4 // functions to convert tag positions to the corresponding tree node and viceversa.
\r
5 // These are implemented in order to be able to change the tree and Tags representations,
\r
6 // without affecting the code so much.
\r
7 // Current implementation corresponds to balanced-parentheses representation for
\r
8 // the tree, and storing 2 tags per tree node (opening and closing tags).
\r
10 // tag position -> tree node
\r
11 inline treeNode tagpos2node(int t) {
\r
15 // tree node -> tag position
\r
16 inline int node2tagpos(treeNode x) {
\r
20 // Save: saves XML tree data structure to file.
\r
21 void XMLTree::Save(unsigned char *filename)
\r
25 char filenameaux[1024];
\r
28 sprintf(filenameaux, "%s.srx", filename);
\r
29 fp = fopen(filenameaux, "w");
\r
31 printf("Error: cannot create file %s to store the tree structure of XML collection\n", filenameaux);
\r
35 // first stores the tree topology
\r
38 // stores the table with tag names
\r
39 fwrite(&ntagnames, sizeof(int), 1, fp);
\r
40 for (i=0; i<ntagnames;i++)
\r
41 fprintf(fp, "%s\n",TagName[i]);
\r
44 fwrite(&indexing_empty_texts, sizeof(bool), 1, fp);
\r
45 fwrite(&initialized, sizeof(bool), 1, fp);
\r
46 fwrite(&finished, sizeof(bool), 1, fp);
\r
48 if (!indexing_empty_texts) EBVector->save(fp);
\r
53 // stores the texts
\r
61 // Load: loads XML tree data structure from file. Returns
\r
62 // a pointer to the loaded data structure
\r
63 XMLTree *XMLTree::Load(unsigned char *filename, int sample_rate_text)
\r
67 char filenameaux[1024];
\r
71 // first load the tree topology
\r
72 sprintf(filenameaux, "%s.srx", filename);
\r
73 fp = fopen(filenameaux, "r");
\r
75 printf("Error: cannot open file %s to load the tree structure of XML collection\n", filenameaux);
\r
79 XML_Tree = new XMLTree();
\r
81 XML_Tree->Par = (bp *)malloc(sizeof(bp));
\r
83 loadTree(XML_Tree->Par, fp);
\r
85 // stores the table with tag names
\r
86 fread(&XML_Tree->ntagnames, sizeof(int), 1, fp);
\r
88 XML_Tree->TagName = (unsigned char **)malloc(XML_Tree->ntagnames*sizeof(unsigned char *));
\r
90 for (i=0; i<XML_Tree->ntagnames;i++) {
\r
92 fscanf(fp, "%s\n",filenameaux);
\r
93 XML_Tree->TagName[i] = (unsigned char *)malloc(sizeof(unsigned char)*(strlen((const char *)filenameaux)+1));
\r
94 strcpy((char *)XML_Tree->TagName[i], (const char *)filenameaux);
\r
98 fread(&(XML_Tree->indexing_empty_texts), sizeof(bool), 1, fp);
\r
99 fread(&(XML_Tree->initialized), sizeof(bool), 1, fp);
\r
100 fread(&(XML_Tree->finished), sizeof(bool), 1, fp);
\r
102 if (!(XML_Tree->indexing_empty_texts)) XML_Tree->EBVector = static_bitsequence_rrr02::load(fp);
\r
105 XML_Tree->Tags = static_sequence_wvtree::load(fp);
\r
107 // loads the texts
\r
108 XML_Tree->Text->Load(fp,sample_rate_text);
\r
116 // ~XMLTree: frees memory of XML tree.
\r
117 XMLTree::~XMLTree()
\r
122 free(Par); // frees the memory of struct Par
\r
124 for (i=0; i<ntagnames;i++)
\r
129 if (!indexing_empty_texts) {
\r
130 EBVector->~static_bitsequence_rrr02();
\r
135 Tags->~static_sequence_wvtree();
\r
139 Text->~TextCollection();
\r
143 initialized = false;
\r
147 // root(): returns the tree root.
\r
148 treeNode XMLTree::Root()
\r
151 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
154 return root_node(Par);
\r
157 // SubtreeSize(x): the number of nodes (and attributes) in the subtree of node x.
