gggeek /
phpxmlrpc
| Total Complexity | 188 |
| Total Lines | 1099 |
| Duplicated Lines | 0 % |
| Coverage | 86.35% |
| Changes | 9 | ||
| Bugs | 1 | Features | 2 |
Complex classes like XMLParser often do a lot of different things. To break such a class down, we need to identify a cohesive component within that class. A common approach to find such a component is to look for fields/methods that share the same prefixes, or suffixes.
Once you have determined the fields that belong together, you can apply the Extract Class refactoring. If the component makes sense as a sub-class, Extract Subclass is also a candidate, and is often faster.
While breaking up the class, it is a good idea to analyze how other classes use XMLParser, and based on these observations, apply Extract Interface, too.
| 1 | <?php |
||
| 24 | class XMLParser |
||
| 25 | { |
||
| 26 | use DeprecationLogger; |
||
| 27 | |||
| 28 | const RETURN_XMLRPCVALS = 'xmlrpcvals'; |
||
| 29 | const RETURN_EPIVALS = 'epivals'; |
||
| 30 | const RETURN_PHP = 'phpvals'; |
||
| 31 | |||
| 32 | const ACCEPT_REQUEST = 1; |
||
| 33 | const ACCEPT_RESPONSE = 2; |
||
| 34 | const ACCEPT_VALUE = 4; |
||
| 35 | const ACCEPT_FAULT = 8; |
||
| 36 | |||
| 37 | /** |
||
| 38 | * @var int |
||
| 39 | * The max length beyond which data will get truncated in error messages |
||
| 40 | */ |
||
| 41 | protected $maxLogValueLength = 100; |
||
| 42 | |||
| 43 | /** |
||
| 44 | * @var array |
||
| 45 | * Used to store state during parsing and to pass parsing results to callers. |
||
| 46 | * Quick explanation of components: |
||
| 47 | * private: |
||
| 48 | * ac - used to accumulate values |
||
| 49 | * stack - array with genealogy of xml elements names, used to validate nesting of xml-rpc elements |
||
| 50 | * valuestack - array used for parsing arrays and structs |
||
| 51 | * lv - used to indicate "looking for a value": implements the logic to allow values with no types to be strings |
||
| 52 | * (values: 0=not looking, 1=looking, 3=found) |
||
| 53 | * public: |
||
| 54 | * isf - used to indicate an xml-rpc response fault (1), invalid xml-rpc fault (2), xml parsing fault (3) |
||
| 55 | * isf_reason - used for storing xml-rpc response fault string |
||
| 56 | * value - used to store the value in responses |
||
| 57 | * method - used to store method name in requests |
||
| 58 | * params - used to store parameters in requests |
||
| 59 | * pt - used to store the type of each received parameter. Useful if parameters are automatically decoded to php values |
||
| 60 | * rt - 'methodcall', 'methodresponse', 'value' or 'fault' (the last one used only in EPI emulation mode) |
||
| 61 | */ |
||
| 62 | protected $_xh = array( |
||
| 63 | 'ac' => '', |
||
| 64 | 'stack' => array(), |
||
| 65 | 'valuestack' => array(), |
||
| 66 | 'lv' => 0, |
||
| 67 | 'isf' => 0, |
||
| 68 | 'isf_reason' => '', |
||
| 69 | 'value' => null, |
||
| 70 | 'method' => false, |
||
| 71 | 'params' => array(), |
||
| 72 | 'pt' => array(), |
||
| 73 | 'rt' => '', |
||
| 74 | ); |
||
| 75 | |||
| 76 | /** |
||
| 77 | * @var array[] |
||
| 78 | */ |
||
| 79 | protected $xmlrpc_valid_parents = array( |
||
| 80 | 'VALUE' => array('MEMBER', 'DATA', 'PARAM', 'FAULT'), |
||
| 81 | 'BOOLEAN' => array('VALUE'), |
||
| 82 | 'I4' => array('VALUE'), |
||
| 83 | 'I8' => array('VALUE'), |
||
| 84 | 'EX:I8' => array('VALUE'), |
||
| 85 | 'INT' => array('VALUE'), |
||
| 86 | 'STRING' => array('VALUE'), |
||
| 87 | 'DOUBLE' => array('VALUE'), |
||
| 88 | 'DATETIME.ISO8601' => array('VALUE'), |
||
| 89 | 'BASE64' => array('VALUE'), |
||
| 90 | 'MEMBER' => array('STRUCT'), |
||
| 91 | 572 | 'NAME' => array('MEMBER'), |
|
| 92 | 'DATA' => array('ARRAY'), |
||
| 93 | 572 | 'ARRAY' => array('VALUE'), |
|
| 94 | 572 | 'STRUCT' => array('VALUE'), |
|
| 95 | 'PARAM' => array('PARAMS'), |
||
| 96 | 'METHODNAME' => array('METHODCALL'), |
||
| 97 | 'PARAMS' => array('METHODCALL', 'METHODRESPONSE'), |
||
| 98 | 'FAULT' => array('METHODRESPONSE'), |
||
| 99 | 'NIL' => array('VALUE'), // only used when extension activated |
||
| 100 | 'EX:NIL' => array('VALUE'), // only used when extension activated |
||
| 101 | ); |
||
| 102 | 712 | ||
| 103 | /** @var array $parsing_options */ |
||
| 104 | 712 | protected $parsing_options = array(); |
|
| 105 | |||
| 106 | /** @var int $accept self::ACCEPT_REQUEST | self::ACCEPT_RESPONSE by default */ |
||
| 107 | //protected $accept = 3; |
||
| 108 | |||
| 109 | /** @var int $maxChunkLength 4 MB by default. Any value below 10MB should be good */ |
||
| 110 | protected $maxChunkLength = 4194304; |
||
| 111 | /** @var array |
||
| 112 | * Used keys: accept, target_charset, methodname_callback, plus the ones set here. |
||
| 113 | * We initialize it partially to help keep BC with subclasses which might have reimplemented `parse()` but not |
||
| 114 | * the element handler methods |
||
| 115 | */ |
||
| 116 | protected $current_parsing_options = array( |
||
| 117 | 712 | 'xmlrpc_null_extension' => false, |
|
| 118 | 'xmlrpc_return_datetimes' => false, |
||
| 119 | 'xmlrpc_reject_invalid_values' => false |
||
| 120 | 712 | ); |
|
| 121 | 2 | ||
| 122 | 2 | /** |
|
| 123 | 2 | * @param array $options integer keys: options passed to the inner xml parser |
|
| 124 | * string keys: |
||
| 125 | * - target_charset (string) |
||
| 126 | 710 | * - methodname_callback (callable) |
|
| 127 | * - xmlrpc_null_extension (bool) |
||
| 128 | 710 | * - xmlrpc_return_datetimes (bool) |
|
| 129 | * - xmlrpc_reject_invalid_values (bool) |
||
| 130 | */ |
||
| 131 | 710 | public function __construct(array $options = array()) |
|
| 134 | } |
||
| 135 | 710 | ||
| 136 | /** |
||
| 137 | 710 | * Parses an xml-rpc xml string. Results of the parsing are found in $this->['_xh']. |
|
| 138 | * Logs to the error log any issues which do not cause the parsing to fail. |
||
| 139 | * |
||
| 140 | 710 | * @param string $data |
|
| 141 | 27 | * @param string $returnType self::RETURN_XMLRPCVALS, self::RETURN_PHP, self::RETURN_EPIVALS |
|
| 142 | 27 | * @param int $accept a bit-combination of self::ACCEPT_REQUEST, self::ACCEPT_RESPONSE, self::ACCEPT_VALUE |
|
| 143 | 708 | * @param array $options integer-key options are passed to the xml parser, string-key options are used independently. |
|
| 144 | * These options are added to options received in the constructor. |
||
| 145 | * Note that if options xmlrpc_null_extension, xmlrpc_return_datetimes and xmlrpc_reject_invalid_values |
