The secure CMS.
Forum Menu
Resources
Topics
- Home
- Development (314)
- Geeklog (209)
- Plugins (108)
- Themes (5)
- News (387)
- Announcements (248)
- Geeklog.net (31)
- Security (76)
- Spam (11)
- Summer of Code (28)
User Functions
Events
There are no upcoming events
What's New
Stories
No new storiesComments last 2 weeks
Trackbacks last 2 weeks
No new trackback commentsLinks last 2 weeks
No recent new linksNEW FILES last 14 days
No new filesWelcome to Geeklog Wednesday, June 19 2013 @ 10:07 AM EDT
|
||||||||
 |
Forum Index > Support > General Geeklog Support |
New Topic
|
Post Reply
|
 error when searching specialchars |
|||
| gape |
|
||||||
 Full Member  Status: offline  Registered: 05/30/02 Posts: 128 |
special - like our native slovenian characters
? and š especially (? stands for c with ? ) edit; now i noticed ... when i change global locale settings to sl_SI (for slovenia) char ž produces the same error (all 3 have problem cšž) the error ocurrs only when i change language to slovenian (in css is defined windows-1250) and it reads: An error has occurred: (This text is only displayed to users in the group 'Root'  2 - preg_replace() [function.preg-replace]: Compilation failed: invalid UTF-8 string at offset 11 @ /snip/httpdocs/lib-common.php line 5889 PHP Formatted Code array(7) {["text"]=> string(24) "Geeklog Project Homepage" ["query"]=> string(1) "?" ["class"]=> string(1) "b" ["mywords"]=> array(1) { [0]=> string(1) "?" } ["searchword"]=> string(1) "?" ["before"]=> string(11) "/(?<!\p{L})" ["after"]=> string(11) "(?!\p{L})/u" } why is there a PHP Formatted Code $_CONF['default_charset'] = 'iso-8859-1';in siteconfig.php ? do i need to change that? |
||||||
|
|||||||
| gape |
|
||||||
|
Full Member Status: offline Registered: 05/30/02 Posts: 128 |
|||||||
|
|||||||
| gape |
|
||||||
|
Full Member Status: offline Registered: 05/30/02 Posts: 128 |
what can i do?
pls |
||||||
|
|||||||
| Laugh |
|
||||||
 Admin  Status: offline  Registered: 09/27/05 Posts: 891 |
I would report it as a bug in the Geeklog Bug Tracker http://project.geeklog.net/. Geeklog 1.8.0 should be released in April.
|
||||||
|
|||||||
| gape |
|
||||||
|
Full Member Status: offline Registered: 05/30/02 Posts: 128 |
did that allready
|
||||||
|
|||||||
| gape |
|
||||||
|
Full Member Status: offline Registered: 05/30/02 Posts: 128 |
since i could not get a fix for that error
i comented out allmost all of the function that makes me trouble so now it reads PHP Formatted Code function COM_highlightQuery( $text, $query, $class = 'highlight' ){ /* // escape PCRE special characters $query = preg_quote($query, '/'); $mywords = explode(' ', $query); foreach ($mywords as $searchword) { if (!empty($searchword)) { $before = "/(?!(?:[^<]+>|[^>]+<\/a>))\b"; $after = "\b/i"; if ($searchword <> utf8_encode($searchword)) { if (@preg_match('/^\pL$/u', urldecode('%C3%B1'))) { // Unicode property support $before = "/(?<!\p{L})"; $after = "(?!\p{L})/u"; } else { $before = "/"; $after = "/u"; } } $text = preg_replace($before . $searchword . $after, "<span class=\"$class\">\\0</span>", '<!-- x -->' . $text . '<!-- x -->' ); } } */ return $text; } so now it just recives text and sends it back i hope if someone understands where the problem is - i will be glad i have few requests for upgrades - but i do not dare anyone sees any problem with what i comented out? will it give me trouble elsewhere? |
||||||
|
|||||||
| Laugh |
|
||||||
|
Admin Status: offline Registered: 09/27/05 Posts: 891 |
Commenting out that code shouldn't give you any problems, it just will not return the requested search query highlighed in the text.