\r
158 int XMLTree::SubtreeSize(treeNode x)
\r
161 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
164 return subtree_size(Par, x);
\r
167 // SubtreeTags(x,tag): the number of occurrences of tag within the subtree of node x.
\r
168 int XMLTree::SubtreeTags(treeNode x, TagType tag)
\r
171 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
175 int s = x + 2*subtree_size(Par, x) - 1;
\r
177 return Tags->rank(tag, s) - Tags->rank(tag, node2tagpos(x)-1);
\r
180 // IsLeaf(x): returns whether node x is leaf or not. In the succinct representation
\r
181 // this is just a bit inspection.
\r
182 bool XMLTree::IsLeaf(treeNode x)
\r
184 return isleaf(Par, x);
\r
187 // IsAncestor(x,y): returns whether node x is ancestor of node y.
\r
188 bool XMLTree::IsAncestor(treeNode x, treeNode y)
\r
191 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
195 return is_ancestor(Par, x, y);
\r
198 // IsChild(x,y): returns whether node x is parent of node y.
\r
199 bool XMLTree::IsChild(treeNode x, treeNode y)
\r
202 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
206 if (!is_ancestor(Par, x, y)) return false;
\r
207 return depth(Par, x) == (depth(Par, y) + 1);
\r
210 // NumChildren(x): number of children of node x. Constant time with the data structure
\r
212 int XMLTree::NumChildren(treeNode x)
\r
215 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
219 return degree(Par, x);
\r
222 // ChildNumber(x): returns i if node x is the i-th children of its parent.
\r
223 int XMLTree::ChildNumber(treeNode x)
\r
226 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
230 return child_rank(Par, x);
\r
233 // Depth(x): depth of node x, a simple binary rank on the parentheses sequence.
\r
234 int XMLTree::Depth(treeNode x)
\r
237 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
241 return depth(Par, x);
\r
244 // Preorder(x): returns the preorder number of node x, just counting the tree
\r
245 // nodes (i.e., tags, it disregards the texts in the tree).
\r
246 int XMLTree::Preorder(treeNode x)
\r
249 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
253 return preorder_rank(Par, x);
\r
256 // Postorder(x): returns the postorder number of node x, just counting the tree
\r
257 // nodes (i.e., tags, it disregards the texts in the tree).
\r
258 int XMLTree::Postorder(treeNode x)
\r
261 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
265 return postorder_rank(Par, x);
\r
268 // Tag(x): returns the tag identifier of node x.
\r
269 TagType XMLTree::Tag(treeNode x)
\r
272 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
276 return Tags->access(node2tagpos(x));
\r
279 // DocIds(x): returns the range of text identifiers that descend from node x.
\r
280 // returns {NULLT, NULLT} when there are no texts descending from x.
\r
281 range XMLTree::DocIds(treeNode x)
\r
284 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
289 if (indexing_empty_texts) { // faster, no rank needed
\r
291 r.max = x+2*subtree_size(Par, x)-2;
\r
293 else { // we are not indexing empty texts, we need rank
\r
294 int min = EBVector->rank1(x-1);
\r
295 int max = EBVector->rank1(x+2*subtree_size(Par, x)-2);
\r
296 if (min==max) { // range is empty, no texts within the subtree of x
\r
300 else { // the range is non-empty, there are texts within the subtree of x
\r
308 // Parent(x): returns the parent node of node x.
\r
309 treeNode XMLTree::Parent(treeNode x)
\r
312 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
316 return parent(Par, x);
\r
319 // Child(x,i): returns the i-th child of node x, assuming it exists.
\r
320 treeNode XMLTree::Child(treeNode x, int i)
\r
323 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
327 if (i <= OPTD) return naive_child(Par, x, i);
\r
328 else return child(Par, x, i);
\r
331 // FirstChild(x): returns the first child of node x, assuming it exists. Very fast in BP.
\r
332 treeNode XMLTree::FirstChild(treeNode x)
\r
335 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
339 return first_child(Par, x);
\r
342 // NextSibling(x): returns the next sibling of node x, assuming it exists.
\r
343 treeNode XMLTree::NextSibling(treeNode x)
\r
346 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
350 return next_sibling(Par, x);
\r
353 // PrevSibling(x): returns the previous sibling of node x, assuming it exists.
