From 5b209950c7b0533b3082fcc2bde5fce1cf2fe632 Mon Sep 17 00:00:00 2001 From: nvalimak Date: Fri, 27 Mar 2009 15:36:47 +0000 Subject: [PATCH] Small fixes git-svn-id: svn+ssh://idea.nguyen.vg/svn/sxsi/trunk/TextCollection@297 3cdefd35-fc62-479d-8e8d-bae585ffb9ca --- TCImplementation.cpp | 320 ++++++++----------------------------------- TCImplementation.h | 55 ++++++-- TextCollection.h | 3 - 3 files changed, 100 insertions(+), 278 deletions(-) diff --git a/TCImplementation.cpp b/TCImplementation.cpp index 290ca9c..723a2b9 100644 --- a/TCImplementation.cpp +++ b/TCImplementation.cpp @@ -38,7 +38,7 @@ namespace SXSI { // Save file version info -const uchar TCImplementation::versionFlag = 3; +const uchar TCImplementation::versionFlag = 4; /** * Constructor inits an empty dynamic FM-index. @@ -46,20 +46,8 @@ const uchar TCImplementation::versionFlag = 3; */ TCImplementation::TCImplementation(uchar * bwt, ulong length, unsigned samplerate_, unsigned numberOfTexts_, ulong maxTextLength_) : n(length), samplerate(samplerate_), alphabetrank(0), sampled(0), suffixes(0), - suffixDocId(0), numberOfTexts(numberOfTexts_), numberOfAllTexts(numberOfTexts_), maxTextLength(maxTextLength_), endmarkerDocId(0), - emptyTextRank(0) + suffixDocId(0), numberOfTexts(numberOfTexts_), maxTextLength(maxTextLength_), Doc(0) { - - // Bitvector of empty texts, FIXME Remove? - { - //std::cout << std::endl << "texts: " << numberOfTexts << ", all texts " << numberOfAllTexts << ", size : " << emptyTextId.size() <::const_iterator it = emptyTextId.begin(); it != emptyTextId.end(); ++it) -// Tools::SetField(tempB, 1, (*it), 1); - emptyTextRank = new BSGAP(tempB, numberOfTexts, true); - } makewavelet(bwt); // Deletes bwt! bwt = 0; @@ -70,25 +58,12 @@ TCImplementation::TCImplementation(uchar * bwt, ulong length, unsigned samplerat bool TCImplementation::EmptyText(DocId k) const { assert(k < (DocId)numberOfTexts); - if (emptyTextRank->IsBitSet(k)) - return true; - return false; + return false; // Empty texts are not indexed } uchar* TCImplementation::GetText(DocId k) const { assert(k < (DocId)numberOfTexts); - - ulong textRank = 0; - if (emptyTextRank->IsBitSet(k, &textRank)) - { - uchar* result = new uchar[1]; - result[0] = 0; - return result; - } - // Map to non-empty text - k -= textRank; //emptyTextRank->rank(k); - TextPosition i = k; string result; @@ -121,15 +96,6 @@ uchar* TCImplementation::GetText(DocId k, TextPosition i, TextPosition j) const assert(i <= j); ulong textRank = 0; - if (emptyTextRank->IsBitSet(k, &textRank)) - { - uchar* result = new uchar[1]; - result[0] = 0; - return result; - } - - // Map to non-empty text - k -= textRank; // emptyTextRank->rank(k); // Start position of k'th text ulong start = (*textStartPos)[k]; @@ -191,11 +157,7 @@ bool TCImplementation::IsEqual(uchar const *pattern) const { TextPosition m = std::strlen((char *)pattern); if (m == 0) - { - if (numberOfAllTexts - numberOfTexts > 0u) - return true; // Empty texts exists return false; // No empty texts exists - } TextPosition sp = 0, ep = 0; // Match including end-marker @@ -211,11 +173,7 @@ bool TCImplementation::IsEqual(uchar const *pattern, DocId begin, DocId end) con { TextPosition m = std::strlen((char *)pattern); if (m == 0) - { - if (numberOfAllTexts - numberOfTexts > 0u) - return true; // Empty texts exists return false; // No empty texts exists - } TextPosition sp = 0, ep = 0; // Match including end-marker @@ -282,7 +240,7 @@ unsigned TCImplementation::CountPrefix(uchar const * pattern) const { TextPosition m = strlen((char *)pattern); if (m == 0) - return numberOfAllTexts; + return numberOfTexts; TextPosition sp = 0, ep = 0; Search(pattern, m, &sp, &ep); @@ -295,7 +253,7 @@ unsigned TCImplementation::CountPrefix(uchar const * pattern, DocId begin, DocId { TextPosition m = strlen((char *)pattern); if (m == 0) - return numberOfAllTexts; + return numberOfTexts; TextPosition sp = 0, ep = 0; Search(pattern, m, &sp, &ep); @@ -308,7 +266,7 @@ unsigned TCImplementation::CountSuffix(uchar const * pattern) const { TextPosition m = strlen((char *)pattern); if (m == 0) - return numberOfAllTexts; + return numberOfTexts; TextPosition sp = 0, ep = 0; // Search with end-marker @@ -321,7 +279,7 @@ unsigned TCImplementation::CountSuffix(uchar const * pattern, DocId begin, DocId { TextPosition m = strlen((char *)pattern); if (m == 0) - return numberOfAllTexts; + return numberOfTexts; TextPosition sp = 0, ep = 0; // Search with end-marker @@ -334,7 +292,7 @@ unsigned TCImplementation::CountEqual(uchar const *pattern) const { TextPosition m = strlen((char const *)pattern); if (m == 0) - return numberOfAllTexts - numberOfTexts; // Empty texts. + return 0; // No empty texts. TextPosition sp = 0, ep = 0; // Match including end-marker @@ -348,7 +306,7 @@ unsigned TCImplementation::CountEqual(uchar const *pattern, DocId begin, DocId e { TextPosition m = strlen((char const *)pattern); if (m == 0) - return numberOfAllTexts - numberOfTexts; // Empty texts. + return 0; // No empty texts. TextPosition sp = 0, ep = 0; // Match including end-marker @@ -362,7 +320,7 @@ unsigned TCImplementation::CountContains(uchar const * pattern) const { TextPosition m = strlen((char const *)pattern); if (m == 0) - return numberOfAllTexts; // Total number of texts. + return numberOfTexts; // Total number of texts. // Here counting is as slow as fetching the result set // because we have to filter out occ's that fall within same document. @@ -374,7 +332,7 @@ unsigned TCImplementation::CountContains(uchar const * pattern, DocId begin, Doc { TextPosition m = strlen((char const *)pattern); if (m == 0) - return numberOfAllTexts; // Total number of texts. + return numberOfTexts; // Total number of texts. // Here counting is as slow as fetching the result set // because we have to filter out occ's that fall within same document. @@ -387,7 +345,7 @@ unsigned TCImplementation::CountLessThan(uchar const * pattern) const { TextPosition m = strlen((char const *)pattern); if (m == 0) - return numberOfAllTexts - numberOfTexts; // Empty texts. + return 0; // No empty texts. TextPosition sp = 0, ep = 0; SearchLessThan(pattern, m, &sp, &ep); @@ -400,7 +358,7 @@ unsigned TCImplementation::CountLessThan(uchar const * pattern, DocId begin, Doc { TextPosition m = strlen((char const *)pattern); if (m == 0) - return numberOfAllTexts - numberOfTexts; // Empty texts. + return 0; // No empty texts. TextPosition sp = 0, ep = 0; SearchLessThan(pattern, m, &sp, &ep); @@ -431,22 +389,7 @@ TextCollection::document_result TCImplementation::Prefix(uchar const * pattern) result.reserve(resultSize); // Try to avoid reallocation. // Iterate through end-markers in [sp,ep]: - unsigned i = 0; - if (sp > 0) - i = alphabetrank->rank(0, sp - 1); - while (resultSize) - { - // End-marker we found belongs to the "preceeding" doc in the collection - DocId docId = ((*endmarkerDocId)[i] + 1) % numberOfTexts; - // Map to doc ID: - docId = emptyTextRank->select0(docId+1); - result.push_back(docId); - - -- resultSize; - ++ i; - } - - return result; + return EnumerateEndmarkers(sp, ep); } TextCollection::document_result TCImplementation::Prefix(uchar const * pattern, DocId begin, DocId end) const @@ -467,24 +410,8 @@ TextCollection::document_result TCImplementation::Prefix(uchar const * pattern, result.reserve(resultSize); // Try to avoid reallocation. - // Iterate through end-markers in [sp,ep] and [begin, end]: - unsigned i = 0; - if (sp > 0) - i = alphabetrank->rank(0, sp - 1); - while (resultSize) - { - // End-marker we found belongs to the "preceeding" doc