httpfs ChangeLog extract.c

[Samuel Thibault]

CVSROOT:        /cvsroot/hurdextras
Module name:    httpfs
Changes by:     Samuel Thibault <sthibaul>      13/01/27 23:16:14

Modified files:
        .              : ChangeLog extract.c 

Log message:
                * extract.c (extract): Use strdupa for temporary strings. Ignore
                absolute URLs. Drop internal refs and form data instead of 
ignoring
                the whole link. Stop trying to resolve directory names.

CVSWeb URLs:
http://cvs.savannah.gnu.org/viewcvs/httpfs/ChangeLog?cvsroot=hurdextras&r1=1.9&r2=1.10
http://cvs.savannah.gnu.org/viewcvs/httpfs/extract.c?cvsroot=hurdextras&r1=1.1.1.1&r2=1.2

Patches:
Index: ChangeLog
===================================================================
RCS file: /cvsroot/hurdextras/httpfs/ChangeLog,v
retrieving revision 1.9
retrieving revision 1.10
diff -u -b -r1.9 -r1.10
--- ChangeLog   27 Jan 2013 23:14:22 -0000      1.9
+++ ChangeLog   27 Jan 2013 23:16:14 -0000      1.10
@@ -1,6 +1,9 @@
 2013-01-28  Samuel Thibault  <address@hidden>
 
        * parsehtml.c (startElement): Use strcasecmp to match "href".
+       * extract.c (extract): Use strdupa for temporary strings. Ignore
+       absolute URLs. Drop internal refs and form data instead of ignoring
+       the whole link. Stop trying to resolve directory names.
 
 2013-01-25  Samuel Thibault  <address@hidden>
 

Index: extract.c
===================================================================
RCS file: /cvsroot/hurdextras/httpfs/extract.c,v
retrieving revision 1.1.1.1
retrieving revision 1.2
diff -u -b -r1.1.1.1 -r1.2
--- extract.c   15 Apr 2002 19:25:41 -0000      1.1.1.1
+++ extract.c   27 Jan 2013 23:16:14 -0000      1.2
@@ -58,12 +58,12 @@
        
 }
 
-void extract(char *string,char *parent)
+void extract(char *_string,char *parent)
 {
        /* This function extracts file/directory info from string and fills
         * into the intermediate data structure *files */
 
-       int check_url=0,unfilled_dir=0;
+       int unfilled_dir=0;
        int i=0;
        const char delimiters[]="/";
        char *temp;
@@ -71,26 +71,42 @@
        char *token;
        char *name;
        int no_of_slashes_here=0;
+       char *string = strdupa(_string);
        
        /* string can be a directory, file, url or path to a file */
-       if ( ! (strcmp((const char*)strndup(string,7),"http://";)) ) 
+       if ( ! (strncasecmp(string,"http://";, 7)) )
        {
+               /* Ignore absolute URLs */
+#if 0
                /* URL */
                string = string + 7;
                if ( check_dups(string,parent) )
                        fill_files(string,parent,0,HTTP_URL);
+#endif
                return;
        }
-       if( strchr(string,'#') != NULL || strchr(string,'?') != NULL || 
string[0] == '.' )
+
+       /* Drop internal link */
+       token = strchr(string, '#');
+       if (token)
+               *token = 0;
+       /* Drop GET form data */
+       token = strchr(string, '?');
+       if (token)
+               *token = 0;
+
+       if( string[0] == '.' && string[1] == '/')
        {
-               /* # internal link leave it */
-               return;
+               /* Skip leading ./// */
+               string++;
+               while (string[0] == '/')
+                       string++;
        }
 
        if(string[0]=='/')
        {
                /* it is a path starting from root directory */
-               temp = strdup(string);
+               temp = strdupa(string);
 
                while ( strchr(temp,'/') != NULL ) 
                {
@@ -120,7 +136,7 @@
                        return;
                }
 
-               token = strdup(string);
+               token = strdupa(string);
                for ( i=0 ; i<no_of_slashes-1 ; i++ ) 
                {
                        /* extract file and directory from the string
@@ -129,15 +145,14 @@
                         * extract temp fill it as a directory under file/
                         * extract a.html fill it as a file under temp/ */
                        
-                       temp = strdup(token); 
+                       temp = strdupa(token); 
                        if ( strcmp(dir_tok[i],strtok(temp,"/")) )
                                return;
                        strcpy(token,strchr(token,'/'));
                        token++;
                }
-               parent = strdup("tmp");
-               string = strdup(token);
-               check_url=1;
+               parent = strdupa("tmp");
+               string = strdupa(token);
        }
 
        if( string[strlen(string)-1]=='/' )
@@ -146,23 +161,12 @@
         * to be filled on demand make a note that its unfilled directory */
                unfilled_dir = 1;
        }
-       token=strdup(string);
+       token=strdupa(string);
        while(strchr(token,'/')!=NULL)
        {
                /* Handles Directories */
-               temp=strdup(token);
+               temp=strdupa(token);
                name=strtok(temp,delimiters);
-               if ( check_url == 0 ) {
-               if(((x=gethostbyname(name))!=NULL))
-               {
-                       /* Check if its an URL */
-                       unfilled_dir = 0;
-                       if ( check_dups(string,parent) ) 
-                               fill_files(name,parent,0,HTTP_URL);
-                       return;
-               }
-               check_url=1;
-               }
                
                strcpy(token,strchr(token,'/'));
                token=token+sizeof(char);
@@ -172,7 +176,7 @@
                        if ( unfilled_dir )
                                this_entry->f_type = HTTP_DIR_NOT_FILLED;
                }
-               parent=strdup(name);
+               parent=strdupa(name);
        }
 
        if(strlen(token)>0)
@@ -183,16 +187,6 @@
                 * Also Handle if there is no slash at all -- just file name 
                 * or it can be a url */
                name=strtok(token,delimiters);
-               if ( check_url == 0 ) 
-               {
-                       if(((x=gethostbyname(name))!=NULL)) 
-                       {
-                               /* It can be a URL check and return */
-                               if ( check_dups(string,parent) )
-                                       fill_files(name,parent,0,HTTP_URL);
-                               return;
-                       }
-               }
                if ( check_dups(name,parent) )
                        fill_files(name,parent,0,HTTP_FILE);
        }