function optimize_keyword_group($key) { + $this->language_data['CACHED_KEYWORD_LISTS'][$key] = + $this->optimize_regexp_list($this->language_data['KEYWORDS'][$key]); } /** @@ -1358,8 +1560,7 @@ class GeSHi { function set_link_target($target) { if (!$target) { $this->link_target = ''; - } - else { + } else { $this->link_target = ' target="' . $target . '" '; } } @@ -1403,23 +1604,32 @@ class GeSHi { * * @param mixed An array of line numbers to highlight, or just a line * number on its own. - * @param string A string specifying the style to use for this line + * @param string A string specifying the style to use for this line. + * If null is specified, the default style is used. + * If false is specified, the line will be removed from + * special highlighting * @since 1.0.2 * @todo Some data replication here that could be cut down on */ function highlight_lines_extra($lines, $style = null) { if (is_array($lines)) { + //Split up the job using single lines at a time foreach ($lines as $line) { $this->highlight_lines_extra($line, $style); } - } - else { + } else { + //Mark the line as being highlighted specially $lines = intval($lines); $this->highlight_extra_lines[$lines] = $lines; - if ($style != null) { - $this->highlight_extra_lines_styles[$lines] = $style; - } else { + + //Decide on which style to use + if ($style === null) { //Check if we should use default style unset($this->highlight_extra_lines_styles[$lines]); + } else if ($style === false) { //Check if to remove this line + unset($this->highlight_extra_lines[$lines]); + unset($this->highlight_extra_lines_styles[$lines]); + } else { + $this->highlight_extra_lines_styles[$lines] = $style; } } } @@ -1477,7 +1687,7 @@ class GeSHi { */ function set_encoding($encoding) { if ($encoding) { - $this->encoding = $encoding; + $this->encoding = strtolower($encoding); } } @@ -1492,6 +1702,226 @@ class GeSHi { } /** + * Setup caches needed for styling. This is automatically called in + * parse_code() and get_stylesheet() when appropriate. This function helps + * stylesheet generators as they rely on some style information being + * preprocessed + * + * @since 1.0.8 + * @access private + */ + function build_style_cache() { + //Build the style cache needed to highlight numbers appropriate + if($this->lexic_permissions['NUMBERS']) { + //First check what way highlighting information for numbers are given + if(!isset($this->language_data['NUMBERS'])) { + $this->language_data['NUMBERS'] = 0; + } + + if(is_array($this->language_data['NUMBERS'])) { + $this->language_data['NUMBERS_CACHE'] = $this->language_data['NUMBERS']; + } else { + $this->language_data['NUMBERS_CACHE'] = array(); + if(!$this->language_data['NUMBERS']) { + $this->language_data['NUMBERS'] = + GESHI_NUMBER_INT_BASIC | + GESHI_NUMBER_FLT_NONSCI; + } + + for($i = 0, $j = $this->language_data['NUMBERS']; $j > 0; ++$i, $j>>=1) { + //Rearrange style indices if required ... + if(isset($this->language_data['STYLES']['NUMBERS'][1<<$i])) { + $this->language_data['STYLES']['NUMBERS'][$i] = + $this->language_data['STYLES']['NUMBERS'][1<<$i]; + unset($this->language_data['STYLES']['NUMBERS'][1<<$i]); + } + + //Check if this bit is set for highlighting + if($j&1) { + //So this bit is set ... + //Check if it belongs to group 0 or the actual stylegroup + if(isset($this->language_data['STYLES']['NUMBERS'][$i])) { + $this->language_data['NUMBERS_CACHE'][$i] = 1 << $i; + } else { + if(!isset($this->language_data['NUMBERS_CACHE'][0])) { + $this->language_data['NUMBERS_CACHE'][0] = 0; + } + $this->language_data['NUMBERS_CACHE'][0] |= 1 << $i; + } + } + } + } + } + } + + /** + * Setup caches needed for parsing. This is automatically called in parse_code() when appropriate. + * This function makes stylesheet generators much faster as they do not need these caches. + * + * @since 1.0.8 + * @access private + */ + function build_parse_cache() { + // cache symbol regexp + //As this is a costy operation, we avoid doing it for multiple groups ... + //Instead we perform it for all symbols at once. + // + //For this to work, we need to reorganize the data arrays. + if ($this->lexic_permissions['SYMBOLS'] && !empty($this->language_data['SYMBOLS'])) { + $this->language_data['MULTIPLE_SYMBOL_GROUPS'] = count($this->language_data['STYLES']['SYMBOLS']) > 1; + + $this->language_data['SYMBOL_DATA'] = array(); + $symbol_preg_multi = array(); // multi char symbols + $symbol_preg_single = array(); // single char symbols + foreach ($this->language_data['SYMBOLS'] as $key => $symbols) { + if (is_array($symbols)) { + foreach ($symbols as $sym) { + $sym = $this->hsc($sym); + if (!isset($this->language_data['SYMBOL_DATA'][$sym])) { + $this->language_data['SYMBOL_DATA'][$sym] = $key; + if (isset($sym[1])) { // multiple chars + $symbol_preg_multi[] = preg_quote($sym, '/'); + } else { // single char + if ($sym == '-') { + // don't trigger range out of order error + $symbol_preg_single[] = '\-'; + } else { + $symbol_preg_single[] = preg_quote($sym, '/'); + } + } + } + } + } else { + $symbols = $this->hsc($symbols); + if (!isset($this->language_data['SYMBOL_DATA'][$symbols])) { + $this->language_data['SYMBOL_DATA'][$symbols] = 0; + if (isset($symbols[1])) { // multiple chars + $symbol_preg_multi[] = preg_quote($symbols, '/'); + } else if ($symbols == '-') { + // don't trigger range out of order error + $symbol_preg_single[] = '\-'; + } else { // single char + $symbol_preg_single[] = preg_quote($symbols, '/'); + } + } + } + } + + //Now we have an array with each possible symbol as the key and the style as the actual data. + //This way we can set the correct style just the moment we highlight ... + // + //Now we need to rewrite our array to get a search string that + $symbol_preg = array(); + if (!empty($symbol_preg_single)) { + $symbol_preg[] = '[' . implode('', $symbol_preg_single) . ']'; + } + if (!empty($symbol_preg_multi)) { + $symbol_preg[] = implode('|', $symbol_preg_multi); + } + $this->language_data['SYMBOL_SEARCH'] = implode("|", $symbol_preg); + } + + // cache optimized regexp for keyword matching + // remove old cache + $this->language_data['CACHED_KEYWORD_LISTS'] = array(); + foreach (array_keys($this->language_data['KEYWORDS']) as $key) { + if (!isset($this->lexic_permissions['KEYWORDS'][$key]) || + $this->lexic_permissions['KEYWORDS'][$key]) { + $this->optimize_keyword_group($key); + } + } + + // brackets + if ($this->lexic_permissions['BRACKETS']) { + $this->language_data['CACHE_BRACKET_MATCH'] = array('[', ']', '(', ')', '{', '}'); + if (!$this->use_classes && isset($this->language_data['STYLES']['BRACKETS'][0])) { + $this->language_data['CACHE_BRACKET_REPLACE'] = array( + '<| style="' . $this->language_data['STYLES']['BRACKETS'][0] . '">[|>', + '<| style="' . $this->language_data['STYLES']['BRACKETS'][0] . '">]|>', + '<| style="' . $this->language_data['STYLES']['BRACKETS'][0] . '">(|>', + '<| style="' . $this->language_data['STYLES']['BRACKETS'][0] . '">)|>', + '<| style="' . $this->language_data['STYLES']['BRACKETS'][0] . '">{|>', + '<| style="' . $this->language_data['STYLES']['BRACKETS'][0] . '">}|>', + ); + } + else { + $this->language_data['CACHE_BRACKET_REPLACE'] = array( + '<| class="br0">[|>', + '<| class="br0">]|>', + '<| class="br0">(|>', + '<| class="br0">)|>', + '<| class="br0">{|>', + '<| class="br0">}|>', + ); + } + } + + //Build the parse cache needed to highlight numbers appropriate + if($this->lexic_permissions['NUMBERS']) { + //Check if the style rearrangements have been processed ... + //This also does some preprocessing to check which style groups are useable ... + if(!isset($this->language_data['NUMBERS_CACHE'])) { + $this->build_style_cache(); + } + + //Number format specification + //All this formats are matched case-insensitively! + static $numbers_format = array( + GESHI_NUMBER_INT_BASIC => + '(?<![0-9a-z_\.%])(?<![\d\.]e[+\-])[1-9]\d*?(?![0-9a-z\.])', + GESHI_NUMBER_INT_CSTYLE => + '(?<![0-9a-z_\.%])(?<![\d\.]e[+\-])[1-9]\d*?l(?![0-9a-z\.])', + GESHI_NUMBER_BIN_SUFFIX => + '(?<![0-9a-z_\.])(?<![\d\.]e[+\-])[01]+?b(?![0-9a-z\.])', + GESHI_NUMBER_BIN_PREFIX_PERCENT => + '(?<![0-9a-z_\.%])(?<![\d\.]e[+\-])%[01]+?(?![0-9a-z\.])', + GESHI_NUMBER_BIN_PREFIX_0B => + '(?<![0-9a-z_\.%])(?<![\d\.]e[+\-])0b[01]+?(?![0-9a-z\.])', + GESHI_NUMBER_OCT_PREFIX => + '(?<![0-9a-z_\.])(?<![\d\.]e[+\-])0[0-7]+?(?![0-9a-z\.])', + GESHI_NUMBER_OCT_SUFFIX => + '(?<![0-9a-z_\.])(?<![\d\.]e[+\-])[0-7]+?o(?![0-9a-z\.])', + GESHI_NUMBER_HEX_PREFIX => + '(?<![0-9a-z_\.])(?<![\d\.]e[+\-])0x[0-9a-f]+?(?![0-9a-z\.])', + GESHI_NUMBER_HEX_SUFFIX => + '(?<![0-9a-z_\.])(?<![\d\.]e[+\-])\d[0-9a-f]*?h(?![0-9a-z\.])', + GESHI_NUMBER_FLT_NONSCI => + '(?<![0-9a-z_\.])(?<![\d\.]e[+\-])\d+?\.\d+?(?![0-9a-z\.])', + GESHI_NUMBER_FLT_NONSCI_F => + '(?<![0-9a-z_\.])(?<![\d\.]e[+\-])(?:\d+?(?:\.\d*?)?|\.\d+?)f(?![0-9a-z\.])', + GESHI_NUMBER_FLT_SCI_SHORT => + '(?<![0-9a-z_\.])(?<![\d\.]e[+\-])\.\d+?(?:e[+\-]?\d+?)?(?![0-9a-z\.])', + GESHI_NUMBER_FLT_SCI_ZERO => + '(?<![0-9a-z_\.])(?<![\d\.]e[+\-])(?:\d+?(?:\.\d*?)?|\.\d+?)(?:e[+\-]?\d+?)?(?![0-9a-z\.])' + ); + + //At this step we have an associative array with flag groups for a + //specific style or an string denoting a regexp given its index. + $this->language_data['NUMBERS_RXCACHE'] = array(); + foreach($this->language_data['NUMBERS_CACHE'] as $key => $rxdata) { + if(is_string($rxdata)) { + $regexp = $rxdata; + } else { + //This is a bitfield of number flags to highlight: + //Build an array, implode them together and make this the actual RX + $rxuse = array(); + for($i = 1; $i <= $rxdata; $i<<=1) { + if($rxdata & $i) { + $rxuse[] = $numbers_format[$i]; + } + } + $regexp = implode("|", $rxuse); + } + + $this->language_data['NUMBERS_RXCACHE'][$key] = + "/(?<!<\|\/NUM!)(?<!\d\/>)($regexp)(?!\|>)/i"; + } + } + + $this->parse_cache_built = true; + } + + /** * Returns the code in $this->source, highlighted and surrounded by the * nessecary HTML. * @@ -1507,29 +1937,36 @@ class GeSHi { // Firstly, if there is an error, we won't highlight if ($this->error) { - $result = GeSHi::hsc($this->source); + //Escape the source for output + $result = $this->hsc($this->source); + + //This fix is related to SF#1923020, but has to be applied regardless of + //actually highlighting symbols. + $result = str_replace(array('<SEMI>', '<PIPE>'), array(';', '|'), $result); + // Timing is irrelevant $this->set_time($start_time, $start_time); - return $this->finalise($result); + $this->finalise($result); + return $result; + } + + // make sure the parse cache is up2date + if (!$this->parse_cache_built) { + $this->build_parse_cache(); } // Replace all newlines to a common form. $code = str_replace("\r\n", "\n", $this->source); $code = str_replace("\r", "\n", $code); + // Add spaces for regular expression matching and line numbers - $code = "\n" . $code . "\n"; +// $code = "\n" . $code . "\n"; // Initialise various stuff $length = strlen($code); $COMMENT_MATCHED = false; - // Turn highlighting on if strict mode doesn't apply to this language - $HIGHLIGHTING_ON = ( !$this->strict_mode ) ? true : ''; - // Whether to highlight inside a block of code - $HIGHLIGHT_INSIDE_STRICT = false; - $HARDQUOTE_OPEN = false; - $STRICTATTRS = ''; $stuff_to_parse = ''; - $result = ''; + $endresult = ''; // "Important" selections are handled like multiline comments // @todo GET RID OF THIS SHIZ @@ -1540,48 +1977,171 @@ class GeSHi { if ($this->strict_mode) { // Break the source into bits. Each bit will be a portion of the code // within script delimiters - for example, HTML between < and > - $parts = array(0 => array(0 => '', 1 => '')); $k = 0; - for ($i = 0; $i < $length; ++$i) { - foreach ($this->language_data['SCRIPT_DELIMITERS'] as $delimiters) { - foreach ($delimiters as $open => $close) { - // Get the next little bit for this opening string - $open_strlen = strlen($open); - $check = substr($code, $i, $open_strlen); - // If it matches... - if ($check == $open) { - // We start a new block with the highlightable - // code in it - ++$k; - $parts[$k][0] = $open; - $close_i = strpos($code, $close, $i + $open_strlen) + strlen($close); - if ($close_i === false) { - $close_i = $length - 1; + $parts = array(); + $matches = array(); + $next_match_pointer = null; + // we use a copy to unset delimiters on demand (when they are not found) + $delim_copy = $this->language_data['SCRIPT_DELIMITERS']; + $i = 0; + while ($i < $length) { + $next_match_pos = $length + 1; // never true + foreach ($delim_copy as $dk => $delimiters) { + if(is_array($delimiters)) { + foreach ($delimiters as $open => $close) { + // make sure the cache is setup properly + if (!isset($matches[$dk][$open])) { + $matches[$dk][$open] = array( + 'next_match' => -1, + 'dk' => $dk, + + 'open' => $open, // needed for grouping of adjacent code blocks (see below) + 'open_strlen' => strlen($open), + + 'close' => $close, + 'close_strlen' => strlen($close), + ); + } + // Get the next little bit for this opening string + if ($matches[$dk][$open]['next_match'] < $i) { + // only find the next pos if it was not already cached + $open_pos = strpos($code, $open, $i); + if ($open_pos === false) { + // no match for this delimiter ever + unset($delim_copy[$dk][$open]); + continue; + } + $matches[$dk][$open]['next_match'] = $open_pos; } - $parts[$k][1] = substr($code, $i, $close_i - $i); - $i = $close_i - 1; - ++$k; - $parts[$k][0] = ''; - $parts[$k][1] = ''; - - // No point going around again... - continue 3; + if ($matches[$dk][$open]['next_match'] < $next_match_pos) { + //So we got a new match, update the close_pos + $matches[$dk][$open]['close_pos'] = + strpos($code, $close, $matches[$dk][$open]['next_match']+1); + + $next_match_pointer =& $matches[$dk][$open]; + $next_match_pos = $matches[$dk][$open]['next_match']; + } + } + } else { + //So we should match an RegExp as Strict Block ... + /** + * The value in $delimiters is expected to be an RegExp + * containing exactly 2 matching groups: + * - Group 1 is the opener + * - Group 2 is the closer + */ + if(!GESHI_PHP_PRE_433 && //Needs proper rewrite to work with PHP >=4.3.0; 4.3.3 is guaranteed to work. + preg_match($delimiters, $code, $matches_rx, PREG_OFFSET_CAPTURE, $i)) { + //We got a match ... + $matches[$dk] = array( + 'next_match' => $matches_rx[1][1], + 'dk' => $dk, + + 'close_strlen' => strlen($matches_rx[2][0]), + 'close_pos' => $matches_rx[2][1], + ); + } else { + // no match for this delimiter ever + unset($delim_copy[$dk]); + continue; + } + + if ($matches[$dk]['next_match'] <= $next_match_pos) { + $next_match_pointer =& $matches[$dk]; + $next_match_pos = $matches[$dk]['next_match']; } } } - // only non-highlightable text reaches this point - $parts[$k][1] .