||
| 146 | * are not set, the default settings from PhpXmlRpc\PhpXmlRpc are used |
||
| 147 | 708 | * @return array see the definition of $this->_xh for the meaning of the results |
|
| 148 | * @throws \Exception this can happen if a callback function is set and it does throw (i.e. we do not catch exceptions) |
||
| 149 | * |
||
| 150 | 710 | * @todo refactor? we could 1. return the parsed data structure, and 2. move $returnType and $accept into options |
|
| 151 | 710 | * @todo feature-creep make it possible to pass in options overriding usage of PhpXmlRpc::$xmlrpc_XXX_format, so |
|
| 152 | * that parsing will be completely independent of global state. Note that it might incur a small perf hit... |
||
| 153 | 710 | */ |
|
| 154 | public function parse($data, $returnType = self::RETURN_XMLRPCVALS, $accept = 3, $options = array()) |
||
| 155 | { |
||
| 156 | 710 | $this->_xh = array( |
|
| 157 | 710 | 'ac' => '', |
|
| 158 | 'stack' => array(), |
||
| 159 | 710 | 'valuestack' => array(), |
|
| 160 | 3 | 'lv' => 0, |
|
| 161 | 3 | 'isf' => 0, |
|
| 162 | 3 | 'isf_reason' => '', |
|
| 163 | 'value' => null, |
||
| 164 | 3 | 'method' => false, // so we can check later if we got a methodname or not |
|
| 165 | 3 | 'params' => false, // so we can check later if we got a params tag or not |
|
| 166 | 3 | 'pt' => array(), |
|
| 167 | 'rt' => '', |
||
| 168 | ); |
||
| 169 | |||
| 170 | 710 | $len = strlen($data); |
|
| 171 | 710 | ||
| 172 | // we test for empty documents here to save on resource allocation and simplify the chunked-parsing loop below |
||
| 173 | if ($len == 0) { |
||
| 174 | $this->_xh['isf'] = 3; |
||
| 175 | $this->_xh['isf_reason'] = 'XML error 5: empty document'; |
||
| 176 | return $this->_xh; |
||
| 177 | } |
||
| 178 | |||
| 179 | $this->current_parsing_options = array('accept' => $accept); |
||
| 180 | |||
| 181 | 710 | $mergedOptions = $this->parsing_options; |
|
| 182 | foreach ($options as $key => $val) { |
||
| 183 | $mergedOptions[$key] = $val; |
||
| 184 | 710 | } |
|
| 185 | |||
| 186 | foreach ($mergedOptions as $key => $val) { |
||
| 187 | 710 | // q: can php be built without ctype? should we use a regexp? |
|
| 188 | if (is_string($key) && !ctype_digit($key)) { |
||
| 189 | /// @todo on invalid options, throw/error-out instead of logging an error message? |
||
| 190 | switch($key) { |
||
| 191 | case 'target_charset': |
||
| 192 | 710 | if (function_exists('mb_convert_encoding')) { |
|
| 193 | 710 | $this->current_parsing_options['target_charset'] = $val; |
|
| 194 | } else { |
||
| 195 | $this->getLogger()->error('XML-RPC: ' . __METHOD__ . ": 'target_charset' option is unsupported without mbstring"); |
||
| 196 | } |
||
| 197 | 710 | break; |
|
| 198 | 708 | ||
| 199 | 3 | case 'methodname_callback': |
|
| 200 | 703 | if (is_callable($val)) { |
|
| 201 | 710 | $this->current_parsing_options['methodname_callback'] = $val; |
|
| 202 | } else { |
||
| 203 | 2 | $this->getLogger()->error('XML-RPC: ' . __METHOD__ . ": Callback passed as 'methodname_callback' is not callable"); |
|
| 204 | 2 | } |
|
| 205 | break; |
||
| 206 | 703 | ||
| 207 | case 'xmlrpc_null_extension': |
||
| 208 | case 'xmlrpc_return_datetimes': |
||
| 209 | case 'xmlrpc_reject_invalid_values': |
||
| 210 | 710 | $this->current_parsing_options[$key] = $val; |
|
| 211 | 710 | break; |
|
| 212 | 2 | ||
| 213 | 2 | default: |
|
| 214 | $this->getLogger()->error('XML-RPC: ' . __METHOD__ . ": unsupported option: $key"); |
||
| 215 | 2 | } |
|
| 216 | unset($mergedOptions[$key]); |
||
| 217 | } |
||
| 218 | } |
||
| 219 | 710 | ||
| 220 | if (!isset($this->current_parsing_options['xmlrpc_null_extension'])) { |
||
| 221 | 710 | $this->current_parsing_options['xmlrpc_null_extension'] = PhpXmlRpc::$xmlrpc_null_extension; |
|
| 222 | } |
||
| 223 | 708 | if (!isset($this->current_parsing_options['xmlrpc_return_datetimes'])) { |
|
| 224 | 708 | $this->current_parsing_options['xmlrpc_return_datetimes'] = PhpXmlRpc::$xmlrpc_return_datetimes; |
|
| 225 | 708 | } |
|
| 226 | 708 | if (!isset($this->current_parsing_options['xmlrpc_reject_invalid_values'])) { |
|
| 227 | 708 | $this->current_parsing_options['xmlrpc_reject_invalid_values'] = PhpXmlRpc::$xmlrpc_reject_invalid_values; |
|
| 228 | 710 | } |
|
| 229 | 710 | ||
| 230 | 1 | // NB: we use '' instead of null to force charset detection from the xml declaration |
|
| 231 | $parser = xml_parser_create(''); |
||
| 232 | |||
| 233 | foreach ($mergedOptions as $key => $val) { |
||
| 234 | xml_parser_set_option($parser, $key, $val); |
||
| 235 | } |
||
| 236 | |||
| 237 | // always set this, in case someone tries to disable it via options... |
||
| 238 | 710 | xml_parser_set_option($parser, XML_OPTION_CASE_FOLDING, 1); |
|
| 239 | 710 | ||
| 240 | 710 | xml_set_object($parser, $this); |
|
| 241 | 710 | ||
| 242 | 710 | switch ($returnType) { |
|
| 243 | 710 | case self::RETURN_PHP: |
|
| 244 | 710 | xml_set_element_handler($parser, 'xmlrpc_se', 'xmlrpc_ee_fast'); |
|
| 245 | 685 | break; |
|
| 246 | case self::RETURN_EPIVALS: |
||
| 247 | 1 | xml_set_element_handler($parser, 'xmlrpc_se', 'xmlrpc_ee_epi'); |
|
| 248 | 1 | break; |
|
| 249 | /// @todo log an error / throw / error-out on unsupported return type |
||
| 250 | 1 | case XMLParser::RETURN_XMLRPCVALS: |
|
| 251 | default: |
||
| 252 | 685 | xml_set_element_handler($parser, 'xmlrpc_se', 'xmlrpc_ee'); |
|
| 253 | 685 | } |
|
| 254 | 710 | ||
| 255 | 710 | xml_set_character_data_handler($parser, 'xmlrpc_cd'); |
|
| 256 | 400 | xml_set_default_handler($parser, 'xmlrpc_dh'); |
|
| 257 | |||
| 258 | 1 | try { |
|
| 259 | 1 | // @see ticket #70 - we have to parse big xml docs in chunks to avoid errors |
|
| 260 | for ($offset = 0; $offset < $len; $offset += $this->maxChunkLength) { |
||
| 261 | 1 | $chunk = substr($data, $offset, $this->maxChunkLength); |
|
| 262 | // error handling: xml not well formed |
||
| 263 | if (!@xml_parse($parser, $chunk, $offset + $this->maxChunkLength >= $len)) { |
||
| 264 | 399 | $errCode = xml_get_error_code($parser); |
|
| 265 | 399 | $errStr = sprintf('XML error %s: %s at line %d, column %d', $errCode, xml_error_string($errCode), |
|
| 266 | 399 | xml_get_current_line_number($parser), xml_get_current_column_number($parser)); |
|
| 267 | $this->_xh['isf'] = 3; |
||
| 268 | $this->_xh['isf_reason'] = $errStr; |
||
| 269 | 399 | } |
|
