avoid further troubles when wild anonymization was performed

[gdcm.git] / src / gdcmSerieHelper.cxx

diff --git a/src/gdcmSerieHelper.cxx b/src/gdcmSerieHelper.cxx
index 28e4218eb151f29cd40373ffb143b7aed42c4b86..4e215e7fb221ca8a96a9c2f8e02b3cb4b9224791 100644 (file)
--- a/src/gdcmSerieHelper.cxx
+++ b/src/gdcmSerieHelper.cxx
@@ -3,8 +3,8 @@
   Program:   gdcm
   Module:    $RCSfile: gdcmSerieHelper.cxx,v $
   Language:  C++
-  Date:      $Date: 2007/10/30 11:11:04 $
-  Version:   $Revision: 1.63 $
+  Date:      $Date: 2010/04/09 15:38:18 $
+  Version:   $Revision: 1.70 $
                                                                                 
   Copyright (c) CREATIS (Centre de Recherche et d'Applications en Traitement de
   l'Image). All rights reserved. See Doc/License.txt or
@@ -13,7 +13,7 @@
      This software is distributed WITHOUT ANY WARRANTY; without even
      the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR
      PURPOSE.  See the above copyright notices for more information.
-                                                                                
+                                                                              
 =========================================================================*/
 
 #include "gdcmSerieHelper.h"
@@ -43,7 +43,7 @@ SerieHelper::SerieHelper()
    ClearAll();
    UserLessThanFunction = 0;
    DirectOrder = true;
-   
+   DropDuplicatePositions = false;   
 }
 
 /**
@@ -161,7 +161,7 @@ bool SerieHelper::AddFile(File *header)
       std::string id = CreateUniqueSeriesIdentifier( header );
       // if id == GDCM_UNFOUND then consistently we should find GDCM_UNFOUND
       // no need here to do anything special
- 
+
       if ( SingleSerieUIDFileSetHT.count(id) == 0 )
       {
          gdcmDebugMacro(" New/gdcmSerieHelper.cxx Serie UID :[" << id << "]");
@@ -258,11 +258,14 @@ void SerieHelper::AddSeriesDetail(uint16_t group, uint16_t elem, bool convert)
 void SerieHelper::SetDirectory(std::string const &dir, bool recursive)
 {
    DirList dirList(dir, recursive); // OS specific
-  
+
    DirListType filenames_list = dirList.GetFilenames();
    for( DirListType::const_iterator it = filenames_list.begin(); 
         it != filenames_list.end(); ++it)
    {
+     // std::cout << "-----------------------------filename [" << *it << "]"
+     //           << std::endl;
+      gdcmDebugMacro("filename [" << *it << "]" );
       AddFileName( *it );
    }
 }
@@ -648,6 +651,8 @@ bool SerieHelper::ImagePositionPatientOrdering( FileList *fileList )
    bool first = true;
    ZSpacing = -1.0;  // will be updated if process doesn't fail
 
+   gdcmDebugMacro("============================================DropDuplicatePositions : " << DropDuplicatePositions );
+    
    std::multimap<double,File *> distmultimap;
    // Use a multimap to sort the distances from 0,0,0
    for ( FileList::const_iterator 
@@ -720,6 +725,8 @@ bool SerieHelper::ImagePositionPatientOrdering( FileList *fileList )
       }
    }
 
+   gdcmDebugMacro("After parsing vector, nb of elements : " << fileList->size() );
+
    // Find out if min/max are coherent
    if ( min == max )
    {
@@ -735,6 +742,9 @@ bool SerieHelper::ImagePositionPatientOrdering( FileList *fileList )
         it2 != distmultimap.end();
         ++it2)
    {
+   
+      gdcmDebugMacro("Check if image shares a common position : " << (*it2).second->GetFileName() );   
+   
       if (distmultimap.count((*it2).first) != 1)
       {
          gdcmWarningMacro("File: ["
@@ -780,9 +790,14 @@ bool SerieHelper::ImagePositionPatientOrdering( FileList *fileList )
          fileList->push_back( (*it3).second );
          if (DropDuplicatePositions)
          {
-            /// \todo ImagePositionPatientOrdering  wrong duplicates are found ???
-   
+            // ImagePositionPatientOrdering  wrong duplicates are found ???
+            // --> fixed. See comment
+
             it3 =  distmultimap.upper_bound((*it3).first); // skip all duplicates
+           // the upper_bound function increments the iterator to the next non-duplicate entry
+           // The for loop iteration also increments the iterator, which causes the code to skip every other image
+           // --> decrement the iterator after the upper_bound function call
+            it3--;
             if (it3 == distmultimap.end() )  // if last image, stop iterate
                break;
          }
@@ -798,7 +813,16 @@ bool SerieHelper::ImagePositionPatientOrdering( FileList *fileList )
          fileList->push_back( (*it4).second );
          if (DropDuplicatePositions)  // skip all duplicates
          {
-           it4 =  distmultimap.upper_bound((*it4).first);
+            // lower_bound finds the next element that is 
+            // less than or *equal to* the current value!
+            //it4 =  distmultimap.lower_bound((*it4).first);
+   
+           // David Feng's fix
+           std::multimap<double, File *>::const_iterator itPrev = it4;
+           while (itPrev->first == it4->first)
+              --itPrev;
+           it4 = itPrev;
+    
            if (it4 == distmultimap.begin() ) // if first image, stop iterate
                break;
          } 
@@ -1014,7 +1038,7 @@ std::string SerieHelper::CreateUniqueSeriesIdentifier( File *inFile )
       }
    }
    // deal with Dicom strings trailing '\0' 
-    if(id[s_size-1] == '_')
+    if(s_size && id[s_size-1] == '_')
       id.erase(s_size-1, 1);
     return id;
     }
@@ -1068,23 +1092,30 @@ std::string SerieHelper::CreateUserDefinedFileIdentifier( File *inFile )
          }
       }
       // Eliminate non-alphanum characters, including whitespace.
+
       unsigned int s_size = s.size();
-      for(unsigned int i=0; i<s_size; i++)
+      if(s_size == 0)
+      { // to avoid further troubles when wild anonymization was performed
+         s = "a";
+      }
+      else
       {
-         while(i<s_size
+         for(unsigned int i=0; i<s_size; i++)
+         {
+            while(i<s_size
                && !( s[i] == '.' || s[i] == '%' || s[i] == '_'
                  || (s[i] >= '+' && s[i] <= '-')       
                  || (s[i] >= 'a' && s[i] <= 'z')
                  || (s[i] >= '0' && s[i] <= '9')
                  || (s[i] >= 'A' && s[i] <= 'Z')))
-         {
-            s.replace(i, 1, "_");  // ImagePositionPatient related stuff will be more human readable
+            {
+               s.replace(i, 1, "_");  // ImagePositionPatient related stuff will be more human readable
+            }
          }
+         // deal with Dicom strings trailing '\0' 
+         if(s[s_size-1] == '_')
+            s.erase(s_size-1, 1);
       }
-      // deal with Dicom strings trailing '\0' 
-      if(s[s_size-1] == '_')
-         s.erase(s_size-1, 1);
-      
       id += s.c_str();
       id += "%%%"; // make the FileIdentifier Tokenizable
    }