| 1 |
<?php |
| 2 |
|
| 3 |
/* |
| 4 |
* This file is part of Mustache.php. |
| 5 |
* |
| 6 |
* (c) 2010-2017 Justin Hileman |
| 7 |
* |
| 8 |
* For the full copyright and license information, please view the LICENSE |
| 9 |
* file that was distributed with this source code. |
| 10 |
*/ |
| 11 |
|
| 12 |
/** |
| 13 |
* Mustache Parser class. |
| 14 |
* |
| 15 |
* This class is responsible for turning a set of Mustache tokens into a parse tree. |
| 16 |
*/ |
| 17 |
class Mustache_Parser |
| 18 |
{ |
| 19 |
private $lineNum; |
| 20 |
private $lineTokens; |
| 21 |
private $pragmas; |
| 22 |
private $defaultPragmas = array(); |
| 23 |
|
| 24 |
private $pragmaFilters; |
| 25 |
private $pragmaBlocks; |
| 26 |
|
| 27 |
/** |
| 28 |
* Process an array of Mustache tokens and convert them into a parse tree. |
| 29 |
* |
| 30 |
* @param array $tokens Set of Mustache tokens |
| 31 |
* |
| 32 |
* @return array Mustache token parse tree |
| 33 |
*/ |
| 34 |
public function parse(array $tokens = array()) |
| 35 |
{ |
| 36 |
$this->lineNum = -1; |
| 37 |
$this->lineTokens = 0; |
| 38 |
$this->pragmas = $this->defaultPragmas; |
| 39 |
|
| 40 |
$this->pragmaFilters = isset($this->pragmas[Mustache_Engine::PRAGMA_FILTERS]); |
| 41 |
$this->pragmaBlocks = isset($this->pragmas[Mustache_Engine::PRAGMA_BLOCKS]); |
| 42 |
|
| 43 |
return $this->buildTree($tokens); |
| 44 |
} |
| 45 |
|
| 46 |
/** |
| 47 |
* Enable pragmas across all templates, regardless of the presence of pragma |
| 48 |
* tags in the individual templates. |
| 49 |
* |
| 50 |
* @internal Users should set global pragmas in Mustache_Engine, not here :) |
| 51 |
* |
| 52 |
* @param string[] $pragmas |
| 53 |
*/ |
| 54 |
public function setPragmas(array $pragmas) |
| 55 |
{ |
| 56 |
$this->pragmas = array(); |
| 57 |
foreach ($pragmas as $pragma) { |
| 58 |
$this->enablePragma($pragma); |
| 59 |
} |
| 60 |
$this->defaultPragmas = $this->pragmas; |
| 61 |
} |
| 62 |
|
| 63 |
/** |
| 64 |
* Helper method for recursively building a parse tree. |
| 65 |
* |
| 66 |
* @throws Mustache_Exception_SyntaxException when nesting errors or mismatched section tags are encountered |
| 67 |
* |
| 68 |
* @param array &$tokens Set of Mustache tokens |
| 69 |
* @param array $parent Parent token (default: null) |
| 70 |
* |
| 71 |
* @return array Mustache Token parse tree |
| 72 |
*/ |
| 73 |
private function buildTree(array &$tokens, ?array $parent = null) |
| 74 |
{ |
| 75 |
$nodes = array(); |
| 76 |
|
| 77 |
while (!empty($tokens)) { |
| 78 |
$token = array_shift($tokens); |
| 79 |
|
| 80 |
if ($token[Mustache_Tokenizer::LINE] === $this->lineNum) { |
| 81 |
$this->lineTokens++; |
| 82 |
} else { |
| 83 |
$this->lineNum = $token[Mustache_Tokenizer::LINE]; |
| 84 |
$this->lineTokens = 0; |
| 85 |
} |
| 86 |
|
| 87 |
if ($this->pragmaFilters && isset($token[Mustache_Tokenizer::NAME])) { |
| 88 |
list($name, $filters) = $this->getNameAndFilters($token[Mustache_Tokenizer::NAME]); |
| 89 |
if (!empty($filters)) { |
| 90 |
$token[Mustache_Tokenizer::NAME] = $name; |
| 91 |
$token[Mustache_Tokenizer::FILTERS] = $filters; |
| 92 |
} |
| 93 |
} |
| 94 |
|
| 95 |
switch ($token[Mustache_Tokenizer::TYPE]) { |
| 96 |
case Mustache_Tokenizer::T_DELIM_CHANGE: |
| 97 |
$this->checkIfTokenIsAllowedInParent($parent, $token); |
| 98 |
$this->clearStandaloneLines($nodes, $tokens); |
| 99 |
break; |
| 100 |
|
| 101 |
case Mustache_Tokenizer::T_SECTION: |
| 102 |
case Mustache_Tokenizer::T_INVERTED: |
| 103 |
$this->checkIfTokenIsAllowedInParent($parent, $token); |
| 104 |
$this->clearStandaloneLines($nodes, $tokens); |
| 105 |