| 270 | 22 | // no need to parse further if we already have a fatal error |
|
| 271 | if ($this->_xh['isf'] >= 2) { |
||
| 272 | 399 | break; |
|
| 273 | 399 | } |
|
| 274 | 399 | } |
|
| 275 | 710 | /// @todo bump minimum php version to 5.5 and use a finally clause instead of doing cleanup 3 times |
|
| 276 | 239 | } catch (\Exception $e) { |
|
| 277 | xml_parser_free($parser); |
||
| 278 | 1 | $this->current_parsing_options = array(); |
|
| 279 | 1 | /// @todo should we set $this->_xh['isf'] and $this->_xh['isf_reason'] ? |
|
| 280 | throw $e; |
||
| 281 | 1 | } catch (\Error $e) { |
|
| 282 | xml_parser_free($parser); |
||
| 283 | 710 | $this->current_parsing_options = array(); |
|
| 284 | 710 | //$this->accept = $prevAccept; |
|
| 285 | 710 | /// @todo should we set $this->_xh['isf'] and $this->_xh['isf_reason'] ? |
|
| 286 | throw $e; |
||
| 287 | 710 | } |
|
| 288 | 710 | ||
| 289 | 710 | xml_parser_free($parser); |
|
| 290 | $this->current_parsing_options = array(); |
||
| 291 | 637 | ||
| 292 | 637 | // BC |
|
| 293 | 710 | if ($this->_xh['params'] === false) { |
|
|
introduced
by
|
|||
| 294 | 109 | $this->_xh['params'] = array(); |
|
| 295 | 109 | } |
|
| 296 | 710 | ||
| 297 | return $this->_xh; |
||
| 298 | 289 | } |
|
| 299 | |||
| 300 | /** |
||
| 301 | 688 | * xml parser handler function for opening element tags. |
|
| 302 | * @internal |
||
| 303 | 710 | * |
|
| 304 | 710 | * @param resource $parser |
|
| 305 | 23 | * @param string $name |
|
| 306 | 23 | * @param $attrs |
|
| 307 | 23 | * @param bool $acceptSingleVals DEPRECATED use the $accept parameter instead |
|
| 308 | 23 | * @return void |
|
| 309 | * |
||
| 310 | * @todo optimization creep: throw when setting $this->_xh['isf'] > 1, to completely avoid further xml parsing |
||
| 311 | * and remove the checking for $this->_xh['isf'] >= 2 everywhere |
||
| 312 | */ |
||
| 313 | public function xmlrpc_se($parser, $name, $attrs, $acceptSingleVals = false) |
||
| 314 | { |
||
| 315 | 23 | // if invalid xml-rpc already detected, skip all processing |
|
| 316 | 23 | if ($this->_xh['isf'] >= 2) { |
|
| 317 | return; |
||
| 318 | } |
||
| 319 | |||
| 320 | // check for correct element nesting |
||
| 321 | if (count($this->_xh['stack']) == 0) { |
||
| 322 | 1 | // top level element can only be of 2 types |
|
| 323 | 1 | /// @todo optimization creep: save this check into a bool variable, instead of using count() every time: |
|
| 324 | 1 | /// there is only a single top level element in xml anyway |
|
| 325 | |||
| 326 | // BC |
||
| 327 | if ($acceptSingleVals === false) { |
||
| 328 | 710 | $accept = $this->current_parsing_options['accept']; |
|
| 329 | } else { |
||
| 330 | $this->logDeprecation('Using argument $acceptSingleVals for method ' . __METHOD__ . ' is deprecated'); |
||
| 331 | 710 | $accept = self::ACCEPT_REQUEST | self::ACCEPT_RESPONSE | self::ACCEPT_VALUE; |
|
| 332 | 710 | } |
|
| 333 | if (($name == 'METHODCALL' && ($accept & self::ACCEPT_REQUEST)) || |
||
| 334 | ($name == 'METHODRESPONSE' && ($accept & self::ACCEPT_RESPONSE)) || |
||
| 335 | 710 | ($name == 'VALUE' && ($accept & self::ACCEPT_VALUE)) || |
|
| 336 | ($name == 'FAULT' && ($accept & self::ACCEPT_FAULT))) { |
||
| 337 | $this->_xh['rt'] = strtolower($name); |
||
| 338 | } else { |
||
| 339 | $this->_xh['isf'] = 2; |
||
| 340 | $this->_xh['isf_reason'] = 'missing top level xmlrpc element. Found: ' . $name; |
||
| 341 | |||
| 342 | return; |
||
| 343 | } |
||
| 344 | } else { |
||
| 345 | // not top level element: see if parent is OK |
||
| 346 | $parent = end($this->_xh['stack']); |
||
| 347 | if (!array_key_exists($name, $this->xmlrpc_valid_parents) || !in_array($parent, $this->xmlrpc_valid_parents[$name])) { |
||
| 348 | $this->_xh['isf'] = 2; |
||
| 349 | $this->_xh['isf_reason'] = "xmlrpc element $name cannot be child of $parent"; |
||
| 350 | |||
| 351 | return; |
||
| 352 | } |
||
| 353 | } |
||
| 354 | |||
| 355 | switch ($name) { |
||
| 356 | // optimize for speed switch cases: most common cases first |
||
| 357 | 710 | case 'VALUE': |
|
| 358 | /// @todo we could check for 2 VALUE elements inside a MEMBER or PARAM element |
||
| 359 | 710 | $this->_xh['vt'] = 'value'; // indicator: no value found yet |
|
| 360 | $this->_xh['ac'] = ''; |
||
| 361 | $this->_xh['lv'] = 1; |
||
| 362 | $this->_xh['php_class'] = null; |
||
| 363 | break; |
||
| 364 | 709 | ||
| 365 | case 'I8': |
||
| 366 | 709 | case 'EX:I8': |
|
| 367 | 709 | if (PHP_INT_SIZE === 4) { |
|
| 368 | // INVALID ELEMENT: RAISE ISF so that it is later recognized!!! |
||
| 369 | 707 | $this->_xh['isf'] = 2; |
|
| 370 | 30 | $this->_xh['isf_reason'] = "Received i8 element but php is compiled in 32 bit mode"; |
|
| 371 | 30 | ||
| 372 | return; |
||
| 373 | } |
||
| 374 | 707 | // fall through voluntarily |
|
| 375 | |||
| 376 | 705 | case 'I4': |
|
| 377 | case 'INT': |
||
| 378 | case 'STRING': |
||
| 379 | 705 | case 'BOOLEAN': |
|
| 380 | 22 | case 'DOUBLE': |
|
| 381 | case 'DATETIME.ISO8601': |
||
| 382 | 705 | case 'BASE64': |
|
| 383 | 27 | if ($this->_xh['vt'] != 'value') { |
|
| 384 | // two data elements inside a value: an error occurred! |
||
| 385 | $this->_xh['isf'] = 2; |
||
| 386 | $this->_xh['isf_reason'] = "$name element following a {$this->_xh['vt']} element inside a single value"; |
||
| 387 | |||
| 388 | return; |
||
| 389 | } |
||
| 390 | $this->_xh['ac'] = ''; // reset the accumulator |
||
| 391 | break; |
||
| 392 | |||
| 393 | case 'STRUCT': |
||
| 394 | case 'ARRAY': |
||
| 395 | if ($this->_xh['vt'] != 'value') { |
||
| 396 | // two data elements inside a value: an error occurred! |
||
| 397 | $this->_xh['isf'] = 2; |
||
| 398 | $this->_xh['isf_reason'] = "$name element following a {$this->_xh['vt']} element inside a single value"; |
||
| 399 | |||
| 400 | return; |
||
| 401 | } |
||
| 402 | // create an empty array to hold child values, and push it onto appropriate stack |
||
| 403 | $curVal = array( |
||
| 404 | 'values' => array(), |
||
| 405 | 'type' => $name, |
||
| 406 | 707 | ); |
|
| 407 | 707 | // check for out-of-band information to rebuild php objs and, in case it is found, save it |
|
| 408 | 239 | if (@isset($attrs['PHP_CLASS'])) { |
|
| 409 | $curVal['php_class'] = $attrs['PHP_CLASS']; |
||
| 410 | 707 | } |
|
| 411 | 709 | $this->_xh['valuestack'][] = $curVal; |
|
| 412 | 709 | $this->_xh['vt'] = 'data'; // be prepared for a data element next |
|
| 413 | 709 | break; |
|
| 414 | 709 | ||
| 415 | 709 | case 'DATA': |