in the collection - DocId docId = ((*endmarkerDocId)[i] + 1) % numberOfTexts; - // Map to doc ID: - docId = emptyTextRank->select0(docId+1); - if (docId >= begin && docId <= end) - result.push_back(docId); - - -- resultSize; - ++ i; - } - - return result; + // Return end-markers in [sp,ep] and [begin, end]: + return EnumerateEndmarkers(sp, ep, begin, end); } TextCollection::document_result TCImplementation::Suffix(uchar const * pattern) const @@ -518,18 +445,11 @@ TextCollection::document_result TCImplementation::Suffix(uchar const * pattern) { // Rank among the end-markers in BWT unsigned endmarkerRank = alphabetrank->rank(0, i) - 1; - - // End-marker that we found belongs to the "preceeding" doc in collection: - DocId docId = ((*endmarkerDocId)[endmarkerRank] + 1) % numberOfTexts; - // Map to doc ID: - docId = emptyTextRank->select0(docId+1); - result.push_back(docId); + result.push_back(Doc->access(endmarkerRank)); } else // Sampled position { DocId docId = (*suffixDocId)[sampled_rank_i-1]; //sampled->rank(i)-1]; - // Map to doc ID: - docId = emptyTextRank->select0(docId+1); result.push_back(docId); } } @@ -568,18 +488,11 @@ TextCollection::document_result TCImplementation::Suffix(uchar const * pattern, { // Rank among the end-markers in BWT unsigned endmarkerRank = alphabetrank->rank(0, i) - 1; - - // End-marker that we found belongs to the "preceeding" doc in collection: - DocId docId = ((*endmarkerDocId)[endmarkerRank] + 1) % numberOfTexts; - // Map to doc ID: - docId = emptyTextRank->select0(docId+1); - result.push_back(docId); + result.push_back(Doc->access(endmarkerRank)); } else // Sampled position { DocId docId = (*suffixDocId)[sampled_rank_i-1]; //sampled->rank(i)-1]; - // Map to doc ID: - docId = emptyTextRank->select0(docId+1); result.push_back(docId); } } @@ -601,28 +514,7 @@ TextCollection::document_result TCImplementation::Equal(uchar const *pattern) co TextCollection::document_result result; // Report end-markers in result interval - unsigned resultSize = CountEndmarkers(sp, ep); - if (resultSize == 0) - return result; - - result.reserve(resultSize); // Try to avoid reallocation. - - unsigned i = 0; - if (sp > 0) - i = alphabetrank->rank(0, sp - 1); - while (resultSize) - { - // End-marker we found belongs to the "previous" doc in the collection - DocId docId = ((*endmarkerDocId)[i] + 1) % numberOfTexts; - // Map to doc ID: - docId = emptyTextRank->select0(docId+1); - result.push_back(docId); - - -- resultSize; - ++ i; - } - - return result; + return EnumerateEndmarkers(sp, ep); } TextCollection::document_result TCImplementation::Equal(uchar const *pattern, DocId begin, DocId end) const @@ -638,28 +530,7 @@ TextCollection::document_result TCImplementation::Equal(uchar const *pattern, Do TextCollection::document_result result; // Report end-markers in result interval - unsigned resultSize = CountEndmarkers(sp, ep); - if (resultSize == 0) - return result; - - result.reserve(resultSize); // Try to avoid reallocation. - - unsigned i = 0; - if (sp > 0) - i = alphabetrank->rank(0, sp - 1); - while (resultSize) - { - // End-marker we found belongs to the "previous" doc in the collection - DocId docId = ((*endmarkerDocId)[i] + 1) % numberOfTexts; - // Map to doc ID: - docId = emptyTextRank->select0(docId+1); - result.push_back(docId); // already within [begin, end] - - -- resultSize; - ++ i; - } - - return result; + return EnumerateEndmarkers(sp, ep, begin, end); } @@ -691,10 +562,7 @@ TextCollection::document_result TCImplementation::Contains(uchar const * pattern { // Rank among the end-markers in BWT unsigned endmarkerRank = alphabetrank->rank(0, i) - 1; - - // End-marker that we found belongs to the "preceeding" doc in collection: - DocId docId = ((*endmarkerDocId)[endmarkerRank] + 