along/secondo
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
dfd1e745480fabd88139d2804acb90d5b6dc055e
secondo/Algebras/SemanticTrajectory/SemanticTrajectoryAlgebra.cpp
along dfd1e74548 firs commit
2026-01-23 17:03:45 +08:00

7409 lines
169 KiB
C++
Raw Blame History

/*
----
This file is part of SECONDO.
Copyright (C) 2004, University in Hagen, Department of Computer Science,
Database Systems for New Applications.
SECONDO is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
SECONDO is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with SECONDO; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
----
//paragraph [1] title: [{\Large \bf ] [}]
[1] SemanticTrajectory Algebra
Author: Catherine Higgins
1 Overview
Briefly describe here what this algebra is for.
2 Defines, includes, and constants
*/
#include "SemanticTrajectoryAlgebra.h"
#include "Algebra.h"
#include "NestedList.h"
#include "QueryProcessor.h"
#include "StandardTypes.h"
#include "Attribute.h"
#include "Symbols.h"
#include "ListUtils.h"
#include "Algebras/Relation-C++/RelationAlgebra.h"
#include "Algebras/Rectangle/CellGrid.h"
#include "Stream.h"
#include "Algebras/FText/FTextAlgebra.h"
extern NestedList* nl;
extern QueryProcessor *qp;
using std::cout;
using std::endl;
namespace semtraj {
/*
3 Implementation of class SemanticTrajectory
3.1 Constructors and Destructor
*/
/*
Non-stardard constructor must be used when using Flobs or DbArrays
Initialize each type of Flob to 0
*/
SemanticTrajectory::SemanticTrajectory(int dummy):
Attribute(true),
coordinates(0),
semantics(0),
semantics_Flob(0),
cells(0),
words(0),
words_Flob(0),
bbox(false)
{
}
SemanticTrajectory::
SemanticTrajectory(const SemanticTrajectory& st) :
Attribute(st.IsDefined()),
coordinates(st.coordinates.Size()),
semantics(st.semantics.Size()),
semantics_Flob(st.semantics_Flob.getSize()),
cells(st.cells.Size()),
words(st.words.Size()),
words_Flob(st.words_Flob.getSize()),
bbox(st.bbox)
{
bool success = false;
success = semantics.Append(st.semantics);
assert(success);
success = coordinates.Append(st.coordinates);
assert(success);
success = semantics_Flob.copyFrom(st.semantics_Flob);
assert(success);
success = cells.Append(st.cells);
assert(success);
success = words.Append(st.words);
success = words_Flob.copyFrom(st.words_Flob);
assert(success);
bbox = st.bbox;
// Maybe add here bbox is defined ?
// Is that part of the rectangle class
}
SemanticTrajectory::~SemanticTrajectory()
{
}
SemanticTrajectory& SemanticTrajectory::
operator=(const SemanticTrajectory& st)
{
SetDefined(st.IsDefined());
bool success = false;
success = semantics.clean();
assert(success);
success = semantics.copyFrom(st.semantics);
assert(success);
success = coordinates.clean();
assert(success);
success = coordinates.copyFrom(st.coordinates);
assert(success);
success = semantics_Flob.clean();
assert(success);
success = semantics_Flob.copyFrom(st.semantics_Flob);
bbox = st.bbox;
success = cells.clean();
assert(success);
success = cells.Append(st.cells);
assert(success);
success = words.clean();
assert(success);
success = words.Append(st.words);
success = words_Flob.clean();
assert(success);
success = words_Flob.copyFrom(st.words_Flob);
return *this;
}
/*
3.2 Functions implementing virtual
function from class
*/
/*
i == 0 return coordinates
i == 1 return st\_TextData
i == 2 return st\_TextFlob
i == 3 returns cells
i == 4 returns words
*/
Flob *SemanticTrajectory::GetFLOB(const int i)
{
Flob* stFlob = 0;
if (i == 0) {
stFlob = &coordinates;
} else if (i == 1) {
stFlob = &semantics;
} else if (i == 2) {
stFlob = &semantics_Flob;
} else if (i == 3) {
stFlob = &cells;
} else if (i == 4) {
stFlob = &words;
} else if (i == 5)
{
stFlob = &words_Flob;
}
return stFlob;
}
void SemanticTrajectory::CopyFrom(
const Attribute* right)
{
if(right != 0)
{
const SemanticTrajectory* st =
static_cast<
const SemanticTrajectory*>(right);
if(st != 0)
{
*this = *st;
}
}
}
SemanticTrajectory*
SemanticTrajectory::Clone() const
{
SemanticTrajectory *st =
new SemanticTrajectory(*this);
assert(st != 0);
return st;
}
/*
3.3 The mandatory set of algebra support functions
3.3.1 Out-function
*/
/*
List represenation
TODO UPDATE THIS INFO AS IT IS NOT UP TO DATE
((id xl xr yl yr)...(id xl xr yl yr))
Nested list of two: first is a 4 coordinates to represent a rectangle
*/
ListExpr
SemanticTrajectory::Out(
ListExpr typeInfo,
Word value )
{
// the addr pointer of value
SemanticTrajectory* st =
(SemanticTrajectory*)(value.addr);
if (!st->IsDefined()) {
return nl->SymbolAtom(Symbol::UNDEFINED());
}
if( st->IsEmpty() )
{
return (nl->TheEmptyList());
}
else
{
std::string holdValue;
bool success = st->GetSemString(0, holdValue);
if (success == false) {
holdValue = "";
}
ListExpr result =
nl->OneElemList(
nl->ThreeElemList(
nl->RealAtom(st->GetCoordinate(0).x),
nl->RealAtom(st->GetCoordinate(0).y),
nl->TextAtom(holdValue)));
ListExpr last = result;
for( int i = 1; i < st->GetNumCoordinates(); i++ )
{
bool success = st->GetSemString(i, holdValue);
if (success == false) {
holdValue = "";
}
last = nl->Append(last,
nl->ThreeElemList(
nl->RealAtom(st->GetCoordinate(i).x),
nl->RealAtom(st->GetCoordinate(i).y),
nl->TextAtom(holdValue)));
}
ListExpr textual = nl->Empty();
ListExpr spatial = nl->Empty();
/* Need to retrieve spatial information if any */
if (!st->IsEmptySpatialSum())
{
spatial =
nl->OneElemList(
nl->ThreeElemList(
nl->IntAtom(st->GetCell(0).tripId),
nl->IntAtom(st->GetCell(0).origx),
nl->IntAtom(st->GetCell(0).origy)
)
);
ListExpr spatiallast = spatial;
for( int i = 1; i < st->GetNumCells(); i++ )
{
spatiallast = nl->Append(spatiallast,
nl->ThreeElemList(
nl->IntAtom(st->GetCell(i).tripId),
nl->IntAtom(st->GetCell(i).origx),
nl->IntAtom(st->GetCell(i).origy)
)
);
}
}
if (!st->IsEmptyTextualSum())
{
std::string holdValue;
bool success = st->GetStringSum(0, holdValue);
if (success == false) {
holdValue = "";
}
textual =
nl->OneElemList(
nl->ThreeElemList(
nl->IntAtom(st->GetWord(0).indexId),
nl->IntAtom(st->GetWord(0).count),
nl->TextAtom(holdValue)
)
);
ListExpr textuallast = textual;
for( int i = 1; i < st->GetNumWords(); i++ )
{
bool success = st->GetStringSum(i, holdValue);
if (success == false) {
holdValue = "";
}
textuallast = nl->Append(textuallast,
nl->ThreeElemList(
nl->IntAtom(st->GetWord(i).indexId),
nl->IntAtom(st->GetWord(i).count),
nl->TextAtom(holdValue)
)
);
}
}
ListExpr final =
nl->FourElemList(
nl->FourElemList(
nl->RealAtom(st->bbox.getMinX()),
nl->RealAtom(st->bbox.getMinY()),
nl->RealAtom(st->bbox.getMaxX()),
nl->RealAtom(st->bbox.getMaxY())
),
spatial,
textual,
result
);
return final;
}
}
/*
3.3.2 In-function
*/
/*
List represenation
Nested list of four:
first nested list: Represents bounding box which is a 4 coordinates
to represent a rectangle
second nested list: Represents the spatial summary
(this can be expected to be empty until operator is makespatialsum
is called)
third nested list: Represents the textual summary
(this can be expected to be empty until operator is called)
fourth nested list: Represents the trajectory data
*/
/* The bouding box get's calculated everytime */
Word
SemanticTrajectory::In(
const ListExpr typeInfo,
const ListExpr instance,
const int errorPos,
ListExpr& errorInfo,
bool& correct )
{
Word word;
correct = false;
SemanticTrajectory* st = new SemanticTrajectory(0);
st->SetDefined(true);
// This is if list is undefined
// ~ setDefined to false
if(listutils::isSymbolUndefined( instance ))
{
st->SetDefined(false);
correct = true;
word.addr = st;
return word;
}
if (nl != 0)
{
int nListLength = nl->ListLength(instance);
// If list has 4 arguments
// and undefined content
//~ setDefined to false
if (nListLength == 4 &&
nl->IsAtom(nl->Fourth(instance)) &&
nl->AtomType(nl->Fourth(instance)) == SymbolType &&
listutils::isSymbolUndefined(nl->Fourth(instance)))
{
correct = true;
st->SetDefined(false);
correct = true;
word.addr = st;
return word;
}
if (nListLength == 4)
{
ListExpr first = nl->Empty();
ListExpr rest = nl->Fourth(instance);
double mind[2];
double maxd[2];
double c_x1;
double c_y1;
double c_x2;
double c_y2;
/* TODO Add check here to make sure
list rest is at least two points
to be valid semantic trajectory (maybe)*/
while (nl->IsEmpty(rest) == false)
{
first = nl->First( rest );
rest = nl->Rest( rest );
if( nl->ListLength(first) == 3 &&
nl->IsAtom(nl->First(first)) &&
nl->AtomType(nl->First(first))
== RealType &&
nl->IsAtom(nl->Second(first)) &&
nl->AtomType(nl->Second(first))
== RealType &&
nl->IsAtom(nl->Third(first)))
{
double x1 = nl->RealValue(nl->First(first));
double y1 = nl->RealValue(nl->Second(first));
Coordinate c(x1, y1);
st->AddCoordinate(c);
bool typeString =
nl->AtomType(nl->Third(first)) == StringType;
bool typeText =
nl->AtomType(nl->Third(first)) == TextType;
std::string stringValue;
if (typeString == true)
{
stringValue =
nl->StringValue(nl->Third(first));
}
if (typeText == true)
{
stringValue =
nl->TextValue(nl->Third(first));
}
st->AddSemString(stringValue);
/* Set bounding box here */
if (st->GetNumCoordinates() == 1)
{
c_x1 = x1;
c_y1 = y1;
c_x2 = x1;
c_y2 = y1;
}
else
{
c_x1 = std::min(c_x1, x1);
c_y1 = std::min(c_y1, y1);
c_x2 = std::max(c_x2, x1);
c_y2 = std::max(c_y2, y1);
}
}
else
{
correct = false;
delete st;
return SetWord( Address(0) );
}
if (c_x1 <= c_x2 && c_y1 <= c_y2)
{
mind[0] = c_x1;
mind[1] = c_y1;
maxd[0] = c_x2;
maxd[1] = c_y2;
st->bbox.Set(true, mind, maxd);
}
}
/* TODO add extra checks and conditions */
/* Extract info for spatial summary if any */
if (!listutils::isSymbolUndefined(
nl->Second(instance)))
{
ListExpr first = nl->Empty();
ListExpr rest = nl->Second(instance);
while (nl->IsEmpty(rest) == false)
{
first = nl->First( rest );
rest = nl->Rest( rest );
if( nl->ListLength(first) == 3 &&
nl->IsAtom(nl->First(first)) &&
nl->AtomType(nl->First(first))
== IntType &&
nl->IsAtom(nl->Second(first)) &&
nl->AtomType(nl->Second(first))
== IntType &&
nl->IsAtom(nl->Third(first)) &&
nl->AtomType(nl->Third(first))
== IntType
)
{
Cell c(nl->IntValue(nl->First(first)),
nl->IntValue(nl->Second(first)),
nl->IntValue(nl->Third(first))
);
st->AddCell(c);
}
else
{
correct = false;
delete st;
return SetWord( Address(0) );
}
} // end of while
} // end of if
/* TODO add extra checks and conditions */
/* Extract info for textual summary if any */
if (!listutils::isSymbolUndefined(nl->Third(instance)))
{
ListExpr first = nl->Empty();
ListExpr rest = nl->Third(instance);
while (nl->IsEmpty(rest) == false)
{
first = nl->First( rest );
rest = nl->Rest( rest );
if( nl->ListLength(first) == 3 &&
nl->IsAtom(nl->First(first)) &&
nl->AtomType(nl->First(first))
== IntType &&
nl->IsAtom(nl->Second(first)) &&
nl->AtomType(nl->Second(first))
== IntType && nl->IsAtom(nl->Third(first))
)
{
bool typeString =
nl->AtomType(nl->Third(first)) == StringType;
std::string stringValue;
if (typeString == true)
{
stringValue =
nl->StringValue(nl->Third(first));
} else {
stringValue =
nl->TextValue(nl->Third(first));
}
st->AddStringSum(stringValue,
nl->IntValue(nl->First(first)),
nl->IntValue(nl->Second(first)));
}
else
{
correct = false;
delete st;
return SetWord( Address(0) );
}
}
}
} // end of if n ===4
correct = true;
return SetWord(st);
} // end of n != 0
return word;
}
/*
3.3.3 Property function
*/
ListExpr
SemanticTrajectory::Property()
{
return
(nl->TwoElemList(
nl->FiveElemList(
nl->StringAtom("Signature"),
nl->StringAtom("Example Type List"),
nl->StringAtom("List Rep"),
nl->StringAtom("Example List"),
nl->StringAtom("Remarks")),
nl->FiveElemList(nl->StringAtom(
"->" + Kind::DATA()
),
nl->StringAtom(
SemanticTrajectory::BasicType()
),
nl->TextAtom(
"((minx miny maxx maxy)"
"((tripid xl1 xr1 yl1 yr1)"
"(tripid xl xr yl yr))"
"()(<x1> <y1> <Text1>)..."
"(<xn> <yn> <Text2>))"),
nl->TextAtom(
"((3.2 15.4 6.34 15.4)"
"()()"
"((3.2 15.4 text1)"
"(6.34 20.8 text2)))"),
nl->TextAtom(
"x- and y-coordinates"
" must be "
"of type real."))));
}
/*
3.3.4 Kind Checking Function
*/
bool
SemanticTrajectory::KindCheck(
ListExpr type,
ListExpr& errorInfo )
{
return (nl->IsEqual(
type,
SemanticTrajectory::BasicType() ));
}
/*
3.3.5 ~Create~-function
*/
Word SemanticTrajectory::Create(
const ListExpr typeInfo)
{
SemanticTrajectory* st = new SemanticTrajectory(0);
return (SetWord(st));
}
/*
3.3.6 ~Destroy~-function
*/
void SemanticTrajectory::Destroy()
{
semantics.Destroy();
semantics_Flob.destroy();
coordinates.Destroy();
cells.Destroy();
words.Destroy();
words_Flob.destroy();
}
/*
3.3.7 ~Delete~-function
*/
void SemanticTrajectory::Delete(
const ListExpr typeInfo,
Word& w)
{
SemanticTrajectory* st =
(SemanticTrajectory*)w.addr;
st->Destroy();
delete st;
}
/*
3.3.8 ~Open~-function
*/
bool
SemanticTrajectory::Open(
SmiRecord& valueRecord,
size_t& offset,
const ListExpr typeInfo,
Word& value )
{
SemanticTrajectory *st =
(SemanticTrajectory*)Attribute::Open(
valueRecord, offset, typeInfo );
value.setAddr( st );
return true;
}
/*
3.3.9 ~Save~-function
*/
bool
SemanticTrajectory::Save( SmiRecord& valueRecord,
size_t& offset,
const ListExpr typeInfo,
Word& value )
{
SemanticTrajectory *st =
(SemanticTrajectory *)value.addr;
Attribute::Save( valueRecord, offset, typeInfo, st );
return true;
}
/*
3.3.10 ~Close~-function
*/
void SemanticTrajectory::Close(
const ListExpr typeInfo,
Word& w)
{
SemanticTrajectory* st = (SemanticTrajectory*)w.addr;
delete st;
}
/*
3.3.11 ~Clone~ funtion
*/
Word SemanticTrajectory::Clone(
const ListExpr typeInfo,
const Word& rWord)
{
Word word;
SemanticTrajectory* pSemanticTrajectory =
static_cast<SemanticTrajectory*>(rWord.addr);
if (pSemanticTrajectory != 0)
{
word.addr =
new SemanticTrajectory(*pSemanticTrajectory);
assert(word.addr != 0);
}
return word;
}
/*
3.3.12 ~SizeOf~-function
*/
int SemanticTrajectory::SizeOfObj()
{
return sizeof(SemanticTrajectory);
}
/*
3.3.13 ~Cast~-function
*/
void* SemanticTrajectory::Cast(void* addr)
{
return (new (addr) SemanticTrajectory);
}
/*
3.4 Implementation of new functions
3.4.1 ~AddCoordinate~ function
Responsible for adding a <x,y> Coordinate object
to DbArray coordinates
*/
void SemanticTrajectory::
AddCoordinate(
const Coordinate& c)
{
coordinates.Append(c);
}
/*
3.4.2 ~GetCoordinate~ function
Retrieve a Coordinate from the DbArray coordinates
*/
Coordinate SemanticTrajectory::
GetCoordinate( int i ) const
{
assert( 0 <= i && i < GetNumCoordinates() );
Coordinate c;
coordinates.Get( i, &c );
return c;
}
/*
3.4.3 ~AddSemString~ function
*/
bool SemanticTrajectory::AddSemString(const std::string& stString)
{
/*
Important that even if a string is Empty
that the information is stored in the
semantics DbArray. If not this might
impact the position of the string in the Out:: function
*/
TextData td;
td.Offset = semantics_Flob.getSize();
td.Length = stString.length();
bool success = semantics.Append(td);
assert(success);
if (!stString.empty())
{
semantics_Flob.write(
stString.c_str(),
td.Length,
td.Offset);
}
return true;
}
/*
3.4.4 ~GetSemString~ function
*/
bool SemanticTrajectory::
GetSemString(int index, std::string& rString) const
{
bool success = false;
int numString = semantics.Size();
if (index < numString)
{
TextData textData;
success = semantics.Get(index, &textData);
if (success == true)
{
SmiSize bufferSize = textData.Length + 1;
char* pBuffer = new char[bufferSize];
if (pBuffer != 0)
{
memset(
pBuffer,
0,
bufferSize * sizeof(char));
success = semantics_Flob.read(
pBuffer,
textData.Length,
textData.Offset);
}
if(success == true)
{
rString = pBuffer;
}
delete [] pBuffer;
}
}
return success;
}
/*
3.4.5 ~AddStringSum~ function
*/
/*
Each trajectory has textual summary which consists of all the
individual unique words found in one semantic SemanticTrajectory
This function adds a string to the Flob X and string info to DbArray X.