|
| 416 | 709 | if ($this->_xh['vt'] != 'data') { |
|
| 417 | 708 | // two data elements inside a value: an error occurred! |
|
| 418 | 708 | $this->_xh['isf'] = 2; |
|
| 419 | 708 | $this->_xh['isf_reason'] = "found two data elements inside an array element"; |
|
| 420 | 685 | ||
| 421 | return; |
||
| 422 | } |
||
| 423 | 685 | ||
| 424 | 594 | case 'METHODCALL': |
|
| 425 | 477 | case 'METHODRESPONSE': |
|
| 426 | 7 | // valid elements that add little to processing |
|
| 427 | break; |
||
| 428 | |||
| 429 | 7 | case 'PARAMS': |
|
| 430 | 7 | $this->_xh['params'] = array(); |
|
| 431 | 472 | break; |
|
| 432 | |||
| 433 | 22 | case 'METHODNAME': |
|
| 434 | 451 | case 'NAME': |
|
| 435 | /// @todo we could check for 2 NAME elements inside a MEMBER element |
||
| 436 | $this->_xh['ac'] = ''; |
||
| 437 | break; |
||
| 438 | |||
| 439 | case 'FAULT': |
||
| 440 | $this->_xh['isf'] = 1; |
||
| 441 | 46 | break; |
|
| 442 | 46 | ||
| 443 | case 'MEMBER': |
||
| 444 | // set member name to null, in case we do not find in the xml later on |
||
| 445 | 24 | $this->_xh['valuestack'][count($this->_xh['valuestack']) - 1]['name'] = null; |
|
| 446 | //$this->_xh['ac']=''; |
||
| 447 | // Drop trough intentionally |
||
| 448 | 46 | ||
| 449 | case 'PARAM': |
||
| 450 | 408 | // clear value type, so we can check later if no value has been passed for this param/member |
|
| 451 | $this->_xh['vt'] = null; |
||
| 452 | break; |
||
| 453 | |||
| 454 | 25 | case 'NIL': |
|
| 455 | case 'EX:NIL': |
||
| 456 | if ($this->current_parsing_options['xmlrpc_null_extension']) { |
||
| 457 | if ($this->_xh['vt'] != 'value') { |
||
| 458 | // two data elements inside a value: an error occurred! |
||
| 459 | $this->_xh['isf'] = 2; |
||
| 460 | 25 | $this->_xh['isf_reason'] = "$name element following a {$this->_xh['vt']} element inside a single value"; |
|
| 461 | |||
| 462 | return; |
||
| 463 | } |
||
| 464 | // reset the accumulator - q: is this necessary at all here? we don't use it on _ee anyway for NILs |
||
| 465 | 387 | $this->_xh['ac'] = ''; |
|
| 466 | |||
| 467 | } else { |
||
| 468 | $this->_xh['isf'] = 2; |
||
| 469 | $this->_xh['isf_reason'] = 'Invalid NIL value received. Support for NIL can be enabled via \\PhpXmlRpc\\PhpXmlRpc::$xmlrpc_null_extension'; |
||
| 470 | |||
| 471 | 387 | return; |
|
| 472 | } |
||
| 473 | break; |
||
| 474 | 685 | ||
| 475 | 685 | default: |
|
| 476 | 708 | // INVALID ELEMENT: RAISE ISF so that it is later recognized |
|
| 477 | 289 | /// @todo feature creep = allow a callback instead |
|
| 478 | 289 | $this->_xh['isf'] = 2; |
|
| 479 | 708 | $this->_xh['isf_reason'] = "found not-xmlrpc xml element $name"; |
|
| 480 | |||
| 481 | return; |
||
| 482 | 289 | } |
|
| 483 | 268 | ||
| 484 | 268 | // Save current element name to stack, to validate nesting |
|
| 485 | $this->_xh['stack'][] = $name; |
||
| 486 | 22 | ||
| 487 | /// @todo optimization creep: move this inside the big switch() above |
||
| 488 | 289 | if ($name != 'VALUE') { |
|
| 489 | 708 | $this->_xh['lv'] = 0; |
|
| 490 | 239 | } |
|
| 491 | 239 | } |
|
| 492 | 707 | ||
| 493 | 707 | /** |
|
| 494 | * xml parser handler function for close element tags. |
||
| 495 | 398 | * @internal |
|
| 496 | 398 | * |
|
| 497 | 398 | * @param resource $parser |
|
| 498 | 398 | * @param string $name |
|
| 499 | 22 | * @param int $rebuildXmlrpcvals >1 for rebuilding xmlrpcvals, 0 for rebuilding php values, -1 for xmlrpc-extension compatibility |
|
| 500 | * @return void |
||
| 501 | 398 | * @throws \Exception this can happen if a callback function is set and it does throw (i.e. we do not catch exceptions) |
|
| 502 | 707 | * |
|
| 503 | * @todo optimization creep: throw when setting $this->_xh['isf'] > 1, to completely avoid further xml parsing |
||
| 504 | * and remove the checking for $this->_xh['isf'] >= 2 everywhere |
||
| 505 | 684 | */ |
|
| 506 | 684 | public function xmlrpc_ee($parser, $name, $rebuildXmlrpcvals = 1) |
|
| 507 | 684 | { |
|
| 508 | if ($this->_xh['isf'] >= 2) { |
||
| 509 | return; |
||
| 510 | } |
||
| 511 | 684 | ||
| 512 | 707 | // push this element name from stack |
|
| 513 | 562 | // NB: if XML validates, correct opening/closing is guaranteed and we do not have to check for $name == $currElem. |
|
| 514 | 562 | // we also checked for proper nesting at start of elements... |
|
| 515 | 706 | $currElem = array_pop($this->_xh['stack']); |
|
| 516 | 706 | ||
| 517 | 23 | switch ($name) { |
|
| 518 | 23 | case 'VALUE': |
|
| 519 | 23 | // If no scalar was inside <VALUE></VALUE>, it was a string value |
|
| 520 | 23 | if ($this->_xh['vt'] == 'value') { |
|
| 521 | 23 | $this->_xh['value'] = $this->_xh['ac']; |
|
| 522 | $this->_xh['vt'] = Value::$xmlrpcString; |
||
| 523 | } |
||
| 524 | 706 | ||
| 525 | 706 | // in case there is charset conversion required, do it here, to catch both cases of string values |
|
| 526 | 706 | if (isset($this->current_parsing_options['target_charset']) && $this->_xh['vt'] === Value::$xmlrpcString) { |
|
| 527 | 705 | $this->_xh['value'] = mb_convert_encoding($this->_xh['value'], $this->current_parsing_options['target_charset'], 'UTF-8'); |
|
| 528 | 706 | } |
|
| 529 | |||
| 530 | if ($rebuildXmlrpcvals > 0) { |
||
| 531 | // build the xml-rpc val out of the data received, and substitute it |
||
| 532 | 705 | $temp = new Value($this->_xh['value'], $this->_xh['vt']); |
|
| 533 | // in case we got info about underlying php class, save it in the object we're rebuilding |
||
| 534 | if (isset($this->_xh['php_class'])) { |
||
| 535 | 710 | $temp->_php_class = $this->_xh['php_class']; |
|
| 536 | } |
||
| 537 | $this->_xh['value'] = $temp; |
||
| 538 | } elseif ($rebuildXmlrpcvals < 0) { |
||
| 539 | if ($this->_xh['vt'] == Value::$xmlrpcDateTime) { |
||
| 540 | $this->_xh['value'] = (object)array( |
||
| 541 | 'xmlrpc_type' => 'datetime', |
||
| 542 | 'scalar' => $this->_xh['value'], |
||
| 543 | 27 | 'timestamp' => \PhpXmlRpc\Helper\Date::iso8601Decode($this->_xh['value']) |
|
| 544 | ); |
||
| 545 | 27 | } elseif ($this->_xh['vt'] == Value::$xmlrpcBase64) { |
|
| 546 | 27 | $this->_xh['value'] = (object)array( |
|
| 547 | 'xmlrpc_type' => 'base64', |
||
| 548 | 'scalar' => $this->_xh['value'] |
||
| 549 | ); |
||
| 550 | } |
||
| 551 | } else { |
||
| 552 | /// @todo this should handle php-serialized objects, since std deserializing is done |
||
| 553 | /// by php_xmlrpc_decode, which we will not be calling... |