\r
354 treeNode XMLTree::PrevSibling(treeNode x)
\r
357 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
361 return prev_sibling(Par, x);
\r
364 // TaggedChild(x,i,tag): returns the i-th child of node x tagged tag, or NULLT if there is none.
\r
365 // Because of the balanced-parentheses representation of the tree, this operation is not supported
\r
366 // efficiently, just iterating among the children of node x until finding the desired child.
\r
367 treeNode XMLTree::TaggedChild(treeNode x, int i, TagType tag)
\r
370 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
376 child = first_child(Par, x); // starts at first child of node x
\r
377 if (child==(treeNode)-1) return NULLT; // node x is a leaf, there is no such child
\r
378 while (child!=(treeNode)-1) {
\r
379 if (Tags->access(node2tagpos(child)) == tag) { // current child is labeled with tag of interest
\r
381 if (i==0) return child; // we have seen i children of x tagged tag, this is the one we are looking for
\r
383 child = next_sibling(Par, x); // OK, let's try with the next child
\r
385 return NULLT; // no such child was found
\r
388 // TaggedDesc(x,tag): returns the first node tagged tag with larger preorder than x and within
\r
389 // the subtree of x. Returns NULLT if there is none.
\r
390 treeNode XMLTree::TaggedDesc(treeNode x, TagType tag)
\r
393 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
399 r = (int) Tags->rank(tag, node2tagpos(x));
\r
400 s = (int) Tags->select(tag, r+1);
\r
401 if (s == -1) return NULLT; // there is no such node
\r
402 y = tagpos2node(s); // transforms the tag position into a node position
\r
403 if (!is_ancestor(Par, x, y)) return NULLT; // the next node tagged tag (in preorder) is not within the subtree of x.
\r
407 // TaggedPrec(x,tag): returns the first node tagged tag with smaller preorder than x and not an
\r
408 // ancestor of x. Returns NULLT if there is none.
\r
409 treeNode XMLTree::TaggedPrec(treeNode x, TagType tag)
\r
412 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
417 treeNode node_s, root;
\r
418 r = (int)Tags->rank(tag, node2tagpos(x)-1);
\r
419 if (r==0) return NULLT; // there is no such node.
\r
420 s = (int)Tags->select(tag, r);
\r
421 root = root_node(Par);
\r
422 node_s = tagpos2node(s);
\r
423 while (is_ancestor(Par, node_s, x) && (node_s!=root)) { // the one that we found is an ancestor of x
\r
425 if (r==0) return NULLT; // there is no such node
\r
426 s = (int)Tags->select(tag, r); // we should use select_prev instead when provided
\r
427 node_s = tagpos2node(s);
\r
429 return NULLT; // there is no such node
\r
432 // TaggedFoll(x,tag): returns the first node tagged tag with larger preorder than x and not in
\r
433 // the subtree of x. Returns NULLT if there is none.
\r
434 treeNode XMLTree::TaggedFoll(treeNode x, TagType tag)
\r
437 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
442 r = (int) Tags->rank(tag, node2tagpos(next_sibling(Par, x))-1);
\r
443 s = (int) Tags->select(tag, r+1); // select returns -1 in case that there is no r+1-th tag.
\r
444 if (s==-1) return NULLT;
\r
445 else return tagpos2node(s);
\r
448 // PrevText(x): returns the document identifier of the text to the left
\r
449 // of node x, or NULLT if x is the root node or the text is empty.
\r
450 // Assumes Doc ids start from 0.
\r
451 DocID XMLTree::PrevText(treeNode x)
\r
454 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
458 if (x == Root()) return NULLT;
\r
459 if (indexing_empty_texts) // faster, no rank needed
\r
461 else { // we are not indexing empty texts, rank is needed
\r
462 if (EBVector->access(x-1) == 0)
\r
463 return (DocID)NULLT; // there is no text to the left of node (text is empty)
\r
465 return (DocID)EBVector->rank1(x-1)-1; //-1 because document ids start from 0
\r
469 // NextText(x): returns the document identifier of the text to the right
\r
470 // of node x, or NULLT if x is the root node. Assumes Doc ids start from 0.