*/
bool SemanticTrajectory::AddStringSum(
const std::string& stString, int id, int count)
{
if (!stString.empty())
{
WordST td;
td.Offset = words_Flob.getSize();
td.Length = stString.length();
td.indexId = id;
td.count = count;
bool success = words.Append(td);
assert(success);
words_Flob.write(
stString.c_str(),
td.Length,
td.Offset);
return true;
}
return false;
}
/*
3.4.6 ~GetStringSum~ function
*/
bool SemanticTrajectory::
GetStringSum(int index, std::string& rString) const
{
bool success = false;
int numString = words.Size();
if (index < numString)
{
WordST textData;
success = words.Get(index, &textData);
if (success == true)
{
SmiSize bufferSize = textData.Length + 1;
char* pBuffer = new char[bufferSize];
if (pBuffer != 0)
{
memset(
pBuffer,
0,
bufferSize * sizeof(char));
success = words_Flob.read(
pBuffer,
textData.Length,
textData.Offset);
}
if(success == true)
{
rString = pBuffer;
}
delete [] pBuffer;
}
}
return success;
}
/*
3.4.7 GetStringArray function
*/
std::list<std::string> SemanticTrajectory::GetStringArray() const
{
std::list<std::string> stringArray;
int nStrings = semantics.Size();
bool bOK = false;
std::string holdString;
for(int i = 0; i < nStrings; i++)
{
bOK = GetSemString(i, holdString);
stringutils::StringTokenizer parse(holdString, " ");
if (bOK == true)
{
while(parse.hasNextToken())
{
std::string eval = parse.nextToken();
stringutils::trim(eval);
if (eval.length() > 0)
{
stringArray.push_back(eval);
}
}
}
}
stringArray.sort(SemanticTrajectory::compare_nocase);
return stringArray;
}
/*
3.4.8 ~AddCell~ function
TODO
*/
void SemanticTrajectory::
AddCell(const Cell& c)
{
cells.Append(c);
}
/*
3.4.9 GetCellList function
TODO
*/
DbArray<Cell> SemanticTrajectory::
GetCellList() const
{
assert(IsDefined());
return cells;
}
/*
3.4.9 ~GetTextSumList~ function
TODO
*/
DbArray<WordST> SemanticTrajectory::
GetTextSumList() const
{
assert(IsDefined());
return words;
}
/*
3.5 Operator helper functions
3.5.1 ~Relevance~ Function
*/
double SemanticTrajectory::Relevance(int i,
std::string& str,
SemanticTrajectory& st,
double diag, double alpha)
{
// Return the max of all return similarity equation
double max = -999999; /*TODO use a max varialble */
int numOfCoor2 = st.GetNumCoordinates();
for (int y = 0; y <numOfCoor2; y++)
{
double temp = Sim(i, y, str, st, diag, alpha);
if (temp > max)
{
max = temp;
}
}
return max;
}
/*
3.5.2 ~Sim~ Function
TODO
*/
double SemanticTrajectory::Sim(int i, int y,
std::string& str, SemanticTrajectory& st,
double diag, double alpha)
{
double dist = SpatialDist(i, y, st);
double normalizedscore = 1 - (double)(dist/diag);
// It's inversly proportional to the
// distance so we need to substract from 1
double LS = alpha * normalizedscore;
double textscore = 0.0;
if (str.length() > 0)
{
textscore = TextualScore(str, y, st);
}
double RH = double (1 - alpha) * textscore;
return LS + RH;
}
/*
3.5.3 ~GetDiagonal~ Function
TODO
*/
double SemanticTrajectory::GetDiagonal(
Rectangle<2>& rec)
{
double x1 = rec.getMinX();
double y1 = rec.getMinY();
double x2 = rec.getMaxX();
double y2 = rec.getMaxY();
double result = sqrt(pow((x1 - x2),2)
+ pow((y1 - y2),2));
return result;
}
/*
3.5.4 ~GetDiag~ Function
TODO
*/
double GetDiag(
Rectangle<2>& rec)
{
double x1 = rec.getMinX();
double y1 = rec.getMinY();
double x2 = rec.getMaxX();
double y2 = rec.getMaxY();
double result = sqrt(pow((x1 - x2),2)
+ pow((y1 - y2),2));
return result;
}
/*
3.5.5 ~SpatialDist~ Function
TODO
*/
double SemanticTrajectory::SpatialDist(int i, int y,
SemanticTrajectory& st)
{
double x1 = GetCoordinate(i).x;
double y1 = GetCoordinate(i).y;
double x2 = st.GetCoordinate(y).x;
double y2 = st.GetCoordinate(y).y;
double result =
sqrt(pow((x1 - x2),2)
+ pow((y1 - y2),2));
return result;
}
/*
3.5.6 TextualScore Function
Used with operator sim to return the textual score between
two trajectories
*/
double SemanticTrajectory::TextualScore(std::string& str, int y,
SemanticTrajectory& st)
{
bool success = false;
std::string s2;
success = st.GetSemString(y, s2);
assert(success);
if (str.length() > 0 && s2.length() > 0)
{
int numMatches = 0;
stringutils::StringTokenizer st1(str, " ");
stringutils::StringTokenizer st2(s2, " ");
int numToken1 = str.length();
int numToken2 = s2.length();
std::string eval = "";
std::string eval2 = "";
bool noToks1 = false;
bool noToks2 = false;
if (st1.hasNextToken())
{
eval = st1.nextToken();
} else {
noToks1 = true;
}
if (st2.hasNextToken())
{
eval2 = st2.nextToken();
} else {
noToks2 = true;
}
while(!noToks2 && !noToks1)
{
if((eval).compare(eval2) == 0)
{
numMatches++;
if (st1.hasNextToken())
{
eval = st1.nextToken();
} else {
noToks1 = true;
}
if (st2.hasNextToken())
{
eval2 = st2.nextToken();
} else {
noToks2 = true;
}
}
else if ((eval).compare(eval2) < 0) {
if (st1.hasNextToken())
{
eval = st1.nextToken();
} else {
noToks1 = true;
}
} else {
if (st2.hasNextToken())
{
eval2 = st2.nextToken();
}
else
{
noToks2 = true;
}
}
}
if (numMatches == 0.0)
{
return 0.0;
}
double uniquewords = (double)
(numToken2 + numToken1 - numMatches);
double result = (double) numMatches / uniquewords;
return result;
}
return 0.0;
}
/*
3.5.7 Placename Function
TODO
*/
double SemanticTrajectory::Similarity(
SemanticTrajectory& st,
double diag,
double alpha)
{
int numCoordinates1 = GetNumCoordinates();
double sumOfRelevance1 = 0;
double leftTotal = 0;
bool success = false;
std::string str;
for (int i = 0; i < numCoordinates1; i++)
{
success = GetSemString(i, str);
assert(success);
sumOfRelevance1 =
sumOfRelevance1 + Relevance(i, str, st, diag, alpha);
}
leftTotal =
sumOfRelevance1 / (double) numCoordinates1;
int numCoordinates2 = st.GetNumCoordinates();
double sumOfRelevance2 = 0;
double rightTotal = 0;
for (int i = 0; i < numCoordinates2; i++)
{
success = st.GetSemString(i, str);
assert(success);
sumOfRelevance2 =
sumOfRelevance2 + st.Relevance(
i, str, *this, diag, alpha);
}
rightTotal =
sumOfRelevance2 / (double) numCoordinates2;
double result = rightTotal + leftTotal;
return result;
}
/*
3.5.8 Placename Function
TODO
*/
double SemanticTrajectory::textualScoreTT(SemanticTrajectory& st2)
{
std::vector<std::string*>** hashTable;
std::vector<std::string*>** hashTable2;
const std::vector<std::string*>* bucket;
std::hash<std::string> str_hash;
unsigned int bucketnum = 100;
hashTable = new std::vector<std::string*>*[bucketnum];
for (unsigned int i = 0; i < bucketnum; i++)
{
hashTable[i] = 0;
}
hashTable2 = new std::vector<std::string*>*[bucketnum];
for (unsigned int i = 0; i < bucketnum; i++)
{
hashTable2[i] = 0;
}
for (int i = 0; i < GetNumWords(); i++)
{
std::string holdValue = "";
bool success = GetStringSum(i, holdValue);
if(success)
{
size_t hash = str_hash(holdValue) % bucketnum;
if (!hashTable[hash])
{
hashTable[hash] = new std::vector<std::string*>();
}
std::string* stn = new std::string(holdValue);
hashTable[hash]->push_back(stn);
}
}
for (int i = 0; i < st2.GetNumWords(); i++)
{
std::string holdValue = "";
bool success = st2.GetStringSum(i, holdValue);
if(success)
{
size_t hash = str_hash(holdValue) % bucketnum;
if (!hashTable2[hash])
{
hashTable2[hash] = new std::vector<std::string*>();
}
std::string* stn = new std::string(holdValue);
hashTable2[hash]->push_back(stn);
}
}
double TSim = 0.0;
unsigned int bucketpos = 0;
for(int i = 0; i < GetNumCoordinates(); i++)
{
std::string holdvalue;
GetSemString(i, holdvalue);
if (holdvalue.length() > 0)
{
stringutils::StringTokenizer parse_st1(holdvalue, " ");
while(parse_st1.hasNextToken())
{
std::string eval = parse_st1.nextToken();
stringutils::trim(eval);
size_t hash = str_hash(eval) % bucketnum;;
bucket = 0;
bucket = hashTable2[hash];
bucketpos = 0;
if(bucket)
{
while(bucketpos < bucket->size())
{
std::string* check = (*bucket)[bucketpos];
if ((check)->compare(eval) == 0)
{
TSim = TSim + ((double) 1/GetNumCoordinates());
break;
}
bucketpos++;
}
}
}
}
}
for(int i = 0; i < st2.GetNumTextData(); i++)
{
std::string holdvalue;
st2.GetSemString(i, holdvalue);
if (holdvalue.length() > 0)
{
stringutils::StringTokenizer parse_st2(holdvalue, " ");
while(parse_st2.hasNextToken())
{
std::string eval = parse_st2.nextToken();
stringutils::trim(eval);
size_t hash = str_hash(eval) % bucketnum;
bucket = 0;
bucket = hashTable[hash];
bucketpos = 0;
if(bucket)
{
while(bucketpos < bucket->size())
{
std::string* check = (*bucket)[bucketpos];
if ((check)->compare(eval) == 0)
{
TSim = TSim + ((double) 1/st2.GetNumCoordinates());
break;
}
bucketpos++;
}
}
}
}
}
for (unsigned int i = 0; i< bucketnum; i++)
{
std::vector<std::string*>* v = hashTable[i];
if (v)
{
for(unsigned int j = 0; j < v->size(); j++)
{
std::string* stn = (*v)[j];
delete stn;
}
(v)->clear();
delete hashTable[i];
hashTable[i] = 0;
}
}
for (unsigned int i = 0; i< bucketnum; i++)
{
std::vector<std::string*>* v = hashTable2[i];
if (v)
{
for(unsigned int j = 0; j < v->size(); j++)
{
std::string* stn = (*v)[j];
delete stn;
}
(v)->clear();
delete hashTable2[i];
hashTable2[i] = 0;
}
}
delete hashTable; delete hashTable2;
return TSim;
}
/*
3.5.9 Placename Function
TODO
*/
double SemanticTrajectory::MinDistAux(
SemanticTrajectory& st2, double wx, double wy)
{
DbArray<Cell>* Ax = new DbArray<Cell>(0);
DbArray<Cell> valuesSt1 = GetCellList();
DbArray<Cell> valuesSt2 = st2.GetCellList();
Ax->Append(valuesSt1);
Ax->Append(valuesSt2);
DbArray<Cell>* Ay = new DbArray<Cell>(0);
Ay->Append(valuesSt1);
Ay->Append(valuesSt2);
Ax->Sort(SemanticTrajectory::CompareX);
Ay->Sort(SemanticTrajectory::CompareY);
int32_t m = Ax->Size();
double result = 0.0;
result = MinDistUtils(*Ax, *Ay, m, wx, wy);
delete Ay;
delete Ax;
return result;
}
/*
3.5.10 MinDistUtils Function
*/
double SemanticTrajectory::MinDistUtils(
DbArray<Cell>& Ax,
DbArray<Cell>& Ay, int32_t m, double wx, double wy)
{
double minD = DBL_MAX;
if (m > 3)
{
int n = m/2;
DbArray<Cell>* AxL = new DbArray<Cell>(0);
Ax.copyTo(*AxL, 0, n, 0);
DbArray<Cell>* AxR = new DbArray<Cell>(0);
Ax.copyTo(*AxR, n, m-n, 0);
DbArray<Cell>* AyL = new DbArray<Cell>(0);
DbArray<Cell>* AyR = new DbArray<Cell>(0);
double minDL = 0.0;
double minDR = 0.0;
for (int i = 0; i < Ay.Size(); i++)
{
Cell celly, cellx;
Ay.Get(i, celly);
Ax.Get(n, cellx);
if(celly.GetX() <= cellx.GetX())
{
AyL->Append(celly);
}
else
{
AyR->Append(celly);
}
}
/* Check to see if AyL has
cells from both C1 and C2 */
if (AyL->Size() != 0){
Cell cellAyL;
AyL->Get(0, cellAyL);
int32_t firstId = cellAyL.GetId();
bool both = false;
for (int i = 1; i < AyL->Size(); i++)
{
AyL->Get(i, cellAyL);
if (cellAyL.GetId() != firstId)
{
both = true;
break;
}
}
if (both == true)
{
minDL = MinDistUtils(*AxL, *AyL, n, wx, wy);
}
else
{
minDL = DBL_MAX;
}
} else {
minDL = DBL_MAX;
}
if (AyR->Size() != 0)
{
Cell cellAyR;
AyR->Get(0, cellAyR);
int32_t firstId = cellAyR.GetId();
bool both = false;
for (int i = 1; i < AyR->Size(); i++)
{
AyR->Get(i, cellAyR);
if (cellAyR.GetId() != firstId)
{
both = true;
break;
}
}
if (both == true)
{
minDR = MinDistUtils(*AxR, *AyR, m-n, wx, wy);
}
else
{
minDR = DBL_MAX;
}
}
else {
minDR = DBL_MAX;
}
minD = minDL > minDR ? minDR : minDL;
DbArray<Cell>* Am = new DbArray<Cell>(0);
for (int i = 0; i < Ay.Size(); i++)
{
Cell x;
Cell y;
Ax.Get(n, x);
Ay.Get(i, y);
if (fabs(y.GetX() - x.GetX()) < minD)
{
Am->Append(y);
}
}
for (int i = 0; i < Am->Size(); i++)
{
for (int j = i + 1; j < Am->Size(); j++)
{
Cell c1;
Cell c2;
Am->Get(i, c1);
Am->Get(j, c2);
if (c1.GetId() != c2.GetId())
{
double tempmin = GetCellDist(c1,c2, wx, wy);
if (tempmin < minD)
{
minD = tempmin;
}
}
}
}
//Don't forget to delete the DbArrays
// that were initialized here
delete Am;
delete AyR;
delete AyL;
delete AxR;
delete AxL;
}
else
{
minD = BruteForce(Ax, m, wx, wy);
}
return minD;
}
/*
3.5.11 GetCellDist Function
TODO
*/
double SemanticTrajectory::GetCellDist(
Cell& c1, Cell& c2, double wx, double wy)
{
double result;
//Are they the same cell
if (c1.GetX() == c2.GetX()
&& c1.GetY() == c2.GetY())
{
return 0.0;
}
//c1 and c2 on the same x Axis
if (c1.GetX() == c2.GetX())
{
// if c1 is above c2
// Take bottom corner of c1 and top corner of c2
if(c1.GetY() > c2.GetY())
{
int32_t y2 = (int32_t)(c2.GetY() + 1);
return
EuclidDist(c1.GetX(), c1.GetY(),
c2.GetX(), y2, wx, wy);
}
// if c1 is below c2
// Take top corner of c1 and bottom corner of c2
else
{
int32_t y1 = (int32_t)(c1.GetY() + 1);
return
EuclidDist(c1.GetX(),
y1,
c2.GetX(),
c2.GetY(), wx, wy);
}
}
//c1 and c2 on the same y axis
if (c1.GetY() == c2.GetY())
{
// if c1 is to the right of c2
if(c1.GetX() > c2.GetX())
{
int32_t x2 = (int32_t)(c2.GetX() + 1);
return
EuclidDist(c1.GetX(), c1.GetY(),
x2, c2.GetY(), wx, wy);
}
// if c1 is to the left of c2
else
{
int32_t x1 = (int32_t)(c1.GetX() + 1);
return
EuclidDist(x1,
c1.GetY(),
c2.GetX(), c2.GetY(), wx, wy);
}
}
// if c1 is above c2
if (c1.GetY() > c2.GetY())
{
// if c1 is to the right of c2
if (c1.GetX() > c2.GetX())
{
int32_t x2 = (int32_t)(c2.GetX() + 1);
int32_t y2 = (int32_t)(c2.GetY() + 1);
return
EuclidDist(c1.GetX(),
c1.GetY(),
x2,
y2, wx, wy);
}
// if c1 is to the left c2
//
else
{
int32_t x1 = (int32_t)(c1.GetX() + 1);
int32_t y2 = (int32_t)(c2.GetY() + 1);
return
EuclidDist(x1, c1.GetY(),
c2.GetX(), y2, wx, wy);
}
}
// if c1 is below c2
else
{
// if c1 is to the right of c2
if (c1.GetX() > c2.GetX())
{
int32_t x1 = (c1.GetX() + 1);
int32_t x2 = (c2.GetX() + 1);
return
EuclidDist(
x1,
c1.GetY(),
x2,
c2.GetY(), wx, wy);
}
// if c1 is to the left c2
else
{
int32_t x1 = (c1.GetX() + 1);
int32_t y1 = (c1.GetY() + 1);
return
EuclidDist(
x1,y1,
c2.GetX(),
c2.GetY(), wx, wy);
}
}
return result;
}
/*
3.5.12 BruteForce Function
TODO
*/
double SemanticTrajectory::BruteForce(
DbArray<Cell>& Ax,
int32_t m, double wx, double wy)
{
double minD = DBL_MAX;
for (int i = 0; i < Ax.Size(); i++)
{
for (int j = i + 1; j < Ax.Size(); j++)
{
Cell c1;
Cell c2;
Ax.Get(i, c1);
Ax.Get(j, c2);
if (c1.GetId() != c2.GetId())
{
double tempmin = GetCellDist(c1,c2, wx, wy);
if (tempmin < minD)
{
minD = tempmin;
}
}
}
}
return minD;
}
/*
3.5.13 Placename Function
TODO
*/
double SemanticTrajectory::EuclidDist(int32_t x1,
int32_t y1,
int32_t x2,
int32_t y2, double wx, double wy) const
{
double x11 = (double) x1*wx;
double x21 = (double) x2*wx;
double y11 = (double) y1*wy;
double y21 = (double) y2*wy;
return sqrt(pow((x11 - x21),2) + pow((y11 - y21),2));
}
/*
4 Implementation of class Batch
4.1 Constructors and Destructor
*/
Batch::Batch(int dummy):
Attribute(true),
batchwords(0),
batchwords_Flob(0),
trips(0),
bcells(0),
bwords(0),
bwords_Flob(0),
bcoordinates(0),
bsemantics(0),
bsemantics_Flob(0),
bbox(false),
batchId(0)
{
}
Batch::
Batch(const Batch& b) :
Attribute(b.IsDefined()),
batchwords(b.batchwords.Size()),
batchwords_Flob(b.batchwords_Flob.getSize()),
trips(b.trips.Size()),
bcells(b.bcells.Size()),
bwords(b.bwords.Size()),
bwords_Flob(b.bwords_Flob.getSize()),
bcoordinates(b.bcoordinates.Size()),
bsemantics(b.bsemantics.Size()),
bsemantics_Flob(b.bsemantics_Flob.getSize()),
bbox(b.bbox),
batchId(b.batchId)
{
bool success = false;
success =
batchwords.Append(b.batchwords);
assert(success);
success =
batchwords_Flob.copyFrom(b.batchwords_Flob);
assert(success);
success = trips.Append(b.trips);
assert(success);
success = bcells.Append(b.bcells);
assert(success);
success = bwords.Append(b.bwords);
assert(success);
success =
bwords_Flob.copyFrom(b.bwords_Flob);
assert(success);
success =
bcoordinates.Append(b.bcoordinates);
assert(success);
success = bsemantics.Append(b.bsemantics);
assert(success);
success =
bsemantics_Flob.copyFrom(b.bsemantics_Flob);
bbox = b.bbox;
}
Batch::~Batch()
{
}
Batch& Batch::
operator=(const Batch& b)
{
SetDefined(b.IsDefined());
bool success = false;
success = batchwords.clean();
assert(success);
success =
batchwords.copyFrom(b.batchwords);
assert(success);
success = batchwords_Flob.clean();
assert(success);
success =
batchwords_Flob.copyFrom(b.batchwords_Flob);
assert(success);
success = trips.clean();
assert(success);
success = trips.copyFrom(b.trips);
assert(success);
success = bcells.clean();
assert(success);
success = bcells.copyFrom(b.bcells);
assert(success);
success = bwords.clean();
assert(success);
success = bwords.copyFrom(b.bwords);
assert(success);
success = bwords_Flob.clean();
assert(success);
success =
bwords_Flob.copyFrom(b.bwords_Flob);
assert(success);
success = bcoordinates.clean();
assert(success);
success =
bcoordinates.copyFrom(b.bcoordinates);
assert(success);
success = bsemantics.clean();
assert(success);
success =
bsemantics.copyFrom(b.bsemantics);
assert(success);
success = bsemantics_Flob.clean();
assert(success);
success =
bsemantics_Flob.copyFrom(b.bsemantics_Flob);
bbox = b.bbox;
batchId = b.batchId;
return *this;
}
/*
4.2 Implementation of virtual functions
*/
Flob *Batch::GetFLOB(const int i)
{
Flob* stFlob = 0;
if (i == 0) {
stFlob = &batchwords;
} else if (i == 1) {
stFlob = &batchwords_Flob;
} else if (i == 2) {
stFlob = &trips;
} else if (i == 3) {
stFlob = &bcells;
} else if (i == 4) {
stFlob = &bwords;
} else if (i == 5) {
stFlob = &bwords_Flob;
} else if (i == 6) {
stFlob = &bcoordinates;
} else if (i ==7) {
stFlob = &bsemantics;
} else if (i == 8) {
stFlob = &bsemantics_Flob;
}
return stFlob;
}
void Batch::CopyFrom(
const Attribute* right)
{
if(right != 0)
{
const Batch* b =
static_cast<
const Batch*>(right);
if(b != 0)
{
*this = *b;
}
}
}
Batch*
Batch::Clone() const
{
Batch *b =
new Batch(*this);
assert(b != 0);
return b;
}
/*
4.3 The mandatory set of algebra support functions
4.3.1 Placename holder
*/
ListExpr
Batch::Property()
{
return
(nl->TwoElemList(
nl->FiveElemList(
nl->StringAtom("Signature"),
nl->StringAtom("Example Type List"),
nl->StringAtom("List Rep"),
nl->StringAtom("Example List"),
nl->StringAtom("Remarks")),
nl->FiveElemList(nl->StringAtom(
"->" + Kind::DATA()
),
nl->StringAtom(
Batch::BasicType()
),
nl->TextAtom(
"((batchId)(minx miny maxx maxy)"
"((1 word1 ctn)"
"(2 word2 ctn)...)"
