diff options
Diffstat (limited to 'Mustache/Parser.php')
-rw-r--r-- | Mustache/Parser.php | 473 |
1 files changed, 385 insertions, 88 deletions
diff --git a/Mustache/Parser.php b/Mustache/Parser.php index 39911d6..5db5824 100644 --- a/Mustache/Parser.php +++ b/Mustache/Parser.php @@ -1,88 +1,385 @@ -<?php
-
-/*
- * This file is part of Mustache.php.
- *
- * (c) 2012 Justin Hileman
- *
- * For the full copyright and license information, please view the LICENSE
- * file that was distributed with this source code.
- */
-
-/**
- * Mustache Parser class.
- *
- * This class is responsible for turning a set of Mustache tokens into a parse tree.
- */
-class Mustache_Parser
-{
-
- /**
- * Process an array of Mustache tokens and convert them into a parse tree.
- *
- * @param array $tokens Set of Mustache tokens
- *
- * @return array Mustache token parse tree
- */
- public function parse(array $tokens = array())
- {
- return $this->buildTree(new ArrayIterator($tokens));
- }
-
- /**
- * Helper method for recursively building a parse tree.
- *
- * @param ArrayIterator $tokens Stream of Mustache tokens
- * @param array $parent Parent token (default: null)
- *
- * @return array Mustache Token parse tree
- *
- * @throws LogicException when nesting errors or mismatched section tags are encountered.
- */
- private function buildTree(ArrayIterator $tokens, array $parent = null)
- {
- $nodes = array();
-
- do {
- $token = $tokens->current();
- $tokens->next();
-
- if ($token === null) {
- continue;
- } else {
- switch ($token[Mustache_Tokenizer::TYPE]) {
- case Mustache_Tokenizer::T_SECTION:
- case Mustache_Tokenizer::T_INVERTED:
- $nodes[] = $this->buildTree($tokens, $token);
- break;
-
- case Mustache_Tokenizer::T_END_SECTION:
- if (!isset($parent)) {
- throw new LogicException('Unexpected closing tag: /'. $token[Mustache_Tokenizer::NAME]);
- }
-
- if ($token[Mustache_Tokenizer::NAME] !== $parent[Mustache_Tokenizer::NAME]) {
- throw new LogicException('Nesting error: ' . $parent[Mustache_Tokenizer::NAME] . ' vs. ' . $token[Mustache_Tokenizer::NAME]);
- }
-
- $parent[Mustache_Tokenizer::END] = $token[Mustache_Tokenizer::INDEX];
- $parent[Mustache_Tokenizer::NODES] = $nodes;
-
- return $parent;
- break;
-
- default:
- $nodes[] = $token;
- break;
- }
- }
-
- } while ($tokens->valid());
-
- if (isset($parent)) {
- throw new LogicException('Missing closing tag: ' . $parent[Mustache_Tokenizer::NAME]);
- }
-
- return $nodes;
- }
-}
+<?php + +/* + * This file is part of Mustache.php. + * + * (c) 2010-2017 Justin Hileman + * + * For the full copyright and license information, please view the LICENSE + * file that was distributed with this source code. + */ + +/** + * Mustache Parser class. + * + * This class is responsible for turning a set of Mustache tokens into a parse tree. + */ +class Mustache_Parser +{ + private $lineNum; + private $lineTokens; + private $pragmas; + private $defaultPragmas = array(); + + private $pragmaFilters; + private $pragmaBlocks; + private $pragmaDynamicNames; + + /** + * Process an array of Mustache tokens and convert them into a parse tree. + * + * @param array $tokens Set of Mustache tokens + * + * @return array Mustache token parse tree + */ + public function parse(array $tokens = array()) + { + $this->lineNum = -1; + $this->lineTokens = 0; + $this->pragmas = $this->defaultPragmas; + + $this->pragmaFilters = isset($this->pragmas[Mustache_Engine::PRAGMA_FILTERS]); + $this->pragmaBlocks = isset($this->pragmas[Mustache_Engine::PRAGMA_BLOCKS]); + $this->pragmaDynamicNames = isset($this->pragmas[Mustache_Engine::PRAGMA_DYNAMIC_NAMES]); + + return $this->buildTree($tokens); + } + + /** + * Enable pragmas across all templates, regardless of the presence of pragma + * tags in the individual templates. + * + * @internal Users should set global pragmas in Mustache_Engine, not here :) + * + * @param string[] $pragmas + */ + public function setPragmas(array $pragmas) + { + $this->pragmas = array(); + foreach ($pragmas as $pragma) { + $this->enablePragma($pragma); + } + $this->defaultPragmas = $this->pragmas; + } + + /** + * Helper method for recursively building a parse tree. + * + * @throws Mustache_Exception_SyntaxException when nesting errors or mismatched section tags are encountered + * + * @param array &$tokens Set of Mustache tokens + * @param array $parent Parent token (default: null) + * + * @return array Mustache Token parse tree + */ + private function buildTree(array &$tokens, array $parent = null) + { + $nodes = array(); + + while (!empty($tokens)) { + $token = array_shift($tokens); + + if ($token[Mustache_Tokenizer::LINE] === $this->lineNum) { + $this->lineTokens++; + } else { + $this->lineNum = $token[Mustache_Tokenizer::LINE]; + $this->lineTokens = 0; + } + + if ($token[Mustache_Tokenizer::TYPE] !