= $code[$i]; + // non-highlightable text + $parts[$k] = array( + 1 => substr($code, $i, $next_match_pos - $i) + ); + ++$k; + + if ($next_match_pos > $length) { + // out of bounds means no next match was found + break; + } + + // highlightable code + $parts[$k][0] = $next_match_pointer['dk']; + + //Only combine for non-rx script blocks + if(is_array($delim_copy[$next_match_pointer['dk']])) { + // group adjacent script blocks, e.g. <foobar><asdf> should be one block, not three! + $i = $next_match_pos + $next_match_pointer['open_strlen']; + while (true) { + $close_pos = strpos($code, $next_match_pointer['close'], $i); + if ($close_pos == false) { + break; + } + $i = $close_pos + $next_match_pointer['close_strlen']; + if ($i == $length) { + break; + } + if ($code[$i] == $next_match_pointer['open'][0] && ($next_match_pointer['open_strlen'] == 1 || + substr($code, $i, $next_match_pointer['open_strlen']) == $next_match_pointer['open'])) { + // merge adjacent but make sure we don't merge things like <tag><!-- comment --> + foreach ($matches as $submatches) { + foreach ($submatches as $match) { + if ($match['next_match'] == $i) { + // a different block already matches here! + break 3; + } + } + } + } else { + break; + } + } + } else { + $close_pos = $next_match_pointer['close_pos'] + $next_match_pointer['close_strlen']; + $i = $close_pos; + } + + if ($close_pos === false) { + // no closing delimiter found! + $parts[$k][1] = substr($code, $next_match_pos); + ++$k; + break; + } else { + $parts[$k][1] = substr($code, $next_match_pos, $i - $next_match_pos); + ++$k; + } } - } - else { + unset($delim_copy, $next_match_pointer, $next_match_pos, $matches); + $num_parts = $k; + + if ($num_parts == 1 && $this->strict_mode == GESHI_MAYBE) { + // when we have only one part, we don't have anything to highlight at all. + // if we have a "maybe" strict language, this should be handled as highlightable code + $parts = array( + 0 => array( + 0 => '', + 1 => '' + ), + 1 => array( + 0 => null, + 1 => $parts[0][1] + ) + ); + $num_parts = 2; + } + + } else { // Not strict mode - simply dump the source into // the array at index 1 (the first highlightable block) $parts = array( - 1 => array( + 0 => array( 0 => '', + 1 => '' + ), + 1 => array( + 0 => null, 1 => $code ) ); + $num_parts = 2; } //Unset variables we won't need any longer @@ -1593,26 +2153,17 @@ class GeSHi { //Preload if line numbers are to be generated afterwards //Added a check if line breaks should be forced even without line numbers, fixes SF#1727398 - $check_linenumbers = $this->line_numbers != GESHI_NO_LINE_NUMBERS - || !empty($this->highlight_extra_lines) || !$this->allow_multiline_span; + $check_linenumbers = $this->line_numbers != GESHI_NO_LINE_NUMBERS || + !empty($this->highlight_extra_lines) || !$this->allow_multiline_span; //preload the escape char for faster checking ... - $escaped_escape_char = GeSHi::hsc($this->language_data['ESCAPE_CHAR']); - - if (!$this->use_classes) { - $string_attributes = ' style="' . $this->language_data['STYLES']['STRINGS'][0] . '"'; - $escape_char_attributes = ' style="' . $this->language_data['STYLES']['ESCAPE_CHAR'][0] . '"'; - } - else { - $string_attributes = ' class="st0"'; - $escape_char_attributes = ' class="es0"'; - } + $escaped_escape_char = $this->hsc($this->language_data['ESCAPE_CHAR']); // this is used for single-line comments $sc_disallowed_before = ""; $sc_disallowed_after = ""; - if(isset($this->language_data['PARSER_CONTROL'])) { + if (isset($this->language_data['PARSER_CONTROL'])) { if (isset($this->language_data['PARSER_CONTROL']['COMMENTS'])) { if (isset($this->language_data['PARSER_CONTROL']['COMMENTS']['DISALLOWED_BEFORE'])) { $sc_disallowed_before = $this->language_data['PARSER_CONTROL']['COMMENTS']['DISALLOWED_BEFORE']; @@ -1623,430 +2174,626 @@ class GeSHi { } } + //Fix for SF#1932083: Multichar Quotemarks unsupported + $is_string_starter = array(); + if ($this->lexic_permissions['STRINGS']) { + foreach ($this->language_data['QUOTEMARKS'] as $quotemark) { + if (!isset($is_string_starter[$quotemark[0]])) { + $is_string_starter[$quotemark[0]] = (string)$quotemark; + } else if (is_string($is_string_starter[$quotemark[0]])) { + $is_string_starter[$quotemark[0]] = array( + $is_string_starter[$quotemark[0]], + $quotemark); + } else { + $is_string_starter[$quotemark[0]][] = $quotemark; + } + } + } + // Now we go through each part. We know that even-indexed parts are // code that shouldn't be highlighted, and odd-indexed parts should // be highlighted - foreach ($parts as $key => $data) { - $part = $data[1]; + for ($key = 0; $key < $num_parts; ++$key) { + $STRICTATTRS = ''; + // If this block should be highlighted... - if ($key % 2) { - if ($this->strict_mode) { - // Find the class key for this block of code - foreach ($this->language_data['SCRIPT_DELIMITERS'] as $script_key => $script_data) { - if (isset($script_data[$data[0]])) { - break; - } - } + if (!($key & 1)) { + // Else not a block to highlight + $endresult .= $this->hsc($parts[$key][1]); + unset($parts[$key]); + continue; + } - if ($this->language_data['STYLES']['SCRIPT'][$script_key] != '' && - $this->lexic_permissions['SCRIPT']) { - // Add a span element around the source to - // highlight the overall source block - if (!$this->use_classes && - $this->language_data['STYLES']['SCRIPT'][$script_key] != '') { - $attributes = ' style="' . $this->language_data['STYLES']['SCRIPT'][$script_key] . '"'; - } - else { - $attributes = ' class="sc' . $script_key . '"'; - } - $result .= "<span$attributes>"; - $STRICTATTRS = $attributes; + $result = ''; + $part = $parts[$key][1]; + + $highlight_part = true; + if ($this->strict_mode && !is_null($parts[$key][0])) { + // get the class key for this block of code + $script_key = $parts[$key][0]; + $highlight_part = $this->language_data['HIGHLIGHT_STRICT_BLOCK'][$script_key]; + if ($this->language_data['STYLES']['SCRIPT'][$script_key] != '' && + $this->lexic_permissions['SCRIPT']) { + // Add a span element around the source to + // highlight the overall source block + if (!$this->use_classes && + $this->language_data['STYLES']['SCRIPT'][$script_key] != '') { + $attributes = ' style="' . $this->language_data['STYLES']['SCRIPT'][$script_key] . '"'; + } else { + $attributes = ' class="sc' . $script_key . '"'; } + $result .= "<span$attributes>"; + $STRICTATTRS = $attributes; } + } - if (!$this->strict_mode || $this->language_data['HIGHLIGHT_STRICT_BLOCK'][$script_key]) { - // Now, highlight the code in this block. This code - // is really the engine of GeSHi (along with the method - // parse_non_string_part). - - // cache comment regexps incrementally - $comment_regexp_cache = array(); - $next_comment_regexp_pos = -1; - $comment_regexp_cache_per_key = array(); - - $length = strlen($part); - for ($i = 0; $i < $length; ++$i) { - // Get the next char - $char = $part[$i]; - - if (in_array($char, $this->language_data['QUOTEMARKS']) && $this->lexic_permissions['STRINGS']) { - // The start of a new string + if ($highlight_part) { + // Now, highlight the code in this block. This code + // is really the engine of GeSHi (along with the method + // parse_non_string_part). + + // cache comment regexps incrementally + $comment_regexp_cache = array(); + $next_comment_regexp_pos = -1; + $next_comment_multi_pos = -1; + $next_comment_single_pos = -1; + $comment_regexp_cache_per_key = array(); + $comment_multi_cache_per_key = array(); + $comment_single_cache_per_key = array(); + $next_open_comment_multi = ''; + $next_comment_single_key = ''; + + $length = strlen($part); + for ($i = 0; $i < $length; ++$i) { + // Get the next char + $char = $part[$i]; + $char_len = 1; + + $string_started = false; + + if (isset($is_string_starter[$char])) { + // Possibly the start of a new string ... + + //Check which starter it was ... + //Fix for SF#1932083: Multichar Quotemarks unsupported + if (is_array($is_string_starter[$char])) { + $char_new = ''; + foreach ($is_string_starter[$char] as $testchar) { + if ($testchar === substr($part, $i, strlen($testchar)) && + strlen($testchar) > strlen($char_new)) { + $char_new = $testchar; + $string_started = true; + } + } + if ($string_started) { + $char = $char_new; + } + } else { + $testchar = $is_string_starter[$char]; + if ($testchar === substr($part, $i, strlen($testchar))) { + $char = $testchar; + $string_started = true; + } + } + $char_len = strlen($char); + } - // parse the stuff before this - $result .= $this->parse_non_string_part($stuff_to_parse); - $stuff_to_parse = ''; + if ($string_started) { + // Hand out the correct style information for this string + $string_key = array_search($char, $this->language_data['QUOTEMARKS']); + if (!isset($this->language_data['STYLES']['STRINGS'][$string_key]) || + !isset($this->language_data['STYLES']['ESCAPE_CHAR'][$string_key])) { + $string_key = 0; + } - // now handle the string - $string = ''; + if (!$this->use_classes) { + $string_attributes = ' style="' . $this->language_data['STYLES']['STRINGS'][$string_key] . '"'; + $escape_char_attributes = ' style="' . $this->language_data['STYLES']['ESCAPE_CHAR'][$string_key] . '"'; + } else { + $string_attributes = ' class="st'.$string_key.'"'; + $escape_char_attributes = ' class="es'.$string_key.'"'; + } - // look for closing quote - $start = $i; - while ($close_pos = strpos($part, $char, $start + 1)) { - $start = $close_pos; - if ($this->lexic_permissions['ESCAPE_CHAR'] && $part[$close_pos - 1] == $this->language_data['ESCAPE_CHAR']) { - // this quote is escaped + // parse the stuff before this + $result .= $this->parse_non_string_part($stuff_to_parse); + $stuff_to_parse = ''; + + // now handle the string + $string = ''; + + // look for closing quote + $start = $i; + while ($close_pos = strpos($part, $char, $start + $char_len)) { + $start = $close_pos; + if ($this->lexic_permissions['ESCAPE_CHAR'] && $part[$close_pos - 1] == $this->language_data['ESCAPE_CHAR']) { + // check wether this quote is escaped or if it is something like '\\' + $escape_char_pos = $close_pos - 1; + while ($escape_char_pos > 0 && + $part[$escape_char_pos - 1] == $this->language_data['ESCAPE_CHAR']) { + --$escape_char_pos; + } + if (($close_pos - $escape_char_pos) & 1) { + // uneven number of escape chars => this quote is escaped continue; } - // found closing quote - break; - } - if (!$close_pos) { - // span till the end of this $part when no closing delimiter is found - $close_pos = $length; } - $string = substr($part, $i, $close_pos - $i + 1); - $i = $close_pos; - - // handle escape chars and encode html chars - // (special because when we have escape chars within our string they may not be escaped) - if ($this->lexic_permissions['ESCAPE_CHAR'] && $this->language_data['ESCAPE_CHAR']) { - $start = 0; - $new_string = ''; - while ($es_pos = strpos($string, $this->language_data['ESCAPE_CHAR'], $start)) { - $new_string .= GeSHi::hsc(substr($string, $start, $es_pos - $start)) - . "<span$escape_char_attributes>" . $escaped_escape_char; - if ($string[$es_pos + 1] == "\n") { - // don't put a newline around newlines - $new_string .= "</span>\n"; + // found closing quote + break; + } + + //Found the closing delimiter? + if (!$close_pos) { + // span till the end of this $part when no closing delimiter is found + $close_pos = $length; + } + + //Get the actual string + $string = substr($part, $i, $close_pos - $i + $char_len); + $i = $close_pos + $char_len - 1; + + // handle escape chars and encode html chars + // (special because when we have escape chars within our string they may not be escaped) + if ($this->lexic_permissions['ESCAPE_CHAR'] && $this->language_data['ESCAPE_CHAR']) { + $start = 0; + $new_string = ''; + while ($es_pos = strpos($string, $this->language_data['ESCAPE_CHAR'], $start)) { + $new_string .= $this->hsc(substr($string, $start, $es_pos - $start)) + . "<span$escape_char_attributes>" . $escaped_escape_char; + $es_char = $string[$es_pos + 1]; + if ($es_char == "\n") { + // don't put a newline around newlines + $new_string .= "</span>\n"; + } else if (ord($es_char) >= 128) { + //This is an non-ASCII char (UTF8 or single byte) + //This code tries to work around SF#2037598 ... + if(function_exists('mb_substr')) { + $es_char_m = mb_substr(substr($string, $es_pos+1, 16), 0, 1, $this->encoding); + $new_string .= $es_char_m . '</span>'; + } else if (!GESHI_PHP_PRE_433 && 'utf-8' == $this->encoding) { + if(preg_match("/[\xC2-\xDF][\x80-\xBF]". + "|\xE0[\xA0-\xBF][\x80-\xBF]". + "|[\xE1-\xEC\xEE\xEF][\x80-\xBF]{2}". + "|\xED[\x80-\x9F][\x80-\xBF]". + "|\xF0[\x90-\xBF][\x80-\xBF]{2}". + "|[\xF1-\xF3][\x80-\xBF]{3}". + "|\xF4[\x80-\x8F][\x80-\xBF]{2}/s", + $string, $es_char_m, null, $es_pos + 1)) { + $es_char_m = $es_char_m[0]; + } else { + $es_char_m = $es_char; + } + $new_string .= $this->hsc($es_char_m) . '</span>'; } else { - $new_string .= GeSHi::hsc($string[$es_pos + 1]) . '</span>'; + $es_char_m = $this->hsc($es_char); } - $start = $es_pos + 2; + $es_pos += strlen($es_char_m) - 1; + } else { + $new_string .= $this->hsc($es_char) . '</span>'; } - $string = $new_string . GeSHi::hsc(substr($string, $start)); - $new_string = ''; - } else { - $string = GeSHi::hsc($string); + $start = $es_pos + 2; } - - if ($check_linenumbers) { - // Are line numbers used? If, we should end the string before - // the newline and begin it again (so when <li>s are put in the source - // remains XHTML compliant) - // note to self: This opens up possibility of config files specifying - // that languages can/cannot have multiline strings??? - $string = str_replace("\n", "</span>\n<span$string_attributes>", $string); - } - - $result .