globbing.c

	/*
				glob_t is a glob structure type defined in glob.h
				gl_pathV is a list of all matched pathnames
				gl_pathc is the total count of paths mathched
				if GLOB_NOCHECK is used the structure returns
				wildcard if no matches are found and
				path count is set to 1

				three state parser for seperating normal keywords from wildcards
				wildcards in this parser are assumed to start with '*' and '[' 
				and ending with a space
				Example *.c *.mp* or *
				or [a-z].c or [1-9].*

	*/


	char * glob_pattern(char * wildcard){

	  char * gfilename;
	  size_t cnt,length;
		glob_t glob_results;
	  char **p;
	
	  glob(wildcard,GLOB_NOCHECK,0,&glob_results);
	
	  for(p=glob_results.gl_pathv,cnt=glob_results.gl_pathc;cnt;p++,cnt--)
	  length += strlen(*p)+ 1;

	  gfilename=(char *) calloc(length,sizeof(char));
	  for(p=glob_results.gl_pathv,cnt=glob_results.gl_pathc;cnt;p++,cnt--)
	  {

	    strcat(gfilename,*p);
	    if(cnt > 1)
	    strcat(gfilename," ");
	  }

	  globfree(&glob_results);
	  return gfilename;

	}

	int glob_parser(char input[]){


	  char * cp;
		char c;
		char word[1024];
		int i=0,j=0;
	  char * results;
		char wild[20];

	
	  enum states { START, IN_WORD, IN_WILD } state = START;

					for (cp = input; *cp != '\0'; cp++) {
									c = (unsigned char) *cp;
								
								
							switch (state) {
									
									case START:

											if(c=='*' || c=='['){
													state=IN_WILD;
													wild[j]=c;
													j++;
											}
											else{

													state=IN_WORD;
													word[i]=c;
													i++;
											}

									continue;



									case IN_WORD:
												if(c =='*' || c=='['){
														state=IN_WILD;
														wild[j]=c;
														j++;
												}
												else{
														word[i]=c;
														i++;
												}
									continue;

									case IN_WILD:
												if(c==' '){
													state=START;
													word[i]=c;
												}
												else{
													wild[j]=c;
													j++;
												}

									}
					}


		word[i]='\0';
		wild[j]='\0';

		//removing the trailing newline from array  
	  strtok(wild,"\n");
	  results=glob_pattern(wild);
	  strcat(word," ");
	  strcat(word,results);

	  parse(word,1024);

	  return 0;
	}