||
| 554 | //if (isset($this->_xh['php_class'])) { |
||
| 555 | //} |
||
| 556 | } |
||
| 557 | |||
| 558 | // check if we are inside an array or struct: |
||
| 559 | // if value just built is inside an array, let's move it into array on the stack |
||
| 560 | $vscount = count($this->_xh['valuestack']); |
||
| 561 | if ($vscount && $this->_xh['valuestack'][$vscount - 1]['type'] == 'ARRAY') { |
||
| 562 | $this->_xh['valuestack'][$vscount - 1]['values'][] = $this->_xh['value']; |
||
| 563 | } |
||
| 564 | break; |
||
| 565 | 710 | ||
| 566 | case 'STRING': |
||
| 567 | $this->_xh['vt'] = Value::$xmlrpcString; |
||
| 568 | 710 | $this->_xh['lv'] = 3; // indicate we've found a value |
|
| 569 | $this->_xh['value'] = $this->_xh['ac']; |
||
| 570 | break; |
||
| 571 | 710 | ||
| 572 | 710 | case 'BOOLEAN': |
|
| 573 | $this->_xh['vt'] = Value::$xmlrpcBoolean; |
||
| 574 | $this->_xh['lv'] = 3; // indicate we've found a value |
||
| 575 | 710 | // We translate boolean 1 or 0 into PHP constants true or false. Strings 'true' and 'false' are accepted, |
|
| 576 | // even though the spec never mentions them (see e.g. Blogger api docs) |
||
| 577 | // NB: this simple checks helps a lot sanitizing input, i.e. no security problems around here |
||
| 578 | // Note the non-strict type check: it will allow ' 1 ' |
||
| 579 | /// @todo feature-creep: use a flexible regexp, the same as we do with int, double and datetime. |
||
| 580 | /// Note that using a regexp would also make this test less sensitive to phpunit shenanigans, and |
||
| 581 | /// to changes in the way php compares strings (since 8.0, leading and trailing newlines are |
||
| 582 | /// accepted when deciding if a string numeric...) |
||
| 583 | if ($this->_xh['ac'] == '1' || strcasecmp($this->_xh['ac'], 'true') === 0) { |
||
| 584 | 696 | $this->_xh['value'] = true; |
|
| 585 | } else { |
||
| 586 | // log if receiving something strange, even though we set the value to false anyway |
||
| 587 | 696 | /// @todo to be consistent with the other types, we should return a value outside the good-value domain, e.g. NULL |
|
| 588 | 696 | if ($this->_xh['ac'] != '0' && strcasecmp($this->_xh['ac'], 'false') !== 0) { |
|
| 589 | if (!$this->handleParsingError('invalid data received in BOOLEAN value: ' . |
||
| 590 | $this->truncateValueForLog($this->_xh['ac']), __METHOD__)) { |
||
| 591 | return; |
||
| 592 | } |
||
| 593 | } |
||
| 594 | 696 | $this->_xh['value'] = false; |
|
| 595 | } |
||
| 596 | break; |
||
| 597 | |||
| 598 | case 'EX:I8': |
||
| 599 | $name = 'i8'; |
||
| 600 | // fall through voluntarily |
||
| 601 | case 'I4': |
||
| 602 | case 'I8': |
||
| 603 | case 'INT': |
||
| 604 | // NB: we build the Value object with the original xml element name found, except for ex:i8. The |
||
| 605 | // `Value::scalarTyp()` function will do some normalization of the data |
||
| 606 | $this->_xh['vt'] = strtolower($name); |
||
| 607 | $this->_xh['lv'] = 3; // indicate we've found a value |
||
| 608 | if (!preg_match(PhpXmlRpc::$xmlrpc_int_format, $this->_xh['ac'])) { |
||
| 609 | if (!$this->handleParsingError('non numeric data received in INT value: ' . |
||
| 610 | $this->truncateValueForLog($this->_xh['ac']), __METHOD__)) { |
||
| 611 | return; |
||
| 612 | } |
||
| 613 | /// @todo: find a better way of reporting an error value than this! Use NaN? |
||
| 614 | $this->_xh['value'] = 'ERROR_NON_NUMERIC_FOUND'; |
||
| 615 | } else { |
||
| 616 | // it's ok, add it on |
||
| 617 | 711 | $this->_xh['value'] = (int)$this->_xh['ac']; |
|
| 618 | } |
||
| 619 | break; |
||
| 620 | |||
| 621 | case 'DOUBLE': |
||
| 622 | $this->_xh['vt'] = Value::$xmlrpcDouble; |
||
| 623 | $this->_xh['lv'] = 3; // indicate we've found a value |
||
| 624 | if (!preg_match(PhpXmlRpc::$xmlrpc_double_format, $this->_xh['ac'])) { |
||
| 625 | if (!$this->handleParsingError('non numeric data received in DOUBLE value: ' . |
||
| 626 | $this->truncateValueForLog($this->_xh['ac']), __METHOD__)) { |
||
| 627 | return; |
||
| 628 | } |
||
| 629 | |||
| 630 | $this->_xh['value'] = 'ERROR_NON_NUMERIC_FOUND'; |
||
| 631 | } else { |
||
| 632 | // it's ok, add it on |
||
| 633 | $this->_xh['value'] = (double)$this->_xh['ac']; |
||
| 634 | 711 | } |
|
| 635 | 711 | break; |
|
| 636 | 695 | ||
| 637 | case 'DATETIME.ISO8601': |
||
| 638 | $this->_xh['vt'] = Value::$xmlrpcDateTime; |
||
| 639 | $this->_xh['lv'] = 3; // indicate we've found a value |
||
| 640 | if (!preg_match(PhpXmlRpc::$xmlrpc_datetime_format, $this->_xh['ac'])) { |
||
| 641 | if (!$this->handleParsingError('invalid data received in DATETIME value: ' . |
||
| 642 | $this->truncateValueForLog($this->_xh['ac']), __METHOD__)) { |
||
| 643 | return; |
||
| 644 | } |
||
| 645 | } |
||
| 646 | 529 | if ($this->current_parsing_options['xmlrpc_return_datetimes']) { |
|
| 647 | try { |
||
| 648 | 529 | $this->_xh['value'] = new \DateTime($this->_xh['ac']); |
|
| 649 | |||
| 650 | 529 | // the default regex used to validate the date string a few lines above should make this case impossible, |
|
| 651 | // but one never knows... |
||
| 652 | } catch(\Exception $e) { |
||
| 653 | // what to do? We can not guarantee that a valid date can be created. We return null... |
||
| 654 | if (!$this->handleParsingError('invalid data received in DATETIME value. Error ' . |
||
| 655 | $e->getMessage(), __METHOD__)) { |
||
| 656 | return; |
||
| 657 | } |
||
| 658 | 529 | } |
|
| 659 | 529 | } else { |
|
| 660 | $this->_xh['value'] = $this->_xh['ac']; |
||
| 661 | 24 | } |
|
| 662 | break; |
||
| 663 | |||
| 664 | case 'BASE64': |
||
| 665 | 506 | $this->_xh['vt'] = Value::$xmlrpcBase64; |
|
| 666 | 506 | $this->_xh['lv'] = 3; // indicate we've found a value |
|
| 667 | 4 | if ($this->current_parsing_options['xmlrpc_reject_invalid_values']) { |
|
| 668 | $v = base64_decode($this->_xh['ac'], true); |
||
| 669 | 506 | if ($v === false) { |
|
| 670 | 4 | $this->_xh['isf'] = 2; |
|
| 671 | $this->_xh['isf_reason'] = 'Invalid data received in BASE64 value: '. $this->truncateValueForLog($this->_xh['ac']); |
||
| 672 | 502 | return; |
|
| 673 | } |
||
| 674 | } else { |
||
| 675 | $v = base64_decode($this->_xh['ac']); |
||
| 676 | 506 | if ($v === '' && $this->_xh['ac'] !== '') { |
|
| 677 | 499 | // only the empty string should decode to the empty string |
|
| 678 | $this->getLogger()->error('XML-RPC: ' . __METHOD__ . ': invalid data received in BASE64 value: ' . |
