diff options
| author | Andrew Dolgov <[email protected]> | 2007-08-13 04:31:57 +0100 |
|---|---|---|
| committer | Andrew Dolgov <[email protected]> | 2007-08-13 04:31:57 +0100 |
| commit | feb3185124bd5f1257c6bdfaeb2300845c923ff5 (patch) | |
| tree | 807d41110225fd53218f138fee3c9e53bf7fffc1 /magpierss | |
| parent | 3a486dec89f5270812f1e699bb8844b359de53d7 (diff) | |
magpie: only force convert encoding if initial parsing failed
Diffstat (limited to 'magpierss')
| -rw-r--r-- | magpierss/rss_fetch.inc | 2 | ||||
| -rw-r--r-- | magpierss/rss_parse.inc | 26 |
2 files changed, 27 insertions, 1 deletions
diff --git a/magpierss/rss_fetch.inc b/magpierss/rss_fetch.inc index dd475e31d..77a1a704f 100644 --- a/magpierss/rss_fetch.inc +++ b/magpierss/rss_fetch.inc @@ -313,7 +313,7 @@ function _convert_entities ($string) { Output: parsed RSS object (see rss_parse) \*=======================================================================*/ function _response_to_rss ($resp) { - $converted_source = _convert_entities(mb_convert_encoding($resp->results, "UTF-8", mb_detect_encoding($resp->results))); + $converted_source = _convert_entities($resp->results); $rss = new MagpieRSS( $converted_source, MAGPIE_OUTPUT_ENCODING, "UTF-8", false); // if RSS parsed successfully diff --git a/magpierss/rss_parse.inc b/magpierss/rss_parse.inc index 0b2ddc83a..8a67523b5 100644 --- a/magpierss/rss_parse.inc +++ b/magpierss/rss_parse.inc @@ -132,7 +132,33 @@ class MagpieRSS { xml_set_character_data_handler( $this->parser, 'feed_cdata' ); $status = xml_parse( $this->parser, $source ); + + # try to force convert everything to UTF-8 and parse again + # to salvage at least some data from the feed + if (! $status) { + $errorcode = xml_get_error_code( $this->parser ); + if ( $errorcode != XML_ERROR_NONE ) { + + xml_parser_free( $this->parser ); + + list($parser, $source) = $this->create_parser($source, + $output_encoding, $input_encoding, $detect_encoding); + + $source = mb_convert_encoding($source, "UTF-8", + mb_detect_encoding($source)); + + $this->parser = $parser; + xml_set_object( $this->parser, $this ); + xml_set_element_handler($this->parser, + 'feed_start_element', 'feed_end_element' ); + + xml_set_character_data_handler( $this->parser, 'feed_cdata' ); + + $status = xml_parse( $this->parser, $source); + } + } + if (! $status ) { $errorcode = xml_get_error_code( $this->parser ); if ( $errorcode != XML_ERROR_NONE ) { |