== Mustache_Tokenizer::T_COMMENT) { + if ($this->pragmaDynamicNames && isset($token[Mustache_Tokenizer::NAME])) { + list($name, $isDynamic) = $this->getDynamicName($token); + if ($isDynamic) { + $token[Mustache_Tokenizer::NAME] = $name; + $token[Mustache_Tokenizer::DYNAMIC] = true; + } + } + + if ($this->pragmaFilters && isset($token[Mustache_Tokenizer::NAME])) { + list($name, $filters) = $this->getNameAndFilters($token[Mustache_Tokenizer::NAME]); + if (!empty($filters)) { + $token[Mustache_Tokenizer::NAME] = $name; + $token[Mustache_Tokenizer::FILTERS] = $filters; + } + } + } + + switch ($token[Mustache_Tokenizer::TYPE]) { + case Mustache_Tokenizer::T_DELIM_CHANGE: + $this->checkIfTokenIsAllowedInParent($parent, $token); + $this->clearStandaloneLines($nodes, $tokens); + break; + + case Mustache_Tokenizer::T_SECTION: + case Mustache_Tokenizer::T_INVERTED: + $this->checkIfTokenIsAllowedInParent($parent, $token); + $this->clearStandaloneLines($nodes, $tokens); + $nodes[] = $this->buildTree($tokens, $token); + break; + + case Mustache_Tokenizer::T_END_SECTION: + if (!isset($parent)) { + $msg = sprintf( + 'Unexpected closing tag: /%s on line %d', + $token[Mustache_Tokenizer::NAME], + $token[Mustache_Tokenizer::LINE] + ); + throw new Mustache_Exception_SyntaxException($msg, $token); + } + + $sameName = $token[Mustache_Tokenizer::NAME] !== $parent[Mustache_Tokenizer::NAME]; + $tokenDynamic = isset($token[Mustache_Tokenizer::DYNAMIC]) && $token[Mustache_Tokenizer::DYNAMIC]; + $parentDynamic = isset($parent[Mustache_Tokenizer::DYNAMIC]) && $parent[Mustache_Tokenizer::DYNAMIC]; + + if ($sameName || ($tokenDynamic !== $parentDynamic)) { + $msg = sprintf( + 'Nesting error: %s%s (on line %d) vs. %s%s (on line %d)', + $parentDynamic ? '*' : '', + $parent[Mustache_Tokenizer::NAME], + $parent[Mustache_Tokenizer::LINE], + $tokenDynamic ? '*' : '', + $token[Mustache_Tokenizer::NAME], + $token[Mustache_Tokenizer::LINE] + ); + throw new Mustache_Exception_SyntaxException($msg, $token); + } + + $this->clearStandaloneLines($nodes, $tokens); + $parent[Mustache_Tokenizer::END] = $token[Mustache_Tokenizer::INDEX]; + $parent[Mustache_Tokenizer::NODES] = $nodes; + + return $parent; + + case Mustache_Tokenizer::T_PARTIAL: + $this->checkIfTokenIsAllowedInParent($parent, $token); + //store the whitespace prefix for laters! + if ($indent = $this->clearStandaloneLines($nodes, $tokens)) { + $token[Mustache_Tokenizer::INDENT] = $indent[Mustache_Tokenizer::VALUE]; + } + $nodes[] = $token; + break; + + case Mustache_Tokenizer::T_PARENT: + $this->checkIfTokenIsAllowedInParent($parent, $token); + $nodes[] = $this->buildTree($tokens, $token); + break; + + case Mustache_Tokenizer::T_BLOCK_VAR: + if ($this->pragmaBlocks) { + // BLOCKS pragma is enabled, let's do this! + if (isset($parent) && $parent[Mustache_Tokenizer::TYPE] === Mustache_Tokenizer::T_PARENT) { + $token[Mustache_Tokenizer::TYPE] = Mustache_Tokenizer::T_BLOCK_ARG; + } + $this->clearStandaloneLines($nodes, $tokens); + $nodes[] = $this->buildTree($tokens, $token); + } else { + // pretend this was just a normal "escaped" token... + $token[Mustache_Tokenizer::TYPE] = Mustache_Tokenizer::T_ESCAPED; + // TODO: figure out how to figure out if there was a space after this dollar: + $token[Mustache_Tokenizer::NAME] = '$' . $token[Mustache_Tokenizer::NAME]; + $nodes[] = $token; + } + break; + + case Mustache_Tokenizer::T_PRAGMA: + $this->enablePragma($token[Mustache_Tokenizer::NAME]); + // no break + + case Mustache_Tokenizer::T_COMMENT: + $this->clearStandaloneLines($nodes, $tokens); + $nodes[] = $token; + break; + + default: + $nodes[] = $token; + break; + } + } + + if (isset($parent)) { + $msg = sprintf( + 'Missing closing tag: %s opened on line %d', + $parent[Mustache_Tokenizer::NAME], + $parent[Mustache_Tokenizer::LINE] + ); + throw