remove autoload section in composer.json

[github/wallabag/wallabag.git] / inc / 3rdparty / htmlpurifier / HTMLPurifier / Strategy / MakeWellFormed.php

<?php\r
\r
/**\r
 * Takes tokens makes them well-formed (balance end tags, etc.)\r
 *\r
 * Specification of the armor attributes this strategy uses:\r
 *\r
 *      - MakeWellFormed_TagClosedError: This armor field is used to\r
 *        suppress tag closed errors for certain tokens [TagClosedSuppress],\r
 *        in particular, if a tag was generated automatically by HTML\r
 *        Purifier, we may rely on our infrastructure to close it for us\r
 *        and shouldn't report an error to the user [TagClosedAuto].\r
 */\r
class HTMLPurifier_Strategy_MakeWellFormed extends HTMLPurifier_Strategy\r
{\r
\r
    /**\r
     * Array stream of tokens being processed.\r
     * @type HTMLPurifier_Token[]\r
     */\r
    protected $tokens;\r
\r
    /**\r
     * Current token.\r
     * @type HTMLPurifier_Token\r
     */\r
    protected $token;\r
\r
    /**\r
     * Zipper managing the true state.\r
     * @type HTMLPurifier_Zipper\r
     */\r
    protected $zipper;\r
\r
    /**\r
     * Current nesting of elements.\r
     * @type array\r
     */\r
    protected $stack;\r
\r
    /**\r
     * Injectors active in this stream processing.\r
     * @type HTMLPurifier_Injector[]\r
     */\r
    protected $injectors;\r
\r
    /**\r
     * Current instance of HTMLPurifier_Config.\r
     * @type HTMLPurifier_Config\r
     */\r
    protected $config;\r
\r
    /**\r
     * Current instance of HTMLPurifier_Context.\r
     * @type HTMLPurifier_Context\r
     */\r
    protected $context;\r
\r
    /**\r
     * @param HTMLPurifier_Token[] $tokens\r
     * @param HTMLPurifier_Config $config\r
     * @param HTMLPurifier_Context $context\r
     * @return HTMLPurifier_Token[]\r
     * @throws HTMLPurifier_Exception\r
     */\r
    public function execute($tokens, $config, $context)\r
    {\r
        $definition = $config->getHTMLDefinition();\r
\r
        // local variables\r
        $generator = new HTMLPurifier_Generator($config, $context);\r
        $escape_invalid_tags = $config->get('Core.EscapeInvalidTags');\r
        // used for autoclose early abortion\r
        $global_parent_allowed_elements = $definition->info_parent_def->child->getAllowedElements($config);\r
        $e = $context->get('ErrorCollector', true);\r
        $i = false; // injector index\r
        list($zipper, $token) = HTMLPurifier_Zipper::fromArray($tokens);\r
        if ($token === NULL) {\r
            return array();\r
        }\r
        $reprocess = false; // whether or not to reprocess the same token\r
        $stack = array();\r
\r
        // member variables\r
        $this->stack =& $stack;\r
        $this->tokens =& $tokens;\r
        $this->token =& $token;\r
        $this->zipper =& $zipper;\r
        $this->config = $config;\r
        $this->context = $context;\r
\r
        // context variables\r
        $context->register('CurrentNesting', $stack);\r
        $context->register('InputZipper', $zipper);\r
        $context->register('CurrentToken', $token);\r
\r
        // -- begin INJECTOR --\r
\r
        $this->injectors = array();\r
\r
        $injectors = $config->getBatch('AutoFormat');\r
        $def_injectors = $definition->info_injector;\r
        $custom_injectors = $injectors['Custom'];\r
        unset($injectors['Custom']); // special case\r
        foreach ($injectors as $injector => $b) {\r
            // XXX: Fix with a legitimate lookup table of enabled filters\r
            if (strpos($injector, '.') !== false) {\r
                continue;\r
            }\r
            $injector = "HTMLPurifier_Injector_$injector";\r
            if (!$b) {\r
                continue;\r
            }\r
            $this->injectors[] = new $injector;\r
        }\r
        foreach ($def_injectors as $injector) {\r
            // assumed to be objects\r
            $this->injectors[] = $injector;\r
        }\r
        foreach ($custom_injectors as $injector) {\r
            if (!$injector) {\r
                continue;\r
            }\r
            if (is_string($injector)) {\r
                $injector = "HTMLPurifier_Injector_$injector";\r
                $injector = new $injector;\r
            }\r
            $this->injectors[] = $injector;\r
        }\r
\r
        // give the injectors references to the definition and context\r
        // variables for performance reasons\r
        foreach ($this->injectors as $ix => $injector) {\r
            $error = $injector->prepare($config, $context);\r
            if (!$error) {\r
                continue;\r
            }\r
            array_splice($this->injectors, $ix, 1); // rm the injector\r
            trigger_error("Cannot enable {$injector->name} injector because $error is not allowed", E_USER_WARNING);\r
        }\r
\r
        // -- end INJECTOR --\r
\r
        // a note on reprocessing:\r
        //      In order to reduce code duplication, whenever some code needs\r
        //      to make HTML changes in order to make things "correct", the\r
        //      new HTML gets sent through the purifier, regardless of its\r