= "<span$string_attributes>" . $string . '</span>'; - $string = ''; - continue; + $string = $new_string . $this->hsc(substr($string, $start)); + $new_string = ''; + } else { + $string = $this->hsc($string); } - else if ($hq && substr($part, $i, $hq_strlen) == $hq && $this->lexic_permissions['STRINGS']) { - // The start of a hard quoted string - // parse the stuff before this - $result .= $this->parse_non_string_part($stuff_to_parse); - $stuff_to_parse = ''; + if ($check_linenumbers) { + // Are line numbers used? If, we should end the string before + // the newline and begin it again (so when <li>s are put in the source + // remains XHTML compliant) + // note to self: This opens up possibility of config files specifying + // that languages can/cannot have multiline strings??? + $string = str_replace("\n", "</span>\n<span$string_attributes>", $string); + } - // now handle the string - $string = ''; - - // look for closing quote - $start = $i + $hq_strlen; - while ($close_pos = strpos($part, $this->language_data['HARDQUOTE'][1], $start + 1)) { - $start = $close_pos; - if ($this->lexic_permissions['ESCAPE_CHAR'] && $part[$close_pos - 1] == $this->language_data['ESCAPE_CHAR']) { - // make sure this quote is not escaped - foreach ($this->language_data['HARDESCAPE'] as $hardescape) { - if (substr($part, $close_pos - 1, strlen($hardescape)) == $hardescape) { - // this quote is escaped + $result .= "<span$string_attributes>" . $string . '</span>'; + $string = ''; + continue; + } else if ($this->lexic_permissions['STRINGS'] && $hq && $hq[0] == $char && + substr($part, $i, $hq_strlen) == $hq) { + // The start of a hard quoted string + if (!$this->use_classes) { + $string_attributes = ' style="' . $this->language_data['STYLES']['STRINGS']['HARDQUOTE'] . '"'; + $escape_char_attributes = ' style="' . $this->language_data['STYLES']['ESCAPE_CHAR']['HARDESCAPE'] . '"'; + } else { + $string_attributes = ' class="st_h"'; + $escape_char_attributes = ' class="es_h"'; + } + // parse the stuff before this + $result .= $this->parse_non_string_part($stuff_to_parse); + $stuff_to_parse = ''; + + // now handle the string + $string = ''; + + // look for closing quote + $start = $i + $hq_strlen; + while ($close_pos = strpos($part, $this->language_data['HARDQUOTE'][1], $start)) { + $start = $close_pos + 1; + if ($this->lexic_permissions['ESCAPE_CHAR'] && $part[$close_pos - 1] == $this->language_data['ESCAPE_CHAR']) { + // make sure this quote is not escaped + foreach ($this->language_data['HARDESCAPE'] as $hardescape) { + if (substr($part, $close_pos - 1, strlen($hardescape)) == $hardescape) { + // check wether this quote is escaped or if it is something like '\\' + $escape_char_pos = $close_pos - 1; + while ($escape_char_pos > 0 + && $part[$escape_char_pos - 1] == $this->language_data['ESCAPE_CHAR']) { + --$escape_char_pos; + } + if (($close_pos - $escape_char_pos) & 1) { + // uneven number of escape chars => this quote is escaped continue 2; } } } - // found closing quote - break; - } - if (!$close_pos) { - // span till the end of this $part when no closing delimiter is found - $close_pos = $length; } - $string = substr($part, $i, $close_pos - $i + 1); - $i = $close_pos; - - // handle escape chars and encode html chars - // (special because when we have escape chars within our string they may not be escaped) - if ($this->lexic_permissions['ESCAPE_CHAR'] && $this->language_data['ESCAPE_CHAR']) { - $start = 0; - $new_string = ''; - while ($es_pos = strpos($string, $this->language_data['ESCAPE_CHAR'], $start)) { - // hmtl escape stuff before - $new_string .= GeSHi::hsc(substr($string, $start, $es_pos - $start)); - // check if this is a hard escape - foreach ($this->language_data['HARDESCAPE'] as $hardescape) { - if (substr($string, $es_pos, strlen($hardescape)) == $hardescape) { - // indeed, this is a hardescape - $new_string .= "<span$escape_char_attributes>" . GeSHi::hsc($hardescape) - . '</span>'; - $start = $es_pos + strlen($hardescape); - continue 2; - } + // found closing quote + break; + } + + //Found the closing delimiter? + if (!$close_pos) { + // span till the end of this $part when no closing delimiter is found + $close_pos = $length; + } + + //Get the actual string + $string = substr($part, $i, $close_pos - $i + 1); + $i = $close_pos; + + // handle escape chars and encode html chars + // (special because when we have escape chars within our string they may not be escaped) + if ($this->lexic_permissions['ESCAPE_CHAR'] && $this->language_data['ESCAPE_CHAR']) { + $start = 0; + $new_string = ''; + while ($es_pos = strpos($string, $this->language_data['ESCAPE_CHAR'], $start)) { + // hmtl escape stuff before + $new_string .= $this->hsc(substr($string, $start, $es_pos - $start)); + // check if this is a hard escape + foreach ($this->language_data['HARDESCAPE'] as $hardescape) { + if (substr($string, $es_pos, strlen($hardescape)) == $hardescape) { + // indeed, this is a hardescape + $new_string .= "<span$escape_char_attributes>" . + $this->hsc($hardescape) . '</span>'; + $start = $es_pos + strlen($hardescape); + continue 2; } - // not a hard escape + } + // not a hard escape, but a normal escape + // they come in pairs of two + $c = 0; + while (isset($string[$es_pos + $c]) && isset($string[$es_pos + $c + 1]) + && $string[$es_pos + $c] == $this->language_data['ESCAPE_CHAR'] + && $string[$es_pos + $c + 1] == $this->language_data['ESCAPE_CHAR']) { + $c += 2; + } + if ($c) { + $new_string .= "<span$escape_char_attributes>" . + str_repeat($escaped_escape_char, $c) . + '</span>'; + $start = $es_pos + $c; + } else { + // this is just a single lonely escape char... $new_string .= $escaped_escape_char; $start = $es_pos + 1; } - $string = $new_string . GeSHi::hsc(substr($string, $start)); - } else { - $string = GeSHi::hsc($string); } + $string = $new_string . $this->hsc(substr($string, $start)); + } else { + $string = $this->hsc($string); + } + + if ($check_linenumbers) { + // Are line numbers used? If, we should end the string before + // the newline and begin it again (so when <li>s are put in the source + // remains XHTML compliant) + // note to self: This opens up possibility of config files specifying + // that languages can/cannot have multiline strings??? + $string = str_replace("\n", "</span>\n<span$string_attributes>", $string); + } + + $result .= "<span$string_attributes>" . $string . '</span>'; + $string = ''; + continue; + } else { + // update regexp comment cache if needed + if (isset($this->language_data['COMMENT_REGEXP']) && $next_comment_regexp_pos < $i) { + $next_comment_regexp_pos = $length; + foreach ($this->language_data['COMMENT_REGEXP'] as $comment_key => $regexp) { + $match_i = false; + if (isset($comment_regexp_cache_per_key[$comment_key]) && + $comment_regexp_cache_per_key[$comment_key] >= $i) { + // we have already matched something + $match_i = $comment_regexp_cache_per_key[$comment_key]; + } else if ( + //This is to allow use of the offset parameter in preg_match and stay as compatible with older PHP versions as possible + (GESHI_PHP_PRE_433 && preg_match($regexp, substr($part, $i), $match, PREG_OFFSET_CAPTURE)) || + (!GESHI_PHP_PRE_433 && preg_match($regexp, $part, $match, PREG_OFFSET_CAPTURE, $i)) + ) { + $match_i = $match[0][1]; + if (GESHI_PHP_PRE_433) { + $match_i += $i; + } + + $comment_regexp_cache[$match_i] = array( + 'key' => $comment_key, + 'length' => strlen($match[0][0]), + ); - if ($check_linenumbers) { - // Are line numbers used? If, we should end the string before - // the newline and begin it again (so when <li>s are put in the source - // remains XHTML compliant) - // note to self: This opens up possibility of config files specifying - // that languages can/cannot have multiline strings??? - $string = str_replace("\n", "</span>\n<span$string_attributes>", $string); + $comment_regexp_cache_per_key[$comment_key] = $match_i; + } else { + $comment_regexp_cache_per_key[$comment_key] = false; + continue; + } + + if ($match_i !== false && $match_i < $next_comment_regexp_pos) { + $next_comment_regexp_pos = $match_i; + if ($match_i === $i) { + break; + } + } } + } + //Have a look for regexp comments + if ($i == $next_comment_regexp_pos) { + $COMMENT_MATCHED = true; + $comment = $comment_regexp_cache[$next_comment_regexp_pos]; + $test_str = $this->hsc(substr($part, $i, $comment['length'])); + + //@todo If remove important do remove here + if ($this->lexic_permissions['COMMENTS']['MULTI']) { + if (!$this->use_classes) { + $attributes = ' style="' . $this->language_data['STYLES']['COMMENTS'][$comment['key']] . '"'; + } else { + $attributes = ' class="co' . $comment['key'] . '"'; + } - $result .= "<span$string_attributes>" . $string . '</span>'; - $string = ''; - continue; + $test_str = "<span$attributes>" . $test_str . "</span>"; + + // Short-cut through all the multiline code + if ($check_linenumbers) { + // strreplace to put close span and open span around multiline newlines + $test_str = str_replace( + "\n", "</span>\n<span$attributes>", + str_replace("\n ", "\n ", $test_str) + ); + } + } + + $i += $comment['length'] - 1; + + // parse the rest + $result .= $this->parse_non_string_part($stuff_to_parse); + $stuff_to_parse = ''; } - else { - // update regexp comment cache if needed - if (isset($this->language_data['COMMENT_REGEXP']) && $next_comment_regexp_pos < $i) { - $next_comment_regexp_pos = $length; - foreach ($this->language_data['COMMENT_REGEXP'] as $comment_key => $regexp) { + + // If we haven't matched a regexp comment, try multi-line comments + if (!$COMMENT_MATCHED) { + // Is this a multiline comment? + if (!empty($this->language_data['COMMENT_MULTI']) && $next_comment_multi_pos < $i) { + $next_comment_multi_pos = $length; + foreach ($this->language_data['COMMENT_MULTI'] as $open => $close) { $match_i = false; - if (isset($comment_regexp_cache_per_key[$comment_key]) && - $comment_regexp_cache_per_key[$comment_key] >= $i) { + if (isset($comment_multi_cache_per_key[$open]) && + $comment_multi_cache_per_key[$open] >= $i) { // we have already matched something - $match_i = $comment_regexp_cache_per_key[$comment_key]; - } - else if (preg_match($regexp, $part, $match, PREG_OFFSET_CAPTURE, $i)) { - $match_i = $match[0][1]; - $comment_regexp_cache[$match_i] = array( - 'key' => $comment_key, - 'length' => strlen($match[0][0]), - ); - $comment_regexp_cache_per_key[$comment_key] = $match_i; + $match_i = $comment_multi_cache_per_key[$open]; + } else if (($match_i = stripos($part, $open, $i)) !== false) { + $comment_multi_cache_per_key[$open] = $match_i; } else { - $comment_regexp_cache_per_key[$comment_key] = false; + $comment_multi_cache_per_key[$open] = false; continue; } - - if ($match_i !== false && $match_i < $next_comment_regexp_pos) { - $next_comment_regexp_pos = $match_i; + if ($match_i !== false && $match_i < $next_comment_multi_pos) { + $next_comment_multi_pos = $match_i; + $next_open_comment_multi = $open; if ($match_i === $i) { break; } } } } - //Have a look for regexp comments - if ($i == $next_comment_regexp_pos) { + if ($i == $next_comment_multi_pos) { + $open = $next_open_comment_multi; + $close = $this->language_data['COMMENT_MULTI'][$open]; + $open_strlen = strlen($open); + $close_strlen = strlen($close); $COMMENT_MATCHED = true; - $comment = $comment_regexp_cache[$next_comment_regexp_pos]; - $test_str = GeSHi::hsc(substr($part, $i, $comment['length'])); - + $test_str_match = $open; //@todo If remove important do remove here - if ($this->lexic_permissions['COMMENTS']['MULTI']) { - if (!$this->use_classes) { - $attributes = ' style="' . $this->language_data['STYLES']['COMMENTS'][$comment['key']] . '"'; - } - else { - $attributes = ' class="co' . $comment['key'] . '"'; - } - $test_str = "<span$attributes>" . $test_str . "</span>"; - - // Short-cut through all the multiline code - if ($check_linenumbers) { - // strreplace to put close span and open span around multiline newlines - $test_str = str_replace( - "\n", "</span>\n<span$attributes>", - str_replace("\n ", "\n ", $test_str) - ); + if ($this->lexic_permissions['COMMENTS']['MULTI'] || + $open == GESHI_START_IMPORTANT) { + if ($open != GESHI_START_IMPORTANT) { + if (!$this->use_classes) { + $attributes = ' style="' . $this->language_data['STYLES']['COMMENTS']['MULTI'] . '"'; + } else { + $attributes = ' class="coMULTI"'; + } + $test_str = "<span$attributes>" . $this->hsc($open); + } else { + if (!$this->use_classes) { + $attributes = ' style="' . $this->important_styles . '"'; + } else { + $attributes = ' class="imp"'; + } + + // We don't include the start of the comment if it's an + // "important" part + $test_str = "<span$attributes>"; } + } else { + $test_str = $this->hsc($open); } - $i += $comment['length'] - 1; + $close_pos = strpos( $part, $close, $i + $open_strlen ); + + if ($close_pos === false) { + $close_pos = $length; + } + + // Short-cut through all the multiline code + $rest_of_comment = $this->hsc(substr($part, $i + $open_strlen, $close_pos - $i - $open_strlen + $close_strlen)); + if (($this->lexic_permissions['COMMENTS']['MULTI'] || + $test_str_match == GESHI_START_IMPORTANT) && + $check_linenumbers) { + + // strreplace to put close span and open span around multiline newlines + $test_str .= str_replace( + "\n", "</span>\n<span$attributes>", + str_replace("\n ", "\n ", $rest_of_comment) + ); + } else { + $test_str .= $rest_of_comment; + } + + if ($this->lexic_permissions['COMMENTS']['MULTI'] || + $test_str_match == GESHI_START_IMPORTANT) { + $test_str .= '</span>'; + } + + $i = $close_pos + $close_strlen - 1; // parse the rest $result .= $this->parse_non_string_part($stuff_to_parse); $stuff_to_parse = ''; } - // If we haven't matched a regexp comment, try multi-line comments - if (!