\r
471 DocID XMLTree::NextText(treeNode x)
\r
474 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
478 if (x == Root()) return NULLT;
\r
479 if (indexing_empty_texts) // faster, no rank needed
\r
480 return (DocID)x+2*subtree_size(Par, x)-1;
\r
481 else { // we are not indexing empty texts, rank is needed
\r
482 int p = x+2*subtree_size(Par, x)-1;
\r
483 if (EBVector->access(p) == 0) // there is no text to the right of node
\r
484 return (DocID)NULLT;
\r
486 return (DocID)EBVector->rank1(p)-1; //-1 because document ids start from 0
\r
490 // MyText(x): returns the document identifier of the text below node x,
\r
491 // or NULLT if x is not a leaf node or the text is empty. Assumes Doc
\r
492 // ids start from 0.
\r
493 DocID XMLTree::MyText(treeNode x)
\r
496 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
500 if (!IsLeaf(x)) return NULLT;
\r
501 if (indexing_empty_texts) // faster, no rank needed
\r
503 else { // we are not indexing empty texts, rank is needed
\r
504 if (EBVector->access(x) == 0) // there is no text below node x
\r
505 return (DocID)NULLT;
\r
507 return (DocID)EBVector->rank1(x)-1; //-1 because document ids start from 0
\r
511 // TextXMLId(d): returns the preorder of document with identifier d in the tree consisting of
\r
512 // all tree nodes and all text nodes. Assumes that the tree root has preorder 1.
\r
513 int XMLTree::TextXMLId(DocID d)
\r
516 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
520 if (indexing_empty_texts)
\r
521 return d + rank_open(Par, d)+1; // +1 because root has preorder 1
\r
522 else { // slower, needs rank and select
\r
523 int s = EBVector->select1(d+1);
\r
524 return rank_open(Par, s) + d + 1; // +1 because root has preorder 1
\r
528 // NodeXMLId(x): returns the preorder of node x in the tree consisting
\r
529 // of all tree nodes and all text nodes. Assumes that the tree root has
\r
531 int XMLTree::NodeXMLId(treeNode x)
\r
534 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
538 if (indexing_empty_texts)
\r
539 return x - 1 + rank_open(Par, x);
\r
541 if (x == Root()) return 1; // root node has preorder 1
\r
543 return rank_open(Par, x) + EBVector->rank1(x-1);
\r
547 // ParentNode(d): returns the parent node of document identifier d.
\r
548 treeNode XMLTree::ParentNode(DocID d)
\r
551 fprintf(stderr, "Error: data structure has not been constructed properly\n");
\r
556 if (indexing_empty_texts) s = d;
\r
557 else s = EBVector->select1(d);
\r
559 if (inspect(Par,s) == CP) // is a closing parenthesis
\r
560 return parent(Par, find_open(Par, s));
\r
561 else // is an opening parenthesis
\r
562 return (treeNode)s;
\r
567 // OpenDocument(empty_texts): it starts the construction of the data structure for
\r
568 // the XML document. Parameter empty_texts indicates whether we index empty texts
\r
569 // in document or not. Returns a non-zero value upon success, NULLT in case of error.
\r
570 int XMLTree::OpenDocument(bool empty_texts, int sample_rate_text)
\r
572 initialized = true;
\r
577 indexing_empty_texts = empty_texts;
\r
579 par_aux = (pb *)malloc(sizeof(pb));
\r
581 fprintf(stderr, "Error: not enough memory\n");
\r
584 setbit(par_aux,npar,OP); // marks a new opening parenthesis for the tree root
\r
587 tags_aux = (TagType *)malloc(sizeof(TagType));
\r
589 fprintf(stderr, "Error: not enough memory\n");
\r
595 if (!indexing_empty_texts) {
\r
596 empty_texts_aux = (unsigned int *)malloc(sizeof(unsigned int));
\r
597 if (!empty_texts_aux) {
\r
598 fprintf(stderr, "Error: not enough memory\n");
\r
603 Text = TextCollection::InitTextCollection((unsigned)sample_rate_text);
\r
605 return 1; // indicates success in the initialization of the data structure
\r
608 // CloseDocument(): it finishes the construction of the data structure for the XML
\r
609 // document. Tree and tags are represented in the final form, dynamic data
\r
610 // structures are made static, and the flag "finished" is set to true. After that,
\r
611 // the data structure can be queried.