1) % numberOfTexts; - resultSet.insert(docId); + resultSet.insert(Doc->access(endmarkerRank)); } else { @@ -706,9 +574,6 @@ TextCollection::document_result TCImplementation::Contains(uchar const * pattern // Convert std::set to std::vector TextCollection::document_result result(resultSet.begin(), resultSet.end()); - // Map to doc ID's - for (document_result::iterator it = result.begin(); it != result.end(); ++it) - *it = emptyTextRank->select0(*it+1); return result; } @@ -740,9 +605,7 @@ TextCollection::document_result TCImplementation::Contains(uchar const * pattern { // Rank among the end-markers in BWT unsigned endmarkerRank = alphabetrank->rank(0, i) - 1; - - // End-marker that we found belongs to the "preceeding" doc in collection: - DocId docId = ((*endmarkerDocId)[endmarkerRank] + 1) % numberOfTexts; + DocId docId = Doc->access(endmarkerRank); if (docId >= begin && docId <= end) resultSet.insert(docId); } @@ -757,9 +620,6 @@ TextCollection::document_result TCImplementation::Contains(uchar const * pattern // Convert std::set to std::vector TextCollection::document_result result(resultSet.begin(), resultSet.end()); - // Map to doc ID's - for (document_result::iterator it = result.begin(); it != result.end(); ++it) - *it = emptyTextRank->select0(*it+1); return result; } @@ -775,29 +635,7 @@ TextCollection::document_result TCImplementation::LessThan(uchar const * pattern TextCollection::document_result result; // Report end-markers in result interval - unsigned resultSize = CountEndmarkers(sp, ep); - if (resultSize == 0) - return result; - - result.reserve(resultSize); // Try to avoid reallocation. - - // Iterate through end-markers in [sp,ep]: - unsigned i = 0; - if (sp > 0) - i = alphabetrank->rank(0, sp - 1); - while (resultSize) - { - // End-marker we found belongs to the "preceeding" doc in the collection - DocId docId = ((*endmarkerDocId)[i] + 1) % numberOfTexts; - // Map to doc ID: - docId = emptyTextRank->select0(docId+1); - result.push_back(docId); - - -- resultSize; - ++ i; - } - - return result; + return EnumerateEndmarkers(sp, ep); } TextCollection::document_result TCImplementation::LessThan(uchar const * pattern, DocId begin, DocId end) const @@ -819,23 +657,7 @@ TextCollection::document_result TCImplementation::LessThan(uchar const * pattern result.reserve(resultSize); // Try to avoid reallocation. // Iterate through end-markers in [sp,ep] and [begin, end]: - unsigned i = 0; - if (sp > 0) - i = alphabetrank->rank(0, sp - 1); - while (resultSize) - { - // End-marker we found belongs to the "preceeding" doc in the collection - DocId docId = ((*endmarkerDocId)[i] + 1) % numberOfTexts; - // Map to doc ID: - docId = emptyTextRank->select0(docId+1); - if (docId >= begin && docId <= end) - result.push_back(docId); - - -- resultSize; - ++ i; - } - - return result; + return EnumerateEndmarkers(sp, ep, begin, end); } /** @@ -871,11 +693,7 @@ TextCollection::full_result TCImplementation::FullContains(uchar const * pattern { // Rank among the end-markers in BWT unsigned endmarkerRank = alphabetrank->rank(0, i) - 1; - - // End-marker that we found belongs to the "preceeding" doc in collection: - DocId docId = ((*endmarkerDocId)[endmarkerRank] + 1) % numberOfTexts; - // Map to doc ID: - docId = emptyTextRank->select0(docId+1); + DocId docId = Doc->access(endmarkerRank); result.push_back(make_pair(docId, dist)); } else @@ -884,8 +702,6 @@ TextCollection::full_result TCImplementation::FullContains(uchar const * pattern DocId docId = (*suffixDocId)[sampled_rank_i-1]; //sampled->rank(i)-1]; // textPos = textPos - (*textStartPos)[docId]; // Offset inside the text - // Map to doc ID: - docId = emptyTextRank->select0(docId+1); result.push_back(make_pair(docId, textPos)); } } @@ -925,9 +741,7 @@ TextCollection::full_result