Sorry we couldn't look into the issue sooner. Someone should before 1.8.0 is released. Tom |
||||||
|
|||||||
| gape |
|
||||||
|
Full Member Status: offline Registered: 05/30/02 Posts: 128 |
pls do
and when u do - do contact me i tried to document the issue as well as i could and i think it goes kinda deep all the way to missrepresenting utf characters - if that is at all possible |
||||||
|
|||||||
| Roccivic |
|
||||||
 Moderator  Status: offline Registered: 05/19/10 Posts: 136 |
Quote by: gape why is there a PHP Formatted Code $_CONF['default_charset'] = 'iso-8859-1';in siteconfig.php ? do i need to change that? Yes, you need to change that. You can do that by wiping out your geeklog installation and then when reinstalling, at step 2 tick the checkbox near "Use UTF-8". I just tested this and it fixed the problem Haven't traced the actual problem though, but only the non-UTF language files are affected... Rouslan |
||||||
|
|||||||
| gape |
|
||||||
|
Full Member Status: offline Registered: 05/30/02 Posts: 128 |
this geeklog was reinstalled once before
i think that was enough ... the q was - why is there (in siteconfig) this line i dont remember putting it there furthermore i would never set such a charset ... i removed it and it did not help ... use utf8 is for multilanguage sites afaik i could use that, but i prefer smaller charsets for one-language sites (in my head thats still faster) furthermore - this is not solution for sites that use something else but utf8 and thats majority of my sites |
||||||
|
|||||||
| Roccivic |
|
||||||
|
Moderator Status: offline Registered: 05/19/10 Posts: 136 |
Well, if you don't want to reinstall, I think that you are stuck with ISO-8859-1. We don't have charset migration for the database, yet...
Meanwhile try the below code, it works just fine for me for both UTF-8 and ISO-8859-1 Rouslan PHP Formatted Code /*** Highlight the words from a search query in a given text string. * * @param string $text the text * @param string $query the search query * @param string $class html class to use to highlight * @return string the text with highlighted search words * */ function COM_highlightQuery($text, $query, $class = 'highlight') { if (!empty($text) && !empty($query)) { $flag = false; if (!mb_check_encoding($text, 'UTF-8')) { // convert strings to UTF if needed $text = utf8_encode($text); $query = utf8_encode($query); $flag = true; } // escape PCRE special characters $query = preg_quote($query, '/'); $mywords = explode(' ', $query); foreach ($mywords as $searchword) { if (!empty($searchword)) { if (@preg_match('/^\pL$/u', urldecode('%C3%B1'))) { // Unicode property support $before = "/(?<!\p{L})"; $after = "(?!\p{L})/u"; } else { $before = "/"; $after = "/u"; } $text = preg_replace($before . $searchword . $after, "<span class=\"$class\">\\0</span>", '<!-- x -->' . $text . '<!-- x -->'); } } if ($flag) { // if we need to convert a string back to ISO-8859-1 $text = utf8_decode($text); } } return $text; } |
||||||
|
|||||||
| ::Ben |
|
||||||
|
Full Member  Status: offline Registered: 01/14/05 Posts: 1369 |
We speak french on http://geeklog.fr |
||||||
|
|||||||
| gape |
|
||||||
|
Full Member Status: offline Registered: 05/30/02 Posts: 128 |
