| 1 |
<?php |
| 2 |
|
| 3 |
/* |
| 4 |
* This file is part of Mustache.php. |
| 5 |
* |
| 6 |
* (c) 2010-2014 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 Tokenizer class. |
| 14 |
* |
| 15 |
* This class is responsible for turning raw template source into a set of Mustache tokens. |
| 16 |
*/ |
| 17 |
class Mustache_Tokenizer |
| 18 |
{ |
| 19 |
// Finite state machine states |
| 20 |
const IN_TEXT = 0; |
| 21 |
const IN_TAG_TYPE = 1; |
| 22 |
const IN_TAG = 2; |
| 23 |
|
| 24 |
// Token types |
| 25 |
const T_SECTION = '#'; |
| 26 |
const T_INVERTED = '^'; |
| 27 |
const T_END_SECTION = '/'; |
| 28 |
const T_COMMENT = '!'; |
| 29 |
const T_PARTIAL = '>'; |
| 30 |
const T_PARTIAL_2 = '<'; |
| 31 |
const T_DELIM_CHANGE = '='; |
| 32 |
const T_ESCAPED = '_v'; |
| 33 |
const T_UNESCAPED = '{'; |
| 34 |
const T_UNESCAPED_2 = '&'; |
| 35 |
const T_TEXT = '_t'; |
| 36 |
const T_PRAGMA = '%'; |
| 37 |
|
| 38 |
// Valid token types |
| 39 |
private static $tagTypes = array( |
| 40 |
self::T_SECTION => true, |
| 41 |
self::T_INVERTED => true, |
| 42 |
self::T_END_SECTION => true, |
| 43 |
self::T_COMMENT => true, |
| 44 |
self::T_PARTIAL => true, |
| 45 |
self::T_PARTIAL_2 => true, |
| 46 |
self::T_DELIM_CHANGE => true, |
| 47 |
self::T_ESCAPED => true, |
| 48 |
self::T_UNESCAPED => true, |
| 49 |
self::T_UNESCAPED_2 => true, |
| 50 |
self::T_PRAGMA => true, |
| 51 |
); |
| 52 |
|
| 53 |
// Interpolated tags |
| 54 |
private static $interpolatedTags = array( |
| 55 |
self::T_ESCAPED => true, |
| 56 |
self::T_UNESCAPED => true, |
| 57 |
self::T_UNESCAPED_2 => true, |
| 58 |
); |
| 59 |
|
| 60 |
// Token properties |
| 61 |
const TYPE = 'type'; |
| 62 |
const NAME = 'name'; |
| 63 |
const OTAG = 'otag'; |
| 64 |
const CTAG = 'ctag'; |
| 65 |
const LINE = 'line'; |
| 66 |
const INDEX = 'index'; |
| 67 |
const END = 'end'; |
| 68 |
const INDENT = 'indent'; |
| 69 |
const NODES = 'nodes'; |
| 70 |
const VALUE = 'value'; |
| 71 |
|
| 72 |
private $state; |
| 73 |
private $tagType; |
| 74 |
private $tag; |
| 75 |
private $buffer; |
| 76 |
private $tokens; |
| 77 |
private $seenTag; |
| 78 |
private $line; |
| 79 |
private $otag; |
| 80 |
private $ctag; |
| 81 |
private $otagLen; |
| 82 |
private $ctagLen; |
| 83 |
|
| 84 |
/** |
| 85 |
* Scan and tokenize template source. |
| 86 |
* |
| 87 |
* @throws Mustache_Exception_SyntaxException when mismatched section tags are encountered. |
| 88 |
* |
| 89 |
* @param string $text Mustache template source to tokenize |
| 90 |
* @param string $delimiters Optionally, pass initial opening and closing delimiters (default: null) |
| 91 |
* |
| 92 |
* @return array Set of Mustache tokens |
| 93 |
*/ |
| 94 |
public function scan($text, $delimiters = null) |
| 95 |
{ |
| 96 |
// Setting mbstring.func_overload makes things *really* slow. |
| 97 |
// Let's do everyone a favor and scan this string as ASCII instead. |
| 98 |
$encoding = null; |
| 99 |
if (function_exists('mb_internal_encoding') && ini_get('mbstring.func_overload') & 2) { |
| 100 |
$encoding = mb_internal_encoding(); |
| 101 |
mb_internal_encoding('ASCII'); |
| 102 |
} |
| 103 |
|
| 104 |
$this->reset(); |
| 105 |
|
| 106 |
if ($delimiters = trim($delimiters)) { |
| 107 |
$this->setDelimiters($delimiters); |
| 108 |
} |
| 109 |
|
| 110 |
$len = strlen($text); |
| 111 |
for ($i = 0; $i < $len; $i++) { |