        //      status. This means that if we add a start token, because it\r
        //      was totally necessary, we don't have to update nesting; we just\r
        //      punt ($reprocess = true; continue;) and it does that for us.\r
\r
        // isset is in loop because $tokens size changes during loop exec\r
        for (;;\r
             // only increment if we don't need to reprocess\r
             $reprocess ? $reprocess = false : $token = $zipper->next($token)) {\r
\r
            // check for a rewind\r
            if (is_int($i)) {\r
                // possibility: disable rewinding if the current token has a\r
                // rewind set on it already. This would offer protection from\r
                // infinite loop, but might hinder some advanced rewinding.\r
                $rewind_offset = $this->injectors[$i]->getRewindOffset();\r
                if (is_int($rewind_offset)) {\r
                    for ($j = 0; $j < $rewind_offset; $j++) {\r
                        if (empty($zipper->front)) break;\r
                        $token = $zipper->prev($token);\r
                        // indicate that other injectors should not process this token,\r
                        // but we need to reprocess it\r
                        unset($token->skip[$i]);\r
                        $token->rewind = $i;\r
                        if ($token instanceof HTMLPurifier_Token_Start) {\r
                            array_pop($this->stack);\r
                        } elseif ($token instanceof HTMLPurifier_Token_End) {\r
                            $this->stack[] = $token->start;\r
                        }\r
                    }\r
                }\r
                $i = false;\r
            }\r
\r
            // handle case of document end\r
            if ($token === NULL) {\r
                // kill processing if stack is empty\r
                if (empty($this->stack)) {\r
                    break;\r
                }\r
\r
                // peek\r
                $top_nesting = array_pop($this->stack);\r
                $this->stack[] = $top_nesting;\r
\r
                // send error [TagClosedSuppress]\r
                if ($e && !isset($top_nesting->armor['MakeWellFormed_TagClosedError'])) {\r
                    $e->send(E_NOTICE, 'Strategy_MakeWellFormed: Tag closed by document end', $top_nesting);\r
                }\r
\r
                // append, don't splice, since this is the end\r
                $token = new HTMLPurifier_Token_End($top_nesting->name);\r
\r
                // punt!\r
                $reprocess = true;\r
                continue;\r
            }\r
\r
            //echo '<br>'; printZipper($zipper, $token);//printTokens($this->stack);\r
            //flush();\r
\r
            // quick-check: if it's not a tag, no need to process\r
            if (empty($token->is_tag)) {\r
                if ($token instanceof HTMLPurifier_Token_Text) {\r
                    foreach ($this->injectors as $i => $injector) {\r
                        if (isset($token->skip[$i])) {\r
                            continue;\r
                        }\r
                        if ($token->rewind !== null && $token->rewind !== $i) {\r
                            continue;\r
                        }\r
                        // XXX fuckup\r
                        $r = $token;\r
                        $injector->handleText($r);\r
                        $token = $this->processToken($r, $i);\r
                        $reprocess = true;\r
                        break;\r
                    }\r
                }\r
                // another possibility is a comment\r
                continue;\r
            }\r
\r
            if (isset($definition->info[$token->name])) {\r
                $type = $definition->info[$token->name]->child->type;\r
            } else {\r
                $type = false; // Type is unknown, treat accordingly\r
            }\r
\r
            // quick tag checks: anything that's *not* an end tag\r
            $ok = false;\r
            if ($type === 'empty' && $token instanceof HTMLPurifier_Token_Start) {\r
                // claims to be a start tag but is empty\r
                $token = new HTMLPurifier_Token_Empty(\r
                    $token->name,\r
                    $token->attr,\r
                    $token->line,\r
                    $token->col,\r
                    $token->armor\r
                );\r
                $ok = true;\r
            } elseif ($type && $type !== 'empty' && $token instanceof HTMLPurifier_Token_Empty) {\r
                // claims to be empty but really is a start tag\r
                // NB: this assignment is required\r
                $old_token = $token;\r
                $token = new HTMLPurifier_Token_End($token->name);\r
                $token = $this->insertBefore(\r
                    new HTMLPurifier_Token_Start($old_token->name, $old_token->attr, $old_token->line, $old_token->col, $old_token->armor)\r
                );\r
                // punt (since we had to modify the input stream in a non-trivial way)\r
                $reprocess = true;\r
                continue;\r
            } elseif ($token instanceof HTMLPurifier_Token_Empty) {\r
