X-Git-Url: http://git.nguyen.vg/gitweb/?a=blobdiff_plain;f=XMLTree.cpp;h=c87c6dc008443bb723a450f7ecef9aff560598a3;hb=f97501e008660c0363f0fe643be09de66efd3533;hp=174c94c76ffe50379cfa1cdce354583e146dbac0;hpb=912ff50e1d38de484b503d8ef877a49a65765ab9;p=SXSI%2FXMLTree.git

diff --git a/XMLTree.cpp b/XMLTree.cpp
index 174c94c..c87c6dc 100644
--- a/XMLTree.cpp
+++ b/XMLTree.cpp
@@ -1,8 +1,11 @@
 #include "basics.h"
-//#include <cstring>
-#include <stack>
 #include "XMLTree.h"
 #include "timings.h"
+#include <errno.h>
+using std::cout;
+using std::endl;
+using std::min;
+using std::string;
 
 // functions to convert tag positions to the corresponding tree node and viceversa. 
 // These are implemented in order to be able to change the tree and Tags representations, 
@@ -53,7 +56,7 @@ static treeNode fast_first_child(bp *Par, treeNode x)
 
 static treeNode fast_next_sibling(bp* Par,treeNode x)
 {
-  x = fwd_excess(Par,x,0);
+  x = fast_find_close(Par,x)+1;
   return (fast_inspect(Par,x) == OP) ? x : NULLT;
 }
 
@@ -107,6 +110,7 @@ XMLTree::XMLTree( pb * const par, uint npar,  vector<string> * const TN,  TagIdM
 		  uint *empty_texts_bmp, TagType *tags,
 		  TextCollection * const TC, bool dis_tc)
  {
+   buffer = 0;
     // creates the data structure for the tree topology
     Par = (bp *)umalloc(sizeof(bp));
     STARTTIMER();
@@ -248,12 +252,13 @@ void XMLTree::Save(int fd)
 // a pointer to the loaded data structure
 XMLTree *XMLTree::Load(int fd, bool load_tc,int sample_factor) 
  {
+
     FILE *fp;
     char buffer[1024];
     XMLTree *XML_Tree;
     int i;
 
-
+    buffer[1023] = '\0';
 
     fp = fdopen(fd, "r");
 
@@ -267,23 +272,21 @@ XMLTree *XMLTree::Load(int fd, bool load_tc,int sample_factor)
     PRINTTIME("Loading parenthesis struct", Loading);
     STARTTIMER();
 
-    XML_Tree->TagName = new vector<string>();
-    XML_Tree->tIdMap = new std::unordered_map<string,int>();
-    
-    string s;
+    XML_Tree->TagName = new std::vector<std::string>();
+    XML_Tree->tIdMap = new std::unordered_map<std::string,int>();
+    std::string s;
     int ntags;
     
     // Load the tag names
     ufread(&ntags, sizeof(int), 1, fp);
 
     for (i=0; i<ntags;i++) {
-      char * r = fgets(buffer,1023,fp);
-       if (r==NULL)
+       if (fgets(buffer,1022,fp) != buffer)
 	 throw "Cannot read tag list";
-       s = (const char*) buffer;
+       s = buffer;
        // remove the trailing \n
        s.erase(s.size()-1);       
-       XML_Tree->TagName->push_back(s);
+       XML_Tree->TagName->push_back(s);       
        XML_Tree->tIdMap->insert(std::make_pair(s,i));
        
     };
@@ -364,7 +367,7 @@ int XMLTree::SubtreeTags(treeNode x, TagType tag)
 
     int s = x + 2*subtree_size(Par, x) - 1;
  
-    return Tags->rank(tag, s) - Tags->rank(tag, node2tagpos(x)-1);
+    return (Tags->rank(tag, s) - Tags->rank(tag, node2tagpos(x)-1))+1;
  }
 int XMLTree::SubtreeElements(treeNode x) 
  {
@@ -536,22 +539,6 @@ treeNode XMLTree::NextElement(treeNode x)
   }
   else return x;  
 }
-value XMLTree::CamlFirstElement(value x)
-{
-  return Val_int(FirstElement(Int_val(x)));
-}
-value XMLTree::CamlNextElement(value x)
-{
-  return Val_int(NextElement(Int_val(x)));
-}
-
-extern "C" value caml_cpp_fast_first_element(value xmltree, value node){
-  return XMLTREE(xmltree)->CamlFirstElement(node);
-}
-
-extern "C" value caml_cpp_fast_next_element(value xmltree, value node){
-  return XMLTREE(xmltree)->CamlNextElement(node);
-}
 
 // LastChild(x): returns the last child of node x.
 treeNode XMLTree::LastChild(treeNode x)
@@ -941,18 +928,18 @@ bool XMLTree::IsOpen(treeNode x) { return fast_inspect(Par,x); }
 