TCImplementation::FullContains(uchar const * pattern unsigned endmarkerRank = alphabetrank->rank(0, i) - 1; // End-marker that we found belongs to the "preceeding" doc in collection: - DocId docId = ((*endmarkerDocId)[endmarkerRank] + 1) % numberOfTexts; - // Map to doc ID: - docId = emptyTextRank->select0(docId+1); + DocId docId = Doc->access(endmarkerRank); if (docId >= begin && docId <= end) result.push_back(make_pair(docId, dist)); } @@ -937,8 +751,6 @@ TextCollection::full_result TCImplementation::FullContains(uchar const * pattern DocId docId = (*suffixDocId)[sampled_rank_i-1]; //sampled->rank(i)-1]; // textPos = textPos - (*textStartPos)[docId]; // Offset inside the text - // Map to doc ID: - docId = emptyTextRank->select0(docId+1); if (docId >= begin && docId <= end) result.push_back(make_pair(docId, textPos)); } @@ -964,10 +776,8 @@ TextCollection::full_result TCImplementation::FullContains(uchar const * pattern BlockArray *suffixes; BlockArray *suffixDocId; unsigned numberOfTexts; - unsigned numberOfAllTexts; ulong maxTextLength; - BlockArray *endmarkerDocId; - BSGAP *emptyTextRank; + static_sequence *docId; */ void TCImplementation::Save(FILE *file) const { @@ -994,13 +804,10 @@ void TCImplementation::Save(FILE *file) const if (std::fwrite(&(this->numberOfTexts), sizeof(unsigned), 1, file) != 1) throw std::runtime_error("TCImplementation::Save(): file write error (numberOfTexts)."); - if (std::fwrite(&(this->numberOfAllTexts), sizeof(unsigned), 1, file) != 1) - throw std::runtime_error("TCImplementation::Save(): file write error (numberOfAllTexts)."); if (std::fwrite(&(this->maxTextLength), sizeof(ulong), 1, file) != 1) throw std::runtime_error("TCImplementation::Save(): file write error (maxTextLength)."); - endmarkerDocId->Save(file); - emptyTextRank->Save(file); + Doc->save(file); fflush(file); } @@ -1013,8 +820,7 @@ void TCImplementation::Save(FILE *file) const */ TCImplementation::TCImplementation(FILE *file, unsigned samplerate_) : n(0), samplerate(samplerate_), alphabetrank(0), sampled(0), suffixes(0), - suffixDocId(0), numberOfTexts(0), numberOfAllTexts(0), maxTextLength(0), endmarkerDocId(0), - emptyTextRank(0) + suffixDocId(0), numberOfTexts(0), maxTextLength(0), Doc(0) { uchar verFlag = 0; if (std::fread(&verFlag, 1, 1, file) != 1) @@ -1044,13 +850,10 @@ TCImplementation::TCImplementation(FILE *file, unsigned samplerate_) if (std::fread(&(this->numberOfTexts), sizeof(unsigned), 1, file) != 1) throw std::runtime_error("TCImplementation::Load(): file read error (numberOfTexts)."); - if (std::fread(&(this->numberOfAllTexts), sizeof(unsigned), 1, file) != 1) - throw std::runtime_error("TCImplementation::Load(): file read error (numberOfAllTexts)."); if (std::fread(&(this->maxTextLength), sizeof(ulong), 1, file) != 1) throw std::runtime_error("TCImplementation::Load(): file read error (maxTextLength)."); - endmarkerDocId = new BlockArray(file); - emptyTextRank = new BSGAP(file); + Doc = static_sequence::load(file); // FIXME Construct data structures with new samplerate //maketables(); @@ -1063,38 +866,7 @@ TCImplementation::TCImplementation(FILE *file, unsigned samplerate_) */ -// FIXME Use 2D-search structure -unsigned TCImplementation::CountEndmarkers(TextPosition sp, TextPosition ep, DocId begin, DocId end) const -{ - if (sp > ep) - return 0; - - ulong ranksp = 0; - if (sp != 0) - ranksp = alphabetrank->rank(0, sp - 1); - - unsigned resultSize = alphabetrank->rank(0, ep) - ranksp; - if (resultSize == 0) - return 0; - // Count end-markers in result interval and within [begin, end] - unsigned count = 0; - unsigned i = ranksp; - while (resultSize) - { - // End-marker we found belongs to the "previous" doc in the collection - DocId docId = ((*endmarkerDocId)[i] + 1) % numberOfTexts; - // Map to