$COMMENT_MATCHED) { - // Is this a multiline comment? - foreach ($this->language_data['COMMENT_MULTI'] as $open => $close) { - $com_len = strlen($open); - $test_str = substr( $part, $i, $com_len ); - $test_str_match = $test_str; - if (strtolower($open) == strtolower($test_str)) { - $COMMENT_MATCHED = true; - //@todo If remove important do remove here - if ($this->lexic_permissions['COMMENTS']['MULTI'] || - $test_str == GESHI_START_IMPORTANT) { - if ($test_str != GESHI_START_IMPORTANT) { - if (!$this->use_classes) { - $attributes = ' style="' . $this->language_data['STYLES']['COMMENTS']['MULTI'] . '"'; - } - else { - $attributes = ' class="coMULTI"'; - } - $test_str = "<span$attributes>" . GeSHi::hsc($test_str); - } - else { - if (!$this->use_classes) { - $attributes = ' style="' . $this->important_styles . '"'; - } - else { - $attributes = ' class="imp"'; - } - // We don't include the start of the comment if it's an - // "important" part - $test_str = "<span$attributes>"; - } - } - else { - $test_str = GeSHi::hsc($test_str); - } - - $close_pos = strpos( $part, $close, $i + strlen($open) ); - - if ($close_pos === false) { - $close_pos = $length; - } - - // Short-cut through all the multiline code - $rest_of_comment = GeSHi::hsc(substr($part, $i + strlen($open), $close_pos - $i - strlen($open) + strlen($close))); - if (($this->lexic_permissions['COMMENTS']['MULTI'] || - $test_str_match == GESHI_START_IMPORTANT) && - $check_linenumbers) { - // strreplace to put close span and open span around multiline newlines - $test_str .= str_replace( - "\n", "</span>\n<span$attributes>", - str_replace("\n ", "\n ", $rest_of_comment) - ); - } - else { - $test_str .= $rest_of_comment; - } + } - if ($this->lexic_permissions['COMMENTS']['MULTI'] || - $test_str_match == GESHI_START_IMPORTANT) { - $test_str .= '</span>'; + // If we haven't matched a multiline comment, try single-line comments + if (!$COMMENT_MATCHED) { + // cache potential single line comment occurances + if (!empty($this->language_data['COMMENT_SINGLE']) && $next_comment_single_pos < $i) { + $next_comment_single_pos = $length; + foreach ($this->language_data['COMMENT_SINGLE'] as $comment_key => $comment_mark) { + $match_i = false; + if (isset($comment_single_cache_per_key[$comment_key]) && + $comment_single_cache_per_key[$comment_key] >= $i) { + // we have already matched something + $match_i = $comment_single_cache_per_key[$comment_key]; + } else if ( + // case sensitive comments + ($this->language_data['CASE_SENSITIVE'][GESHI_COMMENTS] && + ($match_i = stripos($part, $comment_mark, $i)) !== false) || + // non case sensitive + (!$this->language_data['CASE_SENSITIVE'][GESHI_COMMENTS] && + (($match_i = strpos($part, $comment_mark, $i)) !== false))) { + $comment_single_cache_per_key[$comment_key] = $match_i; + } else { + $comment_single_cache_per_key[$comment_key] = false; + continue; + } + if ($match_i !== false && $match_i < $next_comment_single_pos) { + $next_comment_single_pos = $match_i; + $next_comment_single_key = $comment_key; + if ($match_i === $i) { + break; } - $i = $close_pos + strlen($close) - 1; - // parse the rest - $result .= $this->parse_non_string_part($stuff_to_parse); - $stuff_to_parse = ''; - break; } } } - // If we haven't matched a multiline comment, try single-line comments - if (!$COMMENT_MATCHED) { - foreach ($this->language_data['COMMENT_SINGLE'] as $comment_key => $comment_mark) { - $com_len = strlen($comment_mark); - $test_str = substr($part, $i, $com_len); - if ($this->language_data['CASE_SENSITIVE'][GESHI_COMMENTS]) { - $match = ($comment_mark == $test_str); + if ($next_comment_single_pos == $i) { + $comment_key = $next_comment_single_key; + $comment_mark = $this->language_data['COMMENT_SINGLE'][$comment_key]; + $com_len = strlen($comment_mark); + + // This check will find special variables like $# in bash + // or compiler directives of Delphi beginning {$ + if ((empty($sc_disallowed_before) || ($i == 0) || + (false === strpos($sc_disallowed_before, $part[$i-1]))) && + (empty($sc_disallowed_after) || ($length <= $i + $com_len) || + (false === strpos($sc_disallowed_after, $part[$i + $com_len])))) + { + // this is a valid comment + $COMMENT_MATCHED = true; + if ($this->lexic_permissions['COMMENTS'][$comment_key]) { + if (!$this->use_classes) { + $attributes = ' style="' . $this->language_data['STYLES']['COMMENTS'][$comment_key] . '"'; + } else { + $attributes = ' class="co' . $comment_key . '"'; + } + $test_str = "<span$attributes>" . $this->hsc($this->change_case($comment_mark)); + } else { + $test_str = $this->hsc($comment_mark); } - else { - $match = (strtolower($comment_mark) == strtolower($test_str)); + + //Check if this comment is the last in the source + $close_pos = strpos($part, "\n", $i); + $oops = false; + if ($close_pos === false) { + $close_pos = $length; + $oops = true; } - //This check will find special variables like $# in bash or compiler directives of Delphi beginning {$ - if($match) { - $match = $match && (empty($sc_disallowed_before) || ((false === strpos($sc_disallowed_before, $part[$i-1])) && (0 != $i))); - $match = $match && (empty($sc_disallowed_after) || ((false === strpos($sc_disallowed_after, $part[$i+1])) && ($length-1>$i))); + $test_str .= $this->hsc(substr($part, $i + $com_len, $close_pos - $i - $com_len)); + if ($this->lexic_permissions['COMMENTS'][$comment_key]) { + $test_str .= "</span>"; } - if ($match) { - $COMMENT_MATCHED = true; - if ($this->lexic_permissions['COMMENTS'][$comment_key]) { - if (!$this->use_classes) { - $attributes = ' style="' . $this->language_data['STYLES']['COMMENTS'][$comment_key] . '"'; - } - else { - $attributes = ' class="co' . $comment_key . '"'; - } - $test_str = "<span$attributes>" . GeSHi::hsc($this->change_case($test_str)); - } - else { - $test_str = GeSHi::hsc($test_str); - } - $close_pos = strpos($part, "\n", $i); - $oops = false; - if ($close_pos === false) { - $close_pos = $length; - $oops = true; - } - $test_str .= GeSHi::hsc(substr($part, $i + $com_len, $close_pos - $i - $com_len)); - if ($this->lexic_permissions['COMMENTS'][$comment_key]) { - $test_str .= "</span>"; - } - // Take into account that the comment might be the last in the source - if (!$oops) { - $test_str .= "\n"; - } - $i = $close_pos; - // parse the rest - $result .= $this->parse_non_string_part($stuff_to_parse); - $stuff_to_parse = ''; - break; + + // Take into account that the comment might be the last in the source + if (!$oops) { + $test_str .= "\n"; } + + $i = $close_pos; + + // parse the rest + $result .= $this->parse_non_string_part($stuff_to_parse); + $stuff_to_parse = ''; } } } - // Where are we adding this char? - if (!$COMMENT_MATCHED) { - $stuff_to_parse .= $char; - } - else { - $result .= $test_str; - $COMMENT_MATCHED = false; - } } - // Parse the last bit - $result .= $this->parse_non_string_part($stuff_to_parse); - $stuff_to_parse = ''; - } - else { - if ($STRICTATTRS != '') { - $part = str_replace("\n", "</span>\n<span$STRICTATTRS>", GeSHi::hsc($part)); - $STRICTATTRS = ''; + + // Where are we adding this char? + if (!$COMMENT_MATCHED) { + $stuff_to_parse .= $char; + } else { + $result .= $test_str; + unset($test_str); + $COMMENT_MATCHED = false; } - $result .= $part; - } - // Close the <span> that surrounds the block - if ($this->strict_mode && $this->language_data['STYLES']['SCRIPT'][$script_key] != '' && - $this->lexic_permissions['SCRIPT']) { - $result .= '</span>'; } + // Parse the last bit + $result .= $this->parse_non_string_part($stuff_to_parse); + $stuff_to_parse = ''; + } else { + $result .= $this->hsc($part); } - else { - // Else not a block to highlight - $result .= GeSHi::hsc($part); + // Close the <span> that surrounds the block + if ($STRICTATTRS != '') { + $result = str_replace("\n", "</span>\n<span$STRICTATTRS>", $result); + $result .= '</span>'; } + + $endresult .= $result; + unset($part, $parts[$key], $result); } //This fix is related to SF#1923020, but has to be applied regardless of //actually highlighting symbols. - $result = str_replace(array('<SEMI>', '<PIPE>'), array(';', '|'), $result); + /** NOTE: memorypeak #3 */ + $endresult = str_replace(array('<SEMI>', '<PIPE>'), array(';', '|'), $endresult); // // Parse the last stuff (redundant?) // $result .= $this->parse_non_string_part($stuff_to_parse); // Lop off the very first and last spaces - $result = substr($result, 1, -1); +// $result = substr($result, 1, -1); // We're finished: stop timing $this->set_time($start_time, microtime()); - return $this->finalise($result); + $this->finalise($endresult); + return $endresult; } /** * Swaps out spaces and tabs for HTML indentation. Not needed if * the code is in a pre block... * - * @param string The source to indent - * @return string The source with HTML indenting applied + * @param string The source to indent (reference!) * @since 1.0.0 * @access private */ @@ -2079,16 +2826,14 @@ class GeSHi { // html so that < or > would be allowed in user's styles // (e.g. quotes: '<' '>'; or similar) if ($IN_TAG) { - if('>' == $char) { + if ('>' == $char) { $IN_TAG = false; } $lines[$key] .= $char; - } - else if ('<' == $char) { + } else if ('<' == $char) { $IN_TAG = true; $lines[$key] .= '<'; - } - else if ('&' == $char) { + } else if ('&' == $char) { $substr = substr($line, $i + 3, 5); $posi = strpos($substr, ';'); if (false === $posi) { @@ -2097,8 +2842,7 @@ class GeSHi { $pos -= $posi+2; } $lines[$key] .= $char; - } - else if ("\t" == $char) { + } else if ("\t" == $char) { $str = ''; // OPTIMISE - move $strs out. Make an array: // $tabs = array( @@ -2107,7 +2851,7 @@ class GeSHi { // 3 => ' ' etc etc // to use instead of building a string every time $tab_end_width = $tab_width - ($pos % $tab_width); //Moved out of the look as it doesn't change within the loop - if(($pos%2) || 1 == $tab_end_width) { + if (($pos & 1) || 1 == $tab_end_width) { $str .= substr($tab_string, 6, $tab_end_width); } else { $str .= substr($tab_string, 0, $tab_end_width+5); @@ -2119,12 +2863,10 @@ class GeSHi { $lines[$key] .= substr($line, $i + 1); break; } - } - else if (0 == $pos && ' ' == $char) { + } else if (0 == $pos && ' ' == $char) { $lines[$key] .= ' '; ++$pos; - } - else { + } else { $lines[$key] .= $char; ++$pos; } @@ -2145,7 +2887,6 @@ class GeSHi { $result = str_replace("\n", $this->line_ending, $result); } } - return $result; } /** @@ -2168,57 +2909,104 @@ class GeSHi { } /** - * Adds a url to a keyword where needed. + * Handles replacements of keywords to include markup and links if requested * - * @param string The keyword to add the URL HTML to - * @param int What group the keyword is from - * @param boolean Whether to get the HTML for the start or end - * @return The HTML for either the start or end of the HTML <a> tag - * @since 1.0.2 + * @param string The keyword to add the Markup to + * @return The HTML for the match found + * @since 1.0.8 * @access private - * @todo Get rid of ender + * + * @todo Get rid of ender in keyword links */ - function add_url_to_keyword($keyword, $group, $start_or_end) { - if (!$this->keyword_links) { - // Keyword links have been disabled - return; - } - - if (isset($this->language_data['URLS'][$group]) && - $this->language_data['URLS'][$group] != '' && - substr($keyword, 0, 5) != '</') { - // There is a base group for this keyword - if ($start_or_end == 'BEGIN') { - // HTML workaround... not good form (tm) but should work for 1.0.X - if ($keyword != '') { - // Old system: strtolower - //$keyword = ( $this->language_data['CASE_SENSITIVE'][$group] ) ? $keyword : strtolower($keyword); - // New system: get keyword from language file to get correct case - $lower_keyword = strtolower($keyword); - foreach ($this->language_data['KEYWORDS'][$group] as $word) { - if (strtolower($word) == $lower_keyword) { + function handle_keyword_replace($match) { + $k = $this->_kw_replace_group; + $keyword = $match[0]; + + $before = ''; + $after = ''; + + if ($this->keyword_links) { + // Keyword links have been ebabled + + if (isset($this->language_data['URLS'][$k]) && + $this->language_data['URLS'][$k] != '') { + // There is a base group for this keyword + + // Old system: strtolower + //$keyword = ( $this->language_data['CASE_SENSITIVE'][$group] ) ? $keyword : strtolower($keyword); + // New system: get keyword from language file to get correct case + if (!$this->language_data['CASE_SENSITIVE'][$k] && + strpos($this->language_data['URLS'][$k], '{FNAME}') !== false) { + foreach ($this->language_data['KEYWORDS'][$k] as $word) { + if (strcasecmp($word, $keyword) == 0) { break; } } - $word = ( substr($word, 0, 4) == '<' ) ? substr($word, 4) : $word; - $word = ( substr($word, -4) == '>' ) ? substr($word, 0, - 4) : $word; - if (!$word) return ''; - - return '<|UR1|"' . - str_replace( - array('{FNAME}', '{FNAMEL}', '{FNAMEU}', '.'), - array(GeSHi::hsc($word), GeSHi::hsc(strtolower($word)), - GeSHi::hsc(strtoupper($word)), '<DOT>'), - $this->language_data['URLS'][$group] - ) . '">'; + } else { + $word = $keyword; } - return ''; - // HTML fix. Again, dirty hackage... + + $before = '<|UR1|"' . + str_replace( + array('{FNAME}', '{FNAMEL}', '{FNAMEU}', '.'), + array($this->hsc($word), $this->hsc(strtolower($word)), + $this->hsc(strtoupper($word)), '<DOT>'), + $this->language_data['URLS'][$k] + ) . '">'; + $after = '</a>'; } - else if (!($this->language == 'html4strict' && ('>' == $keyword || '<' == $keyword))) { - return '</a>'; + } + + return $before . '<|/'. $k .'