\r
612 int XMLTree::CloseDocument()
\r
614 if (!initialized) { // data structure has not been initialized properly
\r
615 fprintf(stderr, "Error: data structure has not been initialized properly (by calling method OpenDocument)\n");
\r
619 // closing parenthesis for the tree root
\r
620 par_aux = (pb *)realloc(par_aux, sizeof(pb)*(1+npar/(8*sizeof(pb))));
\r
622 fprintf(stderr, "Error: not enough memory\n");
\r
625 setbit(par_aux,npar,CP);
\r
628 // creates the data structure for the tree topology
\r
629 Par = (bp *)malloc(sizeof(bp));
\r
630 bp_construct(Par, npar, par_aux, OPT_DEGREE|0);
\r
631 // creates structure for tags
\r
632 alphabet_mapper * am = new alphabet_mapper_none();
\r
633 static_bitsequence_builder * bmb = new static_bitsequence_builder_rrr02(32);
\r
634 wt_coder * wtc = new wt_coder_huff((uint *)tags_aux,npar-1,am);
\r
635 Tags = new static_sequence_wvtree((uint *) tags_aux, (uint) npar-1, wtc, bmb, am);
\r
637 // makes the text collection static
\r
638 Text->MakeStatic();
\r
640 // creates the data structure marking the non-empty texts (just in the case it is necessary)
\r
641 if (!indexing_empty_texts)
\r
642 EBVector = new static_bitsequence_rrr02((uint *)empty_texts_aux,(ulong)npar,(uint)32);
\r
646 return 1; // indicates success in the inicialization
\r
650 // NewOpenTag(tagname): indicates the event of finding a new opening tag in the document.
\r
651 // Tag name is given. Returns a non-zero value upon success, and returns NULLT
\r
652 // in case of failing when trying to insert the new tag.
\r
653 int XMLTree::NewOpenTag(unsigned char *tagname)
\r
657 if (!initialized) { // data structure has not been initialized properly
\r
658 fprintf(stderr, "Error: you cannot insert a new opening tag without first calling method OpenDocument first\n");
\r
662 // inserts a new opening parentheses in the bit sequence
\r
663 par_aux = (pb *)realloc(par_aux, sizeof(pb)*(1+npar/(8*sizeof(pb))));
\r
665 fprintf(stderr, "Error: not enough memory\n");
\r
668 setbit(par_aux,npar,OP); // marks a new opening parenthesis
\r
670 // transforms the tagname into a tag identifier. If the tag is new, we insert
\r
671 // it in the table.
\r
672 for (i=0; i<ntagnames; i++)
\r
673 if (strcmp((const char *)tagname,(const char *)TagName[i])==0) break;
\r
675 if (i==ntagnames) { // the tag is a new one, then we insert it
\r
676 TagName = (unsigned char **)realloc(TagName, sizeof(char *)*(ntagnames+1));
\r
679 fprintf(stderr, "Error: not enough memory\n");
\r
684 TagName[i] = (unsigned char *)malloc(sizeof(unsigned char)*(strlen((const char *)tagname)+1));
\r
685 strcpy((char *)TagName[i], (const char *)tagname);
\r
688 tags_aux = (TagType *)realloc(tags_aux, sizeof(TagType)*(npar + 1));
\r
691 fprintf(stderr, "Error: not enough memory\n");
\r
695 tags_aux[npar] = i; // inserts the new tag id within the preorder sequence of tags
\r
704 // NewClosingTag(tagname): indicates the event of finding a new closing tag in the document.
\r
705 // Tag name is given. Returns a non-zero value upon success, and returns NULLT
\r
706 // in case of failing when trying to insert the new tag.
\r
707 int XMLTree::NewClosingTag(unsigned char *tagname)
\r
711 if (!initialized) { // data structure has not been initialized properly
\r
712 fprintf(stderr, "Error: you cannot insert a new closing tag without first calling method OpenDocument first\n");
\r
716 // inserts a new closing parentheses in the bit sequence
\r
717 par_aux = (pb *)realloc(par_aux, sizeof(pb)*(1+npar/(8*sizeof(pb))));
\r
719 fprintf(stderr, "Error: not enough memory\n");
\r
722 setbit(par_aux,npar,CP); // marks a new closing opening parenthesis
\r
724 // transforms the tagname into a tag identifier. If the tag is new, we insert
\r
725 // it in the table.
