//this function is a state machine that processes the tweet search json one byte at a time. It works ont he global searchParser structure
//it looks for a tag, and then grabs the tag value into a buffer.
//does OK with escape, removes XMLentities from the text completely, this could be improved.
unsigned char tagSearch(unsigned char inBuf, struct _jsonParser *P ){
#define SEARCHING_FOR_TAG 0
#define GRABBING_TAG 1
switch(P->foundTag){
case SEARCHING_FOR_TAG://search for tag
//compare current byte to next byte in tag
//if no match, need to check one more time to see if the beginning of a new sequence
if( (P->tagCharMatchCnt==0) || (inBuf!=P->searchTag[P->tagCharMatchCnt]) ){
P->tagCharMatchCnt=0; //no match, reset
if(inBuf!=P->searchTag[0]) break; //no match, break
}
P->tagCharMatchCnt++; //match, increment counter
if(P->tagCharMatchCnt==P->searchTagLength){//whole tag matches
P->tagCharMatchCnt=0; //reset the match counter
P->foundTag=GRABBING_TAG;
P->xmlencode=0;
P->escape=0;
P->UTF=0;
P->valueBufferCounter=0;
}
break;
case GRABBING_TAG://tag found, get value
//discard UTF characters, could decode and do a few umlouts and other international characters the LCD can display
if(P->UTF>0){//ignore UTF8 characters (LCD can't display, PIC char can't handle)
P->UTF--;
break;
}
if(P->xmlencode==1){ //already saw &, suck down the code
if(inBuf==';'){//; is end of code, process and clean up
P->xmlencode=0; //done grabbing encoded character
inBuf=XMLdecode(P->xmlenchar[0], P->xmlenchar[1]); //decode it, reuse inBuf character
if(inBuf!=0){//if 0, error, just skip it, else add to valueBuffer
addValueByte(inBuf, P);//add character to the value buffer
}
break;
}else{ //if not end of xmlencode, get another character
if(P->xmlencharcnt<2){ //we only need the first 2 character to identify the ecoding
P->xmlenchar[P->xmlencharcnt]=inBuf;
}
P->xmlencharcnt++;
if(P->xmlencharcnt>5){//quit if it goes over the max of 5 characters
P->xmlencode=0; //quit encoding, there was an error, resume normal on next character
}
break; //don't save byte
}
}
if(P->escape==1){ //ESCAPED, check for /\" and UTF8
P->escape=0; //clear escape
if(inBuf=='u' || inBuf=='U'){//only escape /\", add a \ back to buffer otherwise
P->UTF=4; //next 4 bytes are UFT encoded characters, ignore
break;
}else if(inBuf=='n' || inBuf=='N'){ //\n is line break, ignore....
break;
}
}else{ //NOT ESCAPED, check for end and escape
if(inBuf==P->valueEndChar){ //end on value end character
P->foundTag=SEARCHING_FOR_TAG;
return 1; //tag complete
//break;
}else if(inBuf==0x5c){ //if this is escape char and escape not set, then set escape and break (0x5c= '\')
P->escape=1;
break;
}else if(inBuf=='&'){ //begin XML encoded character
P->xmlencode=1; //set flag
P->xmlencharcnt=0; //clear encoded character counter
break; //don't save this character
}
}
if(inBuf>=0x20 && inBuf<=0x7e){//only allow standard ASCII characters, otherwise ignore
addValueByte(inBuf, P);//add character to the value buffer
}
break;
}
return 0; //tag not complete
//this function is a state machine that processes the tweet search json one byte at a time. It works ont he global searchParser structure
//it looks for a tag, and then grabs the tag value into a buffer.
//does OK with escape, removes XMLentities from the text completely, this could be improved.
void procSearch(unsigned char inBuf){
switch(searchParser.foundTag){
case 0://search for tag
//compare current byte to next byte in tag
//if no match, need to check one more time to see if the beginning of a new sequence
if( (searchParser.tagCharMatchCnt==0) || (inBuf!=jsonSearchTag[searchParser.tagCharMatchCnt]) ){
searchParser.tagCharMatchCnt=0; //no match, reset
#if defined(__18CXX)//there's something about the C18 compiler that wants this or the next instruction is always false..
Nop();
#endif
if(inBuf!=jsonSearchTag[0]) break; //no match, break
}
searchParser.tagCharMatchCnt++; //match, increment counter
if(searchParser.tagCharMatchCnt==(sizeof(jsonSearchTag)-1)){//whole tag matches
searchParser.tagCharMatchCnt=0; //reset the match counter
searchParser.tagTotalCnt++; //increment the total tag counter
searchParser.foundTag=1;
searchParser.xmlencode=0;
searchParser.escape=0;
searchParser.UTF=0;
addToSearchBuffer(' ');
}
break;
case 1://tag found, get value
//discard UTF characters, could decode and do a few umlouts and other international characters the LCD can display
if(searchParser.UTF>0){//ignore UTF8 characters (LCD can't display, PIC char can't handle)
searchParser.UTF--;
break;
}
if(searchParser.xmlencode==1){ //already saw &, suck down the code
if(inBuf==';'){//; is end of code, process and clean up
searchParser.xmlencode=0; //done grabbing encoded character
inBuf=XMLdecode(searchParser.xmlenchar[0], searchParser.xmlenchar[1]); //decode it, reuse inBuf character
if(inBuf!=0) addToSearchBuffer(inBuf); //if 0, error, just skip it, else add to searchBuffer
break;
}else{ //if not end of xmlencode, get another character
if(searchParser.xmlencharcnt<2){ //we only need the first 2 character to identify the ecoding
searchParser.xmlenchar[searchParser.xmlencharcnt]=inBuf;
}
searchParser.xmlencharcnt++;
if(searchParser.xmlencharcnt>5){//quit if it goes over the max of 5 characters
searchParser.xmlencode=0; //quit encoding, there was an error, resume normal on next character
}
break; //don't save byte
}
}
if(searchParser.escape==1){ //ESCAPED, check for /\" and UTF8
searchParser.escape=0; //clear escape
if(inBuf=='u' || inBuf=='U'){//only escape /\", add a \ back to buffer otherwise
searchParser.UTF=4; //next 4 bytes are UFT encoded characters, ignore
break;
}else if(inBuf=='n' || inBuf=='N'){ //\n is line break, ignore....
break;
}
//don't need this, escape character is always escaped
//else if(!(inBuf=='/' || inBuf==0x5c || inBuf=='"')){//only escape /\", add a \ back to buffer otherwise
// addToSearchBuffer('|'); //continue as normal, add current byte to searchbuffer at the end...
//}
}else{ //NOT ESCAPED, check for end and escape
if(inBuf=='"'){ //end on " if not escaped
//add block between entries
addToSearchBuffer(' ');
addToSearchBuffer(0xff);
searchParser.foundTag=0;
break;
}else if(inBuf==0x5c){ //if this is escape char and escape not set, then set escape and break (0x5c= '\')
searchParser.escape=1;
break;
}else if(inBuf=='&'){ //begin XML encoded character
searchParser.xmlencode=1; //set flag
searchParser.xmlencharcnt=0; //clear encoded character counter
break; //don't save this character
}
}
if(inBuf>=0x20 && inBuf<=0x7e){//only allow standard ASCII characters, otherwise ignore
if(inBuf==0x5c) inBuf='|'; //LCD has no '\', use |
if(inBuf=='~') inBuf='-'; //LCD has no "~", use -
addToSearchBuffer(inBuf); //add character to output buffer
}
break;
}
}