| 112 |
switch ($this->state) { |
| 113 |
case self::IN_TEXT: |
| 114 |
if ($this->tagChange($this->otag, $this->otagLen, $text, $i)) { |
| 115 |
$i--; |
| 116 |
$this->flushBuffer(); |
| 117 |
$this->state = self::IN_TAG_TYPE; |
| 118 |
} else { |
| 119 |
$char = $text[$i]; |
| 120 |
$this->buffer .= $char; |
| 121 |
if ($char === "\n") { |
| 122 |
$this->flushBuffer(); |
| 123 |
$this->line++; |
| 124 |
} |
| 125 |
} |
| 126 |
break; |
| 127 |
|
| 128 |
case self::IN_TAG_TYPE: |
| 129 |
$i += $this->otagLen - 1; |
| 130 |
$char = $text[$i + 1]; |
| 131 |
if (isset(self::$tagTypes[$char])) { |
| 132 |
$tag = $char; |
| 133 |
$this->tagType = $tag; |
| 134 |
} else { |
| 135 |
$tag = null; |
| 136 |
$this->tagType = self::T_ESCAPED; |
| 137 |
} |
| 138 |
|
| 139 |
if ($this->tagType === self::T_DELIM_CHANGE) { |
| 140 |
$i = $this->changeDelimiters($text, $i); |
| 141 |
$this->state = self::IN_TEXT; |
| 142 |
} elseif ($this->tagType === self::T_PRAGMA) { |
| 143 |
$i = $this->addPragma($text, $i); |
| 144 |
$this->state = self::IN_TEXT; |
| 145 |
} else { |
| 146 |
if ($tag !== null) { |
| 147 |
$i++; |
| 148 |
} |
| 149 |
$this->state = self::IN_TAG; |
| 150 |
} |
| 151 |
$this->seenTag = $i; |
| 152 |
break; |
| 153 |
|
| 154 |
default: |
| 155 |
if ($this->tagChange($this->ctag, $this->ctagLen, $text, $i)) { |
| 156 |
$token = array( |
| 157 |
self::TYPE => $this->tagType, |
| 158 |
self::NAME => trim($this->buffer), |
| 159 |
self::OTAG => $this->otag, |
| 160 |
self::CTAG => $this->ctag, |
| 161 |
self::LINE => $this->line, |
| 162 |
self::INDEX => ($this->tagType === self::T_END_SECTION) ? $this->seenTag - $this->otagLen : $i + $this->ctagLen |
| 163 |
); |
| 164 |
|
| 165 |
if ($this->tagType === self::T_UNESCAPED) { |
| 166 |
// Clean up `{{{ tripleStache }}}` style tokens. |
| 167 |
if ($this->ctag === '}}') { |
| 168 |
if (($i + 2 < $len) && $text[$i + 2] === '}') { |
| 169 |
$i++; |
| 170 |
} else { |
| 171 |
$msg = sprintf( |
| 172 |
'Mismatched tag delimiters: %s on line %d', |
| 173 |
$token[self::NAME], |
| 174 |
$token[self::LINE] |
| 175 |
); |
| 176 |
|
| 177 |
throw new Mustache_Exception_SyntaxException($msg, $token); |
| 178 |
} |
| 179 |
} else { |
| 180 |
$lastName = $token[self::NAME]; |
| 181 |
if (substr($lastName, -1) === '}') { |
| 182 |
$token[self::NAME] = trim(substr($lastName, 0, -1)); |
| 183 |
} else { |
| 184 |
$msg = sprintf( |
| 185 |
'Mismatched tag delimiters: %s on line %d', |
| 186 |
$token[self::NAME], |
| 187 |
$token[self::LINE] |
| 188 |
); |
| 189 |
|
| 190 |
throw new Mustache_Exception_SyntaxException($msg, $token); |
| 191 |
} |
| 192 |
} |
| 193 |
} |
| 194 |
|
| 195 |
$this->buffer = ''; |
| 196 |
$i += $this->ctagLen - 1; |
| 197 |
$this->state = self::IN_TEXT; |
| 198 |
$this->tokens[] = $token; |
| 199 |
} else { |
| 200 |
$this->buffer .= $text[$i]; |
| 201 |
} |
| 202 |
break; |
| 203 |
} |
| 204 |
} |
| 205 |
|
| 206 |
$this->flushBuffer(); |
| 207 |
|
| 208 |
// Restore the user's encoding... |
| 209 |
if ($encoding) { |
| 210 |
mb_internal_encoding($encoding); |
| 211 |
} |
| 212 |
|
| 213 |
return $this->tokens; |
| 214 |
} |
| 215 |
|
| 216 |
/** |
| 217 |
* Helper function to reset tokenizer internal state. |
| 218 |
*/ |
| 219 |
private function reset() |
| 220 |
{ |
| 221 |
$this->state = self::IN_TEXT; |
| 222 |
$this->tagType = null; |
