PHP Code:
#include <amxmodx>
#pragma semicolon 1
new PLUGIN[] = "";
new AUTHOR[] = "";
new VERSION[] = "0.00";
public plugin_init()
{
register_plugin(PLUGIN, VERSION, AUTHOR);
register_message(get_user_msgid("SayText"), "msgSayText");
}
// Checks if a given string contains Eastern Languages' characters
fnEasternLanguageString(const szString[])
{
new i = 0;
new ch;
while((ch = szString[i]) != EOS)
{
// Information from Unicode wiki page, not sure if this is the actual character range
// Actually, I've never worked with non-ASCII characters in Pawn, not sure if it's even supported
if(0x590 <= ch <= 0x7BF)
return true;
i++;
}
return false;
}
public msgSayText(iId, iDest, pEnt)
{
static szText[192];
// I think the 4th arg is the actual message, can't remember
get_msg_arg_string(4, szText, charsmax(szText));
if(!fnEasternLanguageString(szText))
return PLUGIN_CONTINUE;
new ch; // Cache a character, array indexing is slower
new i = 0, j;
new bool: bSkipWord = false;
new iWordStart = -1;
new iMid;
new bool: isEOS; // Eliminate a need to check for EOS twice
new iTemp;
for(;; i++) // We'll loop until we reach end of string (condition for breaking down below)
{
if(isspace((ch = szText[i])) || (isEOS = (ch == EOS)))
{
bSkipWord = false;
// Reached the end of a word that should get reversed
if(iWordStart != -1)
{
i--; // So I don't have to do "i - 1" each time, restored to original value later
iMid = (i - iWordStart) / 2; // Get number of chars to middle of a word
// Reverse the word
for(j = 0; j < iMid; j++)
{
iTemp = szText[iWordStart];
szText[iWordStart + j] = szText[i - j];
szText[i - j] = iTemp;
}
iWordStart = -1;
i++;
}
// End of string, we're done
if(isEOS)
break;
continue;
}
else if(bSkipWord)
continue;
// Found start position of a word that should get reversed
if(iWordStart != -1)
{
// Should not get reversed if starts with a number (I guess)
if('0' <= ch <= '9')
{
bSkipWord = true;
continue;
}
iWordStart = i;
}
}
set_msg_arg_string(4, szText);
return PLUGIN_CONTINUE;
}
You gave me an example with only 1 word, so I can't guess if you want something like (as appears in a chat message)
to become
or become
The code above does the latter (reverses each word individually, not the whole string). It also checks if the message contains any Arabic characters before reversing it.
It probably could be written more efficient, but I don't have a lot of time to spend right now. I think it's good enough. I haven't tested it though, I don't have HLDS installed on Windows, and I am too lazy to boot to Linux right now. Compiles fine at least.
Linear Mode