||
| 679 | $this->truncateValueForLog($this->_xh['ac'])); |
||
| 680 | 506 | } |
|
| 681 | } |
||
| 682 | $this->_xh['value'] = $v; |
||
| 683 | break; |
||
| 684 | |||
| 685 | case 'NAME': |
||
| 686 | $this->_xh['valuestack'][count($this->_xh['valuestack']) - 1]['name'] = $this->_xh['ac']; |
||
| 687 | break; |
||
| 688 | |||
| 689 | case 'MEMBER': |
||
| 690 | // add to array in the stack the last element built, unless no VALUE or no NAME were found |
||
| 691 | if ($this->_xh['vt']) { |
||
| 692 | $vscount = count($this->_xh['valuestack']); |
||
| 693 | if ($this->_xh['valuestack'][$vscount - 1]['name'] === null) { |
||
| 694 | if (!$this->handleParsingError('missing NAME inside STRUCT in received xml', __METHOD__)) { |
||
| 695 | return; |
||
| 696 | 82 | } |
|
| 697 | $this->_xh['valuestack'][$vscount - 1]['name'] = ''; |
||
| 698 | } |
||
| 699 | $this->_xh['valuestack'][$vscount - 1]['values'][$this->_xh['valuestack'][$vscount - 1]['name']] = $this->_xh['value']; |
||
| 700 | 82 | } else { |
|
| 701 | if (!$this->handleParsingError('missing VALUE inside STRUCT in received xml', __METHOD__)) { |
||
| 702 | 82 | return; |
|
| 703 | } |
||
| 704 | 82 | } |
|
| 705 | break; |
||
| 706 | |||
| 707 | case 'DATA': |
||
| 708 | $this->_xh['vt'] = null; // reset this to check for 2 data elements in a row - even if they're empty |
||
| 709 | break; |
||
| 710 | |||
| 711 | case 'STRUCT': |
||
| 712 | 82 | case 'ARRAY': |
|
| 713 | 82 | // fetch out of stack array of values, and promote it to current value |
|
| 714 | $currVal = array_pop($this->_xh['valuestack']); |
||
| 715 | 78 | $this->_xh['value'] = $currVal['values']; |
|
| 716 | $this->_xh['vt'] = strtolower($name); |
||
| 717 | if (isset($currVal['php_class'])) { |
||
| 718 | 5 | $this->_xh['php_class'] = $currVal['php_class']; |
|
| 719 | } |
||
| 720 | break; |
||
| 721 | |||
| 722 | case 'PARAM': |
||
| 723 | // add to array of params the current value, unless no VALUE was found |
||
| 724 | /// @todo should we also check if there were two VALUE inside the PARAM? |
||
| 725 | if ($this->_xh['vt']) { |
||
| 726 | $this->_xh['params'][] = $this->_xh['value']; |
||
| 727 | $this->_xh['pt'][] = $this->_xh['vt']; |
||
| 728 | } else { |
||
| 729 | if (!$this->handleParsingError('missing VALUE inside PARAM in received xml', __METHOD__)) { |
||
| 730 | return; |
||
| 731 | } |
||
| 732 | } |
||
| 733 | break; |
||
| 734 | |||
| 735 | case 'METHODNAME': |
||
| 736 | if (!preg_match(PhpXmlRpc::$xmlrpc_methodname_format, $this->_xh['ac'])) { |
||
| 737 | if (!$this->handleParsingError('invalid data received in METHODNAME: '. |
||
| 738 | $this->truncateValueForLog($this->_xh['ac']), __METHOD__)) { |
||
| 739 | return; |
||
| 740 | } |
||
| 741 | } |
||
| 742 | $methodName = trim($this->_xh['ac']); |
||
| 743 | $this->_xh['method'] = $methodName; |
||
| 744 | // we allow the callback to f.e. give us back a mangled method name by manipulating $this |
||
| 745 | if (isset($this->current_parsing_options['methodname_callback'])) { |
||
| 746 | call_user_func($this->current_parsing_options['methodname_callback'], $methodName, $this, $parser); |
||
| 747 | } |
||
| 748 | break; |
||
| 749 | |||
| 750 | case 'NIL': |
||
| 751 | case 'EX:NIL': |
||
| 752 | // NB: if NIL support is not enabled, parsing stops at element start. So this If is redundant |
||
| 753 | //if ($this->current_parsing_options['xmlrpc_null_extension']) { |
||
| 754 | $this->_xh['vt'] = 'null'; |
||
| 755 | $this->_xh['value'] = null; |
||
| 756 | $this->_xh['lv'] = 3; |
||
| 757 | //} |
||
| 758 | break; |
||
| 759 | |||
| 760 | /// @todo add extra checking: |
||
| 761 | /// - FAULT should contain a single struct with the 2 expected members (check their name and type) |
||
| 762 | case 'PARAMS': |
||
| 763 | case 'FAULT': |
||
| 764 | break; |
||
| 765 | |||
| 766 | case 'METHODCALL': |
||
| 767 | /// @todo should we allow to accept this case via a call to handleParsingError ? |
||
| 768 | if ($this->_xh['method'] === false) { |
||
| 769 | $this->_xh['isf'] = 2; |
||
| 770 | $this->_xh['isf_reason'] = "missing METHODNAME element inside METHODCALL"; |
||
| 771 | } |
||
| 772 | break; |
||
| 773 | |||
| 774 | case 'METHODRESPONSE': |
||
| 775 | /// @todo should we allow to accept these cases via a call to handleParsingError ? |
||
| 776 | if ($this->_xh['isf'] != 1 && $this->_xh['params'] === false) { |
||
| 777 | $this->_xh['isf'] = 2; |
||
| 778 | $this->_xh['isf_reason'] = "missing both FAULT and PARAMS elements inside METHODRESPONSE"; |
||
| 779 | } elseif ($this->_xh['isf'] == 0 && count($this->_xh['params']) !== 1) { |
||
| 780 | $this->_xh['isf'] = 2; |
||
| 781 | $this->_xh['isf_reason'] = "PARAMS element inside METHODRESPONSE should have exactly 1 PARAM"; |
||
| 782 | } elseif ($this->_xh['isf'] == 1 && $this->_xh['params'] !== false) { |
||
| 783 | $this->_xh['isf'] = 2; |
||
| 784 | $this->_xh['isf_reason'] = "both FAULT and PARAMS elements found inside METHODRESPONSE"; |
||
| 785 | } |
||
| 786 | break; |
||
| 787 | |||
| 788 | default: |
||
| 789 | // End of INVALID ELEMENT |
||
| 790 | // Should we add an assert here for unreachable code? When an invalid element is found in xmlrpc_se, |
||
| 791 | // $this->_xh['isf'] is set to 2... |
||
| 792 | break; |
||
| 793 | } |
||
| 794 | } |
||
| 795 | |||
| 796 | /** |
||
| 797 | * Used in decoding xml-rpc requests/responses without rebuilding xml-rpc Values. |
||
| 798 | * @internal |
||
| 799 | * |
||
| 800 | * @param resource $parser |
||
| 801 | * @param string $name |
||
| 802 | * @return void |
||
| 803 | */ |
||
| 804 | public function xmlrpc_ee_fast($parser, $name) |
||
| 805 | { |
||
| 806 | $this->xmlrpc_ee($parser, $name, 0); |
||
| 807 | } |
||
| 808 | |||
| 809 | /** |
||
| 810 | * Used in decoding xml-rpc requests/responses while building xmlrpc-extension Values (plain php for all but base64 and datetime). |
||
| 811 | * @internal |
||
| 812 | * |
||
| 813 | * @param resource $parser |
||
| 814 | * @param string $name |
||
| 815 | * @return void |
||
| 816 | */ |
||
| 817 | public function xmlrpc_ee_epi($parser, $name) |
||
| 818 | { |
||
| 819 | $this->xmlrpc_ee($parser, $name, -1); |
||
| 820 | } |
||
| 821 | |||
| 822 | /** |
||
| 823 | * xml parser handler function for character data. |
||
| 824 | * @internal |
||
| 825 | * |
||
| 826 | * @param resource $parser |
||
| 827 | * @param string $data |
||
| 828 | * @return void |
||