"((st1) (st2)...))"),
nl->TextAtom(
"((1)(3.2 15.4 6.34 15.4)"
"(1 diversity 2)(2 bear 3)"
"((st data type)"
"(st datatype)))"),
nl->TextAtom(
"batchid, mbr,"
" wordlist, "
"triplist.")
)));
}
/*
4.3.2 Placename holder
*/
bool
Batch::KindCheck(
ListExpr type,
ListExpr& errorInfo )
{
return (nl->IsEqual(
type,
Batch::BasicType() ));
}
/*
4.3.3 Placename holder
*/
Word Batch::Create(
const ListExpr typeInfo)
{
Batch* b = new Batch(0);
return (SetWord(b));
}
/*
4.3.4 Placename holder
*/
void Batch::Destroy()
{
batchwords.Destroy();
batchwords_Flob.destroy();
trips.Destroy();
bcells.Destroy();
bwords.Destroy();
bwords_Flob.destroy();
bcoordinates.Destroy(),
bsemantics.Destroy(),
bsemantics_Flob.destroy();
}
/*
4.3.5 Placename holder
*/
void Batch::Delete(
const ListExpr typeInfo,
Word& w)
{
Batch* b =
(Batch*)w.addr;
b->Destroy();
delete b;
}
/*
4.3.6 Placename holder
*/
bool
Batch::Open(
SmiRecord& valueRecord,
size_t& offset,
const ListExpr typeInfo,
Word& value )
{
Batch *b =
(Batch*)Attribute::Open(
valueRecord, offset, typeInfo );
value.setAddr( b );
return true;
}
/*
4.3.7 Placename holder
*/
bool
Batch::Save( SmiRecord& valueRecord,
size_t& offset,
const ListExpr typeInfo,
Word& value )
{
Batch *b =
(Batch *)value.addr;
Attribute::Save( valueRecord, offset, typeInfo, b );
return true;
}
/*
4.3.8 Placename holder
*/
void Batch::Close(
const ListExpr typeInfo,
Word& w)
{
Batch* b = (Batch*)w.addr;
delete b;
}
/*
4.3.9 Placename holder
*/
Word Batch::Clone(
const ListExpr typeInfo,
const Word& rWord)
{
Word word;
Batch* p =
static_cast<Batch*>(rWord.addr);
if (p != 0)
{
word.addr =
new Batch(*p);
assert(word.addr != 0);
}
return word;
}
/*
4.3.10 Placename holder
*/
int Batch::SizeOfObj()
{
return sizeof(Batch);
}
/*
4.3.11 Placename holder
*/
void* Batch::Cast(void* addr)
{
return (new (addr) Batch);
}
/*
4.3.12 Placename holder
*/
ListExpr
Batch::Out(
ListExpr typeInfo,
Word value )
{
// the addr pointer of value
Batch* b =
(Batch*)(value.addr);
if ( !b->IsDefined() ) {
return nl->SymbolAtom(Symbol::UNDEFINED());
}
if( b->IsEmpty() )
{
return (nl->TheEmptyList());
}
else
{
ListExpr textual = nl->Empty();
/* Need to retrieve equivalent of WordList */
if (!b->IsEmptyWordList())
{
std::string holdValue;
bool success = b->GetBSumString(0, holdValue);
if (success == false) {
holdValue = "";
}
textual =
nl->OneElemList(
nl->ThreeElemList(
nl->IntAtom(b->GetBSumWord(0).indexId),
nl->IntAtom(b->GetBSumWord(0).count),
nl->TextAtom(holdValue)
)
);
ListExpr textuallast = textual;
for( int i = 1; i < b->GetWordListSize(); i++ )
{
bool success = b->GetBSumString(i, holdValue);
if (success == false) {
holdValue = "";
}
textuallast = nl->Append(textuallast,
nl->ThreeElemList(
nl->IntAtom(b->GetBSumWord(i).indexId),
nl->IntAtom(b->GetBSumWord(i).count),
nl->TextAtom(holdValue)
)
);
}
}
/*
Retrive the tripList
*/
ListExpr result;
if (!b->IsEmptyTripList())
{
Trip t = b->GetTrip(0);
/* Retrieve cell information */
int getStartIdx = t.GetStartCellsIdx();
int count = getStartIdx;
int getEndIdx = t.GetEndCellsIdx();
ListExpr spatial = nl->Empty();
if (getStartIdx < getEndIdx)
{
spatial =
nl->OneElemList(
nl->ThreeElemList(
nl->IntAtom(b->GetBCell(count).tripId),
nl->IntAtom(b->GetBCell(count).origx),
nl->IntAtom(b->GetBCell(count).origy)
)
);
ListExpr spatiallast = spatial;
count++;
for( int i = count; i < getEndIdx; i++ )
{
spatiallast = nl->Append(spatiallast,
nl->ThreeElemList(
nl->IntAtom(b->GetBCell(i).tripId),
nl->IntAtom(b->GetBCell(i).origx),
nl->IntAtom(b->GetBCell(i).origy)
)
);
}
} // end of cell information
/* Retrieve Trip textual summary */
int getStartIdx1 = t.GetStartSumWordsIdx();
int getEndIdx1 = t.GetEndSumWordsIdx();
ListExpr tripsum = nl->Empty();
if (getStartIdx1 < getEndIdx1)
{
std::string holdValue;
// Equivalent of GetStingSum
bool success = b->GetBWord(getStartIdx1, holdValue);
if (success == false) {
holdValue = "";
}
tripsum =
nl->OneElemList(
nl->ThreeElemList(
nl->IntAtom(b->GetBWordInfo(getStartIdx1).indexId),
nl->IntAtom(b->GetBWordInfo(getStartIdx1).count),
nl->TextAtom(holdValue)
)
);
ListExpr tripsumlast = tripsum;
for( int i = getStartIdx1 + 1; i < getEndIdx1; i++ )
{
std::string holdValue;
// Equivalent of GetStingSum
bool success = b->GetBWord(i, holdValue);
if (success == false) {
holdValue = "";
}
tripsumlast = nl->Append(tripsumlast,
nl->ThreeElemList(
nl->IntAtom(b->GetBWordInfo(i).indexId),
nl->IntAtom(b->GetBWordInfo(i).count),
nl->TextAtom(holdValue)
)
);
}
} // End of Trip textsummary information
/* Retrive Trip coordinates and Semantics */
int getStartIdx2 = t.GetStartCoordsIdx();
int count2 = getStartIdx2;
int getEndIdx2 = t.GetEndCoordsIdx();
ListExpr tripdata = nl->Empty();
if (getStartIdx2 < getEndIdx2)
{
std::string holdValue;
bool success = b->GetBSemString(count2, holdValue);
if (success == false) {
holdValue = "";
}
tripdata = nl->OneElemList(
nl->ThreeElemList(
nl->RealAtom(b->GetBCoordinate(count2).x),
nl->RealAtom(b->GetBCoordinate(count2).y),
nl->TextAtom(holdValue)));
ListExpr tlast = tripdata;
count2++;
for( int i = count2; i < getEndIdx2; i++ )
{
std::string holdValue;
bool success = b->GetBSemString(i, holdValue);
if (success == false) {
holdValue = "";
}
tlast = nl->Append(tlast,
nl->ThreeElemList(
nl->RealAtom(b->GetBCoordinate(i).x),
nl->RealAtom(b->GetBCoordinate(i).y),
nl->TextAtom(holdValue)));
}
} // End of Trip data
result = nl->OneElemList(
nl->TwoElemList(
nl->OneElemList(
nl->IntAtom(t.GetId())),
nl->FourElemList(
nl->FourElemList(
nl->RealAtom(t.bbox.getMinX()),
nl->RealAtom(t.bbox.getMinY()),
nl->RealAtom(t.bbox.getMaxX()),
nl->RealAtom(t.bbox.getMaxY())
),
spatial,
tripsum,
tripdata
)
)
);
ListExpr last = result;
for (int i = 1; i < b->GetNumTrips(); i++)
{
Trip t = b->GetTrip(i);
/* Get Cell Information */
int getStartIdx = t.GetStartCellsIdx();
int count3 = getStartIdx;
int getEndIdx = t.GetEndCellsIdx();
ListExpr spatial = nl->Empty();
if (getStartIdx < getEndIdx)
{
spatial =
nl->OneElemList(
nl->ThreeElemList(
nl->IntAtom(b->GetBCell(count3).tripId),
nl->IntAtom(b->GetBCell(count3).origx),
nl->IntAtom(b->GetBCell(count3).origy)
)
);
ListExpr spatiallast = spatial;
count3++;
for( int i = count3; i < getEndIdx; i++ )
{
spatiallast = nl->Append(spatiallast,
nl->ThreeElemList(
nl->IntAtom(b->GetBCell(i).tripId),
nl->IntAtom(b->GetBCell(i).origx),
nl->IntAtom(b->GetBCell(i).origy)
)
);
}
} // End of getting cell informtion
int getStartIdx1 = t.GetStartSumWordsIdx();
int count1 = getStartIdx1;
int getEndIdx1 = t.GetEndSumWordsIdx();
ListExpr tripsum = nl->Empty();
if (getStartIdx1 < getEndIdx1)
{
std::string holdValue;
bool success = b->GetBWord(count1, holdValue);
if (success == false) {
holdValue = "";
}
tripsum =
nl->OneElemList(
nl->ThreeElemList(
nl->IntAtom(b->GetBWordInfo(count1).indexId),
nl->IntAtom(b->GetBWordInfo(count1).count),
nl->TextAtom(holdValue)
)
);
ListExpr tripsumlast = tripsum;
count1++;
for( int i = count1; i < getEndIdx1; i++ )
{
std::string holdValue;
// Equivalent of GetStingSum
bool success = b->GetBWord(i, holdValue);
if (success == false) {
holdValue = "";
}
tripsumlast = nl->Append(tripsumlast,
nl->ThreeElemList(
nl->IntAtom(b->GetBWordInfo(i).indexId),
nl->IntAtom(b->GetBWordInfo(i).count),
nl->TextAtom(holdValue)
)
);
}
} // End of Trip information
/* Retrive Trip coordinates and Semantics */
int getStartIdx2 = t.GetStartCoordsIdx();
int count2 = getStartIdx2;
int getEndIdx2 = t.GetEndCoordsIdx();
ListExpr tripdata = nl->Empty();
if (getStartIdx2 < getEndIdx2)
{
std::string holdValue;
bool success = b->GetBSemString(count2, holdValue);
if (success == false) {
holdValue = "";
}
tripdata =
nl->OneElemList(
nl->ThreeElemList(
nl->RealAtom(b->GetBCoordinate(count2).x),
nl->RealAtom(b->GetBCoordinate(count2).y),
nl->TextAtom(holdValue)));
ListExpr tlast = tripdata;
count2++;
for( int i = count2; i < getEndIdx2; i++ )
{
std::string holdValue;
bool success = b->GetBSemString(i, holdValue);
if (success == false) {
holdValue = "";
}
tlast = nl->Append(tlast,
nl->ThreeElemList(
nl->RealAtom(b->GetBCoordinate(i).x),
nl->RealAtom(b->GetBCoordinate(i).y),
nl->TextAtom(holdValue)));
}
} // End of Trip data
last = nl->Append(last,
nl->TwoElemList(
nl->OneElemList(
nl->IntAtom(t.GetId())),
nl->FourElemList(
nl->FourElemList(
nl->RealAtom(t.bbox.getMinX()),
nl->RealAtom(t.bbox.getMinY()),
nl->RealAtom(t.bbox.getMaxX()),
nl->RealAtom(t.bbox.getMaxY())
),
spatial,
tripsum,
tripdata
)
)
);
}
}
ListExpr final = nl->FourElemList(
nl->OneElemList(
nl->IntAtom(b->GetBatchId())),
nl->FourElemList(
nl->RealAtom(b->bbox.getMinX()),
nl->RealAtom(b->bbox.getMinY()),
nl->RealAtom(b->bbox.getMaxX()),
nl->RealAtom(b->bbox.getMaxY())
),
textual,
result
);
return final;
}
}
/*
4.3.13 Placename holder
*/
/*
List represenation
Nested list of 4:
first nl: Represents the batchId
second nested list: Represents bounding box which is a 4 coordinates
to represent a rectangle
third nested list: Represents the wordlist
fourth nested list: Represents the triplist
*/
Word
Batch::In( const ListExpr typeInfo,
const ListExpr instance,
const int errorPos,
ListExpr& errorInfo,
bool& correct )
{
Word word;
correct = false;
Batch* b = new Batch(0);
b->SetDefined(true);
// This is if list is undefined
// ~ setDefined to false
if(listutils::isSymbolUndefined( instance ))
{
b->SetDefined(false);
correct = true;
word.addr = b;
return word;
}
if (nl != 0)
{
int nListLength = nl->ListLength(instance);
// If list has 4 arguments
// and undefined content
//~ setDefined to false
if (nListLength == 4 &&
nl->IsAtom(nl->Fourth(instance)) &&
nl->AtomType(nl->Fourth(instance)) == SymbolType &&
listutils::isSymbolUndefined(nl->Fourth(instance)))
{
correct = true;
b->SetDefined(false);
correct = true;
word.addr = b;
return word;
}
if (nListLength == 4)
{
/* Collect BatchId */
if (!listutils::isSymbolUndefined(nl->First(instance)))
{
ListExpr first = nl->First(instance);
if( nl->ListLength(first) == 1 &&
nl->IsAtom(nl->First(first)) &&
nl->AtomType(nl->First(first))
== IntType)
{
b->batchId = nl->IntValue(nl->First(first));
}
}
/* Extract info for bbox */
if (!listutils::isSymbolUndefined(nl->Second(instance)))
{
ListExpr first = nl->Second(instance);
double mind[2];
double maxd[2];
if( nl->ListLength(first) == 4 &&
nl->IsAtom(nl->First(first)) &&
nl->AtomType(nl->First(first))
== RealType &&
nl->IsAtom(nl->Second(first)) &&
nl->AtomType(nl->Second(first))
== RealType &&
nl->IsAtom(nl->Third(first)) &&
nl->AtomType(nl->Third(first))
== RealType &&
nl->IsAtom(nl->Fourth(first)) &&
nl->AtomType(nl->Fourth(first))
== RealType
)
{
mind[0] = nl->RealValue(nl->First(first));
mind[1] = nl->RealValue(nl->Second(first));
maxd[0] = nl->RealValue(nl->Third(first));
maxd[1] = nl->RealValue(nl->Fourth(first));
b->bbox.Set(true, mind, maxd);
}
}
/* Extract info for textual summary if any */
if (!listutils::isSymbolUndefined(nl->Third(instance)))
{
ListExpr first = nl->Empty();
ListExpr rest = nl->Third(instance);
while (nl->IsEmpty(rest) == false)
{
first = nl->First( rest );
rest = nl->Rest( rest );
if( nl->ListLength(first) == 3 &&
nl->IsAtom(nl->First(first)) &&
nl->AtomType(nl->First(first))
== IntType &&
nl->IsAtom(nl->Second(first)) &&
nl->AtomType(nl->Second(first))
== IntType && nl->IsAtom(nl->Third(first))
)
{
bool typeString =
nl->AtomType(nl->Third(first)) == StringType;
std::string stringValue;
if (typeString == true)
{
stringValue =
nl->StringValue(nl->Third(first));
} else {
stringValue =
nl->TextValue(nl->Third(first));
}
b->AddBSumString(stringValue,
nl->IntValue(nl->First(first)),
nl->IntValue(nl->Second(first)));
}
else
{
correct = false;
delete b;
return SetWord( Address(0) );
}
}
}
/* Extract info for TripList */
if (!listutils::isSymbolUndefined(nl->Fourth(instance)))
{
ListExpr first = nl->Empty();
ListExpr rest = nl->Fourth(instance);
int cellidxcounter = 0;
int tripsumidxcounter = 0;
int tripdataidxcounter = 0;
/* This is the while loop for each trip */
while (nl->IsEmpty(rest) == false)
{
first = nl->First( rest );
rest = nl->Rest( rest );
Trip t(0);
if(!listutils::isSymbolUndefined(nl->First(first)))
{
/* Get first part of two element list */
ListExpr ok = nl->First(first);
t.SetId(nl->IntValue(nl->First(ok)));
}
/*
Extract second part of two element list
Second part consists of 4 parts: BBox, Cell, TextSum, DayTrip
*/
if(!listutils::isSymbolUndefined(nl->Second(first)))
{
ListExpr ok = nl->Second(first);
if( nl->ListLength(ok) == 4)
{
/* Extract Bbox information */
if (!listutils::isSymbolUndefined(nl->First(ok)))
{
ListExpr box = nl->First(ok);
double mind[2];
double maxd[2];
if( nl->ListLength(box) == 4 &&
nl->IsAtom(nl->First(box)) &&
nl->AtomType(nl->First(box))
== RealType &&
nl->IsAtom(nl->Second(box)) &&
nl->AtomType(nl->Second(box))
== RealType &&
nl->IsAtom(nl->Third(box)) &&
nl->AtomType(nl->Third(box))
== RealType &&
nl->IsAtom(nl->Fourth(box)) &&
nl->AtomType(nl->Fourth(box))
== RealType
)
{
mind[0] = nl->RealValue(nl->First(box));
mind[1] = nl->RealValue(nl->Second(box));
maxd[0] = nl->RealValue(nl->Third(box));
maxd[1] = nl->RealValue(nl->Fourth(box));
t.SetBoundingBox(true, mind, maxd);
}
}
/* Extract the Cell info */
if (!listutils::isSymbolUndefined(
nl->Second(ok)))
{
ListExpr first1 = nl->Empty();
ListExpr rest1 = nl->Second(ok);
t.SetStartCellsIdx(cellidxcounter);
while (nl->IsEmpty(rest1) == false)
{
first1 = nl->First( rest1 );
rest1 = nl->Rest( rest1 );
if( nl->ListLength(first1) == 3 &&
nl->IsAtom(nl->First(first1)) &&
nl->AtomType(nl->First(first1))
== IntType &&
nl->IsAtom(nl->Second(first1)) &&
nl->AtomType(nl->Second(first1))
== IntType &&
nl->IsAtom(nl->Third(first1)) &&
nl->AtomType(nl->Third(first1))
== IntType
)
{
Cell c(nl->IntValue(nl->First(first1)),
nl->IntValue(nl->Second(first1)),
nl->IntValue(nl->Third(first1))
);
b->AddBCell(c);
cellidxcounter++;
}
else
{
correct = false;
delete b;
return SetWord( Address(0) );
}
} // end of while
t.SetEndCellsIdx(cellidxcounter);
} // end of if
/* Extract the Textual */
if (!listutils::isSymbolUndefined(
nl->Third(ok)))
{
ListExpr f1 = nl->Empty();
ListExpr r1 = nl->Third(ok);
t.SetStartSumWordsIdx(tripsumidxcounter);
while (nl->IsEmpty(r1) == false) {
f1 = nl->First( r1 );
r1 = nl->Rest( r1 );
if( nl->ListLength(f1) == 3 &&
nl->IsAtom(nl->First(f1)) &&
nl->AtomType(nl->First(f1))
== IntType &&
nl->IsAtom(nl->Second(f1)) &&
nl->AtomType(nl->Second(f1))
== IntType && nl->IsAtom(nl->Third(f1))
)
{
bool typeString =
nl->AtomType(nl->Third(f1)) == StringType;
std::string stringValue;
if (typeString == true)
{
stringValue =
nl->StringValue(nl->Third(f1));
} else {
stringValue =
nl->TextValue(nl->Third(f1));
}
b->AddBWord(stringValue,
nl->IntValue(nl->First(f1)),
nl->IntValue(nl->Second(f1)));
tripsumidxcounter++;
}
else
{
correct = false;
delete b;
return SetWord( Address(0) );
}
}
t.SetEndSumWordsIdx(tripsumidxcounter);
}
// /* Extract the DayTrip */
if (!listutils::isSymbolUndefined(
nl->Fourth(ok)))
{
ListExpr f2 = nl->Empty();
ListExpr r2 = nl->Fourth(ok);
t.SetStartCoordsIdx(tripdataidxcounter);
while (nl->IsEmpty(r2) == false)
{
f2 = nl->First( r2 );
r2 = nl->Rest( r2 );
if( nl->ListLength(f2) == 3 &&
nl->IsAtom(nl->First(f2)) &&
nl->AtomType(nl->First(f2))
== RealType &&
nl->IsAtom(nl->Second(f2)) &&
nl->AtomType(nl->Second(f2))
== RealType &&
nl->IsAtom(nl->Third(f2)))
{
double x1 = nl->RealValue(nl->First(f2));
double y1 = nl->RealValue(nl->Second(f2));
Coordinate c(x1, y1);
b->AddBCoordinate(c);
bool typeString =
nl->AtomType(nl->Third(f2)) == StringType;
bool typeText =
nl->AtomType(nl->Third(f2)) == TextType;
std::string stringValue;
if (typeString == true)
{
stringValue =
nl->StringValue(nl->Third(f2));
}
if (typeText == true)
{
stringValue =
nl->TextValue(nl->Third(f2));
}
b->AddBSemString(stringValue);
tripdataidxcounter++;
}
else
{
correct = false;
delete b;
return SetWord( Address(0) );
}
}
t.SetEndCoordsIdx(tripdataidxcounter);
}
}
b->AddTrip(t);
}
}
}
} // end of if n ===4
correct = true;
return SetWord(b);
} // end of n != 0
return word;
}
/*
4.4 Implementation of helper functions for class Batch
4.4.1 GetBSumString name
*/
bool Batch::
GetBSumString(int index, std::string& rString) const
{
bool success = false;
int numString = batchwords.Size();
if (index < numString)
{
BatchWord textData;
success = batchwords.Get(index, &textData);
if (success == true)
{
SmiSize bufferSize = textData.Length + 1;
char* pBuffer = new char[bufferSize];
if (pBuffer != 0)
{
memset(
pBuffer,
0,
bufferSize * sizeof(char));
success = batchwords_Flob.read(
pBuffer,
textData.Length,
textData.Offset);
}
if(success == true)
{
rString = pBuffer;
}
delete [] pBuffer;
}
}
return success;
}
/*
4.4.2 AddBSumString Function
*/
bool Batch::AddBSumString(