/>' . $this->change_case($keyword) . '|>' . $after; + } + + /** + * handles regular expressions highlighting-definitions with callback functions + * + * @note this is a callback, don't use it directly + * + * @param array the matches array + * @return The highlighted string + * @since 1.0.8 + * @access private + */ + function handle_regexps_callback($matches) { + // before: "' style=\"' . call_user_func(\"$func\", '\\1') . '\"\\1|>'", + return ' style="' . call_user_func($this->language_data['STYLES']['REGEXPS'][$this->_rx_key], $matches[1]) . '"'. $matches[1] . '|>'; + } + + /** + * handles newlines in REGEXPS matches. Set the _hmr_* vars before calling this + * + * @note this is a callback, don't use it directly + * + * @param array the matches array + * @return string + * @since 1.0.8 + * @access private + */ + function handle_multiline_regexps($matches) { + $before = $this->_hmr_before; + $after = $this->_hmr_after; + if ($this->_hmr_replace) { + $replace = $this->_hmr_replace; + $search = array(); + + foreach (array_keys($matches) as $k) { + $search[] = '\\' . $k; } + + $before = str_replace($search, $matches, $before); + $after = str_replace($search, $matches, $after); + $replace = str_replace($search, $matches, $replace); + } else { + $replace = $matches[0]; } + return $before + . '<|!REG3XP' . $this->_hmr_key .'!>' + . str_replace("\n", "|>\n<|!REG3XP" . $this->_hmr_key . '!>', $replace) + . '|>' + . $after; } /** @@ -2230,138 +3018,193 @@ class GeSHi { * @access private * @todo BUGGY! Why? Why not build string and return? */ - function parse_non_string_part(&$stuff_to_parse) { - $stuff_to_parse = ' ' . GeSHi::hsc($stuff_to_parse); - $stuff_to_parse_pregquote = preg_quote($stuff_to_parse, '/'); - $func = '$this->change_case'; - $func2 = '$this->add_url_to_keyword'; + function parse_non_string_part($stuff_to_parse) { + $stuff_to_parse = ' ' . $this->hsc($stuff_to_parse); - // // Regular expressions - // foreach ($this->language_data['REGEXPS'] as $key => $regexp) { if ($this->lexic_permissions['REGEXPS'][$key]) { if (is_array($regexp)) { - $stuff_to_parse = preg_replace( - "/" . - str_replace('/', '\/', $regexp[GESHI_SEARCH]) . - "/{$regexp[GESHI_MODIFIERS]}", - "{$regexp[GESHI_BEFORE]}<|!REG3XP$key!>{$regexp[GESHI_REPLACE]}|>{$regexp[GESHI_AFTER]}", - $stuff_to_parse - ); - } - else { - $stuff_to_parse = preg_replace( "/(" . str_replace('/', '\/', $regexp) . ")/", "<|!REG3XP$key!>\\1|>", $stuff_to_parse); + if ($this->line_numbers != GESHI_NO_LINE_NUMBERS) { + // produce valid HTML when we match multiple lines + $this->_hmr_replace = $regexp[GESHI_REPLACE]; + $this->_hmr_before = $regexp[GESHI_BEFORE]; + $this->_hmr_key = $key; + $this->_hmr_after = $regexp[GESHI_AFTER]; + $stuff_to_parse = preg_replace_callback( + "/" . $regexp[GESHI_SEARCH] . "/{$regexp[GESHI_MODIFIERS]}", + array($this, 'handle_multiline_regexps'), + $stuff_to_parse); + $this->_hmr_replace = false; + $this->_hmr_before = ''; + $this->_hmr_after = ''; + } else { + $stuff_to_parse = preg_replace( + '/' . $regexp[GESHI_SEARCH] . '/' . $regexp[GESHI_MODIFIERS], + $regexp[GESHI_BEFORE] . '<|!REG3XP'. $key .'!>' . $regexp[GESHI_REPLACE] . '|>' . $regexp[GESHI_AFTER], + $stuff_to_parse); + } + } else { + if ($this->line_numbers != GESHI_NO_LINE_NUMBERS) { + // produce valid HTML when we match multiple lines + $this->_hmr_key = $key; + $stuff_to_parse = preg_replace_callback( "/(" . $regexp . ")/", + array($this, 'handle_multiline_regexps'), $stuff_to_parse); + $this->_hmr_key = ''; + } else { + $stuff_to_parse = preg_replace( "/(" . $regexp . ")/", "<|!REG3XP$key!>\\1|>", $stuff_to_parse); + } } } } - // - // Highlight numbers. This regexp sucks... anyone with a regexp that WORKS - // here wins a cookie if they send it to me. At the moment there's two doing - // almost exactly the same thing, except the second one prevents a number - // being highlighted twice (eg <span...><span...>5</span></span>) - // Put /NUM!/ in for the styles, which gets replaced at the end. - // - // NEW ONE: Brice Bernard - // - if ($this->lexic_permissions['NUMBERS'] && preg_match('#[0-9]#', $stuff_to_parse )) { - $stuff_to_parse = preg_replace('/([-+]?\\b(?:[0-9]*\\.)?[0-9]+\\b)/', '<|/NUM!/>\\1|>', $stuff_to_parse); + // Highlight numbers. As of 1.0.8 we support diffent types of numbers + $numbers_found = false; + if ($this->lexic_permissions['NUMBERS'] && preg_match('#\d#', $stuff_to_parse )) { + $numbers_found = true; + + //For each of the formats ... + foreach($this->language_data['NUMBERS_RXCACHE'] as $id => $regexp) { + //Check if it should be highlighted ... + $stuff_to_parse = preg_replace($regexp, "<|/NUM!$id/>\\1|>", $stuff_to_parse); + } } // Highlight keywords - // if there is a couple of alpha symbols there *might* be a keyword - if (preg_match('#[a-zA-Z]{2,}#', $stuff_to_parse)) { - $disallowed_before = "a-zA-Z0-9\$_\|\#;>|^"; - $disallowed_after = "a-zA-Z0-9_\|%\\-&"; - if(isset($this->language_data['PARSER_CONTROL'])) { - if (isset($this->language_data['PARSER_CONTROL']['KEYWORDS'])) { - if (isset($this->language_data['PARSER_CONTROL']['KEYWORDS']['DISALLOWED_BEFORE'])) { - $disallowed_before = $this->language_data['PARSER_CONTROL']['KEYWORDS']['DISALLOWED_BEFORE']; - } - if (isset($this->language_data['PARSER_CONTROL']['KEYWORDS']['DISALLOWED_AFTER'])) { - $disallowed_after = $this->language_data['PARSER_CONTROL']['KEYWORDS']['DISALLOWED_AFTER']; - } + $disallowed_before = "(?<![a-zA-Z0-9\$_\|\#;>|^&"; + $disallowed_after = "(?![a-zA-Z0-9_\|%\\-&;"; + if ($this->lexic_permissions['STRINGS']) { + $quotemarks = preg_quote(implode($this->language_data['QUOTEMARKS']), '/'); + $disallowed_before .= $quotemarks; + $disallowed_after .= $quotemarks; + } + $disallowed_before .= "])"; + $disallowed_after .= "])"; + + $parser_control_pergroup = false; + if (isset($this->language_data['PARSER_CONTROL'])) { + if (isset($this->language_data['PARSER_CONTROL']['KEYWORDS'])) { + $x = 0; // check wether per-keyword-group parser_control is enabled + if (isset($this->language_data['PARSER_CONTROL']['KEYWORDS']['DISALLOWED_BEFORE'])) { + $disallowed_before = $this->language_data['PARSER_CONTROL']['KEYWORDS']['DISALLOWED_BEFORE']; + ++$x; + } + if (isset($this->language_data['PARSER_CONTROL']['KEYWORDS']['DISALLOWED_AFTER'])) { + $disallowed_after = $this->language_data['PARSER_CONTROL']['KEYWORDS']['DISALLOWED_AFTER']; + ++$x; } + $parser_control_pergroup = (count($this->language_data['PARSER_CONTROL']['KEYWORDS']) - $x) > 0; } + } - foreach ($this->language_data['KEYWORDS'] as $k => $keywordset) { - if (!isset($this->lexic_permissions['KEYWORDS'][$k]) || - $this->lexic_permissions['KEYWORDS'][$k]) { - - $case_sensitive = $this->language_data['CASE_SENSITIVE'][$k]; - $modifiers = $case_sensitive ? 'e' : 'ie'; - $styles = "/$k/"; - - foreach ($keywordset as $keyword) { - $keyword = preg_quote($keyword, '/'); - // - // This replacement checks the word is on it's own (except if brackets etc - // are next to it), then highlights it. We don't put the color=" for the span - // in just yet - otherwise languages with the keywords "color" or "or" have - // a fit. - // - if ($case_sensitive) { - $keyword_found = strpos($stuff_to_parse_pregquote, $keyword) !== false; - } else { - $keyword_found = stristr($stuff_to_parse_pregquote, $keyword) !== false; - } - if ($keyword_found) { - // Might make a more unique string for putting the number in soon - // Basically, we don't put the styles in yet because then the styles themselves will - // get highlighted if the language has a CSS keyword in it (like CSS, for example ;)) - - $stuff_to_parse .= ' '; - $stuff_to_parse = preg_replace( - "/([^$disallowed_before])($keyword)(?!\<DOT\>(?:htm|php))(?=[^$disallowed_after])/$modifiers", - "'\\1' . $func2('\\2', '$k', 'BEGIN') . '<|$styles>' . $func('\\2') . '|>' . $func2('\\2', '$k', 'END')", - $stuff_to_parse - ); - - $stuff_to_parse = substr($stuff_to_parse, 0, -1); - } + // if this is changed, don't forget to change it below +// if (!empty($disallowed_before)) { +// $disallowed_before = "(?<![$disallowed_before])"; +// } +// if (!empty($disallowed_after)) { +// $disallowed_after = "(?![$disallowed_after])"; +// } + + foreach (array_keys($this->language_data['KEYWORDS']) as $k) { + if (!isset($this->lexic_permissions['KEYWORDS'][$k]) || + $this->lexic_permissions['KEYWORDS'][$k]) { + + $case_sensitive = $this->language_data['CASE_SENSITIVE'][$k]; + $modifiers = $case_sensitive ? '' : 'i'; + + // NEW in 1.0.8 - per-keyword-group parser control + $disallowed_before_local = $disallowed_before; + $disallowed_after_local = $disallowed_after; + if ($parser_control_pergroup && isset($this->language_data['PARSER_CONTROL']['KEYWORDS'][$k])) { + if (isset($this->language_data['PARSER_CONTROL']['KEYWORDS'][$k]['DISALLOWED_BEFORE'])) { + $disallowed_before_local = + $this->language_data['PARSER_CONTROL']['KEYWORDS'][$k]['DISALLOWED_BEFORE']; + } + + if (isset($this->language_data['PARSER_CONTROL']['KEYWORDS'][$k]['DISALLOWED_AFTER'])) { + $disallowed_after_local = + $this->language_data['PARSER_CONTROL']['KEYWORDS'][$k]['DISALLOWED_AFTER']; } } + + $this->_kw_replace_group = $k; + + //NEW in 1.0.8, the cached regexp list + // since we don't want PHP / PCRE to crash due to too large patterns we split them into smaller chunks + for ($set = 0, $set_length = count($this->language_data['CACHED_KEYWORD_LISTS'][$k]); $set < $set_length; ++$set) { + $keywordset =& $this->language_data['CACHED_KEYWORD_LISTS'][$k][$set]; + // Might make a more unique string for putting the number in soon + // Basically, we don't put the styles in yet because then the styles themselves will + // get highlighted if the language has a CSS keyword in it (like CSS, for example ;)) + $stuff_to_parse = preg_replace_callback( + "/$disallowed_before_local({$keywordset})(?!\<DOT\>(?:htm|php))$disallowed_after_local/$modifiers", + array($this, 'handle_keyword_replace'), + $stuff_to_parse + ); + } } } // // Now that's all done, replace /[number]/ with the correct styles // - foreach ($this->language_data['KEYWORDS'] as $k => $kws) { + foreach (array_keys($this->language_data['KEYWORDS']) as $k) { if (!$this->use_classes) { $attributes = ' style="' . (isset($this->language_data['STYLES']['KEYWORDS'][$k]) ? $this->language_data['STYLES']['KEYWORDS'][$k] : "") . '"'; - } - else { + } else { $attributes = ' class="kw' . $k . '"'; } - $stuff_to_parse = str_replace("/$k/", $attributes, $stuff_to_parse); + $stuff_to_parse = str_replace("<|/$k/>", "<|$attributes>", $stuff_to_parse); } - // Put number styles in - if (!$this->use_classes && $this->lexic_permissions['NUMBERS']) { - $attributes = ' style="' . $this->language_data['STYLES']['NUMBERS'][0] . '"'; - } - else { - $attributes = ' class="nu0"'; + if ($numbers_found) { + // Put number styles in + foreach($this->language_data['NUMBERS_RXCACHE'] as $id => $regexp) { +//Commented out for now, as this needs some review ... +// if ($numbers_permissions & $id) { + //Get the appropriate style ... + //Checking for unset styles is done by the style cache builder ... + if (!$this->use_classes) { + $attributes = ' style="' . $this->language_data['STYLES']['NUMBERS'][$id] . '"'; + } else { + $attributes = ' class="nu'.$id.'"'; + } + + //Set in the correct styles ... + $stuff_to_parse = str_replace("/NUM!$id/", $attributes, $stuff_to_parse); +// } + } } - $stuff_to_parse = str_replace('/NUM!/', $attributes, $stuff_to_parse); - // // Highlight methods and fields in objects - // if ($this->lexic_permissions['METHODS'] && $this->language_data['OOLANG']) { + $oolang_spaces = "[\s]*"; + $oolang_before = ""; + $oolang_after = "[a-zA-Z][a-zA-Z0-9_]*"; + if (isset($this->language_data['PARSER_CONTROL'])) { + if (isset($this->language_data['PARSER_CONTROL']['OOLANG'])) { + if (isset($this->language_data['PARSER_CONTROL']['OOLANG']['MATCH_BEFORE'])) { + $oolang_before = $this->language_data['PARSER_CONTROL']['OOLANG']['MATCH_BEFORE']; + } + if (isset($this->language_data['PARSER_CONTROL']['OOLANG']['MATCH_AFTER'])) { + $oolang_after = $this->language_data['PARSER_CONTROL']['OOLANG']['MATCH_AFTER']; + } + if (isset($this->language_data['PARSER_CONTROL']['OOLANG']['MATCH_SPACES'])) { + $oolang_spaces = $this->language_data['PARSER_CONTROL']['OOLANG']['MATCH_SPACES']; + } + } + } + foreach ($this->language_data['OBJECT_SPLITTERS'] as $key => $splitter) { - if (false !== stristr($stuff_to_parse, $splitter)) { + if (false !== strpos($stuff_to_parse, $splitter)) { if (!$this->use_classes) { $attributes = ' style="' . $this->language_data['STYLES']['METHODS'][$key] . '"'; - } - else { + } else { $attributes = ' class="me' . $key . '"'; } - $stuff_to_parse = preg_replace("/(" . preg_quote($this->language_data['OBJECT_SPLITTERS'][$key], 1) . "[\s]*)([a-zA-Z\*\(][a-zA-Z0-9_\*]*)/", "\\1<|$attributes>\\2|>", $stuff_to_parse); + $stuff_to_parse = preg_replace("/($oolang_before)(" . preg_quote($this->language_data['OBJECT_SPLITTERS'][$key], '/') . ")($oolang_spaces)($oolang_after)/", "\\1\\2\\3<|$attributes>\\4|>", $stuff_to_parse); } } } @@ -2373,132 +3216,95 @@ class GeSHi { // be highlighting regardless // if ($this->lexic_permissions['BRACKETS']) { - $code_entities_match = array('[', ']', '(', ')', '{', '}'); - if (!$this->use_classes && isset($this->language_data['STYLES']['BRACKETS'][0])) { - $code_entities_replace = array( - '<| style="' . $this->language_data['STYLES']['BRACKETS'][0] . '">[|>', - '<| style="' . $this->language_data['STYLES']['BRACKETS'][0] . '">]|>', - '<| style="' . $this->language_data['STYLES']['BRACKETS'][0] . '">(|>', - '<| style="' . $this->language_data['STYLES']['BRACKETS'][0] . '">)|>', - '<| style="' . $this->language_data['STYLES']['BRACKETS'][0] . '">{|>', - '<| style="' . $this->language_data['STYLES']['BRACKETS'][0] . '">}|>', - ); - } - else { - $code_entities_replace = array( - '<| class="br0">[|>', - '<| class="br0">]|>', - '<| class="br0">(|>', - '<| class="br0">)|>', - '<| class="br0">{|>', - '<| class="br0">}|>', - ); - } - $stuff_to_parse = str_replace( $code_entities_match, $code_entities_replace, $stuff_to_parse ); + $stuff_to_parse = str_replace( $this->language_data['CACHE_BRACKET_MATCH'], + $this->language_data['CACHE_BRACKET_REPLACE'], $stuff_to_parse ); } + //FIX for symbol highlighting ... - if($this->lexic_permissions['SYMBOLS'] && !empty($this->language_data['SYMBOLS'])) { - //As this is a costy operation, we avoid doing it for multiple groups ... - //Instead we perform it for all symbols at once. - // - //For this to work, we need to reorganize the data arrays. - $symbol_data = $symbol_preg = array(); - foreach($this->language_data['SYMBOLS'] as $key => $symbols) { - if(is_array($symbols)) { - foreach($symbols as $sym) { - if(!isset($symbol_data[$sym])) { - $symbol_data[GeSHi::hsc($sym)] = $key; - $symbol_preg[] = preg_quote(GeSHi::hsc($sym), '/'); - } - } - } else { - if(!isset($symbol_data[$symbols])) { - $symbol_data[GeSHi::hsc($symbols)] = 0; - $symbol_preg[] = preg_quote(GESHI::hsc($symbols), '/'); - } - } - } - //Now we have an array with each possible symbol as the key and the style as the actual data. - //This way we can set the correct style just the moment we highlight ... - // - //Now we need to rewrite our array to get a search string that - $sym_search = implode("|", $symbol_preg); + if ($this->lexic_permissions['SYMBOLS'] && !empty($this->language_data['SYMBOLS'])) { //Get all matches and throw away those witin a block that is already highlighted... (i.e. matched by a regexp) - preg_match_all("/(?:" . $sym_search . ")+/", $stuff_to_parse, $matches_in_stuff, PREG_OFFSET_CAPTURE | PREG_SET_ORDER); - //Match anything that is a highlighted block ... - preg_match_all("/<\|(?:<DOT>|[^>])+>(?:(?!