$nodes[] = $this->buildTree($tokens, $token); |
| 106 |
break; |
| 107 |
|
| 108 |
case Mustache_Tokenizer::T_END_SECTION: |
| 109 |
if (!isset($parent)) { |
| 110 |
$msg = sprintf( |
| 111 |
'Unexpected closing tag: /%s on line %d', |
| 112 |
$token[Mustache_Tokenizer::NAME], |
| 113 |
$token[Mustache_Tokenizer::LINE] |
| 114 |
); |
| 115 |
throw new Mustache_Exception_SyntaxException($msg, $token); |
| 116 |
} |
| 117 |
|
| 118 |
if ($token[Mustache_Tokenizer::NAME] !== $parent[Mustache_Tokenizer::NAME]) { |
| 119 |
$msg = sprintf( |
| 120 |
'Nesting error: %s (on line %d) vs. %s (on line %d)', |
| 121 |
$parent[Mustache_Tokenizer::NAME], |
| 122 |
$parent[Mustache_Tokenizer::LINE], |
| 123 |
$token[Mustache_Tokenizer::NAME], |
| 124 |
$token[Mustache_Tokenizer::LINE] |
| 125 |
); |
| 126 |
throw new Mustache_Exception_SyntaxException($msg, $token); |
| 127 |
} |
| 128 |
|
| 129 |
$this->clearStandaloneLines($nodes, $tokens); |
| 130 |
$parent[Mustache_Tokenizer::END] = $token[Mustache_Tokenizer::INDEX]; |
| 131 |
$parent[Mustache_Tokenizer::NODES] = $nodes; |
| 132 |
|
| 133 |
return $parent; |
| 134 |
|
| 135 |
case Mustache_Tokenizer::T_PARTIAL: |
| 136 |
$this->checkIfTokenIsAllowedInParent($parent, $token); |
| 137 |
//store the whitespace prefix for laters! |
| 138 |
if ($indent = $this->clearStandaloneLines($nodes, $tokens)) { |
| 139 |
$token[Mustache_Tokenizer::INDENT] = $indent[Mustache_Tokenizer::VALUE]; |
| 140 |
} |
| 141 |
$nodes[] = $token; |
| 142 |
break; |
| 143 |
|
| 144 |
case Mustache_Tokenizer::T_PARENT: |
| 145 |
$this->checkIfTokenIsAllowedInParent($parent, $token); |
| 146 |
$nodes[] = $this->buildTree($tokens, $token); |
| 147 |
break; |
| 148 |
|
| 149 |
case Mustache_Tokenizer::T_BLOCK_VAR: |
| 150 |
if ($this->pragmaBlocks) { |
| 151 |
// BLOCKS pragma is enabled, let's do this! |
| 152 |
if (isset($parent) && $parent[Mustache_Tokenizer::TYPE] === Mustache_Tokenizer::T_PARENT) { |
| 153 |
$token[Mustache_Tokenizer::TYPE] = Mustache_Tokenizer::T_BLOCK_ARG; |
| 154 |
} |
| 155 |
$this->clearStandaloneLines($nodes, $tokens); |
| 156 |
$nodes[] = $this->buildTree($tokens, $token); |
| 157 |
} else { |
| 158 |
// pretend this was just a normal "escaped" token... |
| 159 |
$token[Mustache_Tokenizer::TYPE] = Mustache_Tokenizer::T_ESCAPED; |
| 160 |
// TODO: figure out how to figure out if there was a space after this dollar: |
| 161 |
$token[Mustache_Tokenizer::NAME] = '$' . $token[Mustache_Tokenizer::NAME]; |
| 162 |
$nodes[] = $token; |
| 163 |
} |
| 164 |
break; |
| 165 |
|
| 166 |
case Mustache_Tokenizer::T_PRAGMA: |
| 167 |
$this->enablePragma($token[Mustache_Tokenizer::NAME]); |
| 168 |
// no break |
| 169 |
|
| 170 |
case Mustache_Tokenizer::T_COMMENT: |
| 171 |
$this->clearStandaloneLines($nodes, $tokens); |
| 172 |
$nodes[] = $token; |
| 173 |
break; |
| 174 |
|
| 175 |
default: |
| 176 |
$nodes[] = $token; |
| 177 |
break; |
| 178 |
} |
| 179 |
} |
| 180 |
|
| 181 |
if (isset($parent)) { |
| 182 |
$msg = sprintf( |
| 183 |
'Missing closing tag: %s opened on line %d', |
| 184 |
$parent[Mustache_Tokenizer::NAME], |
| 185 |
$parent[Mustache_Tokenizer::LINE] |
| 186 |
); |
| 187 |
throw new Mustache_Exception_SyntaxException($msg, $parent); |
| 188 |
} |
| 189 |
|
| 190 |
return $nodes; |
| 191 |
} |
| 192 |
|
| 193 |
/** |
| 194 |
* Clear standalone line tokens. |
| 195 |
* |
| 196 |
* Returns a whitespace token for indenting partials, if applicable. |
| 197 |
* |
| 198 |
* @param array $nodes Parsed nodes |
| 199 |
* @param array $tokens Tokens to be parsed |
| 200 |
* |
| 201 |
* @return array|null Resulting indent token, if any |