const std::string& stString, int id, int count)
{
if (!stString.empty()) {
BatchWord td;
td.Offset = batchwords_Flob.getSize();
td.Length = stString.length();
td.indexId = id;
td.count = count;
bool success = batchwords.Append(td);
assert(success);
success = batchwords_Flob.write(
stString.c_str(),
td.Length,
td.Offset);
assert(success);
return true;
}
return false;
}
/*
4.4.3 AddBWord Function
*/
bool Batch::AddBWord(
const std::string& stString, int id, int count)
{
if (!stString.empty()) {
WordST td;
td.Offset = bwords_Flob.getSize();
td.Length = stString.length();
td.indexId = id;
td.count = count;
bool success = bwords.Append(td);
assert(success);
bwords_Flob.write(
stString.c_str(),
td.Length,
td.Offset);
return true;
}
return false;
}
/*
4.4.4 GetBWord Function
*/
bool Batch::
GetBWord(int index, std::string& rString) const
{
bool success = false;
int numString = bwords.Size();
if (index < numString)
{
WordST textData;
success = bwords.Get(index, &textData);
if (success == true)
{
SmiSize bufferSize = textData.Length + 1;
char* pBuffer = new char[bufferSize];
if (pBuffer != 0)
{
memset(
pBuffer,
0,
bufferSize * sizeof(char));
success = bwords_Flob.read(
pBuffer,
textData.Length,
textData.Offset);
}
if(success == true)
{
rString = pBuffer;
}
delete [] pBuffer;
}
}
return success;
}
/*
4.4.5 AddBSemString Function
*/
bool Batch::AddBSemString(const std::string& str)
{
TextData td;
td.Offset = bsemantics_Flob.getSize();
td.Length = str.length();
bool success = bsemantics.Append(td);
assert(success);
if (!str.empty()) {
bsemantics_Flob.write(
str.c_str(),
td.Length,
td.Offset);
}
return true;
}
/*
4.4.6 GetBSemString Function
*/
bool Batch::GetBSemString(int index, std::string& str) const
{
bool success = false;
int numString = bsemantics.Size();
if (index < numString)
{
TextData textData;
success = bsemantics.Get(index, &textData);
if (success == true)
{
SmiSize bufferSize = textData.Length + 1;
char* pBuffer = new char[bufferSize];
if (pBuffer != 0)
{
memset(
pBuffer,
0,
bufferSize * sizeof(char));
success = bsemantics_Flob.read(
pBuffer,
textData.Length,
textData.Offset);
}
if(success == true)
{
str = pBuffer;
}
delete [] pBuffer;
}
}
return success;
}
/*
4.4.7 AddBCoordinate name
*/
void Batch::AddBCoordinate(const Coordinate& c)
{
bcoordinates.Append(c);
}
/*
4.5 Operator helper functions for Batch
4.5.1 RelevanceSumBT function
*/
double Batch::RelevanceSumBT(Rectangle<2>& mbr,
SemanticTrajectory& st,
double alpha,
double diag)
{
double relevancesum = 0.0;
double finalres = 0.0;
int sumWords = 0;
std::vector<std::string*>** hashTable;
const std::vector<std::string*>* bucket;
std::hash<std::string> str_hash;
unsigned int bucketPos = 0;
unsigned int bucketnum = 500;
hashTable = new std::vector<std::string*>*[bucketnum];
for (unsigned int i = 0; i < bucketnum; i++)
{
hashTable[i] = 0;
}
if (!IsEmptyWordList())
{
for (int i = 0; i < GetWordListSize(); i++)
{
std::string holdValue;
bool success = GetBSumString(i, holdValue);
assert(success);
std::string* stn = new std::string(holdValue);
size_t hash = str_hash(holdValue) % bucketnum;
if (!hashTable[hash])
{
hashTable[hash] = new std::vector<std::string*>();
}
hashTable[hash]->push_back(stn);
sumWords++;
}
}
for (int i = 0; i < st.GetNumCoordinates(); i++)
{
double LH = 0;
double RH = 0;
std::string holdvalue;
st.GetSemString(i, holdvalue);
if (holdvalue.length() > 0 && double (1-alpha) > 0)
{
// Do the textual stuff;
int numMatches = 0;
stringutils::StringTokenizer tokenizer(holdvalue, " ");
while(tokenizer.hasNextToken())
{
std::string eval = tokenizer.nextToken();
if (eval.length() > 0)
{
size_t hash = str_hash(eval) % bucketnum;
bucket = 0;
bucket = hashTable[hash];
bucketPos = 0;
// There is a match
if (bucket)
{
// Find the match
while(bucketPos < bucket->size())
{
std::string* p = (*bucket)[bucketPos];
if ((*p).compare(eval) == 0)
{
numMatches++;
break;
}
bucketPos++;
}
}
}
}
double textualscore = 0;
if (!numMatches == 0)
{
double uniquewords = (double)
(sumWords + holdvalue.length() - numMatches);
textualscore = (double) numMatches / uniquewords;
}
RH = double (1 - alpha) * textualscore;
}
if (alpha > 0)
{
double dist =
getDistanceBT(mbr, st.GetCoordinate(i).x, st.GetCoordinate(i).y);
double normalizedScore = 0.0;
if (dist == 0.0)
{
normalizedScore = 1;
}
else
{
normalizedScore = 1 - (double) (dist/diag);
}
LH = alpha * normalizedScore;
}
double localrel = LH + RH;
relevancesum = relevancesum + localrel;
}
if (relevancesum > 0)
{
finalres = relevancesum / st.GetNumCoordinates();
}
for (unsigned int i = 0; i< bucketnum; i++)
{
std::vector<std::string*>* v = hashTable[i];
if (v)
{
for(unsigned int j = 0; j < v->size(); j++)
{
delete (*v)[j];
}
delete hashTable[i];
hashTable[i] = 0;
}
}
delete hashTable;
return finalres;
}
/*
4.5.2 getDistanceBT function
*/
double Batch::getDistanceBT(Rectangle<2>& mbr, double x, double y)
{
double xmin = mbr.getMinX();
double ymin = mbr.getMinY();
double xmax = mbr.getMaxX();
double ymax = mbr.getMaxY();
if (xmin <= x && x <= xmax && ymin <= y && y <= ymax)
{
return 0.0;
}
// Upper quadrant
if (ymin > y)
{
//left
if (xmax < x) {
return EuclidDistRT(x, y, xmax ,ymin);
}
//Center
if (xmin <= x && x <= xmax) {
return EuclidDistRT(x, y, x ,ymin);
}
//Right
if(xmin > x)
{
return EuclidDistRT(x, y, xmin, ymin);
}
}
else
{
//left
if (xmax < x) {
return EuclidDistRT(x, y, xmax ,ymax);
}
//Center
if (xmin <= x && x <= xmax) {
return EuclidDistRT(x, y, x ,ymax);
}
//Right
if(xmin > x)
{
return EuclidDistRT(x, y, xmin, ymax);
}
}
if (ymin <= y && y <= ymax)
{
if (x < xmin)
{
return EuclidDistRT(x, y, xmin, y);
}
if (x > xmax)
{
return EuclidDistRT(x, y, xmax, y);
}
}
return 0.0;
}
/*
4.5.3 getDistanceBT function
*/
double Batch::EuclidDistRT(double x1,
double y1,
double x2,
double y2)
{
return sqrt(pow((x1 - x2),2) + pow((y1 - y2),2));
}
/*
5 Operators
5.1 Operator makesemtraj.
5.1.1 Type map for ~makesemtraj~
*/
ListExpr TypeMapMakeSemtraj(ListExpr args)
{
// Check to see if it's the right number of arguments
if (nl->HasLength(args, 4))
{
// Make sure each param is of right type
std::string err = "stream(tuple) x attr_1 x"
" attr_2 x attr_3 expected";
ListExpr stream = nl->First(args);
ListExpr attrname_longitude = nl->Second(args);
ListExpr attrname_latitude = nl->Third(args);
ListExpr attrname_semantics = nl->Fourth(args);
if(!listutils::isTupleStream(stream)){
return
listutils::typeError(
"first parameter must be a tuple stream");
}
if(!listutils::isSymbol(attrname_longitude)){
return
listutils::typeError("second parameter must"
"be an attribute name");
}
if(!listutils::isSymbol(attrname_latitude)){
return
listutils::typeError("third parameter must"
" be an attribute name");
}
if(!listutils::isSymbol(attrname_semantics)){
return
listutils::typeError("fourth parameter must"
"be an attribute name");
}
ListExpr type;
// extract the attribute list
ListExpr attrList = nl->Second(nl->Second(stream));
// Get the index for the longitude
std::string name =
nl->SymbolValue(attrname_longitude);
int index1 =
listutils::findAttribute(attrList, name, type);
if(index1==0){
return
listutils::typeError(
"attribute name " + name +
" unknown in tuple"
" stream");
}
if(!CcReal::checkType(type)){
return
listutils::typeError("attribute '" + name +
"' must be of type 'real'");
}
name = nl->SymbolValue(attrname_latitude);
int index2 =
listutils::findAttribute(attrList, name, type);
if(index2==0){
return
listutils::typeError("attribute name " + name +
" unknown in tuple stream");
}
if(!CcReal::checkType(type)){
return
listutils::typeError("attribute '" + name +
"' must be of type 'real'");
}
name = nl->SymbolValue(attrname_semantics);
int index3 =
listutils::findAttribute(attrList, name, type);
if(index3==0){
return
listutils::typeError("attribute name " + name +
" unknown in tuple stream");
}
if(!CcString::checkType(type) && !FText::checkType(type)){
return
listutils::typeError("attribute '" + name +
"' must be of type 'string'");
}
std::string restype =
SemanticTrajectory::BasicType();
ListExpr indexes = nl->ThreeElemList(
nl->IntAtom(index1-1),
nl->IntAtom(index2-1),
nl->IntAtom(index3-1));
return
nl->ThreeElemList(nl->SymbolAtom(Symbol::APPEND()),
indexes,
nl->SymbolAtom(restype));
}
return
listutils::typeError("Wrong number of arguments");
}
/*
5.1.2 Value map for ~makesemtraj~
*/
template <class T>
int MakeSemTrajMV(Word* args, Word& result,
int message, Word& local,
Supplier s)
{
result = qp->ResultStorage(s);
SemanticTrajectory* res =
static_cast<SemanticTrajectory*>(result.addr);
int noargs = qp->GetNoSons(s);
int idx1 = ((CcInt*)args[noargs-3].addr)->GetValue();
int idx2 = ((CcInt*)args[noargs-2].addr)->GetValue();
int idx3 = ((CcInt*)args[noargs-1].addr)->GetValue();
double c_x1;
double c_y1;
double c_x2;
double c_y2;
Stream<Tuple> stream(args[0]);
Tuple* tuple;
stream.open();
while((tuple = stream.request()))
{
CcReal* x = (CcReal*) tuple->GetAttribute(idx1);
CcReal* y = (CcReal*) tuple->GetAttribute(idx2);
T* sem = (T*)
tuple->GetAttribute(idx3);
double x1 = x->GetValue();
double y1 = y->GetValue();
Coordinate c(x1, y1);
res->AddCoordinate(c);
std::string str = sem->GetValue();
/*
This section here is to remove any special characters
*/
std::string resu = "";
for (size_t i = 0; i < str.size(); ++i)
{
if ((str[i] >= 'a' && str[i] <= 'z')
|| (str[i] >= 'A' && str[i] <= 'Z') || str[i] == ' ') {
resu = resu + str[i];
}
}
std::list<std::string> tokenlist;
stringutils::StringTokenizer st1(resu, " ");
while(st1.hasNextToken())
{
std::string eval = st1.nextToken();
stringutils::trim(eval);
stringutils::toLower(eval);
if (eval.length() > 1 )
{
tokenlist.push_back(eval);
}
}
std::string finalstr = "";
tokenlist.sort(SemanticTrajectory::compare_nocase);
tokenlist.unique();
int size = tokenlist.size();
int i = 0;
for (std::list<std::string>::iterator it =
tokenlist.begin();
it != tokenlist.end();
++it)
{
if (i < size - 1)
{
finalstr += *it + " ";
}
else
{
finalstr += *it;
}
i++;
}
/*
Double spaces in String seems to not be removed earlier on.
Extra trim needed at the end to remove extra space at the start
of the string.
*/
stringutils::trim(finalstr);
res->AddSemString(finalstr);
if (res->GetNumCoordinates() == 1)
{
c_x1 = x1;
c_y1 = y1;
c_x2 = x1;
c_y2 = y1;
}
else
{
c_x1 = std::min(c_x1, x1);
c_y1 = std::min(c_y1, y1);
c_x2 = std::max(c_x2, x1);
c_y2 = std::max(c_y2, y1);
}
tuple->DeleteIfAllowed();
}
if (c_x1 <= c_x2 && c_y1 <= c_y2)
{
double mind[2];
double maxd[2];
mind[0] = c_x1;
mind[1] = c_y1;
maxd[0] = c_x2;
maxd[1] = c_y2;
res->SetBoundingBox(true, mind, maxd);
}
stream.close();
return 0;
}
ValueMapping makesemtrajFuns[] =
{MakeSemTrajMV<FText>, MakeSemTrajMV<CcString>};
int makesemtrajSelect(ListExpr args)
{
ListExpr type;
ListExpr attrList = nl->Second(nl->Second(nl->First(args)));
std::string name = nl->SymbolValue(nl->Fourth(args));
listutils::findAttribute(attrList, name, type);
if(CcString::checkType(type))
{
return 1;
}
else
{
return 0;
}
}
/*
5.2 Operator ~makesemtraj2~
5.2.1 Type map for ~makesemtraj2~
*/
ListExpr TypeMapMakeSemtraj2(ListExpr args)
{
// Check to see if it's the right number of arguments
if (nl->HasLength(args, 6))
{
// Make sure each param is of right type
std::string err = "stream(tuple) x stream(tuple) x a1 x"
" a2 x a3 x a4 expected";
ListExpr stream = nl->First(args);
ListExpr stream2 = nl->Second(args);
ListExpr attrname_longitude = nl->Third(args);
ListExpr attrname_latitude = nl->Fourth(args);
ListExpr attrname_semantics = nl->Fifth(args);
ListExpr attrname_elem = nl->Sixth(args);
if(!listutils::isTupleStream(stream)){
return
listutils::typeError(
"first parameter must be a tuple stream");
}
if(!listutils::isTupleStream(stream2)){
return
listutils::typeError(
"first parameter must be a tuple stream");
}
if(!listutils::isSymbol(attrname_longitude)){
return
listutils::typeError("second parameter must"
"be an attribute name");
}
if(!listutils::isSymbol(attrname_latitude)){
return
listutils::typeError("third parameter must"
" be an attribute name");
}
if(!listutils::isSymbol(attrname_semantics)){
return
listutils::typeError("fourth parameter must"
"be an attribute name");
}
if(!listutils::isSymbol(attrname_elem)){
return
listutils::typeError("fourth parameter must"
"be an attribute name");
}
ListExpr type;
// extract the attribute list
ListExpr attrList = nl->Second(nl->Second(stream));
ListExpr attrList2 = nl->Second(nl->Second(stream2));
// Get the index for the longitude
std::string name =
nl->SymbolValue(attrname_longitude);
int index1 =
listutils::findAttribute(attrList, name, type);
if(index1==0){
return
listutils::typeError(
"attribute name " + name +
" unknown in tuple"
" stream");
}
if(!CcReal::checkType(type)){
return
listutils::typeError("attribute '" + name +
"' must be of type 'real'");
}
name = nl->SymbolValue(attrname_latitude);
int index2 =
listutils::findAttribute(attrList, name, type);
if(index2==0){
return
listutils::typeError("attribute name " + name +
" unknown in tuple stream");
}
if(!CcReal::checkType(type)){
return
listutils::typeError("attribute '" + name +
"' must be of type 'real'");
}
name = nl->SymbolValue(attrname_semantics);
int index3 =
listutils::findAttribute(attrList, name, type);
if(index3==0){
return
listutils::typeError("attribute name " + name +
" unknown in tuple stream");
}
if(!CcString::checkType(type) && !FText::checkType(type))
{
return
listutils::typeError("attribute '" + name +
"' must be of type 'string'");
}
name = nl->SymbolValue(attrname_elem);
int index4 =
listutils::findAttribute(attrList2, name, type);
if(index4==0){
return
listutils::typeError("attribute elem " + name +
" unknown in tuple stream");
}
if(!CcString::checkType(type)){
return
listutils::typeError("attribute '" + name +
"' must be of type 'string'");
}
std::string restype =
SemanticTrajectory::BasicType();
ListExpr indexes = nl->FourElemList(
nl->IntAtom(index1-1),
nl->IntAtom(index2-1),
nl->IntAtom(index3-1),
nl->IntAtom(index4-1));
return
nl->ThreeElemList(nl->SymbolAtom(Symbol::APPEND()),
indexes,
nl->SymbolAtom(restype));
}
return
listutils::typeError("Wrong number of arguments");
}
/*
5.1.3 Value map for ~makesemtraj2~
*/
int MakeSemTrajMV2(Word* args, Word& result,
int message, Word& local,
Supplier s)
{
result = qp->ResultStorage(s);
SemanticTrajectory* res =
static_cast<SemanticTrajectory*>(result.addr);
int noargs = qp->GetNoSons(s);
int idx1 = ((CcInt*)args[noargs-4].addr)->GetValue();
int idx2 = ((CcInt*)args[noargs-3].addr)->GetValue();
int idx3 = ((CcInt*)args[noargs-2].addr)->GetValue();
int idx4 = ((CcInt*)args[noargs-1].addr)->GetValue();
//Prepare stop list
Stream<Tuple> stream2(args[1]);
unsigned int bucketnum = 800;
std::hash<std::string> str_hash;
std::vector<std::string*>** hashTable;
hashTable = new std::vector<std::string*>*[bucketnum];
for (unsigned int i = 0; i < bucketnum; i++)
{
hashTable[i] = 0;
}
Tuple* tuple2;
stream2.open();
while((tuple2 = stream2.request()))
{
CcString* stopw = (CcString*)
tuple2->GetAttribute(idx4);
std::string stw = stopw->GetValue();
std::string resu = "";
for (size_t i = 0; i < stw.size(); ++i)
{
if ((stw[i] >= 'a' && stw[i] <= 'z') ||
(stw[i] >= 'A' && stw[i] <= 'Z')) {
resu = resu + stw[i];
}
}
stringutils::trim(resu);
stringutils::toLower(resu);
std::string* stn = new std::string(resu);
size_t hash = str_hash(resu) % bucketnum;
if (!hashTable[hash])
{
hashTable[hash] = new std::vector<std::string*>();
}
hashTable[hash]->push_back(stn);
tuple2->DeleteIfAllowed();
}
stream2.close();
double c_x1;
double c_y1;
double c_x2;
double c_y2;
Stream<Tuple> stream(args[0]);
Tuple* tuple;
stream.open();
unsigned int bucketpos = 0;
while((tuple = stream.request()))
{
CcReal* x = (CcReal*) tuple->GetAttribute(idx1);
CcReal* y = (CcReal*) tuple->GetAttribute(idx2);
CcString* sem = (CcString*)
tuple->GetAttribute(idx3);
if ((sem->GetValue()).length() > 0)
{
std::string str = sem->GetValue();
std::string resu = "";
for (size_t i = 0; i < str.size(); ++i)
{
if ((str[i] >= 'a' && str[i] <= 'z') ||
(str[i] >= 'A' && str[i] <= 'Z') || str[i] == ' ') {
resu = resu + str[i];
}
}
std::list<std::string> tokenlist;
stringutils::StringTokenizer st1(resu, " ");
const std::vector<std::string*>* bucket;
size_t hash = 0;
while(st1.hasNextToken())
{
std::string eval = st1.nextToken();
stringutils::trim(eval);
stringutils::toLower(eval);
if (eval.length() > 1)
{
bucket = 0;
hash = str_hash(eval) % bucketnum;
bucket = hashTable[hash];
bucketpos =0;
if(bucket)
{
bool flag = true;
while(bucketpos < bucket->size())
{
std::string* e = (*bucket)[bucketpos];
if ((e)->compare(eval) == 0)
{
flag = false;
break;
}
bucketpos++;
}
if (flag)
{
tokenlist.push_back(eval);
}
}
else
{
tokenlist.push_back(eval);
}
}
}
std::string finalstr = "";