\|>).*?)\|>|<\/a>/", $stuff_to_parse, $highlighted_in_stuff, PREG_OFFSET_CAPTURE); - foreach($matches_in_stuff as $stuff_match_id => $stuff_match_data) { - foreach($highlighted_in_stuff[0] as $highlight_id => $highlight_data) { - //Do a range check of the found highlight identifier and the OOP match ... - if(($highlight_data[1] <= $stuff_match_data[0][1]) && - ($highlight_data[1] + strlen($highlight_data[0]) >= $stuff_match_data[0][1] + strlen($stuff_match_data[0][0]))) - { - //We found a match that was already highlighted ... - unset($matches_in_stuff[$stuff_match_id]); - break; - } + $n_symbols = preg_match_all("/<\|(?:<DOT>|[^>])+>(?:(?!\|>).*?)\|>|<\/a>|(?:" . $this->language_data['SYMBOL_SEARCH'] . ")+/", $stuff_to_parse, $pot_symbols, PREG_OFFSET_CAPTURE | PREG_SET_ORDER); + $global_offset = 0; + for ($s_id = 0; $s_id < $n_symbols; ++$s_id) { + $symbol_match = $pot_symbols[$s_id][0][0]; + if (strpos($symbol_match, '<') !== false || strpos($symbol_match, '>') !== false) { + // already highlighted blocks _must_ include either < or > + // so if this conditional applies, we have to skip this match + continue; } - } - //Rebuild the matches array to be ordered by offset ... - $symbol_offsets = array(); - foreach($matches_in_stuff as $stuff_match_data) { - $symbol_offsets[$stuff_match_data[0][1]] = $stuff_match_data[0][0]; - } - krsort($symbol_offsets); - //Perform the actual replacements ... - foreach($symbol_offsets as $symbol_offset => $symbol_match) { + // if we reach this point, we have a valid match which needs to be highlighted + + $symbol_length = strlen($symbol_match); + $symbol_offset = $pot_symbols[$s_id][0][1]; + unset($pot_symbols[$s_id]); + $symbol_end = $symbol_length + $symbol_offset; $symbol_hl = ""; - $old_sym = -1; - //Split the current stuff to replace into its atomic symbols ... - preg_match_all("/$sym_search/", $symbol_match, $sym_match_syms, PREG_PATTERN_ORDER); - foreach($sym_match_syms[0] as $sym_ms) { - //Check if consequtive symbols belong to the same group to save output ... - if (isset($symbol_data[$sym_ms]) && ($symbol_data[$sym_ms] != $old_sym)) { - if(-1 != $old_sym) { - $symbol_hl .= "|>"; - } - $old_sym = $symbol_data[$sym_ms]; - if (!$this->use_classes) { - $symbol_hl .= '<| style="' . $this->language_data['STYLES']['SYMBOLS'][$old_sym] . '">'; - } - else { - $symbol_hl .= '<| class="sy' . $old_sym . '">'; + + // if we have multiple styles, we have to handle them properly + if ($this->language_data['MULTIPLE_SYMBOL_GROUPS']) { + $old_sym = -1; + // Split the current stuff to replace into its atomic symbols ... + preg_match_all("/" . $this->language_data['SYMBOL_SEARCH'] . "/", $symbol_match, $sym_match_syms, PREG_PATTERN_ORDER); + foreach ($sym_match_syms[0] as $sym_ms) { + //Check if consequtive symbols belong to the same group to save output ... + if (isset($this->language_data['SYMBOL_DATA'][$sym_ms]) + && ($this->language_data['SYMBOL_DATA'][$sym_ms] != $old_sym)) { + if (-1 != $old_sym) { + $symbol_hl .= "|>"; + } + $old_sym = $this->language_data['SYMBOL_DATA'][$sym_ms]; + if (!$this->use_classes) { + $symbol_hl .= '<| style="' . $this->language_data['STYLES']['SYMBOLS'][$old_sym] . '">'; + } else { + $symbol_hl .= '<| class="sy' . $old_sym . '">'; + } } + $symbol_hl .= $sym_ms; } - $symbol_hl .= $sym_ms; - } - //Close remaining tags and insert the replacement at the right position ... - //Take caution if symbol_hl is empty to avoid doubled closing spans. - if (-1 != $old_sym) { - $symbol_hl .= "|>"; + unset($sym_match_syms); + + //Close remaining tags and insert the replacement at the right position ... + //Take caution if symbol_hl is empty to avoid doubled closing spans. + if (-1 != $old_sym) { + $symbol_hl .= "|>"; + } + } else { + if (!$this->use_classes) { + $symbol_hl = '<| style="' . $this->language_data['STYLES']['SYMBOLS'][0] . '">'; + } else { + $symbol_hl = '<| class="sy0">'; + } + $symbol_hl .= $symbol_match . '|>'; } - $stuff_to_parse = substr($stuff_to_parse, 0, $symbol_offset) . $symbol_hl . substr($stuff_to_parse, $symbol_offset + strlen($symbol_match)); + + + $stuff_to_parse = substr_replace($stuff_to_parse, $symbol_hl, $symbol_offset + $global_offset, $symbol_length); + + // since we replace old text with something of different size, + // we'll have to keep track of the differences + $global_offset += strlen($symbol_hl) - $symbol_length; } } //FIX for symbol highlighting ... - // // Add class/style for regexps - // - foreach ($this->language_data['REGEXPS'] as $key => $regexp) { + foreach (array_keys($this->language_data['REGEXPS']) as $key) { if ($this->lexic_permissions['REGEXPS'][$key]) { if (is_callable($this->language_data['STYLES']['REGEXPS'][$key])) { - $func = $this->language_data['STYLES']['REGEXPS'][$key]; - $stuff_to_parse = preg_replace("/!REG3XP$key!(.*)\|>/eU", - "' style=\"' . call_user_func(\"$func\", '\\1') . '\"\\1|>'", $stuff_to_parse); - } - else { + $this->_rx_key = $key; + $stuff_to_parse = preg_replace_callback("/!REG3XP$key!(.*)\|>/U", + array($this, 'handle_regexps_callback'), + $stuff_to_parse); + } else { if (!$this->use_classes) { $attributes = ' style="' . $this->language_data['STYLES']['REGEXPS'][$key] . '"'; - } - else { + } else { if (is_array($this->language_data['REGEXPS'][$key]) && array_key_exists(GESHI_CLASS, $this->language_data['REGEXPS'][$key])) { $attributes = ' class="' . $this->language_data['REGEXPS'][$key][GESHI_CLASS] . '"'; - } - else { + } else { $attributes = ' class="re' . $key . '"'; } } @@ -2513,12 +3319,10 @@ class GeSHi { if (isset($this->link_styles[GESHI_LINK])) { if ($this->use_classes) { $stuff_to_parse = str_replace('<|UR1|', '<a' . $this->link_target . ' href=', $stuff_to_parse); - } - else { + } else { $stuff_to_parse = str_replace('<|UR1|', '<a' . $this->link_target . ' style="' . $this->link_styles[GESHI_LINK] . '" href=', $stuff_to_parse); } - } - else { + } else { $stuff_to_parse = str_replace('<|UR1|', '<a' . $this->link_target . ' href=', $stuff_to_parse); } @@ -2528,7 +3332,6 @@ class GeSHi { $stuff_to_parse = str_replace('<|', '<span', $stuff_to_parse); $stuff_to_parse = str_replace ( '|>', '</span>', $stuff_to_parse ); - return substr($stuff_to_parse, 1); } @@ -2557,6 +3360,45 @@ class GeSHi { } /** + * Merges arrays recursively, overwriting values of the first array with values of later arrays + * + * @since 1.0.8 + * @access private + */ + function merge_arrays() { + $arrays = func_get_args(); + $narrays = count($arrays); + + // check arguments + // comment out if more performance is necessary (in this case the foreach loop will trigger a warning if the argument is not an array) + for ($i = 0; $i < $narrays; $i ++) { + if (!is_array($arrays[$i])) { + // also array_merge_recursive returns nothing in this case + trigger_error('Argument #' . ($i+1) . ' is not an array - trying to merge array with scalar! Returning false!', E_USER_WARNING); + return false; + } + } + + // the first array is in the output set in every case + $ret = $arrays[0]; + + // merege $ret with the remaining arrays + for ($i = 1; $i < $narrays; $i ++) { + foreach ($arrays[$i] as $key => $value) { + if (is_array($value) && isset($ret[$key])) { + // if $ret[$key] is not an array you try to merge an scalar value with an array - the result is not defined (incompatible arrays) + // in this case the call will trigger an E_USER_WARNING and the $ret[$key] will be false. + $ret[$key] = $this->merge_arrays($ret[$key], $value); + } else { + $ret[$key] = $value; + } + } + } + + return $ret; + } + + /** * Gets language information and stores it for later use * * @param string The filename of the language file you want to load @@ -2569,28 +3411,85 @@ class GeSHi { // this file is already loaded! return; } + + //Prepare some stuff before actually loading the language file $this->loaded_language = $file_name; + $this->parse_cache_built = false; $this->enable_highlighting(); $language_data = array(); + + //Load the language file require $file_name; + // Perhaps some checking might be added here later to check that // $language data is a valid thing but maybe not $this->language_data = $language_data; + // Set strict mode if should be set - if ($this->language_data['STRICT_MODE_APPLIES'] == GESHI_ALWAYS) { - $this->strict_mode = true; - } + $this->strict_mode = $this->language_data['STRICT_MODE_APPLIES']; + // Set permissions for all lexics to true // so they'll be highlighted by default - foreach ($this->language_data['KEYWORDS'] as $key => $words) { - $this->lexic_permissions['KEYWORDS'][$key] = true; + foreach (array_keys($this->language_data['KEYWORDS']) as $key) { + if (!empty($this->language_data['KEYWORDS'][$key])) { + $this->lexic_permissions['KEYWORDS'][$key] = true; + } else { + $this->lexic_permissions['KEYWORDS'][$key] = false; + } } - foreach ($this->language_data['COMMENT_SINGLE'] as $key => $comment) { + + foreach (array_keys($this->language_data['COMMENT_SINGLE']) as $key) { $this->lexic_permissions['COMMENTS'][$key] = true; } - foreach ($this->language_data['REGEXPS'] as $key => $regexp) { + foreach (array_keys($this->language_data['REGEXPS']) as $key) { $this->lexic_permissions['REGEXPS'][$key] = true; } + + // for BenBE and future code reviews: + // we can use empty here since we only check for existance and emptiness of an array + // if it is not an array at all but rather false or null this will work as intended as well + // even if $this->language_data['PARSER_CONTROL'] is undefined this won't trigger a notice + if (!empty($this->language_data['PARSER_CONTROL']['ENABLE_FLAGS'])) { + foreach ($this->language_data['PARSER_CONTROL']['ENABLE_FLAGS'] as $flag => $value) { + // it's either true or false and maybe is true as well + $perm = $value !== GESHI_NEVER; + if ($flag == 'ALL') { + $this->enable_highlighting($perm); + continue; + } + if (!isset($this->lexic_permissions[$flag])) { + // unknown lexic permission + continue; + } + if (is_array($this->lexic_permissions[$flag])) { + foreach ($this->lexic_permissions[$flag] as $key => $val) { + $this->lexic_permissions[$flag][$key] = $perm; + } + } else { + $this->lexic_permissions[$flag] = $perm; + } + } + unset($this->language_data['PARSER_CONTROL']['ENABLE_FLAGS']); + } + + //NEW in 1.0.8: Allow styles to be loaded from a separate file to override defaults + $style_filename = substr($file_name, 0, -4) . '.style.php'; + if (is_readable($style_filename)) { + //Clear any style_data that could have been set before ... + if (isset($style_data)) { + unset($style_data); + } + + //Read the Style Information from the style file + include $style_filename; + + //Apply the new styles to our current language styles + if (isset($style_data) && is_array($style_data)) { + $this->language_data['STYLES'] = + $this->merge_arrays($this->language_data['STYLES'], $style_data); + } + } + // Set default class for CSS $this->overall_class = $this->language; } @@ -2599,8 +3498,7 @@ class GeSHi { * Takes the parsed code and various options, and creates the HTML * surrounding it to make it look nice. * - * @param string The code already parsed - * @return string The code nicely finalised + * @param string The code already parsed (reference!) * @since 1.0.0 * @access private */ @@ -2609,18 +3507,18 @@ class GeSHi { // This is BUGGY!! My fault for bad code: fix coming in 1.2 // @todo Remove this crap if ($this->enable_important_blocks && - (strpos($parsed_code, GeSHi::hsc(GESHI_START_IMPORTANT)) === false)) { - $parsed_code = str_replace(GeSHi::hsc(GESHI_END_IMPORTANT), '', $parsed_code); + (strpos($parsed_code, $this->hsc(GESHI_START_IMPORTANT)) === false)) { + $parsed_code = str_replace($this->hsc(GESHI_END_IMPORTANT), '', $parsed_code); } // Add HTML whitespace stuff if we're using the <div> header - if ($this->header_type != GESHI_HEADER_PRE) { - $parsed_code = $this->indent($parsed_code); + if ($this->header_type != GESHI_HEADER_PRE && $this->header_type != GESHI_HEADER_PRE_VALID) { + $this->indent($parsed_code); } // purge some unnecessary stuff + /** NOTE: memorypeak #1 */ $parsed_code = preg_replace('#<span[^>]+>(\s*)</span>#', '\\1', $parsed_code); - $parsed_code = preg_replace('#<div[^>]+>(\s*)</div>#', '\\1', $parsed_code); // If we are using IDs for line numbers, there needs to be an overall // ID set to prevent collisions. @@ -2629,15 +3527,16 @@ class GeSHi { } // Get code into lines + /** NOTE: memorypeak #2 */ $code = explode("\n", $parsed_code); - $parsed_code = ''; + $parsed_code = $this->header(); // If we're using line numbers, we insert <li>s and appropriate // markup to style them (otherwise we don't need to do anything) - if ($this->line_numbers != GESHI_NO_LINE_NUMBERS) { + if ($this->line_numbers != GESHI_NO_LINE_NUMBERS && $this->header_type != GESHI_HEADER_PRE_TABLE) { // If we're using the <pre> header, we shouldn't add newlines because // the <pre> will line-break them (and the <li>s already do this for us) - $ls = ($this->header_type != GESHI_HEADER_PRE) ? "\n" : ''; + $ls = ($this->header_type != GESHI_HEADER_PRE && $this->header_type != GESHI_HEADER_PRE_VALID) ? "\n" : ''; // Set vars to defaults for following loop $i = 0; @@ -2661,8 +3560,7 @@ class GeSHi { //$attr = ' class="li2"'; $attrs['class'][] = 'li2'; $def_attr = ' class="de2"'; - } - else { + } else { //$attr = ' style="' . $this->line_style2 . '"'; $attrs['style'][] = $this->line_style2; // This style "covers up" the special styles set for special lines @@ -2670,21 +3568,24 @@ class GeSHi { // code on that line $def_attr = ' style="' . $this->code_style . '"'; } - // Span or div? - $start = "<div$def_attr>"; - $end = '</div>'; - } - else { + } else { if ($this->use_classes) { //$attr = ' class="li1"'; $attrs['class'][] = 'li1'; $def_attr = ' class="de1"'; - } - else { + } else { //$attr = ' style="' . $this->line_style1 . '"'; $attrs['style'][] = $this->line_style1; $def_attr = ' style="' . $this->code_style . '"'; } + } + + //Check which type of tag to insert for this line + if ($this->header_type == GESHI_HEADER_PRE_VALID) { + $start = "<pre$def_attr>"; + $end = '</pre>'; + } else { + // Span or div? $start = "<div$def_attr>"; $end = '</div>'; } @@ -2696,6 +3597,7 @@ class GeSHi { $attrs['id'][] = "$this->overall_id-$i"; } + //Is this some line with extra styles??? if (in_array($i, $this->highlight_extra_lines)) { if ($this->use_classes) { if (isset($this->highlight_extra_lines_styles[$i])) { @@ -2714,38 +3616,98 @@ class GeSHi { $attr_string .