merge conflicts

Makefile

@@ -31,7 +31,7 @@ SUBDIRS = $(SRC_DIR)/annotateBed \
 		  $(SRC_DIR)/bedpeToBam \
 		  $(SRC_DIR)/bedToIgv \
 		  $(SRC_DIR)/bed12ToBed6 \
-		  $(SRC_DIR)/closestBed \
+		  $(SRC_DIR)/closestFile \
 		  $(SRC_DIR)/clusterBed \
 		  $(SRC_DIR)/complementBed \
 		  $(SRC_DIR)/coverageBed \

src/bedtools.cpp

@@ -141,9 +141,9 @@ int main(int argc, char *argv[])
     else if (sub_cmd == "makewindows") return windowmaker_main(argc-1, argv+1);
     else if (sub_cmd == "groupby")     return groupby_main(argc-1, argv+1);
     else if (sub_cmd == "expand")      return expand_main(argc-1, argv+1);
-    else if (sub_cmd == "sample")		return sample_main(argc-1, argv+1);
-    else if (sub_cmd == "neksb1")	   	return nek_sandbox1_main(argc-1, argv+1);
-    else if (sub_cmd == "regresstest")	return regress_test_main(argc, argv); //this command does need all the orig args.
+    else if (sub_cmd == "sample")       return sample_main(argc-1, argv+1);
+    else if (sub_cmd == "neksb1")       return nek_sandbox1_main(argc-1, argv+1);
+    else if (sub_cmd == "regresstest")  return regress_test_main(argc, argv); //this command does need all the orig args.
     // help
     else if (sub_cmd == "-h" || sub_cmd == "--help" ||
              sub_cmd == "-help")

src/closestBed/closestBed.h

-/*****************************************************************************
-  closestBed.h
-
-  (c) 2009 - Aaron Quinlan
-  Hall Laboratory
-  Department of Biochemistry and Molecular Genetics
-  University of Virginia
-  aaronquinlan@gmail.com
-
-  Licenced under the GNU General Public License 2.0 license.
-******************************************************************************/
-#ifndef CLOSESTBED_H
-#define CLOSESTBED_H
-
-#include "bedFile.h"
-#include <vector>
-#include <iostream>
-#include <fstream>
-
-using namespace std;
-
-//************************************************
-// Class methods and elements
-//************************************************
-class BedClosest {
-
-public:
-
-    // constructor
-    BedClosest(string &bedAFile, string &bedBFile, 
-               bool sameStrand, bool diffStrand, string &tieMode, 
-               bool reportDistance, bool signDistance, string &strandedDistMode,
-               bool ignoreOverlaps, bool ignoreUpstream, bool ignoreDownstream,
-               bool printHeader, bool diffNames);
-
-    // destructor
-    ~BedClosest(void);
-
-    // find the closest feature in B to A
-    void FindClosestBed();
-
-private:
-
-    // data
-    string _bedAFile;
-    string _bedBFile;
-    string _tieMode;
-    bool   _sameStrand;
-    bool   _diffStrand;
-    bool   _reportDistance;
-    bool   _signDistance;
-    string _strandedDistMode;
-    bool   _ignoreOverlaps;
-    bool   _ignoreUpstream;
-    bool   _ignoreDownstream;
-    bool   _printHeader;
-    bool   _diffNames;
-
-    BedFile *_bedA, *_bedB;
-
-    // methods
-    void FindWindowOverlaps(BED &, vector<BED> &);
-    void ReportClosestNotFound(const BED &a);
-
-};
-#endif /* CLOSEST_H */