tokenlist.sort(SemanticTrajectory::compare_nocase);
tokenlist.unique();
int size = tokenlist.size();
int i = 0;
for(const auto &word : tokenlist)
{
if (i != size - 1)
{
finalstr += word + " ";
}
else
{
finalstr += word;
}
i++;
}
res->AddSemString(finalstr);
}
else
{
res->AddSemString("");
}
double x1 = x->GetValue();
double y1 = y->GetValue();
Coordinate c(x1, y1);
res->AddCoordinate(c);
if (res->GetNumCoordinates() == 1)
{
c_x1 = x1;
c_y1 = y1;
c_x2 = x1;
c_y2 = y1;
}
else
{
c_x1 = std::min(c_x1, x1);
c_y1 = std::min(c_y1, y1);
c_x2 = std::max(c_x2, x1);
c_y2 = std::max(c_y2, y1);
}
tuple->DeleteIfAllowed();
}
if (c_x1 < c_x2 && c_y1 < c_y2)
{
double mind[2];
double maxd[2];
mind[0] = c_x1;
mind[1] = c_y1;
maxd[0] = c_x2;
maxd[1] = c_y2;
res->SetBoundingBox(true, mind, maxd);
}
for (unsigned int i = 0; i< bucketnum; i++)
{
std::vector<std::string*>* v = hashTable[i];
if (v)
{
for(unsigned int j = 0; j < v->size(); j++)
{
std::string* stn = (*v)[j];
delete stn;
}
(v)->clear();
delete hashTable[i];
hashTable[i] = 0;
}
}
delete hashTable;
stream.close();
return 0;
}
/*
5.3 Operator ~stbox~
5.3.1 Type map for ~stbox~
*/
ListExpr STboxTM(ListExpr args)
{
std::string errmsg =
"Expected datatype semanticTrajectory";
if(!SemanticTrajectory::checkType(nl->First(args)))
{
return listutils::typeError(errmsg);
}
return (nl->SymbolAtom( Rectangle<2>::BasicType()));
}
/*
5.3.2 Value map for ~stbox~
*/
int STbboxMapValue(Word* args, Word& result,
int message,
Word& local,
Supplier s)
{
result = qp->ResultStorage( s );
Rectangle<2> *box =
static_cast<Rectangle<2>*>(result.addr);
const SemanticTrajectory* st =
static_cast<SemanticTrajectory*>(args[0].addr);
(*box) = st->GetBoundingBox();
box->SetDefined(true);
return 0;
}
/*
5.4 Operator ~extractkeywords~
5.4.1 Type map for ~extractkeywords~
*/
ListExpr
extractkeywordsTM( ListExpr args )
{
std::string err =
"must be of type uniquestringarray or semantic";
if ( nl->ListLength(args) == 1
&& (SemanticTrajectory::checkType(nl->First(args))))
{
return nl->TwoElemList( nl->SymbolAtom(Symbol::STREAM()),
nl->SymbolAtom(CcString::BasicType()));
}
return nl->SymbolAtom(Symbol::TYPEERROR());
}
/*
5.4.2 Value map for ~extractkeywords~
*/
int extractkeywordMapV(Word* args, Word& result,
int message,
Word& local,
Supplier s)
{
// An auxiliary type which keeps of
// which coordinate is being
// examine between requests
struct TheWords {
int visitedwords = 0;
std::list<std::string>::iterator it;
int numOfWords = 0;
std::list<std::string> wordList;
TheWords(std::list<std::string>& list)
{
wordList = list;
numOfWords = wordList.size();
it = wordList.begin();
}
bool getNextWord(std::string& el)
{
el = *it;
stringutils::trim(el);
it++;
visitedwords++;
return true;
}
};
TheWords* localword =
static_cast<TheWords*>(local.addr);
std::list<std::string> list;
switch(message)
{
case OPEN: // Initialize the local storage
{
// If localword already exists
// need to delete and restart
if(localword)
{
delete localword;
localword = 0;
}
SemanticTrajectory* strarr =
static_cast<SemanticTrajectory*>(args[0].addr);
list = strarr->GetStringArray();
localword =
new TheWords(list);
local.addr = localword;
return 0;
}
case REQUEST: // returnthe next stream element
{
if (localword->visitedwords <
localword->numOfWords)
{
std::string str;
bool success = localword->getNextWord(str);
if (success == true)
{
CcString* elem = new CcString(true, str);
result.addr = elem;
}
else
{
result.addr = 0;
}
return YIELD;
}
else
{
result.addr = 0;
return CANCEL;
}
}
case CLOSE:
{
if (localword != 0)
{
delete localword;
local.addr = 0;
list.clear();
}
return 0;
}
default:
{
// This should never happen
return -1;
}
}
return 0;
}
/*
5.5 Operator ~makesummaries~
5.5.1 Type map for ~makesummaries~
*/
ListExpr MakesummariesTM( ListExpr args )
{
if( (nl->ListLength(args) == 7)
&& listutils::isTupleStream(nl->First(args))
&& CcInt::checkType(nl->Second(args))
&& SemanticTrajectory::checkType(nl->Third(args))
&& CellGrid2D::checkType(nl->Fourth(args))
&& listutils::isSymbol(nl->Fifth(args))
&& listutils::isSymbol(nl->Sixth(args))
&& listutils::isSymbol(nl->Seventh(args))
)
{
ListExpr stream = nl->First(args);
ListExpr attrname_WiD = nl->Fifth(args);
ListExpr attrname_Ctn = nl->Sixth(args);
ListExpr attrname_word = nl->Seventh(args);
ListExpr attrList = nl->Second(nl->Second(stream));
ListExpr type;
std::string name =
nl->SymbolValue(attrname_WiD);
int index1 =
listutils::findAttribute(attrList, name, type);
if(index1==0){
return
listutils::typeError(
"attribute name " + name +
" unknown in tuple"
" stream");
}
if(!CcInt::checkType(type)){
return
listutils::typeError("attribute '" + name +
"' must be of type 'real'");
}
name = nl->SymbolValue(attrname_Ctn);
int index2 =
listutils::findAttribute(attrList, name, type);
if(index2==0){
return
listutils::typeError("attribute name " + name +
" unknown in tuple stream");
}
name = nl->SymbolValue(attrname_word);
int index3 =
listutils::findAttribute(attrList, name, type);
if(index3==0){
return
listutils::typeError("attribute name " + name +
" unknown in tuple stream");
}
ListExpr indexes = nl->ThreeElemList(
nl->IntAtom(index1-1),
nl->IntAtom(index2-1),
nl->IntAtom(index3-1));
return
nl->ThreeElemList(nl->SymbolAtom(Symbol::APPEND()),
indexes,
nl->SymbolAtom(SemanticTrajectory::BasicType()));
}
return
listutils::typeError("Expected "
"(stream(tuple) x int x semantictrajectory x "
"cellgrid2d x attr_WordId x attr_Ctn x attr_Word).");
}
/*
5.5.2 Value map for ~makesummaries~
*/
int MakesummariesMV(Word* args, Word& result,
int message,
Word& local,
Supplier s)
{
result = qp->ResultStorage(s);
int noargs = qp->GetNoSons(s);
int idx1 = ((CcInt*)args[noargs-3].addr)->GetValue();
int idx2 = ((CcInt*)args[noargs-2].addr)->GetValue();
int idx3 = ((CcInt*)args[noargs-1].addr)->GetValue();
Stream<Tuple> stream(args[0]);
Tuple* tuple;
stream.open();
CcInt* id = static_cast<CcInt*>( args[1].addr );
SemanticTrajectory* st =
static_cast<SemanticTrajectory*>(args[2].addr);
CellGrid2D* grid =
static_cast<CellGrid2D*>(args[3].addr);
SemanticTrajectory* res =
static_cast<SemanticTrajectory*>(result.addr);
res->Clear();
res->InitializeST();
/*
Create vector to hold already found cell numbers
*/
std::vector<std::pair<int, int>> cellsresult;
/*
Iterate over each coordinate in semantictrajectory
Find the cell and return origin of cell
See if that cell origin is already present
in if not add to spatialsummary
*/
for(int i = 0; i < st->GetNumCoordinates(); i++)
{
Coordinate c = st->GetCoordinate(i);
if(!grid->onGrid(c.x, c.y))
{
return listutils::typeError("Coordinate outside of grid"
", please enter a new grid dimension in operator");
}
int32_t cellIndexX = grid->getXIndex(c.x);
int32_t cellIndexY = grid->getYIndex(c.y);
if(std::find(
cellsresult.begin(),
cellsresult.end(),
std::make_pair(cellIndexX,cellIndexY))
== cellsresult.end())
{
Cell cell(id->GetValue(),
cellIndexX,
cellIndexY);
res->AddCell(cell);
cellsresult.push_back(std::make_pair(
cellIndexX,
cellIndexY));
}
}
while((tuple = stream.request()))
{
CcInt* id = (CcInt*) tuple->GetAttribute(idx1);
CcInt* ctn = (CcInt*) tuple->GetAttribute(idx2);
CcString* str = (CcString*) tuple->GetAttribute(idx3);
res->AddStringSum(str->GetValue(),id->GetIntval(), ctn->GetIntval());
tuple->DeleteIfAllowed();
}
res->EndST(*st);
res->Finalize();
stream.close();
return 0;
}
/*
5.6 Operator ~filtersim~
5.6.1 Type map for ~filtersim~
*/
ListExpr FilterSimTypeMap(ListExpr args)
{
std::string err = "stream(tuple) x attr1 x "
"attr2 x real x real x rect";
if(!nl->HasLength(args, 6))
{
return listutils::typeError(err);
}
ListExpr stream1 = nl->First(args);
ListExpr attr1 = nl->Second(args); // semtraj1
ListExpr attr2 = nl->Third(args); // semtraj2
ListExpr attr4 = nl->Fourth(args); // real
ListExpr attr5 = nl->Fifth(args); // Real
ListExpr attr6 = nl->Sixth(args); // Rect
if (!Stream<Tuple>::checkType(stream1))
{
return listutils::typeError(err + "(first arg is not a tuple sream)");
}
if(!listutils::isSymbol(attr1)){
return listutils::typeError(err + "(first attrname is not valid)");
}
if(!listutils::isSymbol(attr2)){
return listutils::typeError(err + "(second attrname is not valid)");
}
if (!CcReal::checkType(attr4))
{
return listutils::typeError(err + "4th arg must be a real");
}
if (!CcReal::checkType(attr5))
{
return listutils::typeError(err + "5th arg must be a real");
}
if (!Rectangle<2>::checkType(attr6))
{
return listutils::typeError(err + "6th arg must be a rectangle");
}
if(!listutils::isSymbol(attr1)){
return listutils::typeError(err + "(first attrname is not valid)");
}
if(!listutils::isSymbol(attr2)){
return listutils::typeError(err + "(second attrname is not valid)");
}
ListExpr attrList1 = nl->Second(nl->Second(stream1));
std::string attrname1 = nl->SymbolValue(attr1);
std::string attrname2 = nl->SymbolValue(attr2);
ListExpr attrType1;
int index1 = listutils::findAttribute(attrList1,attrname1,attrType1);
if(index1==0){
return listutils::typeError(attrname1+
" is not an attribute of the first stream");
}
int index2 = listutils::findAttribute(attrList1,attrname2,attrType1);
if(index2==0){
return listutils::typeError(attrname2+
" is not an attribute of the second stream");
}
ListExpr indexList = nl->TwoElemList(
nl->IntAtom(index1-1),
nl->IntAtom(index2-1));
/*
Last argument to this ThreeElemList
Return types Stream<Tuple>
Tuple Type and attribute List
*/
return nl->ThreeElemList( nl->SymbolAtom(Symbols::APPEND()),
indexList,
nl->TwoElemList(
nl->SymbolAtom(Stream<Tuple>::BasicType()),
nl->TwoElemList(
nl->SymbolAtom(Tuple::BasicType()),
attrList1)));
}
/*
5.6.2 Class for ~filtersim~
*/
class SIMData
{
public:
SIMData(Word& _stream1,
const ListExpr _resType,
const int _index1, const int _index2,
double _alpha, double _threshold, double _diag) :
stream1(_stream1),
tt(0),
index1(_index1), index2(_index2),
alpha(_alpha), threshold(_threshold),
diag(_diag)
{
tt = new TupleType(_resType);
stream1.open();
}
~SIMData()
{
stream1.close();
tt->DeleteIfAllowed();
}
Tuple* nextTuple()
{
Tuple* res = new Tuple(tt);
while((res = stream1.request()))
{
SemanticTrajectory* st1 =
(SemanticTrajectory*) res->GetAttribute(index1);
SemanticTrajectory* st2 =
(SemanticTrajectory*) res->GetAttribute(index2);
double result = st1->Similarity(*st2, diag, alpha);
if (result > threshold)
{
return res;
}
}
return 0;
}
private:
Stream<Tuple> stream1;
TupleType* tt;
int index1;
int index2;
double alpha;
double threshold;
double diag;
};
/*
5.6.3 Value map for ~filtersim~
*/
int FilterSimMapValue( Word* args, Word& result,
int message, Word& local, Supplier s )
{
SIMData* li = (SIMData*) local.addr;
switch(message) {
case OPEN: { if(li){
delete li;
}
ListExpr ttype = nl->Second(GetTupleResultType(s));
int noargs = qp->GetNoSons(s);
int idx1 = ((CcInt*)args[noargs-2].addr)->GetValue();
int idx2 = ((CcInt*)args[noargs-1].addr)->GetValue();
CcReal * alpha = static_cast<CcReal*>(args[3].addr);
CcReal * thresh = static_cast<CcReal*>(args[4].addr);
Rectangle<2>* rec = static_cast<Rectangle<2>*>(args[5].addr);
double diagonal = GetDiag(*rec);
local.addr = new SIMData(
args[0],
ttype,
idx1,
idx2,
alpha->GetValue(),
thresh->GetValue(),
diagonal);
return 0;
}
case REQUEST: { if(!li){
return CANCEL;
}
result.addr = li->nextTuple();;
return result.addr ? YIELD:CANCEL;
}
case CLOSE : {
if(li){
delete li;
local.addr=0;
}
return 0;
}
}
return 0;
}
/*
5.7 Operator ~filterttsim~
5.7.1 Type map for ~filterttsim~
*/
ListExpr FilterTTSimTypeMap(ListExpr args)
{
std::string err = "stream(tuple) x attr1 x "
"attr2 x real x real x rect x Grid";
if(!nl->HasLength(args, 7))
{
return listutils::typeError(err);
}
ListExpr stream1 = nl->First(args);
ListExpr attr1 = nl->Second(args); // semtraj1
ListExpr attr2 = nl->Third(args); // semtraj2
ListExpr attr4 = nl->Fourth(args); // real
ListExpr attr5 = nl->Fifth(args); // Real
ListExpr attr6 = nl->Sixth(args); // Rect
ListExpr attr7 = nl->Seventh(args); // Grid
if (!Stream<Tuple>::checkType(stream1))
{
return listutils::typeError(err + "(first arg is not a tuple sream)");
}
if(!listutils::isSymbol(attr1)){
return listutils::typeError(err + "(first attrname is not valid)");
}
if(!listutils::isSymbol(attr2)){
return listutils::typeError(err + "(second attrname is not valid)");
}
if (!CcReal::checkType(attr4))
{
return listutils::typeError(err + "4th arg must be a real");
}
if (!CcReal::checkType(attr5))
{
return listutils::typeError(err + "5th arg must be a real");
}
if (!Rectangle<2>::checkType(attr6))
{
return listutils::typeError(err + "6th arg must be a rectangle");
}
if(!CellGrid2D::checkType(attr7))
{
return listutils::typeError(err + "(7th arg must be CellGrid2D");
}
if(!listutils::isSymbol(attr1)){
return listutils::typeError(err + "(first attrname is not valid)");
}
if(!listutils::isSymbol(attr2)){
return listutils::typeError(err + "(second attrname is not valid)");
}
ListExpr attrList1 = nl->Second(nl->Second(stream1));
std::string attrname1 = nl->SymbolValue(attr1);
std::string attrname2 = nl->SymbolValue(attr2);
ListExpr attrType1;
int index1 = listutils::findAttribute(attrList1,attrname1,attrType1);
if(index1==0){
return listutils::typeError(attrname1+
" is not an attribute of the first stream");
}
int index2 = listutils::findAttribute(attrList1,attrname2,attrType1);
if(index2==0){
return listutils::typeError(attrname2+
" is not an attribute of the second stream");
}
ListExpr indexList = nl->TwoElemList(
nl->IntAtom(index1-1),
nl->IntAtom(index2-1));
/*
Last argument to this ThreeElemList
Return types Stream<Tuple>
Tuple Type and attribute List
*/
return nl->ThreeElemList( nl->SymbolAtom(Symbols::APPEND()),
indexList,
nl->TwoElemList(
nl->SymbolAtom(Stream<Tuple>::BasicType()),
nl->TwoElemList(
nl->SymbolAtom(Tuple::BasicType()),
attrList1)));
}
/*
5.7.2 Class for ~filterttsim~
*/
class TTInfo
{
public:
TTInfo(unsigned int bucknum, Word& _stream1,
const ListExpr _resType,
const int _index1, const int _index2,
double _alpha, double _threshold, double _diag, double _wx, double _wy) :
hashTable(0), hashTable2(0), bucket(0),
bucketnum(bucknum),
stream1(_stream1),
tt(0),
index1(_index1), index2(_index2),
alpha(_alpha), threshold(_threshold),
diag(_diag), wx(_wx), wy(_wy)
{
tt = new TupleType(_resType);
stream1.open();
if (double (1-alpha) > 0)
{
InitializeTables();
}
}
~TTInfo()
{
stream1.close();
tt->DeleteIfAllowed();
}
size_t getBucket(std::string& str)
{
return str_hash(str) % bucketnum;
}
void InitializeTables()
{
if(!hashTable)
{
hashTable = new std::vector<std::string*>*[bucketnum];
for (unsigned int i = 0; i < bucketnum; i++)
{
hashTable[i] = 0;
}
}
if(!hashTable2)
{
hashTable2 = new std::vector<std::string*>*[bucketnum];
for (unsigned int i = 0; i < bucketnum; i++)
{
hashTable2[i] = 0;
}
}
}
void getTablesReady(SemanticTrajectory& st1, SemanticTrajectory& st2)
{
for (int i = 0; i < st1.GetNumWords(); i++)
{
std::string holdValue = "";
bool success = st1.GetStringSum(i, holdValue);
if(success)
{
size_t hash = getBucket(holdValue);
if (!hashTable[hash])
{
hashTable[hash] = new std::vector<std::string*>();
}
std::string* stn = new std::string(holdValue);
hashTable[hash]->push_back(stn);
}
}
for (int i = 0; i < st2.GetNumWords(); i++)
{
std::string holdValue = "";
bool success = st2.GetStringSum(i, holdValue);
if(success)
{
size_t hash = getBucket(holdValue);
if (!hashTable2[hash])
{
hashTable2[hash] = new std::vector<std::string*>();
}
std::string* stn = new std::string(holdValue);
hashTable2[hash]->push_back(stn);
}
}
}
Tuple* nextTuple()
{
Tuple* res = new Tuple(tt);
while((res = stream1.request()))
{
SemanticTrajectory* st1 =
(SemanticTrajectory*) res->GetAttribute(index1);
SemanticTrajectory* st2 =
(SemanticTrajectory*) res->GetAttribute(index2);
double result1 = 0.0;
double result2 = 0.0;
double result = 0.0;
double ts = 0.0;
double textalpha = double (1 - alpha);
if (textalpha > 0)
{
getTablesReady(*st1, *st2);
ts = textualScore(*st1, *st2);
result2 = double (1 - alpha ) * ts;
}
double dist = 0.0;
if (alpha > 0)
{
dist =
MinDistAux(*st1, *st2, wx, wy);
}
double normalizedScore = 0.0;
if (dist != 0.0)
{
normalizedScore = 1 - (double)(dist/diag);
}
else {
normalizedScore = 1;
}
result1 = alpha * normalizedScore;
result = result1 + result2;
clearContent();
if (result > threshold)
{
return res;
}
}
return 0;
}
void clearContent()
{
for (unsigned int i = 0; i< bucketnum; i++)
{
std::vector<std::string*>* v = hashTable[i];
if (v)
{
for(unsigned int j = 0; j < v->size(); j++)
{
std::string* stn = (*v)[j];
delete stn;
}
(v)->clear();
delete hashTable[i];
hashTable[i] = 0;
}
}
for (unsigned int i = 0; i< bucketnum; i++)
{
std::vector<std::string*>* v = hashTable2[i];
if (v)
{
for(unsigned int j = 0; j < v->size(); j++)