 //WARNING this uses directly the underlying implementation for plain text
 
-void XMLTree::Print(int fd,treeNode x){
+
+void XMLTree::Print(int fd,treeNode x, bool no_text){
   
   int newfd = dup(fd);
   stream = fdopen(newfd,"wa");
-  /*  if (stream_fd != fd){
-    if (stream != NULL)
-      fclose(stream);
-    int newfd = dup(fd);
-    stream = fdopen(newfd,"wa");
-    stream_fd = fd;
-    };
-  */
+  if (stream == 0){
+    perror(NULL);
+    return;
+  };
+
+  if (buffer == 0)
+    buffer = new string();
 
   FILE* fp = stream;
   treeNode fin = fast_find_close(Par,x);
@@ -961,8 +948,8 @@ void XMLTree::Print(int fd,treeNode x){
   uchar * tagstr;
   range r = DocIds(x);
   treeNode first_idx;
-  treeNode first_text = (tag == PCDATA_TAG_ID ?  x : TaggedDescendant(x,PCDATA_TAG_ID));
-  treeNode first_att =  NULLT;//TaggedDesc(x,ATTRIBUTE_DATA_TAG_ID);
+  treeNode first_text = (tag == PCDATA_TAG_ID ?  x : ParentNode(r.min-1));
+  treeNode first_att =  NULLT;
   
   if (first_att  == NULLT)
   first_idx = first_text;
@@ -974,52 +961,66 @@ void XMLTree::Print(int fd,treeNode x){
    uchar * current_text=NULL;
    if (first_idx != NULLT)
    current_text = GetText(MyText(first_idx));
-   int read = 0;
-
- std::stack<uchar*> st;
+   size_t read = 0;
+   std::vector<uchar*> st;
  while (n <= fin){
    if (fast_inspect(Par,n)){
-     if (tag == PCDATA_TAG_ID) {       
-       //       fputs((const char*) (GetText(MyTextUnsafe(n))),fp);
-      
-       read = fprintf(fp,"%s",(const char*) current_text);
-       current_text += (read + 1);
-
+     if (tag == PCDATA_TAG_ID  ) {       
+
+       if (no_text)
+	 myfputs("<$/>",fp);
+       else{
+	 read = myfprintf((const char*) current_text, fp);
+	 current_text += (read + 1);
+       };
        n+=2; // skip closing $
        tag = Tag(n);
+      
      }
      else {
-
-       fputc('<',fp);
+       myfputc('<',fp);
        tagstr = (uchar*) GetTagNameByRef(tag);
-       fputs((const char*) tagstr ,fp);
+       myfputs((const char*) tagstr ,fp);
        n++;
        if (fast_inspect(Par,n)) {
-	 st.push(tagstr);
+	 st.push_back(tagstr);
 	 tag = Tag(n);
 	 if (tag == ATTRIBUTE_TAG_ID){
 	   n++;
+	   if (no_text) myfputs("><@@>",fp);
 	   while (fast_inspect(Par,n)){
-	     fputc(' ',fp);
-	     fputs((const char*) &(GetTagNameByRef(Tag(n))[3]),fp);
-	     fputs("=\"",fp);
-	     n++;
-	     read = fprintf(fp,"%s",(const char*) current_text);
-	     current_text += (read + 1);
-	     //fputs((const char*) GetText(MyTextUnsafe(n)),fp);
-	     fputc('"',fp);
-	     n+=3; //close @$ @@	     	     
+	     if (no_text) {
+	       myfputc('<',fp);
+	       myfputs((const char*) &(GetTagNameByRef(Tag(n))[3]),fp);
+	       myfputc('>',fp);
+	       myfputs("<$@/></",fp);
+	       myfputs((const char*) &(GetTagNameByRef(Tag(n))[3]),fp);
+	       myfputc('>',fp);
+	       n+= 4;
+	     }
+	     else {
+	       myfputc(' ',fp);
+	       myfputs((const char*) &(GetTagNameByRef(Tag(n))[3]),fp);
+	       n++;
+	       myfputs("=\"",fp);
+	       read = myfprintf((const char*) current_text,fp);
+	       current_text += (read + 1);
+	       myfputc('"',fp);
+	       n+=3;
+	     }
 	   };
-	   fputc('>',fp);
+	   if (no_text) 
+	     myfputs("</@@>",fp);
+	   else myfputc('>',fp);
 	   n++;
 	   tag=Tag(n);
 	 }
 	 else {
-	   fputc('>',fp);
+	   myfputc('>',fp);
 	 };
        }
        else {// <foo /> tag
-	 fputs("/>",fp);
+	 myfputs("/>",fp);
 	 n++;
 	 tag=Tag(n);	 
        };     
@@ -1027,15 +1028,16 @@ void XMLTree::Print(int fd,treeNode x){
    }
    else
      do {
-       fputs("</",fp);
-       fputs((const char*)st.top(),fp);
-       fputc('>', fp);
-       st.pop();
+       myfputs("</",fp);
+       myfputs((const char*)st.back(),fp);
+       myfputc('>', fp);
+       st.pop_back();
        n++;
      }while (!fast_inspect(Par,n) && !st.empty());
    tag=Tag(n);
  };
- fputc('\n',fp);
- fflush(fp);
+ myfputc('\n',fp);
+ mybufferflush(fp);
+ //fflush(fp);
  fclose(fp);
 }