| 829 | */ |
||
| 830 | public function xmlrpc_cd($parser, $data) |
||
| 831 | { |
||
| 832 | // skip processing if xml fault already detected |
||
| 833 | if ($this->_xh['isf'] >= 2) { |
||
| 834 | return; |
||
| 835 | } |
||
| 836 | |||
| 837 | // "lookforvalue == 3" means that we've found an entire value and should discard any further character data |
||
| 838 | if ($this->_xh['lv'] != 3) { |
||
| 839 | $this->_xh['ac'] .= $data; |
||
| 840 | } |
||
| 841 | } |
||
| 842 | |||
| 843 | /** |
||
| 844 | * xml parser handler function for 'other stuff', i.e. not char data or element start/end tag. |
||
| 845 | * In fact, it only gets called on unknown entities... |
||
| 846 | * @internal |
||
| 847 | * |
||
| 848 | * @param $parser |
||
| 849 | * @param string data |
||
| 850 | * @return void |
||
| 851 | */ |
||
| 852 | public function xmlrpc_dh($parser, $data) |
||
| 853 | { |
||
| 854 | // skip processing if xml fault already detected |
||
| 855 | if ($this->_xh['isf'] >= 2) { |
||
| 856 | return; |
||
| 857 | } |
||
| 858 | |||
| 859 | if (substr($data, 0, 1) == '&' && substr($data, -1, 1) == ';') { |
||
| 860 | $this->_xh['ac'] .= $data; |
||
| 861 | } |
||
| 862 | } |
||
| 863 | |||
| 864 | /** |
||
| 865 | * xml charset encoding guessing helper function. |
||
| 866 | * Tries to determine the charset encoding of an XML chunk received over HTTP. |
||
| 867 | * |
||
| 868 | * NB: according to the spec (RFC 3023), if text/xml content-type is received over HTTP without a content-type, |
||
| 869 | * we SHOULD assume it is strictly US-ASCII. But we try to be more tolerant of non-conforming (legacy?) clients/servers, |
||
| 870 | * which will be most probably using UTF-8 anyway... |
||
| 871 | * In order of importance checks: |
||
| 872 | * 1. http headers |
||
| 873 | * 2. BOM |
||
| 874 | * 3. XML declaration |
||
| 875 | * 4. guesses using mb_detect_encoding() |
||
| 876 | * |
||
| 877 | * @param string $httpHeader the http Content-type header |
||
| 878 | * @param string $xmlChunk xml content buffer |
||
| 879 | * @param string $encodingPrefs comma separated list of character encodings to be used as default (when mb extension is enabled). |
||
| 880 | * This can also be set globally using PhpXmlRpc::$xmlrpc_detectencodings |
||
| 881 | * @return string the encoding determined. Null if it can't be determined and mbstring is enabled, |
||
| 882 | * PhpXmlRpc::$xmlrpc_defencoding if it can't be determined and mbstring is not enabled |
||
| 883 | * |
||
| 884 | * @todo as of 2023, the relevant RFC for XML Media Types is now 7303, and for HTTP it is 9110. Check if the order of |
||
| 885 | * precedence implemented here is still correct |
||
| 886 | * @todo explore usage of mb_http_input(): does it detect http headers + post data? if so, use it instead of hand-detection!!! |
||
| 887 | * @todo feature-creep make it possible to pass in options overriding usage of PhpXmlRpc static variables, to make |
||
| 888 | * the method independent of global state |
||
| 889 | */ |
||
| 890 | public static function guessEncoding($httpHeader = '', $xmlChunk = '', $encodingPrefs = null) |
||
| 891 | { |
||
| 892 | // discussion: see http://www.yale.edu/pclt/encoding/ |
||
| 893 | // 1 - test if encoding is specified in HTTP HEADERS |
||
| 894 | |||
| 895 | // Details: |
||
| 896 | // LWS: (\13\10)?( |\t)+ |
||
| 897 | // token: (any char but excluded stuff)+ |
||
| 898 | // quoted string: " (any char but double quotes and control chars)* " |
||
| 899 | // header: Content-type = ...; charset=value(; ...)* |
||
| 900 | // where value is of type token, no LWS allowed between 'charset' and value |
||
| 901 | // Note: we do not check for invalid chars in VALUE: |
||
| 902 | // this had better be done using pure ereg as below |
||
| 903 | // Note 2: we might be removing whitespace/tabs that ought to be left in if |
||
| 904 | // the received charset is a quoted string. But nobody uses such charset names... |
||
| 905 | |||
| 906 | /// @todo this test will pass if ANY header has charset specification, not only Content-Type. Fix it? |
||
| 907 | $matches = array(); |
||
| 908 | if (preg_match('/;\s*charset\s*=([^;]+)/i', $httpHeader, $matches)) { |
||
| 909 | return strtoupper(trim($matches[1], " \t\"")); |
||
| 910 | } |
||
| 911 | |||
| 912 | // 2 - scan the first bytes of the data for a UTF-16 (or other) BOM pattern |
||
| 913 | // (source: http://www.w3.org/TR/2000/REC-xml-20001006) |
||
| 914 | // NOTE: actually, according to the spec, even if we find the BOM and determine |
||
| 915 | // an encoding, we should check if there is an encoding specified |
||
| 916 | // in the xml declaration, and verify if they match. |
||
| 917 | /// @todo implement check as described above? |
||
| 918 | /// @todo implement check for first bytes of string even without a BOM? (It sure looks harder than for cases WITH a BOM) |
||
| 919 | if (preg_match('/^(?:\x00\x00\xFE\xFF|\xFF\xFE\x00\x00|\x00\x00\xFF\xFE|\xFE\xFF\x00\x00)/', $xmlChunk)) { |
||
| 920 | return 'UCS-4'; |
||
| 921 | } elseif (preg_match('/^(?:\xFE\xFF|\xFF\xFE)/', $xmlChunk)) { |
||
| 922 | return 'UTF-16'; |
||
| 923 | } elseif (preg_match('/^(?:\xEF\xBB\xBF)/', $xmlChunk)) { |
||
| 924 | return 'UTF-8'; |
||
| 925 | } |
||
| 926 | |||
| 927 | // 3 - test if encoding is specified in the xml declaration |
||
| 928 | /// @todo this regexp will fail if $xmlChunk uses UTF-32/UCS-4, and most likely UTF-16/UCS-2 as well. In that |
||
| 929 | /// case we leave the guesswork up to mbstring - which seems to be able to detect it, starting with php 5.6. |
||
| 930 | /// For lower versions, we could attempt usage of mb_ereg... |
||
| 931 | // Details: |
||
| 932 | // SPACE: (#x20 | #x9 | #xD | #xA)+ === [ \x9\xD\xA]+ |
||
| 933 | // EQ: SPACE?=SPACE? === [ \x9\xD\xA]*=[ \x9\xD\xA]* |
||
| 934 | // We could be stricter on version number: VersionNum ::= '1.' [0-9]+ |
||
| 935 | if (preg_match('/^<\?xml\s+version\s*=\s*' . "((?:\"[a-zA-Z0-9_.:-]+\")|(?:'[a-zA-Z0-9_.:-]+'))" . |
||
| 936 | '\s+encoding\s*=\s*' . "((?:\"[A-Za-z][A-Za-z0-9._-]*\")|(?:'[A-Za-z][A-Za-z0-9._-]*'))/", |
||
| 937 | $xmlChunk, $matches)) { |
||
| 938 | return strtoupper(substr($matches[2], 1, -1)); |
||
| 939 | } |
||
| 940 | |||
| 941 | // 4 - if mbstring is available, let it do the guesswork |
||
| 942 | if (function_exists('mb_detect_encoding')) { |
||
| 943 | if ($encodingPrefs == null && PhpXmlRpc::$xmlrpc_detectencodings != null) { |