{
std::string* stn = (*v)[j];
delete stn;
}
(v)->clear();
delete hashTable2[i];
hashTable2[i] = 0;
}
}
}
double MinDistAux(SemanticTrajectory& st1,
SemanticTrajectory& st2, double wx, double wy)
{
DbArray<Cell>* Ax = new DbArray<Cell>(0);
DbArray<Cell> valuesSt1 = st1.GetCellList();
DbArray<Cell> valuesSt2 = st2.GetCellList();
Ax->Append(valuesSt1);
Ax->Append(valuesSt2);
DbArray<Cell>* Ay = new DbArray<Cell>(0);
Ay->Append(valuesSt1);
Ay->Append(valuesSt2);
Ax->Sort(SemanticTrajectory::CompareX);
Ay->Sort(SemanticTrajectory::CompareY);
int32_t m = Ax->Size();
double result = 0.0;
for (int i = 0; i < Ax->Size(); i++)
{
Cell c;
Ax->Get(i, &c);
}
for (int i = 0; i < Ay->Size(); i++)
{
Cell c;
Ay->Get(i, &c);
}
result = MinDistUtils(*Ax, *Ay, m, wx, wy);
return result;
}
double MinDistUtils(
DbArray<Cell>& Ax,
DbArray<Cell>& Ay, int32_t m, double wx, double wy)
{
double minD = DBL_MAX;
if (m > 3)
{
int n = m/2;
DbArray<Cell>* AxL = new DbArray<Cell>(0);
Cell cell1;
Ax.Get(0,cell1);
Ax.copyTo(*AxL, 0, n, 0);
DbArray<Cell>* AxR = new DbArray<Cell>(0);
Ax.copyTo(*AxR, n, m-n, 0);
DbArray<Cell>* AyL = new DbArray<Cell>(0);
DbArray<Cell>* AyR = new DbArray<Cell>(0);
double minDL = 0.0;
double minDR = 0.0;
for (int i = 0; i < Ay.Size(); i++)
{
Cell celly, cellx;
Ay.Get(i, celly);
Ax.Get(n, cellx);
if(celly.GetX() <= cellx.GetX())
{
AyL->Append(celly);
}
else
{
AyR->Append(celly);
}
}
/* Check to see if AxL has
cells from both C1 and C2 */
if (AyL->Size() != 0){
Cell cellAyL;
AyL->Get(0, cellAyL);
int32_t firstId = cellAyL.GetId();
bool both = false;
for (int i = 1; i < AyL->Size(); i++)
{
AyL->Get(i, cellAyL);
if (cellAyL.GetId() != firstId)
{
both = true;
break;
}
}
if (both == true)
{
minDL = MinDistUtils(*AxL, *AyL, n, wx, wy);
}
else
{
minDL = DBL_MAX;
}
} else {
minDL = DBL_MAX;
}
if (AyR->Size() != 0)
{
Cell cellAyR;
AyR->Get(0, cellAyR);
int32_t firstId = cellAyR.GetId();
bool both = false;
for (int i = 1; i < AyR->Size(); i++)
{
AyR->Get(i, cellAyR);
if (cellAyR.GetId() != firstId)
{
both = true;
break;
}
}
if (both == true)
{
minDR = MinDistUtils(*AxR, *AyR, m-n, wx, wy);
}
else
{
minDR = DBL_MAX;
}
}
else {
minDR = DBL_MAX;
}
minD = minDL > minDR ? minDR : minDL;
DbArray<Cell>* Am = new DbArray<Cell>(0);
for (int i = 0; i < Ay.Size(); i++)
{
Cell x;
Cell y;
Ax.Get(n, x);
Ay.Get(i, y);
if (fabs(y.GetX() - x.GetX()) < minD)
{
Am->Append(y);
}
}
for (int i = 0; i < Am->Size(); i++)
{
for (int j = i + 1; j < Am->Size(); j++)
{
Cell c1;
Cell c2;
Am->Get(i, c1);
Am->Get(j, c2);
if (c1.GetId() != c2.GetId())
{
double tempmin = GetCellDist(c1,c2, wx, wy);
if (tempmin < minD)
{
minD = tempmin;
}
}
}
}
//Don't forget to delete the DbArrays
// that were initialized here
delete Am;
delete AyR;
delete AyL;
delete AxR;
delete AxL;
}
else
{
minD = BruteForce(Ax, m, wx, wy);
}
return minD;
}
double GetCellDist(
Cell& c1, Cell& c2, double wx, double wy)
{
double result;
//Are they the same cell
if (c1.GetX() == c2.GetX()
&& c1.GetY() == c2.GetY())
{
return 0.0;
}
//c1 and c2 on the same x Axis
if (c1.GetX() == c2.GetX())
{
// if c1 is above c2
// Take bottom corner of c1 and top corner of c2
if(c1.GetY() > c2.GetY())
{
int32_t y2 = (int32_t)(c2.GetY() + 1);
return
EuclidDist(c1.GetX(), c1.GetY(),
c2.GetX(), y2, wx, wy);
}
// if c1 is below c2
// Take top corner of c1 and bottom corner of c2
else
{
int32_t y1 = (int32_t)(c1.GetY() + 1);
return
EuclidDist(c1.GetX(),
y1,
c2.GetX(),
c2.GetY(), wx, wy);
}
}
//c1 and c2 on the same y axis
if (c1.GetY() == c2.GetY())
{
// if c1 is to the right of c2
if(c1.GetX() > c2.GetX())
{
int32_t x2 = (int32_t)(c2.GetX() + 1);
return
EuclidDist(c1.GetX(), c1.GetY(),
x2, c2.GetY(), wx, wy);
}
// if c1 is to the left of c2
else
{
int32_t x1 = (int32_t)(c1.GetX() + 1);
return
EuclidDist(x1,
c1.GetY(),
c2.GetX(), c2.GetY(), wx, wy);
}
}
// if c1 is above c2
if (c1.GetY() > c2.GetY())
{
// if c1 is to the right of c2
if (c1.GetX() > c2.GetX())
{
int32_t x2 = (int32_t)(c2.GetX() + 1);
int32_t y2 = (int32_t)(c2.GetY() + 1);
return
EuclidDist(c1.GetX(),
c1.GetY(),
x2,
y2, wx, wy);
}
// if c1 is to the left c2
//
else
{
int32_t x1 = (int32_t)(c1.GetX() + 1);
int32_t y2 = (int32_t)(c2.GetY() + 1);
return
EuclidDist(x1, c1.GetY(),
c2.GetX(), y2, wx, wy);
}
}
// if c1 is below c2
else
{
// if c1 is to the right of c2
if (c1.GetX() > c2.GetX())
{
int32_t x1 = (c1.GetX() + 1);
int32_t x2 = (c2.GetX() + 1);
return
EuclidDist(
x1,
c1.GetY(),
x2,
c2.GetY(), wx, wy);
}
// if c1 is to the left c2
else
{
int32_t x1 = (c1.GetX() + 1);
int32_t y1 = (c1.GetY() + 1);
return
EuclidDist(
x1,y1,
c2.GetX(),
c2.GetY(), wx, wy);
}
}
return result;
}
double BruteForce(
DbArray<Cell>& Ax,
int32_t m, double wx, double wy)
{
double minD = DBL_MAX;
for (int i = 0; i < Ax.Size(); i++)
{
for (int j = i + 1; j < Ax.Size(); j++)
{
Cell c1;
Cell c2;
Ax.Get(i, c1);
Ax.Get(j, c2);
if (c1.GetId() != c2.GetId())
{
double tempmin = GetCellDist(c1,c2, wx, wy);
if (tempmin < minD)
{
minD = tempmin;
}
}
}
}
return minD;
}
double EuclidDist(int32_t x1,
int32_t y1,
int32_t x2,
int32_t y2, double wx, double wy) const
{
double x11 = (double) x1*wx;
double x21 = (double) x2*wx;
double y11 = (double) y1*wy;
double y21 = (double) y2*wy;
return sqrt(pow((x11 - x21),2) + pow((y11 - y21),2));
}
double textualScore(SemanticTrajectory& st1, SemanticTrajectory& st2)
{
double TSim = 0.0;
unsigned int bucketpos = 0;
for(int i = 0; i < st1.GetNumCoordinates(); i++)
{
std::string holdvalue;
st1.GetSemString(i, holdvalue);
if (holdvalue.length() > 0)
{
stringutils::StringTokenizer parse_st1(holdvalue, " ");
while(parse_st1.hasNextToken())
{
std::string eval = parse_st1.nextToken();
stringutils::trim(eval);
size_t hash = getBucket(eval);
bucket = 0;
bucket = hashTable2[hash];
bucketpos = 0;
if(bucket)
{
while(bucketpos < bucket->size())
{
std::string* check = (*bucket)[bucketpos];
if ((check)->compare(eval) == 0)
{
TSim = TSim + ((double) 1/st1.GetNumCoordinates());
break;
}
bucketpos++;
}
}
}
}
}
for(int i = 0; i < st2.GetNumTextData(); i++)
{
std::string holdvalue;
st2.GetSemString(i, holdvalue);
if (holdvalue.length() > 0)
{
stringutils::StringTokenizer parse_st2(holdvalue, " ");
while(parse_st2.hasNextToken())
{
std::string eval = parse_st2.nextToken();
stringutils::trim(eval);
size_t hash = getBucket(eval);
bucket = 0;
bucket = hashTable[hash];
bucketpos = 0;
if(bucket)
{
while(bucketpos < bucket->size())
{
std::string* check = (*bucket)[bucketpos];
if ((check)->compare(eval) == 0)
{
TSim = TSim + ((double) 1/st2.GetNumCoordinates());
break;
}
bucketpos++;
}
}
}
}
}
return TSim;
}
private:
std::vector<std::string*>** hashTable;
std::vector<std::string*>** hashTable2;
const std::vector<std::string*>* bucket;
unsigned int bucketnum;
Stream<Tuple> stream1;
TupleType* tt;
int index1;
int index2;
double alpha;
double threshold;
double diag;
std::hash<std::string> str_hash;
double wx;
double wy;
};
/*
5.7.3 Value map for ~filterttsim~
*/
int FilterTTSimMapValue( Word* args, Word& result,
int message, Word& local, Supplier s )
{
TTInfo* li = (TTInfo*) local.addr;
switch(message){
case OPEN: { if(li){
delete li;
}
ListExpr ttype = nl->Second(GetTupleResultType(s));
int noargs = qp->GetNoSons(s);
int idx1 = ((CcInt*)args[noargs-2].addr)->GetValue();
int idx2 = ((CcInt*)args[noargs-1].addr)->GetValue();
CellGrid2D* grid = static_cast<CellGrid2D*>(args[6].addr);
CcReal * alpha = static_cast<CcReal*>(args[3].addr);
CcReal * thresh = static_cast<CcReal*>(args[4].addr);
Rectangle<2>* rec = static_cast<Rectangle<2>*>(args[5].addr);
double diagonal = GetDiag(*rec);
double wx = grid->getXw();
double wy = grid->getYw();
local.addr = new TTInfo(
100,
args[0],
ttype,
idx1,
idx2,
alpha->GetValue(),
thresh->GetValue(),
diagonal,
wx,
wy);
return 0;
}
case REQUEST: { if(!li){
return CANCEL;
}
result.addr = li->nextTuple();;
return result.addr ? YIELD:CANCEL;
}
case CLOSE : {
if(li){
delete li;
local.addr=0;
}
return 0;
}
}
return 0;
}
/*
5.8 Operator ~filterbbsim~
5.8.1 Type map for ~filterbbsim~
*/
ListExpr FilterBBSimTypeMap(ListExpr args)
{
std::string err = "stream(tuple) x attr1 x "
"attr2 x attr3 x attr4 x CcReal";
if(!nl->HasLength(args, 8))
{
return listutils::typeError(err);
}
ListExpr stream1 = nl->First(args);
ListExpr attr1 = nl->Second(args); // batch 1 id
ListExpr attr2 = nl->Third(args); // batch 2 id
ListExpr attr3 = nl->Fourth(args); // batch mbr 1
ListExpr attr4 = nl->Fifth(args); // batch mbr2
if (!Stream<Tuple>::checkType(stream1))
{
return listutils::typeError(err + "(first arg is not a tuple sream)");
}
if (!CcReal::checkType(nl->Sixth(args)))
{
return listutils::typeError(err + "6th arg must be a real");
}
if (!CcReal::checkType(nl->Seventh(args)))
{
return listutils::typeError(err + "7th arg must be a real");
}
if (!Rectangle<2>::checkType(nl->Eigth(args)))
{
return listutils::typeError(err + "8th arg must be a rectangle");
}
if(!listutils::isSymbol(attr1)){
return listutils::typeError(err + "(first attrname is not valid)");
}
if(!listutils::isSymbol(attr2)){
return listutils::typeError(err + "(second attrname is not valid)");
}
if(!listutils::isSymbol(attr3)){
return listutils::typeError(err + "(first attrname is not valid)");
}
if(!listutils::isSymbol(attr4)){
return listutils::typeError(err + "(second attrname is not valid)");
}
ListExpr attrList1 = nl->Second(nl->Second(stream1));
std::string attrname1 = nl->SymbolValue(attr1);
std::string attrname2 = nl->SymbolValue(attr2);
std::string attrname3 = nl->SymbolValue(attr3);
std::string attrname4 = nl->SymbolValue(attr4);
ListExpr attrType1;
int index1 = listutils::findAttribute(attrList1,attrname1,attrType1);
if(index1==0){
return listutils::typeError(attrname1+
" is not an attribute of the first stream");
}
int index2 = listutils::findAttribute(attrList1,attrname2,attrType1);
if(index2==0){
return listutils::typeError(attrname2+
" is not an attribute of the second stream");
}
int index3 = listutils::findAttribute(attrList1,attrname3,attrType1);
if(index3==0){
return listutils::typeError(attrname3+
" is not an attribute of the first stream");
}
int index4 = listutils::findAttribute(attrList1,attrname4,attrType1);
if(index4==0){
return listutils::typeError(attrname4+
" is not an attribute of the second stream");
}
ListExpr indexList = nl->FourElemList(
nl->IntAtom(index1-1),
nl->IntAtom(index2-1),
nl->IntAtom(index3-1),
nl->IntAtom(index4-1));
/*
Last argument to this ThreeElemList
Return types Stream<Tuple>
Tuple Type and attribute List
*/
return nl->ThreeElemList( nl->SymbolAtom(Symbols::APPEND()),
indexList,
nl->TwoElemList(
nl->SymbolAtom(Stream<Tuple>::BasicType()),
nl->TwoElemList(
nl->SymbolAtom(Tuple::BasicType()),
attrList1)));
}
class BatchBatchInfo
{
public:
BatchBatchInfo(Word& _stream1,
const ListExpr _resType,
const int _index1, const int _index2,
const int _index3, const int _index4,
double _alpha, double _threshold, double _diag) :
stream1(_stream1),
tt(0),
index1(_index1), index2(_index2),
index3(_index3), index4(_index4),
alpha(_alpha), threshold(_threshold),
diag(_diag)
{
tt = new TupleType(_resType);
stream1.open();
}
~BatchBatchInfo()
{
stream1.close();
tt->DeleteIfAllowed();
}
Tuple* nextTuple()
{
Tuple* res = new Tuple(tt);
while((res = stream1.request()))
{
CcInt* id1 = (CcInt*) res->GetAttribute(index1);
CcInt* id2 = (CcInt*) res->GetAttribute(index2);
if(id1->GetIntval() < id2->GetIntval()
|| id1->GetIntval() == id2->GetIntval())
{
Rectangle<2>* r1 = (Rectangle<2>*) res->GetAttribute(index3);
Rectangle<2>* r2 = (Rectangle<2>*) res->GetAttribute(index4);
const Geoid* geoid = 0;
double distance = 0.0;
if (alpha > 0)
{
distance = r1->Distance(*r2, geoid);
}
double normalizedScore = 0.0;
if (distance == 0.0) {
normalizedScore = 1;
}
else
{
normalizedScore = 1 - (double)(distance/diag);
}
double result = alpha * normalizedScore * 2 + double (1-alpha) * 2;
if (result > threshold)
{
if (r1->Area() < r2->Area())
{
Tuple* t = reverseBatches(res);
res->DeleteIfAllowed();
return t;
} else {
return res;
}
}
}
}
return 0;
}
private:
Stream<Tuple> stream1;
TupleType* tt;
int index1;
int index2;
int index3;
int index4;
double alpha;
double threshold;
double diag;
Tuple* reverseBatches(Tuple* t1)
{
Tuple* res = new Tuple(tt);
int no1 = t1->GetNoAttributes();
int half = no1 / 2;
for (int i = 0; i < no1; i++)
{
if (i < half)
{
res->CopyAttribute(half+i, t1, i);
}
else
{
res->CopyAttribute(i-half, t1, i);
}
}
return res;
}
};
int FilterBBSimMapValue( Word* args, Word& result,
int message, Word& local, Supplier s )
{
BatchBatchInfo* li = (BatchBatchInfo*) local.addr;
switch(message){
case OPEN: { if(li){
delete li;
}
ListExpr ttype = nl->Second(GetTupleResultType(s));
int noargs = qp->GetNoSons(s);
int idx1 = ((CcInt*)args[noargs-4].addr)->GetValue();
int idx2 = ((CcInt*)args[noargs-3].addr)->GetValue();
int idx3 = ((CcInt*)args[noargs-2].addr)->GetValue();
int idx4 = ((CcInt*)args[noargs-1].addr)->GetValue();
CcReal * alpha = static_cast<CcReal*>(args[5].addr);
CcReal * thresh = static_cast<CcReal*>(args[6].addr);
Rectangle<2>* rec = static_cast<Rectangle<2>*>(args[7].addr);
double diagonal = GetDiag(*rec);
local.addr = new BatchBatchInfo(args[0], ttype,
idx1,
idx2,
idx3,
idx4,
alpha->GetValue(),
thresh->GetValue(),
diagonal);
return 0;
}
case REQUEST: { if(!li){
return CANCEL;
}
result.addr = li->nextTuple();;
return result.addr ? YIELD:CANCEL;
}
case CLOSE : {
if(li){
delete li;
local.addr=0;
}
return 0;
}
}
return 0;
}
/*
5.9 Operator ~batches~
5.9.1 Type map for ~batches~
*/
ListExpr BatchesTM(ListExpr args)
{
std::string err = "stream(tuple) x attr1 x "
"attr2";
if(!nl->HasLength(args, 4))
{
return listutils::typeError(err);
}
ListExpr stream1 = nl->First(args);
ListExpr attr1 = nl->Second(args);
ListExpr attr2 = nl->Third(args);
if (!Stream<Tuple>::checkType(stream1))
{
return listutils::typeError(err + "(first arg is not a tuple sream)");
}
if (!CcReal::checkType(nl->Fourth(args)))
{
return listutils::typeError(err + "4th arg must be a real");
}
if(!listutils::isSymbol(attr1)){
return listutils::typeError(err + "(first attrname is not valid)");
}
if(!listutils::isSymbol(attr2)){
return listutils::typeError(err + "(second attrname is not valid)");
}
ListExpr attrList1 = nl->Second(nl->Second(stream1));
std::string attrname1 = nl->SymbolValue(attr1);
std::string attrname2 = nl->SymbolValue(attr2);
ListExpr attrType1;
int index1 = listutils::findAttribute(attrList1,attrname1,attrType1);
if(index1==0){
return listutils::typeError(attrname1+
" is not an attribute of the first stream");
}
int index2 = listutils::findAttribute(attrList1,attrname2,attrType1);
if(index2==0){
return listutils::typeError(attrname2+
" is not an attribute of the second stream");
}
ListExpr indexList = nl->TwoElemList(
nl->IntAtom(index1-1),
nl->IntAtom(index2-1));
/*
Last argument to this ThreeElemList
Return types Stream<Tuple>
Tuple Type and attribute List
*/
return nl->ThreeElemList( nl->SymbolAtom(Symbols::APPEND()),
indexList,
nl->TwoElemList(
nl->SymbolAtom(Stream<Tuple>::BasicType()),
nl->TwoElemList(
nl->SymbolAtom(Tuple::BasicType()),
attrList1)));
}
class HoldBatchesInfo
{
public:
HoldBatchesInfo(Word& _stream1,
const ListExpr _resType,
const int _index1, const int _index2, double _threshold) :
stream1(_stream1),
tt(0), batchidcounter(0), currentBatches(0),
index1(_index1), index2(_index2), diathreshold(_threshold)
{
tt = new TupleType(_resType);
batchidcounter = 0;
stream1.open();
}
~HoldBatchesInfo()
{
stream1.close();
tt->DeleteIfAllowed();
}
Tuple* nextTuple()
{
std::vector<BatchGroup>::iterator it;
Tuple* res = new Tuple(tt);
int assignBatchId = -1;
if ((res = stream1.request()))
{
int bmin = -1;
double diagmin = DBL_MAX;
int i = 0;
SemanticTrajectory* st =
(SemanticTrajectory*) res->GetAttribute(index1);
for (it=currentBatches.begin(); it!=currentBatches.end(); ++it)
{
BatchGroup tempBatch(st->GetBoundingBox());
tempBatch.addnewMBR((*it).GetMBR());
Rectangle<2> mbr = tempBatch.GetMBR();
double x1 = mbr.getMinX();
double y1 = mbr.getMinY();
double x2 = mbr.getMaxX();
double y2 = mbr.getMaxY();
double dgbox = sqrt(pow((x1 - x2),2)
+ pow((y1 - y2),2));
if(dgbox < diagmin)
{
diagmin = dgbox;
bmin = i; // hold the index of
}
i++;
}
if (diagmin <= diathreshold)
{
it = currentBatches.begin();
std::advance(it, bmin);
(*it).addnewMBR(st->GetBoundingBox());
assignBatchId = bmin;
}
else
{
currentBatches.push_back(BatchGroup(st->GetBoundingBox()));
assignBatchId = batchidcounter;
batchidcounter++;
}
res->PutAttribute(index2, new CcInt(true, assignBatchId));
return res;
} else
{
return 0;
}
return res;
}
private:
Stream<Tuple> stream1;
TupleType* tt;
int batchidcounter;
std::vector<BatchGroup> currentBatches;
int index1;
int index2;
double diathreshold;