                // real empty token\r
                $ok = true;\r
            } elseif ($token instanceof HTMLPurifier_Token_Start) {\r
                // start tag\r
\r
                // ...unless they also have to close their parent\r
                if (!empty($this->stack)) {\r
\r
                    // Performance note: you might think that it's rather\r
                    // inefficient, recalculating the autoclose information\r
                    // for every tag that a token closes (since when we\r
                    // do an autoclose, we push a new token into the\r
                    // stream and then /process/ that, before\r
                    // re-processing this token.)  But this is\r
                    // necessary, because an injector can make an\r
                    // arbitrary transformations to the autoclosing\r
                    // tokens we introduce, so things may have changed\r
                    // in the meantime.  Also, doing the inefficient thing is\r
                    // "easy" to reason about (for certain perverse definitions\r
                    // of "easy")\r
\r
                    $parent = array_pop($this->stack);\r
                    $this->stack[] = $parent;\r
\r
                    $parent_def = null;\r
                    $parent_elements = null;\r
                    $autoclose = false;\r
                    if (isset($definition->info[$parent->name])) {\r
                        $parent_def = $definition->info[$parent->name];\r
                        $parent_elements = $parent_def->child->getAllowedElements($config);\r
                        $autoclose = !isset($parent_elements[$token->name]);\r
                    }\r
\r
                    if ($autoclose && $definition->info[$token->name]->wrap) {\r
                        // Check if an element can be wrapped by another\r
                        // element to make it valid in a context (for\r
                        // example, <ul><ul> needs a <li> in between)\r
                        $wrapname = $definition->info[$token->name]->wrap;\r
                        $wrapdef = $definition->info[$wrapname];\r
                        $elements = $wrapdef->child->getAllowedElements($config);\r
                        if (isset($elements[$token->name]) && isset($parent_elements[$wrapname])) {\r
                            $newtoken = new HTMLPurifier_Token_Start($wrapname);\r
                            $token = $this->insertBefore($newtoken);\r
                            $reprocess = true;\r
                            continue;\r
                        }\r
                    }\r
\r
                    $carryover = false;\r
                    if ($autoclose && $parent_def->formatting) {\r
                        $carryover = true;\r
                    }\r
\r
                    if ($autoclose) {\r
                        // check if this autoclose is doomed to fail\r
                        // (this rechecks $parent, which his harmless)\r
                        $autoclose_ok = isset($global_parent_allowed_elements[$token->name]);\r
                        if (!$autoclose_ok) {\r
                            foreach ($this->stack as $ancestor) {\r
                                $elements = $definition->info[$ancestor->name]->child->getAllowedElements($config);\r
                                if (isset($elements[$token->name])) {\r
                                    $autoclose_ok = true;\r
                                    break;\r
                                }\r
                                if ($definition->info[$token->name]->wrap) {\r
                                    $wrapname = $definition->info[$token->name]->wrap;\r
                                    $wrapdef = $definition->info[$wrapname];\r
                                    $wrap_elements = $wrapdef->child->getAllowedElements($config);\r
                                    if (isset($wrap_elements[$token->name]) && isset($elements[$wrapname])) {\r
                                        $autoclose_ok = true;\r
                                        break;\r
                                    }\r
                                }\r
                            }\r
                        }\r
                        if ($autoclose_ok) {\r
                            // errors need to be updated\r
                            $new_token = new HTMLPurifier_Token_End($parent->name);\r
                            $new_token->start = $parent;\r
                            // [TagClosedSuppress]\r
                            if ($e && !isset($parent->armor['MakeWellFormed_TagClosedError'])) {\r
                                if (!$carryover) {\r
                                    $e->send(E_NOTICE, 'Strategy_MakeWellFormed: Tag auto closed', $parent);\r
                                } else {\r
                                    $e->send(E_NOTICE, 'Strategy_MakeWellFormed: Tag carryover', $parent);\r
                                }\r
                            }\r
                            if ($carryover) {\r
                                $element = clone $parent;\r