= ' ' . $key . '="' . implode(' ', $attr) . '"'; } - $parsed_code .= "<li$attr_string>$start{$code[$i - 1]}$end</li>$ls"; + $parsed_code .= "<li$attr_string>$start{$code[$i-1]}$end</li>$ls"; + unset($code[$i - 1]); + } + } else { + $n = count($code); + if ($this->use_classes) { + $attributes = ' class="de1"'; + } else { + $attributes = ' style="'. $this->code_style .'"'; + } + if ($this->header_type == GESHI_HEADER_PRE_VALID) { + $parsed_code .= '<pre'. $attributes .'>'; + } elseif ($this->header_type == GESHI_HEADER_PRE_TABLE) { + if ($this->line_numbers != GESHI_NO_LINE_NUMBERS) { + if ($this->use_classes) { + $attrs = ' class="ln"'; + } else { + $attrs = ' style="'. $this->table_linenumber_style .'"'; + } + $parsed_code .= '<td'.$attrs.'><pre>'; + // get linenumbers + // we don't merge it with the for below, since it should be better for + // memory consumption this way + for ($i = 1; $i <= $n; ++$i) { + $parsed_code .= $i; + if ($i != $n) { + $parsed_code .= "\n"; + } + } + $parsed_code .= '</pre></td><td>'; + } + $parsed_code .= '<pre'. $attributes .'>'; } - } - else { // No line numbers, but still need to handle highlighting lines extra. // Have to use divs so the full width of the code is highlighted - for ($i = 0, $n = count($code); $i < $n; ++$i) { + $close = 0; + for ($i = 0; $i < $n; ++$i) { // Make lines have at least one space in them if they're empty // BenBE: Checking emptiness using trim instead of relying on blanks if ('' == trim($code[$i])) { $code[$i] = ' '; } + // fancy lines + if ($this->line_numbers == GESHI_FANCY_LINE_NUMBERS && + $i % $this->line_nth_row == ($this->line_nth_row - 1)) { + // Set the attributes to style the line + if ($this->use_classes) { + $parsed_code .= '<span class="xtra li2"><span class="de2">'; + } else { + // This style "covers up" the special styles set for special lines + // so that styles applied to special lines don't apply to the actual + // code on that line + $parsed_code .= '<span style="display:block;' . $this->line_style2 . '">' + .'<span style="' . $this->code_style .'">'; + } + $close += 2; + } + //Is this some line with extra styles??? if (in_array($i + 1, $this->highlight_extra_lines)) { if ($this->use_classes) { if (isset($this->highlight_extra_lines_styles[$i])) { - $parsed_code .= "<div class=\"lx$i\">"; + $parsed_code .= "<span class=\"xtra lx$i\">"; } else { - $parsed_code .= "<div class=\"ln-xtra\">"; + $parsed_code .= "<span class=\"xtra ln-xtra\">"; } } else { - $parsed_code .= "<div style=\"" . $this->get_line_style($i) . "\">"; + $parsed_code .= "<span style=\"display:block;" . $this->get_line_style($i) . "\">"; } - // Remove \n because it stuffs up <pre> header - $parsed_code .= $code[$i] . "</div>"; - } else { - $parsed_code .= $code[$i] . "\n"; + ++$close; } + + $parsed_code .= $code[$i]; + + if ($close) { + $parsed_code .= str_repeat('</span>', $close); + $close = 0; + } + elseif ($i + 1 < $n) { + $parsed_code .= "\n"; + } + unset($code[$i]); + } + + if ($this->header_type == GESHI_HEADER_PRE_VALID || $this->header_type == GESHI_HEADER_PRE_TABLE) { + $parsed_code .= '</pre>'; + } + if ($this->header_type == GESHI_HEADER_PRE_TABLE && $this->line_numbers != GESHI_NO_LINE_NUMBERS) { + $parsed_code .= '</td>'; } } - unset($code); - return $this->header() . chop($parsed_code) . $this->footer(); + $parsed_code .= $this->footer(); } /** @@ -2757,7 +3719,22 @@ class GeSHi { */ function header() { // Get attributes needed - $attributes = $this->get_attributes(); + /** + * @todo Document behaviour change - class is outputted regardless of whether + * we're using classes or not. Same with style + */ + $attributes = ' class="' . $this->language; + if ($this->overall_class != '') { + $attributes .= " ".$this->overall_class; + } + $attributes .= '"'; + + if ($this->overall_id != '') { + $attributes .= " id=\"{$this->overall_id}\""; + } + if ($this->overall_style != '') { + $attributes .= ' style="' . $this->overall_style . '"'; + } $ol_attributes = ''; @@ -2766,31 +3743,47 @@ class GeSHi { } // Get the header HTML - $header = $this->format_header_content(); + $header = $this->header_content; + if ($header) { + if ($this->header_type == GESHI_HEADER_PRE || $this->header_type == GESHI_HEADER_PRE_VALID) { + $header = str_replace("\n", '', $header); + } + $header = $this->replace_keywords($header); + + if ($this->use_classes) { + $attr = ' class="head"'; + } else { + $attr = " style=\"{$this->header_content_style}\""; + } + if ($this->header_type == GESHI_HEADER_PRE_TABLE && $this->line_numbers != GESHI_NO_LINE_NUMBERS) { + $header = "<thead><tr><td colspan=\"2\" $attr>$header</td></tr></thead>"; + } else { + $header = "<div$attr>$header</div>"; + } + } if (GESHI_HEADER_NONE == $this->header_type) { if ($this->line_numbers != GESHI_NO_LINE_NUMBERS) { - return "$header<ol$ol_attributes>"; + return "$header<ol$attributes$ol_attributes>"; } - return $header . - ($this->force_code_block ? '<div>' : ''); + return $header . ($this->force_code_block ? '<div>' : ''); } // Work out what to return and do it if ($this->line_numbers != GESHI_NO_LINE_NUMBERS) { if ($this->header_type == GESHI_HEADER_PRE) { return "<pre$attributes>$header<ol$ol_attributes>"; - } - else if ($this->header_type == GESHI_HEADER_DIV) { + } else if ($this->header_type == GESHI_HEADER_DIV || + $this->header_type == GESHI_HEADER_PRE_VALID) { return "<div$attributes>$header<ol$ol_attributes>"; + } else if ($this->header_type == GESHI_HEADER_PRE_TABLE) { + return "<table$attributes>$header<tbody><tr class=\"li1\">"; } - } - else { + } else { if ($this->header_type == GESHI_HEADER_PRE) { return "<pre$attributes>$header" . ($this->force_code_block ? '<div>' : ''); - } - else if ($this->header_type == GESHI_HEADER_DIV) { + } else { return "<div$attributes>$header" . ($this->force_code_block ? '<div>' : ''); } @@ -2798,83 +3791,56 @@ class GeSHi { } /** - * Returns the header content, formatted for output + * Returns the footer for the code block. * - * @return string The header content, formatted for output - * @since 1.0.2 + * @return string The footer for the code block + * @since 1.0.0 * @access private */ - function format_header_content() { - $header = $this->header_content; - if ($header) { + function footer() { + $footer = $this->footer_content; + if ($footer) { if ($this->header_type == GESHI_HEADER_PRE) { - $header = str_replace("\n", '', $header); + $footer = str_replace("\n", '', $footer);; } - $header = $this->replace_keywords($header); + $footer = $this->replace_keywords($footer); if ($this->use_classes) { - $attr = ' class="head"'; + $attr = ' class="foot"'; + } else { + $attr = " style=\"{$this->footer_content_style}\""; } - else { - $attr = " style=\"{$this->header_content_style}\""; + if ($this->header_type == GESHI_HEADER_PRE_TABLE && $this->linenumbers != GESHI_NO_LINE_NUMBERS) { + $footer = "<tfoot><tr><td colspan=\"2\">$footer</td></tr></tfoot>"; + } else { + $footer = "<div$attr>$footer</div>"; } - return "<div$attr>$header</div>"; } - } - - /** - * Returns the footer for the code block. - * - * @return string The footer for the code block - * @since 1.0.0 - * @access private - */ - function footer() { - $footer_content = $this->format_footer_content(); if (GESHI_HEADER_NONE == $this->header_type) { - return ($this->line_numbers != GESHI_NO_LINE_NUMBERS) ? '</ol>' . $footer_content - : $footer_content; + return ($this->line_numbers != GESHI_NO_LINE_NUMBERS) ? '</ol>' . $footer : $footer; } - if ($this->header_type == GESHI_HEADER_DIV) { + if ($this->header_type == GESHI_HEADER_DIV || $this->header_type == GESHI_HEADER_PRE_VALID) { if ($this->line_numbers != GESHI_NO_LINE_NUMBERS) { - return "</ol>$footer_content</div>"; + return "</ol>$footer</div>"; } return ($this->force_code_block ? '</div>' : '') . - "$footer_content</div>"; + "$footer</div>"; } - else { + elseif ($this->header_type == GESHI_HEADER_PRE_TABLE) { if ($this->line_numbers != GESHI_NO_LINE_NUMBERS) { - return "</ol>$footer_content</pre>"; + return "</tr></tbody>$footer</table>"; } return ($this->force_code_block ? '</div>' : '') . - "$footer_content</pre>"; + "$footer</div>"; } - } - - /** - * Returns the footer content, formatted for output - * - * @return string The footer content, formatted for output - * @since 1.0.2 - * @access private - */ - function format_footer_content() { - $footer = $this->footer_content; - if ($footer) { - if ($this->header_type == GESHI_HEADER_PRE) { - $footer = str_replace("\n", '', $footer);; - } - $footer = $this->replace_keywords($footer); - - if ($this->use_classes) { - $attr = ' class="foot"'; - } - else { - $attr = " style=\"{$this->footer_content_style}\""; + else { + if ($this->line_numbers != GESHI_NO_LINE_NUMBERS) { + return "</ol>$footer</pre>"; } - return "<div$attr>$footer</div>"; + return ($this->force_code_block ? '</div>' : '') . + "$footer</pre>"; } } @@ -2892,41 +3858,31 @@ class GeSHi { $keywords[] = '<TIME>'; $keywords[] = '{TIME}'; - $replacements[] = $replacements[] = number_format($this->get_time(), 3); + $replacements[] = $replacements[] = number_format($time = $this->get_time(), 3); $keywords[] = '<LANGUAGE>'; $keywords[] = '{LANGUAGE}'; - $replacements[] = $replacements[] = $this->language; + $replacements[] = $replacements[] = $this->language_data['LANG_NAME']; $keywords[] = '<VERSION>'; $keywords[] = '{VERSION}'; $replacements[] = $replacements[] = GESHI_VERSION; - return str_replace($keywords, $replacements, $instr); - } - - /** - * Gets the CSS attributes for this code - * - * @return The CSS attributes for this code - * @since 1.0.0 - * @access private - * @todo Document behaviour change - class is outputted regardless of whether we're using classes or not. - * Same with style - */ - function get_attributes() { - $attributes = ''; - - if ($this->overall_class != '') { - $attributes .= " class=\"{$this->overall_class}\""; - } - if ($this->overall_id != '') { - $attributes .= " id=\"{$this->overall_id}\""; - } - if ($this->overall_style != '') { - $attributes .= ' style="' . $this->overall_style . '"'; + $keywords[] = '<SPEED>'; + $keywords[] = '{SPEED}'; + if ($time <= 0) { + $speed = 'N/A'; + } else { + $speed = strlen($this->source) / $time; + if ($speed >= 1024) { + $speed = sprintf("%.2f KB/s", $speed / 1024.0); + } else { + $speed = sprintf("%.0f B/s", $speed); + } } - return $attributes; + $replacements[] = $replacements[] = $speed; + + return str_replace($keywords, $replacements, $instr); } /** @@ -2982,9 +3938,9 @@ class GeSHi { * @return string converted string * @since 1.0.7.18 */ - function hsc($string, $quote_style=ENT_COMPAT) { + function hsc($string, $quote_style = ENT_COMPAT) { // init - $aTransSpecchar = array( + static $aTransSpecchar = array( '&' => '&', '"' => '"', '<' => '<', @@ -3000,17 +3956,17 @@ class GeSHi { '|' => '<PIPE>' // Force | to be processed as entity ); // ENT_COMPAT set - if (ENT_NOQUOTES == $quote_style) // don't convert double quotes - { - unset($aTransSpecchar['"']); - } - elseif (ENT_QUOTES == $quote_style) // convert single quotes as well - { - $aTransSpecchar["'"] = '''; // (apos) htmlspecialchars() uses ''' + switch ($quote_style) { + case ENT_NOQUOTES: // don't convert double quotes + unset($aTransSpecchar['"']); + break; + case ENT_QUOTES: // convert single quotes as well + $aTransSpecchar["'"] = '''; // (apos) htmlspecialchars() uses ''' + break; } // return translated string - return strtr($string,$aTransSpecchar); + return strtr($string, $aTransSpecchar); } /** @@ -3029,11 +3985,25 @@ class GeSHi { if ($this->error) { return ''; } + + //Check if the style rearrangements have been processed ... + //This also does some preprocessing to check which style groups are useable ... + if(!isset($this->language_data['NUMBERS_CACHE'])) { + $this->build_style_cache(); + } + // First, work out what the selector should be. If there's an ID, // that should be used, the same for a class. Otherwise, a selector // of '' means that these styles will be applied anywhere - $selector = ($this->overall_id != '') ? "#{$this->overall_id} " : ''; - $selector = ($selector == '' && $this->overall_class != '') ? ".{$this->overall_class} " : $selector; + if ($this->overall_id) { + $selector = '#' . $this->overall_id; + } else { + $selector = '.' . $this->language; + if ($this->overall_class) { + $selector .= '.' . $this->overall_class; + } + } + $selector .= ' '; // Header of the stylesheet if (!