doc ID: - docId = emptyTextRank->select0(docId+1); - if (docId >= begin && docId <= end) - ++ count; - - -- resultSize; - ++ i; - } - return count; -} - ulong TCImplementation::Search(uchar const * pattern, TextPosition m, TextPosition *spResult, TextPosition *epResult) const { // use the FM-search replacing function Occ(c,1,i) with alphabetrank->rank(c,i) @@ -1168,16 +940,11 @@ ulong TCImplementation::SearchLessThan(uchar const * pattern, TextPosition m, Te TCImplementation::~TCImplementation() { -#ifdef CSA_TEST_BWT - delete dynFMI; -#endif delete alphabetrank; delete sampled; delete suffixes; delete suffixDocId; - - delete endmarkerDocId; - delete emptyTextRank; + delete Doc; } void TCImplementation::makewavelet(uchar *bwt) @@ -1251,7 +1018,7 @@ void TCImplementation::maketables() BlockArray* tmpSuffix = new BlockArray(sampleLength, ceilLog2n); // Mapping from end-markers to doc ID's: - endmarkerDocId = new BlockArray(numberOfTexts, Tools::CeilLog2(numberOfTexts)); + BlockArray *endmarkerDocId = new BlockArray(numberOfTexts, Tools::CeilLog2(numberOfTexts)); ulong *sampledpositions = new ulong[n/W+1]; for (ulong i=0;iaccess(1, 2, 0, 1); + cout << "result: "; + for (document_result::iterator it = res.begin(); it != res.end(); ++it) + cout << *it << ", "; + cout << endl;*/ } diff --git a/TCImplementation.h b/TCImplementation.h index 5e488db..b74f705 100644 --- a/TCImplementation.h +++ b/TCImplementation.h @@ -43,8 +43,6 @@ namespace SXSI { -// Un-comment to compare BWT against a BWT generated from class dynFMI: -//#define CSA_TEST_BWT /** * Implementation of the TextCollection interface @@ -125,18 +123,12 @@ private: // Total number of texts in the collection unsigned numberOfTexts; - // Total number of texts including empty texts - unsigned numberOfAllTexts; // Length of the longest text ulong maxTextLength; // Array of document id's in the order of end-markers in BWT // Access by endmarkerDocId[rank_$(L, p) - 1]. - BlockArray *endmarkerDocId; - - // FIXME Replace with a more succinct data structure - // Note: Empty texts are already handled inside XMLTree class. - BSGAP *emptyTextRank; // FIXME Remove + static_sequence *Doc; // Following are not part of the public API uchar * BWT(uchar *); @@ -161,7 +153,50 @@ private: return alphabetrank->rank(0, ep) - ranksp; } - unsigned CountEndmarkers(TextPosition, TextPosition, DocId, DocId) const; + + /** + * Count end-markers in given interval and + * within docIds [min,max] + */ + inline unsigned CountEndmarkers(TextPosition sp, TextPosition ep, DocId min, DocId max) const + { + if (sp != 0) + sp = alphabetrank->rank(0, sp - 1); + ep = alphabetrank->rank(0, ep); + if (ep == 0) + return 0; + + return Doc->count(sp, ep-1, min, max); + } + + /** + * Enumerate all end-markers in given interval + */ + inline document_result EnumerateEndmarkers(TextPosition sp, TextPosition ep) const + { + if (sp != 0) + sp = alphabetrank->rank(0, sp - 1); + ep = alphabetrank->rank(0, ep); + if (ep == 0) + return document_result(); + + return Doc->accessAll(sp, ep-1); + } + + /** + * Enumerate end-markers in given interval and + * within docIds [min,max] + */ + inline document_result EnumerateEndmarkers(TextPosition sp, TextPosition ep, DocId min, DocId max) const + { + if (sp != 0) + sp = alphabetrank->rank(0, sp - 1); + ep = alphabetrank->rank(0, ep); + if (ep == 0) + return document_result(); + + return Doc->access(sp, ep-1, min, max); + } }; // class TCImplementation } // namespace SXSI diff --git a/TextCollection.h b/TextCollection.h index dad5c88..1e34ab0 100644 --- a/TextCollection.h +++ b/TextCollection.h @@ -25,9 +25,6 @@ #include #include // Defines std::pair. -// Default samplerate for suffix array samples -#define TEXTCOLLECTION_DEFAULT_SAMPLERATE 64 - namespace SXSI { -- 2.17.1