src/closestBed/Makefile → src/closestFile/Makefile

-UTILITIES_DIR = ../utils/
-OBJ_DIR = ../../obj/
-BIN_DIR = ../../bin/
-
-# -------------------
-# define our includes
-# -------------------
-INCLUDES = -I$(UTILITIES_DIR)/bedFile/ \
-           -I$(UTILITIES_DIR)/lineFileUtilities/ \
-           -I$(UTILITIES_DIR)/gzstream/ \
-           -I$(UTILITIES_DIR)/fileType/ \
-           -I$(UTILITIES_DIR)/version/
-# ----------------------------------
-# define our source and object files
-# ----------------------------------
-SOURCES= closestMain.cpp closestBed.cpp closestBed.h
-OBJECTS= closestMain.o closestBed.o
-BUILT_OBJECTS= $(patsubst %,$(OBJ_DIR)/%,$(OBJECTS))
-
-all: $(BUILT_OBJECTS)
-
-.PHONY: all
-
-$(BUILT_OBJECTS): $(SOURCES)
-	@echo "  * compiling" $(*F).cpp
-	@$(CXX) -c -o $@ $(*F).cpp $(LDFLAGS) $(CXXFLAGS) $(INCLUDES)
-
-clean:
-	@echo "Cleaning up."
-	@rm -f $(OBJ_DIR)/closestMain.o  $(OBJ_DIR)/closestBed.o
-	
-.PHONY: clean
+UTILITIES_DIR = ../utils/
+OBJ_DIR = ../../obj/
+BIN_DIR = ../../bin/
+
+# -------------------
+# define our includes
+# -------------------
+INCLUDES = -I$(UTILITIES_DIR)/Contexts/ \
+			-I$(UTILITIES_DIR)/general/ \
+			-I$(UTILITIES_DIR)/fileType/ \
+			-I$(UTILITIES_DIR)/gzstream/ \
+           -I$(UTILITIES_DIR)/GenomeFile/ \
+           -I$(UTILITIES_DIR)/BamTools/include \
+           -I$(UTILITIES_DIR)/BamTools/src \
+           -I$(UTILITIES_DIR)/BlockedIntervals \
+           -I$(UTILITIES_DIR)/BamTools-Ancillary \
+           -I$(UTILITIES_DIR)/FileRecordTools/ \
+           -I$(UTILITIES_DIR)/FileRecordTools/FileReaders/ \
+           -I$(UTILITIES_DIR)/FileRecordTools/Records/ \
+			-I$(UTILITIES_DIR)/KeyListOps/ \
+           -I$(UTILITIES_DIR)/RecordOutputMgr/ \
+           -I$(UTILITIES_DIR)/NewChromsweep \
+           -I$(UTILITIES_DIR)/BinTree \
+           -I$(UTILITIES_DIR)/version/
+# ----------------------------------
+# define our source and object files
+# ----------------------------------
+SOURCES= closestMain.cpp closestFile.cpp closestFile.h
+OBJECTS= closestMain.o closestFile.o
+BUILT_OBJECTS= $(patsubst %,$(OBJ_DIR)/%,$(OBJECTS))
+
+all: $(BUILT_OBJECTS)
+
+.PHONY: all
+
+$(BUILT_OBJECTS): $(SOURCES)
+	@echo "  * compiling" $(*F).cpp
+	@$(CXX) -c -o $@ $(*F).cpp $(LDFLAGS) $(CXXFLAGS) $(INCLUDES)
+
+clean:
+	@echo "Cleaning up."
+	@rm -f $(OBJ_DIR)/closestMain.o  $(OBJ_DIR)/closestFile.o
+	
+.PHONY: clean

src/closestFile/closestFile.cpp

+/*
+ * newClosestFile.cpp
+ *
+ *  Created on: Sep 25, 2014
+ *      Author: nek3d
+ */
+
+#include "FileRecordMgr.h"
+#include "RecordOutputMgr.h"
+#include "closestFile.h"
+#include "CloseSweep.h"
+
+ClosestFile::ClosestFile(ContextClosest *context)
+: _context(context),
+ _recordOutputMgr(NULL)
+{
+	_recordOutputMgr = new RecordOutputMgr();
+	_recordOutputMgr->init(_context);
+}
+
+ClosestFile::~ClosestFile() {
+	delete _recordOutputMgr;
+}
+
+bool ClosestFile::getClosest() {
+    CloseSweep sweep(_context);
+    if (!sweep.init()) {
+      return false;
+    }
+    RecordKeyVector hitSet;
+    while (sweep.next(hitSet)) {
+    	if (_context->reportDistance()) {
+    		_recordOutputMgr->printClosest(hitSet, &(sweep.getDistances()));
+    	} else {
+    		_recordOutputMgr->printClosest(hitSet, NULL);
+
+    	}
+    }
+    return true;
+
+}

src/closestFile/closestFile.h

+/*
+ * newClosestFile.h
+ *
+ *  Created on: Sep 25, 2014
+ *      Author: nek3d
+ */
+
+#ifndef NEWCLOSESTFILE_H_
+#define NEWCLOSESTFILE_H_
+
+#include "ContextClosest.h"
+
+using namespace std;
+
+class RecordOutputMgr;
+
+class ClosestFile {
+
+public:
+    ClosestFile(ContextClosest *context);
+    ~ClosestFile(void);
+
+    bool getClosest();
+
+private:
+    ContextClosest *_context;
+    RecordOutputMgr *_recordOutputMgr;
+};
+
+
+#endif /* NEWCLOSESTFILE_H_ */

