getmetadataforfile.c

It is a chm file viewer lib with source code

#include <CoreFoundation/CoreFoundation.h>
#include <CoreServices/CoreServices.h> 
#include "chm_lib.h"
#include <stdio.h>
#include <stdlib.h>
#include <strings.h>

#include <unistd.h>
#include <sys/stat.h>
#include <sys/types.h>
/* -----------------------------------------------------------------------------
Step 1
Set the UTI types the importer supports

Modify the CFBundleDocumentTypes entry in Info.plist to contain
an array of Uniform Type Identifiers (UTI) for the LSItemContentTypes 
that your importer can handle

----------------------------------------------------------------------------- */

/* -----------------------------------------------------------------------------
Step 2 
Implement the GetMetadataForFile function

Implement the GetMetadataForFile function below to scrape the relevant
metadata from your document and return it as a CFDictionary using standard keys
(defined in MDItem.h) whenever possible.
----------------------------------------------------------------------------- */

/* -----------------------------------------------------------------------------
Step 3 (optional) 
If you have defined new attributes, update the schema.xml file

Edit the schema.xml file to include the metadata keys that your importer returns.
Add them to the <allattrs> and <displayattrs> elements.

Add any custom types that your importer requires to the <attributes> element

<attribute name="com_mycompany_metadatakey" type="CFString" multivalued="true"/>

----------------------------------------------------------------------------- */
CFMutableStringRef text;

int _extract_callback(struct chmFile *h,
					  struct chmUnitInfo *ui,
					  void *context)
{
	
    if (ui->length != 0)
    {
        LONGINT64 len, remain=ui->length;
        LONGUINT64 offset = 0;
		
		if(strlen(ui->path)!=0){
			if(!strstr(ui->path,".htm"))
				return CHM_ENUMERATOR_CONTINUE;
		}
		else
			return CHM_ENUMERATOR_CONTINUE;
			Boolean skip=FALSE, space=FALSE;
		int i;
		int j;
		unsigned char newbuffer[32769];
		while (remain != 0)
        {
			
			unsigned char buffer[32768];
			len = chm_retrieve_object(h, ui, buffer, offset, 32768);

			j=0;
            if (len > 0)
            {
				char *c;
				for(c=buffer;*c;c++)
				{
					switch(*c){
						case '<': skip=TRUE; break;
						case '>': skip=FALSE; break;
						case 0xd:  
						case 0xa: if(!space){newbuffer[j++]= ' ';space=TRUE;};break;
						case '&': if(!skip){
							i=1;
							while(*(c+i)){
								if(*(c+i)==' ')
									break;
								else if(*(c+i)==';'){
									c+=i;
									break;
								}
								i++;
							}
						}; break;	
						case ' ': if(!space && !skip){
							newbuffer[j++]= *c;
							space=TRUE;}
							break;
						default: if(!skip && ((*c>='A' && *c<='Z') || (*c>='a' && *c<='z')) ){
		
							 newbuffer[j++]= *c;
							 if(space)
							 space=FALSE;
						}; break;
					}
				}
				newbuffer[j++]='\0';
				CFStringAppendCString(text,newbuffer,kCFStringEncodingMacRoman);
				
                offset += len;
                remain -= len;
            }
			else
                break;
        }
		
    }	
    return CHM_ENUMERATOR_CONTINUE;
}


/* -----------------------------------------------------------------------------
Get metadata attributes from file

This function's job is to extract useful information your file format supports
and return it as a dictionary
----------------------------------------------------------------------------- */

Boolean GetMetadataForFile(void* thisInterface, 
						   CFMutableDictionaryRef attributes, 
						   CFStringRef contentTypeUTI,
						   CFStringRef pathToFile)
{
    /* Pull any available metadata from the file at the specified path */
    /* Return the attribute keys and attribute values in the dict */
    /* Return TRUE if successful, FALSE if there was no data provided */
	
	char buffer[CFStringGetLength(pathToFile)+1];
	CFStringGetCString(pathToFile,buffer,CFStringGetLength(pathToFile)+1,kCFStringEncodingMacRoman);
	
	struct chmFile *h;			
    h = chm_open(buffer);
    if (h == NULL)
		return FALSE; 
	
	text=CFStringCreateMutable(0,kCFStringEncodingMacRoman);
    if (! chm_enumerate(h,CHM_ENUMERATE_ALL,_extract_callback,NULL))
        return FALSE;
	
    chm_close(h);
	
	CFDictionaryAddValue(attributes, kMDItemTextContent, text);
	
	CFRelease(text);
    return TRUE;
}