identify OLE-position in object

Mathias Mamsch - Tue Sep 14 13:44:07 EDT 2010

Hmm you are right, it IS really slow :-) Maybe you can optimize your replace code to be faster, by explicitly checking for "{\object" or "{\pict" phrases, then search their ending by finding an approriate "}" using two regular expressions:

The example code below does a loop over all characters of the buffer, thereby skipping OLE objects ... Hope this helps, regards, Mathias
 

pragma runLim, 10000
 
Regexp reObj = regexp "({\\\\object)|({\\\\pict)"
Regexp reBrace = regexp "([{}])"
 
Buffer b = create(); 
b = richTextWithOle (current Object)."Object Text"
 
// print tempStringOf b
 
// halt
 
int pos = 0; 
int iLen = length b
int oleStart = 0
int oleEnd   = 0 
 
Buffer result = create() 
int nextOle = -1
 
while (pos < iLen) {
   if (nextOle < pos) {
    // find the next ole object
        if ( search (reObj, b, pos) ) {
                oleStart = oleStart + start 0
                int braceLevel = 1
                oleEnd = oleStart+1
                print "Searching End of OLE from:" oleStart "\n"
                while (braceLevel > 0) {
                        if ( search (reBrace, b, oleEnd) ) {
                                string found = b[oleEnd+(start 0):oleEnd+(end 0)]
                                print "Found brace at:" (start 0) "(" found ")\n"
                                oleEnd = (end 0)+1+oleEnd
                                print "oleEnd:" (oleEnd) "-" found "\n"
                                if found == "{" then braceLevel++
                                if found == "}" then braceLevel--                          
                        } else break
                }            
                nextOle = oleStart
        } else { nextOle = iLen + 1 }
        print "Found OLE object: " oleStart "\n"
   }
   // skip ole object
   if (nextOle == pos) { 
        pos = oleEnd; continue
        print "Skipping OLE object: " oleEnd "\n"
   }
 
   // your replace code here: 
   pos ++
}

Mathias Mamsch, IT-QBase GmbH, Consultant for Requirement Engineering and D00RS

Mathias Mamsch - Tue Sep 14 13:44:07 EDT 2010

Hmm you are right, it IS really slow :-) Maybe you can optimize your replace code to be faster, by explicitly checking for "{\object" or "{\pict" phrases, then search their ending by finding an approriate "}" using two regular expressions:

The example code below does a loop over all characters of the buffer, thereby skipping OLE objects ... Hope this helps, regards, Mathias
 

pragma runLim, 10000
 
Regexp reObj = regexp "({\\\\object)|({\\\\pict)"
Regexp reBrace = regexp "([{}])"
 
Buffer b = create(); 
b = richTextWithOle (current Object)."Object Text"
 
// print tempStringOf b
 
// halt
 
int pos = 0; 
int iLen = length b
int oleStart = 0
int oleEnd   = 0 
 
Buffer result = create() 
int nextOle = -1
 
while (pos < iLen) {
   if (nextOle < pos) {
    // find the next ole object
        if ( search (reObj, b, pos) ) {
                oleStart = oleStart + start 0
                int braceLevel = 1
                oleEnd = oleStart+1
                print "Searching End of OLE from:" oleStart "\n"
                while (braceLevel > 0) {
                        if ( search (reBrace, b, oleEnd) ) {
                                string found = b[oleEnd+(start 0):oleEnd+(end 0)]
                                print "Found brace at:" (start 0) "(" found ")\n"
                                oleEnd = (end 0)+1+oleEnd
                                print "oleEnd:" (oleEnd) "-" found "\n"
                                if found == "{" then braceLevel++
                                if found == "}" then braceLevel--                          
                        } else break
                }            
                nextOle = oleStart
        } else { nextOle = iLen + 1 }
        print "Found OLE object: " oleStart "\n"
   }
   // skip ole object
   if (nextOle == pos) { 
        pos = oleEnd; continue
        print "Skipping OLE object: " oleEnd "\n"
   }
 
   // your replace code here: 
   pos ++
}

Mathias Mamsch, IT-QBase GmbH, Consultant for Requirement Engineering and D00RS

Mr.Ferrari - Wed Oct 06 09:43:12 EDT 2010

Hey...
Your code is working so far. It finds the beginnig and the end braces but it will not enter the 'skip ole' part...
 

...
// skip ole object
if (nextOle == pos) { 
  pos = oleEnd; continue
  print "Skipping OLE object: " oleEnd "\n"
}
...

What can still be wrong?

THX

Mr.Ferrari - Wed Oct 06 09:43:12 EDT 2010

Hey...
Your code is working so far. It finds the beginnig and the end braces but it will not enter the 'skip ole' part...
 

...
// skip ole object
if (nextOle == pos) { 
  pos = oleEnd; continue
  print "Skipping OLE object: " oleEnd "\n"
}
...

What can still be wrong?

THX

Mathias Mamsch - Wed Oct 06 10:06:25 EDT 2010

Not sure. It used to work. Can you attach the richText you are trying this with in a file and post the code you are using (if you modified it)?
 

print richTextWithOle (current Object)."Object Text"

Regards, Mathias

Mathias Mamsch, IT-QBase GmbH, Consultant for Requirement Engineering and D00RS

Mr.Ferrari - Wed Oct 06 10:30:05 EDT 2010
I did not modify the code.

I am getting an error: null buffer parameter was passed into argument position 1

argh...