$economy_mode) { @@ -3046,7 +4016,7 @@ class GeSHi { " * (http://qbnz.com/highlighter/ and http://geshi.org/)\n". " * --------------------------------------\n". " */\n"; - } else { + } else { $stylesheet = "/**\n". " * GeSHi (C) 2004 - 2007 Nigel McNie, 2007 - 2008 Benny Baumann\n" . " * (http://qbnz.com/highlighter/ and http://geshi.org/)\n". @@ -3063,115 +4033,132 @@ class GeSHi { } // Add overall styles - if (!$economy_mode || $this->overall_style != '') { + // note: neglect economy_mode, empty styles are meaningless + if ($this->overall_style != '') { $stylesheet .= "$selector {{$this->overall_style}}\n"; } // Add styles for links + // note: economy mode does not make _any_ sense here + // either the style is empty and thus no selector is needed + // or the appropriate key is given. foreach ($this->link_styles as $key => $style) { - if (!$economy_mode || $key == GESHI_LINK && $style != '') { - $stylesheet .= "{$selector}a:link {{$style}}\n"; - } - if (!$economy_mode || $key == GESHI_HOVER && $style != '') { - $stylesheet .= "{$selector}a:hover {{$style}}\n"; - } - if (!$economy_mode || $key == GESHI_ACTIVE && $style != '') { - $stylesheet .= "{$selector}a:active {{$style}}\n"; - } - if (!$economy_mode || $key == GESHI_VISITED && $style != '') { - $stylesheet .= "{$selector}a:visited {{$style}}\n"; + if ($style != '') { + switch ($key) { + case GESHI_LINK: + $stylesheet .= "{$selector}a:link {{$style}}\n"; + break; + case GESHI_HOVER: + $stylesheet .= "{$selector}a:hover {{$style}}\n"; + break; + case GESHI_ACTIVE: + $stylesheet .= "{$selector}a:active {{$style}}\n"; + break; + case GESHI_VISITED: + $stylesheet .= "{$selector}a:visited {{$style}}\n"; + break; + } } } // Header and footer - if (!$economy_mode || $this->header_content_style != '') { + // note: neglect economy_mode, empty styles are meaningless + if ($this->header_content_style != '') { $stylesheet .= "$selector.head {{$this->header_content_style}}\n"; } - if (!$economy_mode || $this->footer_content_style != '') { + if ($this->footer_content_style != '') { $stylesheet .= "$selector.foot {{$this->footer_content_style}}\n"; } // Styles for important stuff - if (!$economy_mode || $this->important_styles != '') { + // note: neglect economy_mode, empty styles are meaningless + if ($this->important_styles != '') { $stylesheet .= "$selector.imp {{$this->important_styles}}\n"; } // Simple line number styles - if (!$economy_mode || ($this->line_numbers != GESHI_NO_LINE_NUMBERS && $this->line_style1 != '')) { - $stylesheet .= "{$selector}li, {$selector}li.li1 {{$this->line_style1}}\n"; + if ((!$economy_mode || $this->line_numbers != GESHI_NO_LINE_NUMBERS) && $this->line_style1 != '') { + $stylesheet .= "{$selector}li, {$selector}.li1 {{$this->line_style1}}\n"; + } + if ((!$economy_mode || $this->line_numbers != GESHI_NO_LINE_NUMBERS) && $this->table_linenumber_style != '') { + $stylesheet .= "{$selector}.ln {{$this->table_linenumber_style}}\n"; } // If there is a style set for fancy line numbers, echo it out - if (!$economy_mode || ($this->line_numbers == GESHI_FANCY_LINE_NUMBERS && $this->line_style2 != '')) { - $stylesheet .= "{$selector}li.li2 {{$this->line_style2}}\n"; + if ((!$economy_mode || $this->line_numbers == GESHI_FANCY_LINE_NUMBERS) && $this->line_style2 != '') { + $stylesheet .= "{$selector}.li2 {{$this->line_style2}}\n"; } + // note: empty styles are meaningless foreach ($this->language_data['STYLES']['KEYWORDS'] as $group => $styles) { - if (!$economy_mode || ($economy_mode && $styles != '') && + if ($styles != '' && (!$economy_mode || (isset($this->lexic_permissions['KEYWORDS'][$group]) && - $this->lexic_permissions['KEYWORDS'][$group])) { + $this->lexic_permissions['KEYWORDS'][$group]))) { $stylesheet .= "$selector.kw$group {{$styles}}\n"; } } foreach ($this->language_data['STYLES']['COMMENTS'] as $group => $styles) { - if (!$economy_mode || ($economy_mode && $styles != '') && + if ($styles != '' && (!$economy_mode || (isset($this->lexic_permissions['COMMENTS'][$group]) && - $this->lexic_permissions['COMMENTS'][$group])) { + $this->lexic_permissions['COMMENTS'][$group]) || + (!empty($this->language_data['COMMENT_REGEXP']) && + !empty($this->language_data['COMMENT_REGEXP'][$group])))) { $stylesheet .= "$selector.co$group {{$styles}}\n"; } } foreach ($this->language_data['STYLES']['ESCAPE_CHAR'] as $group => $styles) { - if (!$economy_mode || !($economy_mode && $styles == '') && !($economy_mode && - !$this->lexic_permissions['ESCAPE_CHAR'])) { + if ($styles != '' && (!$economy_mode || $this->lexic_permissions['ESCAPE_CHAR'])) { + // NEW: since 1.0.8 we have to handle hardescapes + if ($group == 'HARD') { + $group = '_h'; + } $stylesheet .= "$selector.es$group {{$styles}}\n"; } } foreach ($this->language_data['STYLES']['BRACKETS'] as $group => $styles) { - if (!$economy_mode || !($economy_mode && $styles == '') && !($economy_mode && - !$this->lexic_permissions['BRACKETS'])) { + if ($styles != '' && (!$economy_mode || $this->lexic_permissions['BRACKETS'])) { $stylesheet .= "$selector.br$group {{$styles}}\n"; } } foreach ($this->language_data['STYLES']['SYMBOLS'] as $group => $styles) { - if (!$economy_mode || !($economy_mode && $styles == '') && !($economy_mode && - !$this->lexic_permissions['SYMBOLS'])) { + if ($styles != '' && (!$economy_mode || $this->lexic_permissions['SYMBOLS'])) { $stylesheet .= "$selector.sy$group {{$styles}}\n"; } } foreach ($this->language_data['STYLES']['STRINGS'] as $group => $styles) { - if (!$economy_mode || !($economy_mode && $styles == '') && !($economy_mode && - !$this->lexic_permissions['STRINGS'])) { + if ($styles != '' && (!$economy_mode || $this->lexic_permissions['STRINGS'])) { + // NEW: since 1.0.8 we have to handle hardquotes + if ($group === 'HARD') { + $group = '_h'; + } $stylesheet .= "$selector.st$group {{$styles}}\n"; } } foreach ($this->language_data['STYLES']['NUMBERS'] as $group => $styles) { - if (!$economy_mode || !($economy_mode && $styles == '') && !($economy_mode && - !$this->lexic_permissions['NUMBERS'])) { + if ($styles != '' && (!$economy_mode || $this->lexic_permissions['NUMBERS'])) { $stylesheet .= "$selector.nu$group {{$styles}}\n"; } } foreach ($this->language_data['STYLES']['METHODS'] as $group => $styles) { - if (!$economy_mode || !($economy_mode && $styles == '') && !($economy_mode && - !$this->lexic_permissions['METHODS'])) { + if ($styles != '' && (!$economy_mode || $this->lexic_permissions['METHODS'])) { $stylesheet .= "$selector.me$group {{$styles}}\n"; } } + // note: neglect economy_mode, empty styles are meaningless foreach ($this->language_data['STYLES']['SCRIPT'] as $group => $styles) { - if (!$economy_mode || !($economy_mode && $styles == '')) { + if ($styles != '') { $stylesheet .= "$selector.sc$group {{$styles}}\n"; } } foreach ($this->language_data['STYLES']['REGEXPS'] as $group => $styles) { - if (!$economy_mode || ($economy_mode && $styles != '') && + if ($styles != '' && (!$economy_mode || (isset($this->lexic_permissions['REGEXPS'][$group]) && - $this->lexic_permissions['REGEXPS'][$group])) { + $this->lexic_permissions['REGEXPS'][$group]))) { if (is_array($this->language_data['REGEXPS'][$group]) && - array_key_exists(GESHI_CLASS, - $this->language_data['REGEXPS'][$group])) { + array_key_exists(GESHI_CLASS, $this->language_data['REGEXPS'][$group])) { $stylesheet .= "$selector."; $stylesheet .= $this->language_data['REGEXPS'][$group][GESHI_CLASS]; $stylesheet .= " {{$styles}}\n"; - } - else { + } else { $stylesheet .= "$selector.re$group {{$styles}}\n"; } } @@ -3180,6 +4167,7 @@ class GeSHi { if (!$economy_mode || (count($this->highlight_extra_lines)!=count($this->highlight_extra_lines_styles))) { $stylesheet .= "{$selector}.ln-xtra, {$selector}li.ln-xtra, {$selector}div.ln-xtra {{$this->highlight_extra_lines_style}}\n"; } + $stylesheet .= "{$selector}span.xtra { display:block; }\n"; foreach ($this->highlight_extra_lines_styles as $lineid => $linestyle) { $stylesheet .= "{$selector}.lx$lineid, {$selector}li.lx$lineid, {$selector}div.lx$lineid {{$linestyle}}\n"; } @@ -3205,6 +4193,162 @@ class GeSHi { return $style; } + + /** + * this functions creates an optimized regular expression list + * of an array of strings. + * + * Example: + * <code>$list = array('faa', 'foo', 'foobar'); + * => string 'f(aa|oo(bar)?)'</code> + * + * @param $list array of (unquoted) strings + * @param $regexp_delimiter your regular expression delimiter, @see preg_quote() + * @return string for regular expression + * @author Milian Wolff <mail@milianw.de> + * @since 1.0.8 + * @access private + */ + function optimize_regexp_list($list, $regexp_delimiter = '/') { + $regex_chars = array('.', '\\', '+', '*', '?', '[', '^', ']', '$', + '(', ')', '{', '}', '=', '!', '<', '>', '|', ':', $regexp_delimiter); + sort($list); + $regexp_list = array(''); + $num_subpatterns = 0; + $list_key = 0; + + // the tokens which we will use to generate the regexp list + $tokens = array(); + $prev_keys = array(); + // go through all entries of the list and generate the token list + for ($i = 0, $i_max = count($list); $i < $i_max; ++$i) { + $level = 0; + $entry = preg_quote((string) $list[$i], $regexp_delimiter); + $pointer = &$tokens; + // properly assign the new entry to the correct position in the token array + // possibly generate smaller common denominator keys + while (true) { + // get the common denominator + if (isset($prev_keys[$level])) { + if ($prev_keys[$level] == $entry) { + // this is a duplicate entry, skip it + continue 2; + } + $char = 0; + while (isset($entry[$char]) && isset($prev_keys[$level][$char]) + && $entry[$char] == $prev_keys[$level][$char]) { + ++$char; + } + if ($char > 0) { + // this entry has at least some chars in common with the current key + if ($char == strlen($prev_keys[$level])) { + // current key is totally matched, i.e. this entry has just some bits appended + $pointer = &$pointer[$prev_keys[$level]]; + } else { + // only part of the keys match + $new_key_part1 = substr($prev_keys[$level], 0, $char); + $new_key_part2 = substr($prev_keys[$level], $char); + if (in_array($new_key_part1[0], $regex_chars) + || in_array($new_key_part2[0], $regex_chars)) { + // this is bad, a regex char as first character + $pointer[$entry] = array('' => true); + array_splice($prev_keys, $level, count($prev_keys), $entry); + continue; + } else { + // relocate previous tokens + $pointer[$new_key_part1] = array($new_key_part2 => $pointer[$prev_keys[$level]]); + unset($pointer[$prev_keys[$level]]); + $pointer = &$pointer[$new_key_part1]; + // recreate key index + array_splice($prev_keys, $level, count($prev_keys), array($new_key_part1, $new_key_part2)); + } + } + ++$level; + $entry = substr($entry, $char); + continue; + } + // else: fall trough, i.e. no common denominator was found + } + if ($level == 0 && !empty($tokens)) { + // we can dump current tokens into the string and throw them away afterwards + $new_entry = $this->_optimize_regexp_list_tokens_to_string($tokens); + $new_subpatterns = substr_count($new_entry, '(?:'); + if (GESHI_MAX_PCRE_SUBPATTERNS && $num_subpatterns + $new_subpatterns > GESHI_MAX_PCRE_SUBPATTERNS) { + $regexp_list[++$list_key] = $new_entry; + $num_subpatterns = $new_subpatterns; + } else { + if (!empty($regexp_list[$list_key])) { + $new_entry = '|' . $new_entry; + } + $regexp_list[$list_key] .= $new_entry; + $num_subpatterns += $new_subpatterns; + } + $tokens = array(); + } + // no further common denominator found + $pointer[$entry] = array('' => true); + array_splice($prev_keys, $level, count($prev_keys), $entry); + break; + } + unset($list[$i]); + } + // make sure the last tokens get converted as well + $new_entry = $this->_optimize_regexp_list_tokens_to_string($tokens); + if (GESHI_MAX_PCRE_SUBPATTERNS && $num_subpatterns + substr_count($new_entry, '(?:') > GESHI_MAX_PCRE_SUBPATTERNS) { + $regexp_list[++$list_key] = $new_entry; + } else { + if (!empty($regexp_list[$list_key])) { + $new_entry = '|' . $new_entry; + } + $regexp_list[$list_key] .= $new_entry; + } + return $regexp_list; + } + /** + * this function creates the appropriate regexp string of an token array + * you should not call this function directly, @see $this->optimize_regexp_list(). + * + * @param &$tokens array of tokens + * @param $recursed bool to know wether we recursed or not + * @return string + * @author Milian Wolff <mail@milianw.de> + * @since 1.0.8 + * @access private + */ + function _optimize_regexp_list_tokens_to_string(&$tokens, $recursed = false) { + $list = ''; + foreach ($tokens as $token => $sub_tokens) { + $list .= $token; + $close_entry = isset($sub_tokens['']); + unset($sub_tokens['']); + if (!empty($sub_tokens)) { + $list .= '(?:' . $this->_optimize_regexp_list_tokens_to_string($sub_tokens, true) . ')'; + if ($close_entry) { + // make sub_tokens optional + $list .= '?'; + } + } + $list .= '|'; + } + if (!$recursed) { + // do some optimizations + // common trailing strings + // BUGGY! + //$list = preg_replace_callback('#(?<=^|\:|\|)\w+?(\w+)(?:\|.+\1)+(?=\|)#', create_function( + // '$matches', 'return "(?:" . preg_replace("#" . preg_quote($matches[1], "#") . "(?=\||$)#", "", $matches[0]) . ")" . $matches[1];'), $list); + // (?:p)? => p? + $list = preg_replace('#\(\?\:(.)\)\?#', '\1?', $list); + // (?:a|b|c|d|...)? => [abcd...]? + // TODO: a|bb|c => [ac]|bb + static $callback_2; + if (!isset($callback_2)) { + $callback_2 = create_function('$matches', 'return "[" . str_replace("|", "", $matches[1]) . "]";'); + } + $list = preg_replace_callback('#\(\?\:((?:.\|)+.)\)#', $callback_2, $list); + } + // return $list without trailing pipe + return substr($list, 0, -1); + } } // End Class GeSHi @@ -3223,10 +4367,13 @@ if (!function_exists('geshi_highlight')) { function geshi_highlight($string, $language, $path = null, $return = false) { $geshi = new GeSHi($string, $language, $path); $geshi->set_header_type(GESHI_HEADER_NONE); + if ($return) { return '<code>' . $geshi->parse_code() . '</code>'; } + echo '<code>' . $geshi->parse_code() . '</code>'; + if ($geshi->error()) { return false; } |