                                // [TagClosedAuto]\r
                                $element->armor['MakeWellFormed_TagClosedError'] = true;\r
                                $element->carryover = true;\r
                                $token = $this->processToken(array($new_token, $token, $element));\r
                            } else {\r
                                $token = $this->insertBefore($new_token);\r
                            }\r
                        } else {\r
                            $token = $this->remove();\r
                        }\r
                        $reprocess = true;\r
                        continue;\r
                    }\r
\r
                }\r
                $ok = true;\r
            }\r
\r
            if ($ok) {\r
                foreach ($this->injectors as $i => $injector) {\r
                    if (isset($token->skip[$i])) {\r
                        continue;\r
                    }\r
                    if ($token->rewind !== null && $token->rewind !== $i) {\r
                        continue;\r
                    }\r
                    $r = $token;\r
                    $injector->handleElement($r);\r
                    $token = $this->processToken($r, $i);\r
                    $reprocess = true;\r
                    break;\r
                }\r
                if (!$reprocess) {\r
                    // ah, nothing interesting happened; do normal processing\r
                    if ($token instanceof HTMLPurifier_Token_Start) {\r
                        $this->stack[] = $token;\r
                    } elseif ($token instanceof HTMLPurifier_Token_End) {\r
                        throw new HTMLPurifier_Exception(\r
                            'Improper handling of end tag in start code; possible error in MakeWellFormed'\r
                        );\r
                    }\r
                }\r
                continue;\r
            }\r
\r
            // sanity check: we should be dealing with a closing tag\r
            if (!$token instanceof HTMLPurifier_Token_End) {\r
                throw new HTMLPurifier_Exception('Unaccounted for tag token in input stream, bug in HTML Purifier');\r
            }\r
\r
            // make sure that we have something open\r
            if (empty($this->stack)) {\r
                if ($escape_invalid_tags) {\r
                    if ($e) {\r
                        $e->send(E_WARNING, 'Strategy_MakeWellFormed: Unnecessary end tag to text');\r
                    }\r
                    $token = new HTMLPurifier_Token_Text($generator->generateFromToken($token));\r
                } else {\r
                    if ($e) {\r
                        $e->send(E_WARNING, 'Strategy_MakeWellFormed: Unnecessary end tag removed');\r
                    }\r
                    $token = $this->remove();\r
                }\r
                $reprocess = true;\r
                continue;\r
            }\r
\r
            // first, check for the simplest case: everything closes neatly.\r
            // Eventually, everything passes through here; if there are problems\r
            // we modify the input stream accordingly and then punt, so that\r
            // the tokens get processed again.\r
            $current_parent = array_pop($this->stack);\r
            if ($current_parent->name == $token->name) {\r
                $token->start = $current_parent;\r
                foreach ($this->injectors as $i => $injector) {\r
                    if (isset($token->skip[$i])) {\r
                        continue;\r
                    }\r
                    if ($token->rewind !== null && $token->rewind !== $i) {\r
                        continue;\r
                    }\r
                    $r = $token;\r
                    $injector->handleEnd($r);\r
                    $token = $this->processToken($r, $i);\r
                    $this->stack[] = $current_parent;\r
                    $reprocess = true;\r
                    break;\r
                }\r
                continue;\r
            }\r
\r
            // okay, so we're trying to close the wrong tag\r
\r
            // undo the pop previous pop\r
            $this->stack[] = $current_parent;\r
\r
            // scroll back the entire nest, trying to find our tag.\r
            // (feature could be to specify how far you'd like to go)\r
            $size = count($this->stack);\r
            // -2 because -1 is the last element, but we already checked that\r
            $skipped_tags = false;\r
            for ($j = $size - 2; $j >= 0; $j--) {\r
                if ($this->stack[$j]->name == $token->name) {\r
                    $skipped_tags = array_slice($this->stack, $j);\r
                    break;\r
                }\r
            }\r
\r
            // we didn't find the tag, so remove\r
            if ($skipped_tags === false) {\r
                if ($escape_invalid_tags) {\r
                    if ($e) {\r
                        $e->send(E_WARNING, 'Strategy_MakeWellFormed: Stray end tag to text');\r
                    }\r
                    $token = new HTMLPurifier_Token_Text($generator->generateFromToken($token));\r
                } else {\r