\r
726 for (i=0; i<ntagnames; i++)
\r
727 if (strcmp((const char *)tagname,(const char *)TagName[i])==0) break;
\r
729 if (i==ntagnames) { // the tag is a new one, then we insert it
\r
730 TagName = (unsigned char **)realloc(TagName, sizeof(char *)*(ntagnames+1));
\r
733 fprintf(stderr, "Error: not enough memory\n");
\r
738 TagName[i] = (unsigned char *)malloc(sizeof(char)*(strlen((const char *)tagname)+2));
\r
739 TagName[i][0] = '/';
\r
740 strcpy((char *)&(TagName[i][1]), (const char *)tagname);
\r
743 tags_aux = (TagType *)realloc(tags_aux, sizeof(TagType)*(npar + 1));
\r
746 fprintf(stderr, "Error: not enough memory\n");
\r
750 tags_aux[npar] = i; // inserts the new tag id within the preorder sequence of tags
\r
754 return 1; // success
\r
759 // NewText(s): indicates the event of finding a new (non-empty) text s in the document.
\r
760 // The new text is inserted within the text collection. Returns a non-zero value upon
\r
761 // success, NULLT in case of error.
\r
762 int XMLTree::NewText(unsigned char *s)
\r
764 if (!initialized) { // data structure has not been initialized properly
\r
765 fprintf(stderr, "Error: you cannot insert a new text without first calling method OpenDocument first\n");
\r
769 if (!indexing_empty_texts) {
\r
770 empty_texts_aux = (unsigned int *)realloc(empty_texts_aux, sizeof(pb)*(1+(npar-1)/(8*sizeof(pb))));
\r
771 if (!empty_texts_aux) {
\r
772 fprintf(stderr, "Error: not enough memory\n");
\r
776 bitset(empty_texts_aux, npar-1); // marks the non-empty text with a 1 in the bit vector
\r
779 Text->InsertText(s);
\r
781 return 1; // success
\r
784 // NewEmptyText(): indicates the event of finding a new empty text in the document.
\r
785 // In case of indexing empty and non-empty texts, we insert the empty texts into the
\r
786 // text collection. In case of indexing only non-empty texts, it just indicates an
\r
787 // empty text in the bit vector of empty texts. Returns a non-zero value upon
\r
788 // success, NULLT in case of error.
\r
789 int XMLTree::NewEmptyText()
\r
791 unsigned char c = 0;
\r
792 if (!initialized) { // data structure has not been initialized properly
\r
793 fprintf(stderr, "Error: you cannot insert a new empty text without first calling method OpenDocument first\n");
\r
797 if (!indexing_empty_texts) {
\r
798 empty_texts_aux = (unsigned int *)realloc(empty_texts_aux, sizeof(pb)*(1+(npar-1)/(8*sizeof(pb))));
\r
799 if (!empty_texts_aux) {
\r
800 fprintf(stderr, "Error: not enough memory\n");
\r
804 bitclean(empty_texts_aux, npar-1); // marks the empty text with a 0 in the bit vector
\r
806 else Text->InsertText(&c); // we insert the empty text just in case we index all the texts
\r
808 return 1; // success
\r
812 // GetTagId: returns the tag identifier corresponding to a given tag name.
\r
813 // Returns NULLT in case that the tag name does not exists.
\r
814 TagType XMLTree::GetTagId(unsigned char *tagname)
\r
817 // this should be changed for more efficient processing
\r
818 for (i=0; i<ntagnames; i++)
\r
819 if (strcmp((const char *)tagname,(const char *)TagName[i])==0) break;
\r
820 if (i==ntagnames) return (TagType)NULLT; // tagname does not exists in the table
\r
825 // GetTagName(tagid): returns the tag name of a given tag identifier.
\r
826 // Returns NULL in case that the tag identifier is not valid.
\r
827 unsigned char *XMLTree::GetTagName(TagType tagid)
\r
831 if (tagid >= ntagnames) return NULL; // invalid tag identifier
\r
832 s = (unsigned char *)malloc((strlen((const char *)TagName[tagid])+1)*sizeof(unsigned char));
\r
833 strcpy((char *)s, (const char *)TagName[tagid]);
\r