};
int BatchesVM( Word* args, Word& result,
int message, Word& local, Supplier s )
{
HoldBatchesInfo* li = (HoldBatchesInfo*) local.addr;
switch(message){
case OPEN: { if(li){
delete li;
}
ListExpr ttype = nl->Second(GetTupleResultType(s));
int noargs = qp->GetNoSons(s);
int idx1 = ((CcInt*)args[noargs-2].addr)->GetValue();
int idx2 = ((CcInt*)args[noargs-1].addr)->GetValue();
CcReal * thresh = static_cast<CcReal*>(args[3].addr);
local.addr = new HoldBatchesInfo(args[0], ttype,
idx1,
idx2,
thresh->GetValue());
return 0;
}
case REQUEST: { if(!li){
return CANCEL;
}
result.addr = li->nextTuple();;
return result.addr ? YIELD:CANCEL;
}
case CLOSE : {
if(li){
delete li;
local.addr=0;
}
return 0;
}
}
return 0;
}
/*
5.10 Operator ~btsim~
5.10.1 Type map for ~btsim~
*/
ListExpr BTSimTypeMap(ListExpr args)
{
std::string err = "stream(tuple) x x "
"";
if(!nl->HasLength(args, 4))
{
return listutils::typeError(err);
}
if(!Batch::checkType(nl->First(args)))
{
return listutils::typeError("expecting one batch type");
}
if(!SemanticTrajectory::checkType(nl->Second(args)))
{
return listutils::typeError("expecting semanticTrajectory type");
}
if (!CcReal::checkType(nl->Third(args)) ||
!Rectangle<2>::checkType(nl->Fourth(args)))
{
return
listutils::typeError("3rd and 4th arg of real and rectangle");
}
return NList(CcReal::BasicType()).listExpr();
}
int BTSimMapValue( Word* args, Word& result,
int message, Word& local, Supplier s )
{
Batch* b = static_cast<Batch*>( args[0].addr );
SemanticTrajectory* st = static_cast<SemanticTrajectory*>( args[1].addr );
CcReal * alpha = static_cast<CcReal*>(args[2].addr);
Rectangle<2>* rec = static_cast<Rectangle<2>*>(args[3].addr);
double alpha1 = alpha->GetValue();
double diag = GetDiag(*rec);
const Geoid* geoid = 0;
Rectangle<2> bmbr = b->GetBoundingBox();
double distance = bmbr.Distance(st->GetBoundingBox(), geoid);
double normalizedScore = 1 - (double)(distance/diag);
double RH = alpha1 * normalizedScore + double (1-alpha1);
// double LH = 0.0;
double LH = b->RelevanceSumBT(bmbr, *st, alpha1, diag);
result = qp->ResultStorage(s);
double answer = LH + RH;
((CcReal*) result.addr)->Set(answer);
return 0;
}
/*
5.11 Operator ~filterbtsim~
5.11.1 Type map for ~filterbtsim~
*/
ListExpr FilterBTSimTypeMap(ListExpr args)
{
std::string err = "stream(tuple) x x "
"";
if(!nl->HasLength(args, 9))
{
return listutils::typeError(err);
}
ListExpr stream1 = nl->First(args);
ListExpr stream2 = nl->Second(args);
ListExpr batchMBR = nl->Sixth(args);
ListExpr st1 = nl->Third(args);
ListExpr word = nl->Fourth(args);
ListExpr Ctn = nl->Fifth(args);
if (!Stream<Tuple>::checkType(stream1))
{
return listutils::typeError(err + "(first arg is not a tuple stream)");
}
if (!Stream<Tuple>::checkType(stream2))
{
return listutils::typeError(err + "(second arg is not a tuple stream)");
}
if(!listutils::isSymbol(st1)){
return listutils::typeError(err + "(first attrname is not valid)");
}
if(!listutils::isSymbol(word)){
return listutils::typeError(err + "(second attrname is not valid)");
}
if(!listutils::isSymbol(Ctn)){
return listutils::typeError(err + "(third attrname is not valid)");
}
if (!Rectangle<2>::checkType(batchMBR))
{
return listutils::typeError(err + "6th arg must be a rectangle");
}
// THE PARAMS FOR NORMALIZATION ~ ALPHA ~ _threshold
if (!CcReal::checkType(nl->Seventh(args)))
{
return listutils::typeError(err + "6th arg must be a real");
}
if (!CcReal::checkType(nl->Eigth(args)))
{
return listutils::typeError(err + "7th arg must be a real");
}
if (!Rectangle<2>::checkType(nl->Ninth(args)))
{
return listutils::typeError(err + "8th arg must be a rectangle");
}
std::string attrname1 = nl->SymbolValue(st1);
std::string attrname2 = nl->SymbolValue(word);
std::string attrname3 = nl->SymbolValue(Ctn);
ListExpr attrType1;
ListExpr attrList1 = nl->Second(nl->Second(stream1));
ListExpr attrList2 = nl->Second(nl->Second(stream2));
int index1 = listutils::findAttribute(attrList1, attrname1,attrType1);
if(index1==0)
{
return listutils::typeError(attrname1+
" is not an attribute of the first stream");
}
int index2 = listutils::findAttribute(attrList2, attrname2,attrType1);
if(index2==0)
{
return listutils::typeError(attrname2+
" is not an attribute of the first stream");
}
int index3 = listutils::findAttribute(attrList2, attrname3,attrType1);
if(index3==0)
{
return listutils::typeError(attrname3+
" is not an attribute of the first stream");
}
// Temporarily here because I will be adding more
ListExpr indexList = nl->ThreeElemList(
nl->IntAtom(index1-1),
nl->IntAtom(index2-1),
nl->IntAtom(index3-1)
);
/*
Last argument to this ThreeElemList
Return types Stream<Tuple>
Tuple Type and attribute List
*/
return nl->ThreeElemList( nl->SymbolAtom(Symbols::APPEND()),
indexList,
nl->TwoElemList(
nl->SymbolAtom(Stream<Tuple>::BasicType()),
nl->TwoElemList(
nl->SymbolAtom(Tuple::BasicType()),
attrList1)));
}
class BatchTrajInfo
{
struct WordInfo
{
CcString* str;
CcInt* count;
};
public:
BatchTrajInfo(unsigned int bucknum,
Rectangle<2>& _rect, Word& _stream1, Word& _stream2,
const ListExpr _resType,
const int _index1, const int _index2, const int _index3,
double _alpha, double _threshold, double _diag) :
hashTable(0), bucket(0), bucketPos(0), bucketnum(bucknum), bmbr(_rect),
stream1(_stream1),
stream2(_stream2),
tt(0),
index1(_index1), index2(_index2),
index3(_index3),
alpha(_alpha), threshold(_threshold),
diag(_diag), wordTuple(0), sumOfBoth(0)
{
// this->stream2 = stream2;
wordTuple =0;
tt = new TupleType(_resType);
stream1.open();
qp->Open(stream2.addr);
readBatchTSum();
}
~BatchTrajInfo()
{
stream1.close();
qp->Close(stream2.addr);
clearTable();
tt->DeleteIfAllowed();
}
void clearTable(){
for (unsigned int i = 0; i< bucketnum; i++)
{
std::vector<WordInfo*>* v = hashTable[i];
if (v)
{
for(unsigned int j = 0; j < v->size(); j++)
{
delete (*v)[j];
}
delete hashTable[i];
hashTable[i] = 0;
}
}
}
size_t getBucket(CcString* str)
{
return str->HashValue() % bucketnum;
}
void readBatchTSum()
{
if (!hashTable)
{
hashTable = new std::vector<WordInfo*>*[bucketnum];
for (unsigned int i = 0; i < bucketnum; i++)
{
hashTable[i] = 0;
}
}
Word tuple;
qp->Request(stream2.addr, tuple);
while (qp->Received(stream2.addr))
{
wordTuple = static_cast<Tuple*>(tuple.addr);
CcString* str1 = (CcString*) wordTuple->GetAttribute(index2);
CcInt* ctn = (CcInt*) wordTuple->GetAttribute(index3);
size_t hash = getBucket(str1);
if (!hashTable[hash])
{
hashTable[hash] = new std::vector<WordInfo*>();
}
WordInfo* rInfo = new WordInfo;
rInfo->count = ctn;
rInfo->str = str1;
hashTable[hash]->push_back(rInfo);
sumOfBoth = sumOfBoth + 1;
qp->Request(stream2.addr, tuple);
}
}
Tuple* nextTuple()
{
Tuple* res = new Tuple(tt);
while ((res = stream1.request()))
{
SemanticTrajectory* st =
(SemanticTrajectory*) res->GetAttribute(index1);
const Geoid* geoid = 0;
double distance = bmbr.Distance((st->GetBoundingBox()), geoid);
double normalizedScore = 1 - (double)(distance/diag);
double result1 = alpha * normalizedScore + double (1-alpha);
double result2 = 0.0;
double sum = 0.0;
for (int i = 0; i < st->GetNumCoordinates(); i++)
{
std::string holdvalue;
st->GetSemString(i, holdvalue);
double rel = RelevanceBT(bmbr, st->GetCoordinate(i).x,
st->GetCoordinate(i).y, holdvalue,alpha, diag);
sum = sum + rel;
}
if (sum > 0)
{
result2 = sum / st->GetNumCoordinates();
}
double result3 = result1 + result2;
if (result3 > threshold)
{
return res;
}
}
return 0;
}
private:
std::vector<WordInfo*>** hashTable;
const std::vector<WordInfo*>* bucket;
unsigned int bucketPos;
unsigned int bucketnum;
Rectangle<2> bmbr;
Stream<Tuple> stream1;
Word stream2;
TupleType* tt;
int index1;
int index2;
int index3;
double alpha;
double threshold;
double diag;
Tuple* wordTuple;
int sumOfBoth = 0;
double EuclidDistRT(double x1,
double y1,
double x2,
double y2)
{
return sqrt(pow((x1 - x2),2) + pow((y1 - y2),2));
}
double RelevanceBT(Rectangle<2>& mbr, double x,
double y,
std::string& objectwords, double alpha,
double diag)
{
double result1 = 0.0;
double result2 = 0.0;
double textscore = 0.0;
if (objectwords.length() > 0)
{
textscore = getTextualScoreBT(objectwords);
}
result2 = double (1 - alpha) * textscore;
double dist = getDistanceBT(mbr, x, y);
double normalizedScore = 0.0;
if (dist == 0.0)
{
normalizedScore = 1;
}
else
{
normalizedScore = 1 - (double) (dist/diag);
}
result1 = alpha * normalizedScore;
return result1 + result2;
}
double getTextualScoreBT(std::string& objectwords)
{
int numMatches = 0;
int count = objectwords.length();
stringutils::StringTokenizer parse1(objectwords, " ");
while(parse1.hasNextToken())
{
std::string eval = parse1.nextToken();
CcString* stn = new CcString(true, eval);
size_t hash = getBucket(stn);
bucket = 0;
bucket = hashTable[hash];
bucketPos = 0;
// There is a match
if (bucket)
{
// Find the match
while(bucketPos < bucket->size())
{
WordInfo* p = (*bucket)[bucketPos];
if (((p->str)->GetValue()).compare(eval) == 0)
{
numMatches++;
break;
}
bucketPos++;
}
}
}
if (numMatches == 0)
{
return 0.0;
}
double uniquewords = (double)
(sumOfBoth + count - numMatches);
return (double) numMatches / uniquewords;
}
double getDistanceBT(Rectangle<2>& mbr, double x, double y)
{
double xmin = mbr.getMinX();
double ymin = mbr.getMinY();
double xmax = mbr.getMaxX();
double ymax = mbr.getMaxY();
if (xmin <= x && x <= xmax && ymin <= y && y <= ymax)
{
return 0.0;
}
// Upper quadrant
if (ymin > y)
{
//left
if (xmax < x) {
return EuclidDistRT(x, y, xmax ,ymin);
}
//Center
if (xmin <= x && x <= xmax) {
return EuclidDistRT(x, y, x ,ymin);
}
//Right
if(xmin > x)
{
return EuclidDistRT(x, y, xmin, ymin);
}
}
else
{
//left
if (xmax < x) {
return EuclidDistRT(x, y, xmax ,ymax);
}
//Center
if (xmin <= x && x <= xmax) {
return EuclidDistRT(x, y, x ,ymax);
}
//Right
if(xmin > x)
{
return EuclidDistRT(x, y, xmin, ymax);
}
}
if (ymin <= y && y <= ymax)
{
if (x < xmin)
{
return EuclidDistRT(x, y, xmin, y);
}
if (x > xmax)
{
return EuclidDistRT(x, y, xmax, y);
}
}
return 0.0;
}
};
int FilterBTSimMapValue( Word* args, Word& result,
int message, Word& local, Supplier s )
{
BatchTrajInfo* li = (BatchTrajInfo*) local.addr;
switch(message){
case OPEN: { if(li){
delete li;
}
ListExpr ttype = nl->Second(GetTupleResultType(s));
int noargs = qp->GetNoSons(s);
int idx1 = ((CcInt*)args[noargs-3].addr)->GetValue();
int idx2 = ((CcInt*)args[noargs-2].addr)->GetValue();
int idx3 = ((CcInt*)args[noargs-1].addr)->GetValue();
Rectangle<2> * mbr = static_cast<Rectangle<2>*>(args[5].addr);
CcReal * alpha = static_cast<CcReal*>(args[6].addr);
CcReal * thresh = static_cast<CcReal*>(args[7].addr);
Rectangle<2>* rec = static_cast<Rectangle<2>*>(args[8].addr);
double diagonal = GetDiag(*rec);
local.addr = new BatchTrajInfo(
9999,
*(mbr),
args[0],
args[1],
ttype,
idx1,
idx2,
idx3,
alpha->GetValue(),
thresh->GetValue(),
diagonal);
return 0;
}
case REQUEST: { if(!li){
return CANCEL;
}
result.addr = li->nextTuple();
return result.addr ? YIELD:CANCEL;
}
case CLOSE : {
if(li){
delete li;
local.addr=0;
}
return 0;
}
}
return 0;
}
/*
5.12 Operator ~sim~
5.12.1 Type map for ~sim~
*/
ListExpr SimTypeMap(ListExpr args)
{
NList type(args);
if(type != NList(SemanticTrajectory::BasicType(),
SemanticTrajectory::BasicType(),
Rectangle<2>::BasicType(),
CcReal::BasicType()))
{
return NList::typeError("Expecting two semantic"
"trajectories, a rectangle<2> and a real value for the alpha");
}
return NList(CcReal::BasicType()).listExpr();
}
int SimMapValue(Word* args,
Word& result,
int message,
Word& local,
Supplier s)
{
SemanticTrajectory* st1 = static_cast<SemanticTrajectory*>(args[0].addr);
SemanticTrajectory* st2 = static_cast<SemanticTrajectory*>(args[1].addr);
Rectangle<2>* rec = static_cast<Rectangle<2>*>(args[2].addr);
CcReal * alpha = static_cast<CcReal*>(args[3].addr);
result = qp->ResultStorage(s);
double diag = st1->GetDiagonal(*rec);
double answer = st1->Similarity(*st2, diag, alpha->GetValue());
((CcReal*) result.addr)->Set(answer);
// cout << answer << endl;
return 0;
}
/*
5.13 Operator ~bbsim~
5.13.1 Type map for ~bbsim~
*/
ListExpr BBSimTypeMap(ListExpr args)
{
if(!Batch::checkType(nl->First(args))
&& !Batch::checkType(nl->Second(args)))
{
return listutils::
typeError("expecting two tuples for first and second arguments");
}
if (!CcReal::checkType(nl->Third(args)) ||
!Rectangle<2>::checkType(nl->Fourth(args))){
return listutils::
typeError("third and fourth argument of real and rectangle");
}
return NList(CcReal::BasicType()).listExpr();
}
int BBSimMapValue(Word* args,
Word& result,
int message,
Word& local,
Supplier s)
{
Batch* t1 = static_cast<Batch*>( args[0].addr );
Batch* t2 = static_cast<Batch*>( args[1].addr );
CcReal * alpha = static_cast<CcReal*>(args[2].addr);
Rectangle<2>* rec = static_cast<Rectangle<2>*>(args[3].addr);
result = qp->ResultStorage(s);
double diag = GetDiag(*rec);
const Rectangle<2> r1 = t1->GetBoundingBox();
const Rectangle<2> r2 = t2->GetBoundingBox();
const Geoid* geoid = 0;
double distance = 0.0;
double alpha1 = alpha->GetValue();
if (alpha1 > 0)
{
distance = r1.Distance(r2, geoid);
}
double normalizedScore = 0.0;
if (distance == 0.0) {
normalizedScore = 1;
}
else
{
normalizedScore = 1 - (double)(distance/diag);
}
double answer = alpha1 * normalizedScore * 2 + double (1-alpha1) * 2;
((CcReal*) result.addr)->Set(answer);
return 0;
}
/*
5.14 Operator ~ttsim~
5.14.1 Type map for ~ttsim~
*/
ListExpr TTSimTypeMap(ListExpr args)
{
std::string err = "Should have 5 arguments";
if(!nl->HasLength(args, 5))
{
return listutils::typeError(err);
}
if(!SemanticTrajectory::checkType(nl->First(args))
&& !SemanticTrajectory::checkType(nl->Second(args)))
{
return listutils::
typeError("expecting two first arguments to first"
"arguments to be semantictrajectory");
}
if (!CcReal::checkType(nl->Third(args))){
return listutils::
typeError("third argument should be a real");
}
if (!Rectangle<2>::checkType(nl->Fourth(args)))
{
return listutils::
typeError("fourth argument should be a rectangle");
}
if(!CellGrid2D::checkType(nl->Fifth(args)))
{
return listutils::
typeError(err + "(5th arg must be CellGrid2D");
}
return NList(CcReal::BasicType()).listExpr();
}
int TTSimMapValue(Word* args,
Word& result,
int message,
Word& local,
Supplier s)
{
result = qp->ResultStorage(s);
CcReal* res = static_cast<CcReal*>(result.addr);
SemanticTrajectory* s1 = static_cast<SemanticTrajectory*>( args[0].addr );
SemanticTrajectory* s2 = static_cast<SemanticTrajectory*>( args[1].addr );
CcReal * alpha = static_cast<CcReal*>(args[2].addr);
Rectangle<2>* rec = static_cast<Rectangle<2>*>(args[3].addr);
CellGrid2D* grid = static_cast<CellGrid2D*>(args[4].addr);
double LH = 0.0;
double RH = 0.0;
double textualscore = 0.0;
double distancescore = 0.0;
double alpha1 = alpha->GetValue();
double textalpha = 1.0 - alpha1;
if (textalpha > 0)
{
textualscore = s1->textualScoreTT(*s2);
RH = (1.0 - textalpha) * textualscore;
}
if (alpha1 > 0)
{
double wx = grid->getXw();
double wy = grid->getYw();
distancescore = s1->MinDistAux(*s2, wx, wy);
}
double normalizedScore = 0.0;
if (distancescore != 0.0)
{
double diag = s1->GetDiagonal(*rec);
normalizedScore = 1 - (double)(distancescore/diag);
}
else {
normalizedScore = 1;
}
LH = alpha1 * normalizedScore;
double answer = LH + RH;
res->Set(true, answer);
return 0;
}
/*
5.15 Operator ~buildbatch~
5.15.1 Type map for ~buildbatch~
*/
ListExpr BuildBatchTypeMap(ListExpr args)
{
// Check to see if it's the right number of arguments
if (nl->HasLength(args, 8))
{
// Make sure each param is of right type
std::string err = "stream(tuple) x stream(tuple) x Ccint x"
" attr_2 x attr_3 x attr_4 x attr_5 x rect expected";
ListExpr stream = nl->First(args);
ListExpr stream2 = nl->Second(args);
ListExpr BatchId = nl->Third(args);
ListExpr attrname_word = nl->Fourth(args);
ListExpr attrname_ctn = nl->Fifth(args);
ListExpr attrname_tripid = nl->Sixth(args);
ListExpr attrname_st = nl->Seventh(args);
ListExpr r1 = nl->Eigth(args);
if(!listutils::isTupleStream(stream)){
return
listutils::typeError(
"first parameter must be a tuple stream");
}
if(!listutils::isTupleStream(stream2)){
return
listutils::typeError(
"second parameter must be a tuple stream");
}
if (!CcInt::checkType(BatchId))
{
return listutils::typeError(err + "3rd arg must be a int");
}
if(!listutils::isSymbol(attrname_word)){
return
listutils::typeError("4th parameter must"
"be an attribute name");
}
if(!listutils::isSymbol(attrname_ctn)){
return
listutils::typeError("5th parameter must"
" be an attribute name");
}
if(!listutils::isSymbol(attrname_tripid)){
return
listutils::typeError("5th parameter must"
"be an attribute name");
}
if(!listutils::isSymbol(attrname_st)){
return
listutils::typeError("6th parameter must"
" be an attribute name");