| 202 |
*/ |
| 203 |
private function clearStandaloneLines(array &$nodes, array &$tokens) |
| 204 |
{ |
| 205 |
if ($this->lineTokens > 1) { |
| 206 |
// this is the third or later node on this line, so it can't be standalone |
| 207 |
return; |
| 208 |
} |
| 209 |
|
| 210 |
$prev = null; |
| 211 |
if ($this->lineTokens === 1) { |
| 212 |
// this is the second node on this line, so it can't be standalone |
| 213 |
// unless the previous node is whitespace. |
| 214 |
if ($prev = end($nodes)) { |
| 215 |
if (!$this->tokenIsWhitespace($prev)) { |
| 216 |
return; |
| 217 |
} |
| 218 |
} |
| 219 |
} |
| 220 |
|
| 221 |
if ($next = reset($tokens)) { |
| 222 |
// If we're on a new line, bail. |
| 223 |
if ($next[Mustache_Tokenizer::LINE] !== $this->lineNum) { |
| 224 |
return; |
| 225 |
} |
| 226 |
|
| 227 |
// If the next token isn't whitespace, bail. |
| 228 |
if (!$this->tokenIsWhitespace($next)) { |
| 229 |
return; |
| 230 |
} |
| 231 |
|
| 232 |
if (count($tokens) !== 1) { |
| 233 |
// Unless it's the last token in the template, the next token |
| 234 |
// must end in newline for this to be standalone. |
| 235 |
if (substr($next[Mustache_Tokenizer::VALUE], -1) !== "\n") { |
| 236 |
return; |
| 237 |
} |
| 238 |
} |
| 239 |
|
| 240 |
// Discard the whitespace suffix |
| 241 |
array_shift($tokens); |
| 242 |
} |
| 243 |
|
| 244 |
if ($prev) { |
| 245 |
// Return the whitespace prefix, if any |
| 246 |
return array_pop($nodes); |
| 247 |
} |
| 248 |
} |
| 249 |
|
| 250 |
/** |
| 251 |
* Check whether token is a whitespace token. |
| 252 |
* |
| 253 |
* True if token type is T_TEXT and value is all whitespace characters. |
| 254 |
* |
| 255 |
* @param array $token |
| 256 |
* |
| 257 |
* @return bool True if token is a whitespace token |
| 258 |
*/ |
| 259 |
private function tokenIsWhitespace(array $token) |
| 260 |
{ |
| 261 |
if ($token[Mustache_Tokenizer::TYPE] === Mustache_Tokenizer::T_TEXT) { |
| 262 |
return preg_match('/^\s*$/', $token[Mustache_Tokenizer::VALUE]); |
| 263 |
} |
| 264 |
|
| 265 |
return false; |
| 266 |
} |
| 267 |
|
| 268 |
/** |
| 269 |
* Check whether a token is allowed inside a parent tag. |
| 270 |
* |
| 271 |
* @throws Mustache_Exception_SyntaxException if an invalid token is found inside a parent tag |
| 272 |
* |
| 273 |
* @param array|null $parent |
| 274 |
* @param array $token |
| 275 |
*/ |
| 276 |
private function checkIfTokenIsAllowedInParent($parent, array $token) |
| 277 |
{ |
| 278 |
if (isset($parent) && $parent[Mustache_Tokenizer::TYPE] === Mustache_Tokenizer::T_PARENT) { |
| 279 |
throw new Mustache_Exception_SyntaxException('Illegal content in < parent tag', $token); |
| 280 |
} |
| 281 |
} |
| 282 |
|
| 283 |
/** |
| 284 |
* Split a tag name into name and filters. |
| 285 |
* |
| 286 |
* @param string $name |
| 287 |
* |
| 288 |
* @return array [Tag name, Array of filters] |
| 289 |
*/ |
| 290 |
private function getNameAndFilters($name) |
| 291 |
{ |
| 292 |
$filters = array_map('trim', explode('|', $name)); |
| 293 |
$name = array_shift($filters); |
| 294 |
|
| 295 |
return array($name, $filters); |
| 296 |
} |
| 297 |
|
| 298 |
/** |
| 299 |
* Enable a pragma. |
| 300 |
* |
| 301 |
* @param string $name |
| 302 |
*/ |
| 303 |
private function enablePragma($name) |
| 304 |
{ |
| 305 |
$this->pragmas[$name] = true; |
| 306 |
|
| 307 |
switch ($name) { |
| 308 |
case Mustache_Engine::PRAGMA_BLOCKS: |
| 309 |
$this->pragmaBlocks = true; |
| 310 |
break; |
| 311 |
|
| 312 |
case Mustache_Engine::PRAGMA_FILTERS: |
| 313 |
$this->pragmaFilters = true; |
| 314 |
break; |
| 315 |
} |
| 316 |
} |
| 317 |
} |
| 318 |
|