Quote by: Roccivic Well, if you don't want to reinstall, I think that you are stuck with ISO-8859-1. We don't have charset migration for the database, yet... PHP Formatted Code /*** Highlight the words from a search query in a given text string. * * @param string $text the text * @param string $query the search query * @param string $class html class to use to highlight * @return string the text with highlighted search words * */ function COM_highlightQuery($text, $query, $class = 'highlight') { if (!empty($text) && !empty($query)) { $flag = false; if (!mb_check_encoding($text, 'UTF-8')) { // convert strings to UTF if needed $text = utf8_encode($text); $query = utf8_encode($query); $flag = true; } // escape PCRE special characters $query = preg_quote($query, '/'); $mywords = explode(' ', $query); foreach ($mywords as $searchword) { if (!empty($searchword)) { if (@preg_match('/^pL$/u', urldecode('%C3%B1'))) { // Unicode property support $before = "/(?<!p{L})"; $after = "(?!p{L})/u"; } else { $before = "/"; $after = "/u"; } $text = preg_replace($before . $searchword . $after, "<span class="$class">\0</span>", '<!-- x -->' . $text . '<!-- x -->'); } } if ($flag) { // if we need to convert a string back to ISO-8859-1 $text = utf8_decode($text); } } return $text; } sry but it produces a PHP Formatted Code 2 - preg_replace() [function.preg-replace]: Compilation failed: invalid UTF-8 string at offset 11 @ /snip/httpdocs/lib-common.php line 5895line reads: PHP Formatted Code '<!-- x -->' . $text . '<!-- x -->'); |
||||||
|
|||||||
| Roccivic |
|
||||||
|
Moderator Status: offline Registered: 05/19/10 Posts: 136 |
@gape: The code that you posted is not an exact copy of the code that I posted. Maybe that could be a problem...
I checked again, it still works for me. Anyway, that's probably not the best solution because it would replace stuff inside links... Did you know that you can comment out just a part of that function, like below, so that it will only highlight ASCII search words? Maybe it could be the best temporary solution, while we fix it properly in a later release... Rouslan PHP Formatted Code function COM_highlightQuery($text, $query, $class = 'highlight'){ if (!empty($text) && !empty($query)) { // escape PCRE special characters $query = preg_quote($query, '/'); $mywords = explode(' ', $query); foreach ($mywords as $searchword) { if (!empty($searchword)) { $before = "/(?!(?:[^<]+>|[^>]+<\/a>))\b"; $after = "\b/i"; /*if ($searchword <> utf8_encode($searchword)) { if (@preg_match('/^\pL$/u', urldecode('%C3%B1'))) { // Unicode property support $before = "/(?<!\p{L})"; $after = "(?!\p{L})/u"; } else { $before = "/"; $after = "/u"; } }*/ $text = preg_replace($before . $searchword . $after, "<span class=\"$class\">\\0</span>", '<!-- x -->' . $text . '<!-- x -->'); } } } return $text; } |
||||||
|
|||||||
| gape |
|
||||||
|
Full Member Status: offline Registered: 05/30/02 Posts: 128 |
Quote by: Roccivic @gape: The code that you posted is not an exact copy of the code that I posted. Maybe that could be a problem... im not sure if i understand this im using (1.7.1sr1) on that particular site (this) code was changed in 172? |
||||||
|
|||||||
| Roccivic |
|
||||||
|
Moderator Status: offline Registered: 05/19/10 Posts: 136 |
Quote by: gape Quote by: Roccivic @gape: The code that you posted is not an exact copy of the code that I posted. Maybe that could be a problem... im not sure if i understand this im using (1.7.1sr1) on that particular site (this) code was changed in 172? I just meant that some quotes are unescaped in your copy, that would cause a syntax error, look near the line: "<span class=\"$class\">\\0</span>", Just a bad copy/paste perhaps... |
||||||
|
|||||||
| gape |
|
||||||
|
Full Member Status: offline Registered: 05/30/02 Posts: 128 |
i c what u mean now
when i quoted ur code one \ was lost i didnt remove it - maybe the forum in my first code there are 2 // anyways your last code works well for me for now tyvm |
||||||
|
|||||||
02/13/11 08:55AM (Read 1248 times)
Quote| Content generated in: 2.47 seconds |
|
|
|