}
if (!Rectangle<2>::checkType(r1))
{
return listutils::typeError(err + "6th arg must be a rectangle");
}
ListExpr type;
// extract the attribute list
ListExpr attrList = nl->Second(nl->Second(stream));
ListExpr attrList2 = nl->Second(nl->Second(stream2));
// Get the index for the longitude
std::string name =
nl->SymbolValue(attrname_word);
int index1 =
listutils::findAttribute(attrList, name, type);
if(index1==0)
{
return
listutils::typeError(
"attribute name " + name +
" unknown in tuple"
" stream");
}
if(!CcString::checkType(type))
{
return
listutils::typeError("attribute "
+ name +
" must be of type real");
}
name = nl->SymbolValue(attrname_ctn);
int index2 =
listutils::findAttribute(attrList, name, type);
if(index2==0){
return
listutils::typeError("attribute name " + name +
" unknown in tuple stream");
}
if(!CcInt::checkType(type)){
return
listutils::typeError("attribute '" + name +
"' must be of type 'real'");
}
name = nl->SymbolValue(attrname_tripid);
int index3 =
listutils::findAttribute(attrList2, name, type);
if(index3==0){
return
listutils::typeError("attribute name " + name +
" unknown in tuple stream");
}
if(!CcInt::checkType(type)){
return
listutils::typeError("attribute '" + name +
"' must be of type 'real'");
}
name = nl->SymbolValue(attrname_st);
int index4 =
listutils::findAttribute(attrList2, name, type);
if(index4==0){
return
listutils::typeError("attribute name " + name +
" unknown in tuple stream");
}
if(!SemanticTrajectory::checkType(type)){
return
listutils::typeError("attribute '" + name +
"' must be of type 'real'");
}
std::string restype =
Batch::BasicType();
ListExpr indexes = nl->FourElemList(
nl->IntAtom(index1-1),
nl->IntAtom(index2-1),
nl->IntAtom(index3-1),
nl->IntAtom(index4-1)
);
return
nl->ThreeElemList(nl->SymbolAtom(Symbol::APPEND()),
indexes,
nl->SymbolAtom(restype));
}
return
listutils::typeError("Wrong number of arguments");
}
int BuildBatchMapValue(Word* args, Word& result,
int message, Word& local,
Supplier s)
{
result = qp->ResultStorage(s);
Batch* b =
static_cast<Batch*>(result.addr);
b->Clear();
int noargs = qp->GetNoSons(s);
int idx1 = ((CcInt*)args[noargs-4].addr)->GetValue();
int idx2 = ((CcInt*)args[noargs-3].addr)->GetValue();
int idx3 = ((CcInt*)args[noargs-2].addr)->GetValue();
int idx4 = ((CcInt*)args[noargs-1].addr)->GetValue();
Stream<Tuple> stream(args[0]);
Stream<Tuple> stream2(args[1]);
CcInt * batchId = static_cast<CcInt*>(args[2].addr);
Rectangle<2>* rec = static_cast<Rectangle<2>*>(args[7].addr);
b->SetBatchId(batchId->GetValue());
Tuple* tuple;
stream.open();
while((tuple = stream.request()))
{
CcString* x = (CcString*) tuple->GetAttribute(idx1);
CcInt* y = (CcInt*) tuple->GetAttribute(idx2);
b->AddBSumString(x->GetValue(), -1, y->GetValue());
tuple->DeleteIfAllowed();
}
double mind[2];
double maxd[2];
mind[0] = rec->getMinX();
mind[1] = rec->getMinY();
maxd[0] = rec->getMaxX();
maxd[1] = rec->getMaxY();
b->SetBoundingBox(true, mind, maxd);
/*
Retrieval of Trip information for batch starts here
*/
stream2.open();
int coordcounters = 0;
int cellcounters = 0;
int sumwordscounter = 0;
while((tuple = stream2.request()))
{
SemanticTrajectory* st = (SemanticTrajectory*) tuple->GetAttribute(idx4);
CcInt* id = (CcInt*) tuple->GetAttribute(idx3);
Trip t;
double mind[2];
double maxd[2];
const Rectangle<2> r = st->GetBoundingBox();
mind[0] = r.getMinX();
mind[1] = r.getMinY();
maxd[0] = r.getMaxX();
maxd[1] = r.getMaxY();
t.SetBoundingBox(true, mind, maxd);
t.SetId(id->GetValue());
/* TODO Collect the Coordinate information */
t.SetStartCoordsIdx(coordcounters);
for (int i = 0; i < st->GetNumCoordinates(); i++)
{
Coordinate c = st->GetCoordinate(i);
b->AddBCoordinate(c);
std::string holdValue = "";
st->GetSemString(i, holdValue);
b->AddBSemString(holdValue);
coordcounters++;
}
t.SetEndCoordsIdx(coordcounters);
/* TODO Collect the Cell information */
t.SetStartCellsIdx(cellcounters);
for (int i = 0; i < st->GetNumCells(); i++)
{
Cell c = st->GetCell(i);
b->AddBCell(c);
cellcounters++;
}
t.SetEndCellsIdx(cellcounters);
/* TODO Collect the Summary Information */
t.SetStartSumWordsIdx(sumwordscounter);
for (int i = 0; i < st->GetNumWords(); i++)
{
std::string holdValue = "";
bool success = st->GetStringSum(i, holdValue);
assert(success);
int id = st->GetWord(i).indexId;
int count = st->GetWord(i).count;
b->AddBWord(holdValue, id, count);
sumwordscounter++;
}
t.SetEndSumWordsIdx(sumwordscounter);
b->AddTrip(t);
tuple->DeleteIfAllowed();
}
stream.close();
stream2.close();
return 0;
}
/*
5.16 Operator ~getTrips~
5.16.1 Type map for ~getTrips~
*/
ListExpr GetTripsTypeMap(ListExpr args)
{
if (!nl->HasLength(args, 1)) {
return listutils::typeError("Only takes one argument");
}
if (!Batch::checkType(nl->First(args))) {
return listutils::typeError("Needs a batch");
}
return nl->TwoElemList(
nl->SymbolAtom(Symbol::STREAM()),
nl->TwoElemList(
nl->SymbolAtom(Tuple::BasicType()),
nl->TwoElemList(
nl->TwoElemList(nl->SymbolAtom("TripId"),
nl->SymbolAtom(CcInt::BasicType())
),
nl->TwoElemList(nl->SymbolAtom("Trip"),
nl->SymbolAtom(SemanticTrajectory::BasicType())
)
)
));
}
class TripInfo
{
public:
TripInfo(const ListExpr _resType,
int numTrips, Batch& bt) :
tt(0),visitedtrip(0),
numOfTrips(numTrips),
curidx(0), b(bt)
{
tt = new TupleType(_resType);
}
~TripInfo() {
tt->DeleteIfAllowed();
}
int getIndex()
{
if (visitedtrip < numOfTrips)
{
curidx = visitedtrip;
visitedtrip++;
return curidx;
}
return -1;
}
int getTripId() {
Trip t = b.GetTrip(curidx);
return t.GetId();
}
Tuple* nextTuple()
{
Tuple* res = new Tuple(tt);
/* Create ST */
Trip t = b.GetTrip(curidx);
int id = t.GetId();
res->PutAttribute(0, new CcInt(true, id));
SemanticTrajectory* st = new SemanticTrajectory(0);
double mind[2];
double maxd[2];
Rectangle<2> bbox = b.GetBoundingBox();
mind[0] = bbox.getMinX();
mind[1] = bbox.getMinY();
maxd[0] = bbox.getMaxX();
maxd[1] = bbox.getMaxY();
st->SetBoundingBox(true, mind, maxd);
int startCellIdx = t.GetStartCellsIdx();
int endCellIdx = t.GetEndCellsIdx();
if (startCellIdx < endCellIdx)
{
for (int i = startCellIdx; i < endCellIdx; i++)
{
Cell c = b.GetBCell(i);
st->AddCell(c);
}
}
int startTSumIdx = t.GetStartSumWordsIdx();
int endTSumIdx = t.GetEndSumWordsIdx();
if (startTSumIdx < endTSumIdx)
{
for (int i = startTSumIdx; i < endTSumIdx; i++)
{
WordST wordInfo = b.GetBWordInfo(i);
std::string holdValue = "";
b.GetBWord(i, holdValue);
st->AddStringSum(holdValue, wordInfo.indexId, wordInfo.count);
}
}
int startCoordsIdx = t.GetStartCoordsIdx();
int endCoordsIdx = t.GetEndCoordsIdx();
if (startCoordsIdx < endCoordsIdx)
{
for (int i = startCoordsIdx; i < endCoordsIdx; i++)
{
std::string holdvalue = "";
b.GetBSemString(i, holdvalue);
Coordinate c = b.GetBCoordinate(i);
st->AddCoordinate(c);
st->AddSemString(holdvalue);
}
}
res->PutAttribute(1, st);
return res;
}
Rectangle<2> getBoundingBox() {
return b.GetBoundingBox();
}
private:
TupleType* tt;
int visitedtrip;
int numOfTrips;
int curidx;
Batch b;
};
int GetTripsValueMap(Word* args, Word& result,
int message,
Word& local,
Supplier s)
{
TripInfo* localtrip =
(TripInfo*) local.addr;
switch(message)
{
case OPEN: // Initialize the local storage
{
if(localtrip)
{
delete localtrip;
localtrip = 0;
}
Batch* b =
static_cast<Batch*>(args[0].addr);
int numoftrips = b->GetNumTrips();
local.addr =
new TripInfo(nl->Second(GetTupleResultType(s)), numoftrips, *b);
return 0;
}
case REQUEST: // returnthe next stream element
{
int localidx = localtrip->getIndex();
if (localidx >= 0)
{
result.addr = localtrip->nextTuple();
return YIELD;
}
else
{
result.addr = 0;
return CANCEL;
}
}
case CLOSE:
{
if (localtrip != 0)
{
delete localtrip;
local.addr = 0;
}
return 0;
}
default:
{
// This should never happen
return -1;
}
}
return 0;
}
/*
5.17 Operator ~largerbatch~
5.17.1 Type map for ~largerbatch~
*/
ListExpr LargerBatchTypeMap(ListExpr args)
{
if (!nl->HasLength(args, 2)) {
return listutils::typeError("Only takes one argument");
}
if (!Batch::checkType(nl->First(args)) &&
!Batch::checkType(nl->Second(args))) {
return listutils::typeError("Needs a batch");
}
return NList(CcBool::BasicType()).listExpr();
}
int LargerBatchValueMap(Word* args, Word& result,
int message,
Word& local,
Supplier s)
{
Batch *b1 = static_cast<Batch*>( args[0].addr );
Batch *b2 = static_cast<Batch*>( args[1].addr );
Rectangle<2> r1 = b1->GetBoundingBox();
Rectangle<2> r2 = b2->GetBoundingBox();
result = qp->ResultStorage(s);
CcBool* b = static_cast<CcBool*>( result.addr );
b->Set(true, r1.Area() > r2.Area());
return 0;
}
/*
6 Operator Info
*/
struct LargerBatchInfo : OperatorInfo
{
LargerBatchInfo()
{
name = "largerbatch";
signature = "batch x batch -> bool";
syntax = "largerbatch(_,_)";
meaning =
"Compares area of two batches and"
"returns true if first param is the largest batch";
}
};
struct GetTripsInfo : OperatorInfo
{
GetTripsInfo()
{
name = "getTrips";
signature = "batch -> Stream(x)"
"batch -> Stream(x)";
syntax = "_ getTrips";
meaning =
"Get all trips data from a batch as a tuple with id and semantictrajectory";
}
};
struct BuildBatchInfo : OperatorInfo
{
BuildBatchInfo()
{
name = "buildbatch";
signature = "stream(tulple(X)) x stream(tuple(y))"
"int x a1 x a2 x a3 x a4 x rectangle -> batch";
syntax = "_ _ buildbatch[_,_,_,_,_,_]";
meaning =
"Pass a tuple stream to convert object into Batch Datatype";
}
};
struct BatchesInfo : OperatorInfo
{
BatchesInfo()
{
name = "batches";
signature = "stream(tulple(X)) x "
"a1 x a2 x real -> Stream (tuple(X))";
syntax = "_ batches[_,_,_]";
meaning =
"Assigns batchid to a list of semantictrajectory "
"~attribute a1 should be the ST"
"attribute a2 should batchId field"
" and attribute a3 should be the diagonal threshold value";
}
};
struct BBSimInfo : OperatorInfo
{
BBSimInfo()
{
name = "bbsim";
signature = "Batch x Batch x CcReal x Rectangle"
" -> CcReal";
syntax = "bbsim(_,_,_,_)";
meaning = "Evaluates two batch type"
"in order to calculate the spatial-textual distance and returns score";
}
};
struct BTSimInfo : OperatorInfo
{
BTSimInfo()
{
name = "btsim";
signature = "Batch x SemanticTrajectory "
"x CcReal x Rectangle"
" -> CcReal";
syntax = "btsim(_,_,_,_)";
meaning = "Evaluates trajectory-batch pair"
"using an uppberbound evaluation and returns score";
}
};
struct FilterBBSimInfo : OperatorInfo
{
FilterBBSimInfo()
{
name = "filterbbsim";
signature = "Stream(Tuple(x)) x id1 x id2"
"x MBR1 x MBR2 x CcReal x CcReal x Rectangle"
" -> Stream(Tuple(x))";
syntax = "_ filterbbsim[_,_,_,_,_,_,_]";
meaning = "Filters out batch-batch pair that are below the threshold"
"using an uppberbound evaluation"
"and return the batch pair tuple"
"with the batch with the largest Area 1st in tuple";
}
};
struct FilterBTSimInfo : OperatorInfo
{
FilterBTSimInfo()
{
name = "filterbtsim";
signature = "Stream(Tuple(x)) x Stream(Tuple(x)) "
"x attr x attr x attr x rect x real x real x rect"
" -> Stream(Tuple(x))";
syntax = "_ _ filterbtsim[_,_,_,_,_,_,_]";
meaning = "Filters out trajectory-batch pair"
"that are below the threshold"
"using an uppberbound evaluation";
}
};
struct FilterTTSimInfo : OperatorInfo
{
FilterTTSimInfo()
{
name = "filterttsim";
signature = "Stream(Tuple(x)) x attr x attr"
" x real x real x rect x CellGrid2D"
" -> Stream(Tuple(x))";
syntax = "_ filterttsim[_,_,_,_,_,_]";
meaning = "Filters out trajectory-trajectory pair"
" that are below the threshold"
"using an uppberbound evaluation";
}
};
struct FilterSimInfo : OperatorInfo
{
FilterSimInfo()
{
name = "filtersim";
signature = "Stream(Tuple(x)) x attr x attr x "
" real x real x rect"
" -> Stream(Tuple(x))";
syntax = "_ filtersim[_,_,_,_,_]";
meaning =
"Filter out trajectory-trajectory pair that are below threshold"
"by evaluating all the points";
}
};
struct TTSimInfo : OperatorInfo
{
TTSimInfo()
{
name = "ttsim";
signature = "semantictrajectory x semantictrajectory x real x "
" rectangle x CellGrid2D -> real";
syntax = "ttsim (_,_,_,_,_)";
meaning = "Returns the upperbound score between"
"two values semantictrajectory";
}
};
struct SimInfo : OperatorInfo
{
SimInfo()
{
name = "sim";
signature = SemanticTrajectory::BasicType() + " x "
+ SemanticTrajectory::BasicType()
+ " x " + Rectangle<2>::BasicType() + " x "
+ CcReal::BasicType() + " -> " + CcReal::BasicType();
syntax = "sim(_,_,_,_)";
meaning = "Get similarity score of two semantic Trajectories";
}
};
struct STBboxInfo : OperatorInfo
{
STBboxInfo()
{
name = "stbox";
signature = SemanticTrajectory::BasicType()
+ " -> " + Rectangle<2>::BasicType();
syntax = "stbox(_)";
meaning = "Returns the bounding box"
"of a semantic trajectory";
}
};
struct MakeSemTrajInfo : OperatorInfo
{
MakeSemTrajInfo()
{
name = "makesemtraj";
signature =
"stream(tuple(a1 t1) ...(an tn) ) "
" x ai x aj x ak-> semantictrajectory";
syntax = "_ makesemtraj [_,_,_]";
meaning =
"Convert stream of tuples"
" into a semantictrajectory datatype";
}
};
struct MakeSemTrajInfo2 : OperatorInfo
{
MakeSemTrajInfo2()
{
name = "makesemtraj2";
signature =
"stream(tuple(a1 t1) ...(an tn) ) "
" x ai x aj x ak-> semantictrajectory";
syntax = "_ _ makesemtraj2 [_,_,_,_]";
meaning =
"Convert stream of tuples"
" into a semantictrajectory datatype"
" + removes stopwords";
}
};
struct MakeSummariesInfo : OperatorInfo
{
MakeSummariesInfo()
{
name = "makesum";
signature =
"stream(tuple) x int x semantictrajectory x "
"cellgrid2d x attr x attr x attr"
" -> semantictrajectory";
syntax = "_ makesum [_,_,_,_,_,_]";
meaning =
"Retrieves cell origin coordinates from"
"a semantictrajectory to create spatial and textual summary";
}
};
struct ExtractKeywordsInfo : OperatorInfo
{
ExtractKeywordsInfo()
{
name = "extractkeywords";
signature =
"semantictrajectory -> stream(string)";
syntax = "_ extractkeywords";
meaning =
"Extracts semanticTrajectory string and"
" converts it into a"
"a stream of words";
}
};
/*
7 Type Constructor for SemanticTrajectory
*/
TypeConstructor semantictrajectory (
SemanticTrajectory::BasicType(),
//name
SemanticTrajectory::Property,
//property function
SemanticTrajectory::Out,
SemanticTrajectory::In,
//Out and In functions
0,
0,
//SaveTo and RestoreFrom functions
SemanticTrajectory::Create,
SemanticTrajectory::Delete,
//object creation and deletion
SemanticTrajectory::Open,
SemanticTrajectory::Save,
//object open and save
SemanticTrajectory::Close,
SemanticTrajectory::Clone,
//object close and clone
SemanticTrajectory::Cast,
//cast function
SemanticTrajectory::SizeOfObj,
//sizeof function
SemanticTrajectory::KindCheck);
//kind checking function
/*
8 Type Constructor for Batch
*/
TypeConstructor batch (
Batch::BasicType(),
//name
Batch::Property,
//property function
Batch::Out,
Batch::In,
//Out and In functions
0,
0,
//SaveTo and RestoreFrom functions
Batch::Create,
Batch::Delete,
//object creation and deletion
Batch::Open,
Batch::Save,
//object open and save
Batch::Close,
Batch::Clone,
//object close and clone
Batch::Cast,
//cast function
Batch::SizeOfObj,
//sizeof function
Batch::KindCheck);
//kind checking function
/*
9 SemanticTrajectoryAlgebra
*/
class SemanticTrajectoryAlgebra : public Algebra
{
public:
SemanticTrajectoryAlgebra() : Algebra()
{
AddTypeConstructor(&semantictrajectory);
semantictrajectory.AssociateKind(Kind::DATA());
AddTypeConstructor(&batch);
batch.AssociateKind(Kind::DATA());
AddOperator(
BuildBatchInfo(),
BuildBatchMapValue,
BuildBatchTypeMap);
AddOperator(
LargerBatchInfo(),
LargerBatchValueMap,
LargerBatchTypeMap
);
AddOperator(
GetTripsInfo(),
GetTripsValueMap,
GetTripsTypeMap
);
AddOperator(
FilterBBSimInfo(),
FilterBBSimMapValue,
FilterBBSimTypeMap
);
AddOperator(
BBSimInfo(),
BBSimMapValue,
BBSimTypeMap
);
AddOperator(
FilterBTSimInfo(),
FilterBTSimMapValue,
FilterBTSimTypeMap
);
AddOperator(
BTSimInfo(),
BTSimMapValue,
BTSimTypeMap
);
AddOperator(
FilterTTSimInfo(),
FilterTTSimMapValue,
FilterTTSimTypeMap
);
AddOperator(
TTSimInfo(),
TTSimMapValue,
TTSimTypeMap
);
AddOperator(
FilterSimInfo(),
FilterSimMapValue,
FilterSimTypeMap);
AddOperator(
SimInfo(),
SimMapValue,
SimTypeMap);
AddOperator(
MakeSemTrajInfo(),
makesemtrajFuns,
makesemtrajSelect,
TypeMapMakeSemtraj);
AddOperator(
MakeSemTrajInfo2(),
MakeSemTrajMV2,
TypeMapMakeSemtraj2);
AddOperator(
STBboxInfo(),
STbboxMapValue,
STboxTM);
AddOperator(
MakeSummariesInfo(),
MakesummariesMV,
MakesummariesTM );
AddOperator(
ExtractKeywordsInfo(),
extractkeywordMapV,
extractkeywordsTM);
AddOperator (
BatchesInfo(),
BatchesVM,
BatchesTM);
}
~SemanticTrajectoryAlgebra() {};
};
/*
10 Initialization
*/
extern "C"
Algebra*
InitializeSemanticTrajectoryAlgebra(NestedList *nlRef,
QueryProcessor *qpRef)
{
nl = nlRef;
qp = qpRef;
return (new SemanticTrajectoryAlgebra());
}
} // end of namespace
Reference in New Issue View Git Blame Copy Permalink