new Mustache_Exception_SyntaxException($msg, $parent); + } + + return $nodes; + } + + /** + * Clear standalone line tokens. + * + * Returns a whitespace token for indenting partials, if applicable. + * + * @param array $nodes Parsed nodes + * @param array $tokens Tokens to be parsed + * + * @return array|null Resulting indent token, if any + */ + private function clearStandaloneLines(array &$nodes, array &$tokens) + { + if ($this->lineTokens > 1) { + // this is the third or later node on this line, so it can't be standalone + return; + } + + $prev = null; + if ($this->lineTokens === 1) { + // this is the second node on this line, so it can't be standalone + // unless the previous node is whitespace. + if ($prev = end($nodes)) { + if (!$this->tokenIsWhitespace($prev)) { + return; + } + } + } + + if ($next = reset($tokens)) { + // If we're on a new line, bail. + if ($next[Mustache_Tokenizer::LINE] !== $this->lineNum) { + return; + } + + // If the next token isn't whitespace, bail. + if (!$this->tokenIsWhitespace($next)) { + return; + } + + if (count($tokens) !== 1) { + // Unless it's the last token in the template, the next token + // must end in newline for this to be standalone. + if (substr($next[Mustache_Tokenizer::VALUE], -1) !== "\n") { + return; + } + } + + // Discard the whitespace suffix + array_shift($tokens); + } + + if ($prev) { + // Return the whitespace prefix, if any + return array_pop($nodes); + } + } + + /** + * Check whether token is a whitespace token. + * + * True if token type is T_TEXT and value is all whitespace characters. + * + * @param array $token + * + * @return bool True if token is a whitespace token + */ + private function tokenIsWhitespace(array $token) + { + if ($token[Mustache_Tokenizer::TYPE] === Mustache_Tokenizer::T_TEXT) { + return preg_match('/^\s*$/', $token[Mustache_Tokenizer::VALUE]); + } + + return false; + } + + /** + * Check whether a token is allowed inside a parent tag. + * + * @throws Mustache_Exception_SyntaxException if an invalid token is found inside a parent tag + * + * @param array|null $parent + * @param array $token + */ + private function checkIfTokenIsAllowedInParent($parent, array $token) + { + if (isset($parent) && $parent[Mustache_Tokenizer::TYPE] === Mustache_Tokenizer::T_PARENT) { + throw new Mustache_Exception_SyntaxException('Illegal content in < parent tag', $token); + } + } + + /** + * Parse dynamic names. + * + * @throws Mustache_Exception_SyntaxException when a tag does not allow * + * @throws Mustache_Exception_SyntaxException on multiple *s, or dots or filters with * + */ + private function getDynamicName(array $token) + { + $name = $token[Mustache_Tokenizer::NAME]; + $isDynamic = false; + + if (preg_match('/^\s*\*\s*/', $name)) { + $this->ensureTagAllowsDynamicNames($token); + $name = preg_replace('/^\s*\*\s*/', '', $name); + $isDynamic = true; + } + + return array($name, $isDynamic); + } + + /** + * Check whether the given token supports dynamic tag names. + * + * @throws Mustache_Exception_SyntaxException when a tag does not allow * + * + * @param array $token + */ + private function ensureTagAllowsDynamicNames(array $token) + { + switch ($token[Mustache_Tokenizer::TYPE]) { + case Mustache_Tokenizer::T_PARTIAL: + case Mustache_Tokenizer::T_PARENT: + case Mustache_Tokenizer::T_END_SECTION: + return; + } + + $msg = sprintf( + 'Invalid dynamic name: %s in %s tag', + $token[Mustache_Tokenizer::NAME], + Mustache_Tokenizer::getTagName($token[Mustache_Tokenizer::TYPE]) + ); + + throw new Mustache_Exception_SyntaxException($msg, $token); + } + + + /** + * Split a tag name into name and filters. + * + * @param string $name + * + * @return array [Tag name, Array of filters] + */ + private function getNameAndFilters($name) + { + $filters = array_map('trim', explode('|', $name)); + $name = array_shift($filters); + + return array($name, $filters); + } + + /** + * Enable a pragma. + * + * @param string $name + */ + private function enablePragma($name) + { + $this->pragmas[$name] = true; + + switch ($name) { + case Mustache_Engine::PRAGMA_BLOCKS: + $this->pragmaBlocks = true; + break; + + case Mustache_Engine::PRAGMA_FILTERS: + $this->pragmaFilters = true; + break; + + case Mustache_Engine::PRAGMA_DYNAMIC_NAMES: + $this->pragmaDynamicNames = true; + break; + } + } +} |