                    if ($e) {\r
                        $e->send(E_WARNING, 'Strategy_MakeWellFormed: Stray end tag removed');\r
                    }\r
                    $token = $this->remove();\r
                }\r
                $reprocess = true;\r
                continue;\r
            }\r
\r
            // do errors, in REVERSE $j order: a,b,c with </a></b></c>\r
            $c = count($skipped_tags);\r
            if ($e) {\r
                for ($j = $c - 1; $j > 0; $j--) {\r
                    // notice we exclude $j == 0, i.e. the current ending tag, from\r
                    // the errors... [TagClosedSuppress]\r
                    if (!isset($skipped_tags[$j]->armor['MakeWellFormed_TagClosedError'])) {\r
                        $e->send(E_NOTICE, 'Strategy_MakeWellFormed: Tag closed by element end', $skipped_tags[$j]);\r
                    }\r
                }\r
            }\r
\r
            // insert tags, in FORWARD $j order: c,b,a with </a></b></c>\r
            $replace = array($token);\r
            for ($j = 1; $j < $c; $j++) {\r
                // ...as well as from the insertions\r
                $new_token = new HTMLPurifier_Token_End($skipped_tags[$j]->name);\r
                $new_token->start = $skipped_tags[$j];\r
                array_unshift($replace, $new_token);\r
                if (isset($definition->info[$new_token->name]) && $definition->info[$new_token->name]->formatting) {\r
                    // [TagClosedAuto]\r
                    $element = clone $skipped_tags[$j];\r
                    $element->carryover = true;\r
                    $element->armor['MakeWellFormed_TagClosedError'] = true;\r
                    $replace[] = $element;\r
                }\r
            }\r
            $token = $this->processToken($replace);\r
            $reprocess = true;\r
            continue;\r
        }\r
\r
        $context->destroy('CurrentToken');\r
        $context->destroy('CurrentNesting');\r
        $context->destroy('InputZipper');\r
\r
        unset($this->injectors, $this->stack, $this->tokens);\r
        return $zipper->toArray($token);\r
    }\r
\r
    /**\r
     * Processes arbitrary token values for complicated substitution patterns.\r
     * In general:\r
     *\r
     * If $token is an array, it is a list of tokens to substitute for the\r
     * current token. These tokens then get individually processed. If there\r
     * is a leading integer in the list, that integer determines how many\r
     * tokens from the stream should be removed.\r
     *\r
     * If $token is a regular token, it is swapped with the current token.\r
     *\r
     * If $token is false, the current token is deleted.\r
     *\r
     * If $token is an integer, that number of tokens (with the first token\r
     * being the current one) will be deleted.\r
     *\r
     * @param HTMLPurifier_Token|array|int|bool $token Token substitution value\r
     * @param HTMLPurifier_Injector|int $injector Injector that performed the substitution; default is if\r
     *        this is not an injector related operation.\r
     * @throws HTMLPurifier_Exception\r
     */\r
    protected function processToken($token, $injector = -1)\r
    {\r
        // normalize forms of token\r
        if (is_object($token)) {\r
            $token = array(1, $token);\r
        }\r
        if (is_int($token)) {\r
            $token = array($token);\r
        }\r
        if ($token === false) {\r
            $token = array(1);\r
        }\r
        if (!is_array($token)) {\r
            throw new HTMLPurifier_Exception('Invalid token type from injector');\r
        }\r
        if (!is_int($token[0])) {\r
            array_unshift($token, 1);\r
        }\r
        if ($token[0] === 0) {\r
            throw new HTMLPurifier_Exception('Deleting zero tokens is not valid');\r
        }\r
\r
        // $token is now an array with the following form:\r
        // array(number nodes to delete, new node 1, new node 2, ...)\r
\r
        $delete = array_shift($token);\r
        list($old, $r) = $this->zipper->splice($this->token, $delete, $token);\r
\r
        if ($injector > -1) {\r
            // determine appropriate skips\r
            $oldskip = isset($old[0]) ? $old[0]->skip : array();\r
            foreach ($token as $object) {\r
                $object->skip = $oldskip;\r
                $object->skip[$injector] = true;\r
            }\r
        }\r
\r
        return $r;\r
\r
    }\r
\r
    /**\r
     * Inserts a token before the current token. Cursor now points to\r
     * this token.  You must reprocess after this.\r
     * @param HTMLPurifier_Token $token\r
     */\r
    private function insertBefore($token)\r
    {\r
        // NB not $this->zipper->insertBefore(), due to positioning\r
        // differences\r
        $splice = $this->zipper->splice($this->token, 0, array($token));\r
\r
        return $splice[1];\r
    }\r
\r
    /**\r
     * Removes current token. Cursor now points to new token occupying previously\r
     * occupied space.  You must reprocess after this.\r
     */\r
    private function remove()\r
    {\r
        return $this->zipper->delete();\r
    }\r
}\r
\r
// vim: et sw=4 sts=4\r