| 223 |
$this->tag = null; |
| 224 |
$this->buffer = ''; |
| 225 |
$this->tokens = array(); |
| 226 |
$this->seenTag = false; |
| 227 |
$this->line = 0; |
| 228 |
$this->otag = '{{'; |
| 229 |
$this->ctag = '}}'; |
| 230 |
$this->otagLen = 2; |
| 231 |
$this->ctagLen = 2; |
| 232 |
} |
| 233 |
|
| 234 |
/** |
| 235 |
* Flush the current buffer to a token. |
| 236 |
*/ |
| 237 |
private function flushBuffer() |
| 238 |
{ |
| 239 |
if (strlen($this->buffer) > 0) { |
| 240 |
$this->tokens[] = array( |
| 241 |
self::TYPE => self::T_TEXT, |
| 242 |
self::LINE => $this->line, |
| 243 |
self::VALUE => $this->buffer |
| 244 |
); |
| 245 |
$this->buffer = ''; |
| 246 |
} |
| 247 |
} |
| 248 |
|
| 249 |
/** |
| 250 |
* Change the current Mustache delimiters. Set new `otag` and `ctag` values. |
| 251 |
* |
| 252 |
* @param string $text Mustache template source |
| 253 |
* @param int $index Current tokenizer index |
| 254 |
* |
| 255 |
* @return int New index value |
| 256 |
*/ |
| 257 |
private function changeDelimiters($text, $index) |
| 258 |
{ |
| 259 |
$startIndex = strpos($text, '=', $index) + 1; |
| 260 |
$close = '='.$this->ctag; |
| 261 |
$closeIndex = strpos($text, $close, $index); |
| 262 |
|
| 263 |
$this->setDelimiters(trim(substr($text, $startIndex, $closeIndex - $startIndex))); |
| 264 |
|
| 265 |
$this->tokens[] = array( |
| 266 |
self::TYPE => self::T_DELIM_CHANGE, |
| 267 |
self::LINE => $this->line, |
| 268 |
); |
| 269 |
|
| 270 |
return $closeIndex + strlen($close) - 1; |
| 271 |
} |
| 272 |
|
| 273 |
/** |
| 274 |
* Set the current Mustache `otag` and `ctag` delimiters. |
| 275 |
* |
| 276 |
* @param string $delimiters |
| 277 |
*/ |
| 278 |
private function setDelimiters($delimiters) |
| 279 |
{ |
| 280 |
list($otag, $ctag) = explode(' ', $delimiters); |
| 281 |
$this->otag = $otag; |
| 282 |
$this->ctag = $ctag; |
| 283 |
$this->otagLen = strlen($otag); |
| 284 |
$this->ctagLen = strlen($ctag); |
| 285 |
} |
| 286 |
|
| 287 |
/** |
| 288 |
* Add pragma token. |
| 289 |
* |
| 290 |
* Pragmas are hoisted to the front of the template, so all pragma tokens |
| 291 |
* will appear at the front of the token list. |
| 292 |
* |
| 293 |
* @param string $text |
| 294 |
* @param int $index |
| 295 |
* |
| 296 |
* @return int New index value |
| 297 |
*/ |
| 298 |
private function addPragma($text, $index) |
| 299 |
{ |
| 300 |
$end = strpos($text, $this->ctag, $index); |
| 301 |
$pragma = trim(substr($text, $index + 2, $end - $index - 2)); |
| 302 |
|
| 303 |
// Pragmas are hoisted to the front of the template. |
| 304 |
array_unshift($this->tokens, array( |
| 305 |
self::TYPE => self::T_PRAGMA, |
| 306 |
self::NAME => $pragma, |
| 307 |
self::LINE => 0, |
| 308 |
)); |
| 309 |
|
| 310 |
return $end + $this->ctagLen - 1; |
| 311 |
} |
| 312 |
|
| 313 |
/** |
| 314 |
* Test whether it's time to change tags. |
| 315 |
* |
| 316 |
* @param string $tag Current tag name |
| 317 |
* @param int $tagLen Current tag name length |
| 318 |
* @param string $text Mustache template source |
| 319 |
* @param int $index Current tokenizer index |
| 320 |
* |
| 321 |
* @return boolean True if this is a closing section tag |
| 322 |
*/ |
| 323 |
private function tagChange($tag, $tagLen, $text, $index) |
| 324 |
{ |
| 325 |
return substr($text, $index, $tagLen) === $tag; |
| 326 |
} |
| 327 |
} |
| 328 |
|