||
| 944 | $encodingPrefs = PhpXmlRpc::$xmlrpc_detectencodings; |
||
| 945 | } |
||
| 946 | if ($encodingPrefs) { |
||
| 947 | $enc = mb_detect_encoding($xmlChunk, $encodingPrefs); |
||
| 948 | } else { |
||
| 949 | $enc = mb_detect_encoding($xmlChunk); |
||
| 950 | } |
||
| 951 | // NB: mb_detect likes to call it ascii, xml parser likes to call it US_ASCII... |
||
| 952 | // IANA also likes better US-ASCII, so go with it |
||
| 953 | if ($enc == 'ASCII') { |
||
| 954 | $enc = 'US-' . $enc; |
||
| 955 | } |
||
| 956 | |||
| 957 | return $enc; |
||
| 958 | } else { |
||
| 959 | // No encoding specified: assume it is iso-8859-1, as per HTTP1.1? |
||
| 960 | // Both RFC 2616 (HTTP 1.1) and RFC 1945 (HTTP 1.0) clearly state that for text/xxx content types |
||
| 961 | // this should be the standard. And we should be getting text/xml as request and response. |
||
| 962 | // BUT we have to be backward compatible with the lib, which always used UTF-8 as default. Moreover, |
||
| 963 | // RFC 7231, which obsoletes the two RFC mentioned above, has changed the rules. It says: |
||
| 964 | // "The default charset of ISO-8859-1 for text media types has been removed; the default is now whatever |
||
| 965 | // the media type definition says." |
||
| 966 | return PhpXmlRpc::$xmlrpc_defencoding; |
||
| 967 | } |
||
| 968 | } |
||
| 969 | |||
| 970 | /** |
||
| 971 | * Helper function: checks if an xml chunk has a charset declaration (BOM or in the xml declaration). |
||
| 972 | * |
||
| 973 | * @param string $xmlChunk |
||
| 974 | * @return bool |
||
| 975 | * |
||
| 976 | * @todo rename to hasEncodingDeclaration |
||
| 977 | */ |
||
| 978 | public static function hasEncoding($xmlChunk) |
||
| 979 | { |
||
| 980 | // scan the first bytes of the data for a UTF-16 (or other) BOM pattern |
||
| 981 | // (source: http://www.w3.org/TR/2000/REC-xml-20001006) |
||
| 982 | if (preg_match('/^(?:\x00\x00\xFE\xFF|\xFF\xFE\x00\x00|\x00\x00\xFF\xFE|\xFE\xFF\x00\x00)/', $xmlChunk)) { |
||
| 983 | return true; |
||
| 984 | } elseif (preg_match('/^(?:\xFE\xFF|\xFF\xFE)/', $xmlChunk)) { |
||
| 985 | return true; |
||
| 986 | } elseif (preg_match('/^(?:\xEF\xBB\xBF)/', $xmlChunk)) { |
||
| 987 | return true; |
||
| 988 | } |
||
| 989 | |||
| 990 | // test if encoding is specified in the xml declaration |
||
| 991 | // Details: |
||
| 992 | // SPACE: (#x20 | #x9 | #xD | #xA)+ === [ \x9\xD\xA]+ |
||
| 993 | // EQ: SPACE?=SPACE? === [ \x9\xD\xA]*=[ \x9\xD\xA]* |
||
| 994 | // We could be stricter on version number: VersionNum ::= '1.' [0-9]+ |
||
| 995 | if (preg_match('/^<\?xml\s+version\s*=\s*' . "((?:\"[a-zA-Z0-9_.:-]+\")|(?:'[a-zA-Z0-9_.:-]+'))" . |
||
| 996 | '\s+encoding\s*=\s*' . "((?:\"[A-Za-z][A-Za-z0-9._-]*\")|(?:'[A-Za-z][A-Za-z0-9._-]*'))/", |
||
| 997 | $xmlChunk)) { |
||
| 998 | return true; |
||
| 999 | } |
||
| 1000 | |||
| 1001 | return false; |
||
| 1002 | } |
||
| 1003 | |||
| 1004 | /** |
||
| 1005 | * @param string $message |
||
| 1006 | * @param string $method method/file/line info |
||
| 1007 | * @return bool false if the caller has to stop parsing |
||
| 1008 | */ |
||
| 1009 | protected function handleParsingError($message, $method = '') |
||
| 1010 | { |
||
| 1011 | if ($this->current_parsing_options['xmlrpc_reject_invalid_values']) { |
||
| 1012 | $this->_xh['isf'] = 2; |
||
| 1013 | $this->_xh['isf_reason'] = ucfirst($message); |
||
| 1014 | return false; |
||
| 1015 | } else { |
||
| 1016 | $this->getLogger()->error('XML-RPC: ' . ($method != '' ? $method . ': ' : '') . $message); |
||
| 1017 | return true; |
||
| 1018 | } |
||
| 1019 | } |
||
| 1020 | |||
| 1021 | /** |
||
| 1022 | * Truncates unsafe data |
||
| 1023 | * @param string $data |
||
| 1024 | * @return string |
||
| 1025 | */ |
||
| 1026 | protected function truncateValueForLog($data) |
||
| 1027 | { |
||
| 1028 | if (strlen($data) > $this->maxLogValueLength) { |
||
| 1029 | return substr($data, 0, $this->maxLogValueLength - 3) . '...'; |
||
| 1030 | } |
||
| 1031 | |||
| 1032 | return $data; |
||
| 1033 | } |
||
| 1034 | |||
| 1035 | // *** BC layer *** |
||
| 1036 | |||
| 1037 | /** |
||
| 1038 | * xml parser handler function for opening element tags. |
||
| 1039 | * Used in decoding xml chunks that might represent single xml-rpc values as well as requests, responses. |
||
| 1040 | * @deprecated |
||
| 1041 | * |
||
| 1042 | * @param resource $parser |
||
| 1043 | * @param $name |
||
| 1044 | * @param $attrs |
||
| 1045 | * @return void |
||
| 1046 | */ |
||
| 1047 | public function xmlrpc_se_any($parser, $name, $attrs) |
||
| 1048 | { |
||
| 1049 | // this will be spamming the log if this method is in use... |
||
| 1050 | $this->logDeprecation('Method ' . __METHOD__ . ' is deprecated'); |
||
| 1051 | |||
| 1052 | $this->xmlrpc_se($parser, $name, $attrs, true); |
||
| 1053 | } |
||
| 1054 | |||
| 1055 | public function &__get($name) |
||
| 1056 | { |
||
| 1057 | switch ($name) { |
||
| 1058 | case '_xh': |
||
| 1059 | case 'xmlrpc_valid_parents': |
||
| 1060 | $this->logDeprecation('Getting property XMLParser::' . $name . ' is deprecated'); |
||
| 1061 | return $this->$name; |
||
| 1062 | default: |
||
| 1063 | /// @todo throw instead? There are very few other places where the lib trigger errors which can potentially reach stdout... |
||
| 1064 | $trace = debug_backtrace(DEBUG_BACKTRACE_IGNORE_ARGS, 1); |
||
| 1065 | trigger_error('Undefined property via __get(): ' . $name . ' in ' . $trace[0]['file'] . ' on line ' . $trace[0]['line'], E_USER_WARNING); |
||
| 1066 | $result = null; |
||
| 1067 | return $result; |
||
| 1068 | } |
||
| 1069 | } |
||
| 1070 | |||
| 1071 | public function __set($name, $value) |
||
| 1088 | } |
||
| 1089 | } |
||
| 1090 | |||
| 1091 | public function __isset($name) |
||
| 1092 | { |
||
| 1093 | switch ($name) { |
||
| 1094 | case 'accept': |
||
| 1095 | $this->logDeprecation('Checking property XMLParser::' . $name . ' is deprecated'); |
||
| 1096 | return isset($this->current_parsing_options['accept']); |
||
| 1097 | case '_xh': |
||
| 1098 | case 'xmlrpc_valid_parents': |
||
| 1099 | $this->logDeprecation('Checking property XMLParser::' . $name . ' is deprecated'); |
||
| 1100 | return isset($this->$name); |
||
| 1101 | default: |
||
| 1102 | return false; |
||
| 1103 | } |
||
| 1104 | } |
||
| 1105 | |||
| 1106 | public function __unset($name) |
||
| 1123 | } |
||
| 1124 | } |
||
| 1125 | } |
||
| 1126 |
Loading history...