src/utils/Contexts/ContextClosest.cpp

+/*
+ * ContextClosest.cpp
+ *
+ *  Created on: Sep 25, 2014
+ *      Author: nek3d
+ */
+
+#include "ContextClosest.h"
+
+ContextClosest::ContextClosest()
+: 	_haveTieMode(false),
+	_ignoreOverlaps(false),
+	_ignoreUpstream(false),
+	_ignoreDownstream(false),
+	_reportDistance(false),
+	_signDistance(false),
+	_haveStrandedDistMode(false),
+	_diffNames(false),
+	_tieMode(ALL_TIES),
+	_strandedDistMode(REF_DIST),
+	_multiDbMode(EACH_DB)
+{
+	// closest requires sorted input
+	setSortedInput(true);
+
+}
+
+ContextClosest::~ContextClosest(){
+
+}
+
+bool ContextClosest::parseCmdArgs(int argc, char **argv, int skipFirstArgs){
+	_argc = argc;
+	_argv = argv;
+	_skipFirstArgs = skipFirstArgs;
+	if (_argc < 2) {
+		setShowHelp(true);
+		return false;
+	}
+
+	setProgram(_programNames[argv[0]]);
+
+	_argsProcessed.resize(_argc - _skipFirstArgs, false);
+
+	for (_i=_skipFirstArgs; _i < argc; _i++) {
+		if (isUsed(_i - _skipFirstArgs)) {
+			continue;
+		}
+		if (strcmp(_argv[_i], "-c") == 0) {
+			//bypass intersect's use of the -c option, because -c
+			//means writeCount for intersect, but means columns for map.
+			if (!ContextBase::handle_c()) return false;
+		}
+        else if (strcmp(_argv[_i], "-d") == 0) {
+           if (!handle_d()) return false;
+        }
+        else if (strcmp(_argv[_i], "-D") == 0) {
+        	if (!handle_D()) return false;
+        }
+        else if (strcmp(_argv[_i], "-io") == 0) {
+        	if (!handle_io()) return false;
+        }
+        else if (strcmp(_argv[_i], "-iu") == 0) {
+        	if (!handle_iu()) return false;
+        }
+        else if (strcmp(_argv[_i], "-id") == 0) {
+        	if (!handle_id()) return false;
+        }
+        else if (strcmp(_argv[_i], "-N") == 0) {
+        	if (!handle_N()) return false;
+        }
+        else if (strcmp(_argv[_i], "-t") == 0) {
+        	if (!handle_t()) return false;
+        }
+        else if (strcmp(_argv[_i], "-mdb") == 0) {
+        	if (!handle_mdb()) return false;
+        }
+
+	}
+	return ContextIntersect::parseCmdArgs(argc, argv, _skipFirstArgs);
+}
+
+bool ContextClosest::isValidState(){
+	if (!ContextIntersect::isValidState()) return false;
+
+   // make sure we have both input files
+	if (_haveTieMode && (_tieMode != ALL_TIES) && (_tieMode != FIRST_TIE)
+					&& (_tieMode != LAST_TIE)) {
+		_errorMsg = "\n*****\n*****ERROR: Request \"all\" or \"first\" or \"last\" for Tie Mode (-t)\n*****\n";
+		return false;
+	}
+
+	if (_haveStrandedDistMode && (_strandedDistMode != A_DIST) && (_strandedDistMode != B_DIST)
+							 && (_strandedDistMode != REF_DIST)) {
+		_errorMsg = "\n*****\n*****ERROR: Request \"a\" or \"b\" or \"ref\" for Stranded Distance Mode (-D)\n*****\n";
+		return false;
+	}
+
+	if (_ignoreUpstream && _ignoreDownstream) {
+		_errorMsg = "\n*****\n*****ERROR: Request either -iu OR -id, not both.\n*****\n";
+		return false;
+	}
+
+	if ((_ignoreUpstream || _ignoreDownstream) && ! _haveStrandedDistMode) {
+		_errorMsg  = "\n*****\n*****ERROR: When requesting -iu or -id, you also need to specify -D.\n*****\n";
+		return false;
+	}
+
+	return true;
+}
+
+bool ContextClosest::handle_d() {
+    _reportDistance = true;
+    markUsed(_i - _skipFirstArgs);
+    return true;
+}
+
+bool ContextClosest::handle_D() {
+	bool strandError = false;
+    if ((_i+1) < _argc) {
+        _reportDistance = true;
+        _signDistance = true;
+        _haveStrandedDistMode = true;
+        QuickString modeStr(_argv[_i + 1]);
+        if (modeStr == "ref") {
+        	_strandedDistMode = REF_DIST;
+        } else if (modeStr == "a") {
+        	_strandedDistMode = A_DIST;
+        } else if (modeStr == "b") {
+        	_strandedDistMode = B_DIST;
+        } else {
+        	strandError = true;
+        }
+    } else {
+    	strandError = true;
+    }
+    if (!strandError) {
+        markUsed(_i - _skipFirstArgs);
+        _i++;
+        markUsed(_i - _skipFirstArgs);
+        return true;
+    }
+    _errorMsg = "*****ERROR: -D option must be followed with \"ref\", \"a\", or \"b\"";
+    return false;
+}
+
+bool ContextClosest::handle_io() {
+    _ignoreOverlaps = true;
+    markUsed(_i - _skipFirstArgs);
+    return true;
+}
+
+bool ContextClosest::handle_iu() {
+	_ignoreUpstream = true;
+    markUsed(_i - _skipFirstArgs);
+    return true;
+}
+
+bool ContextClosest::handle_id() {
+	_ignoreUpstream = true;
+    markUsed(_i - _skipFirstArgs);
+    return true;
+}
+
+bool ContextClosest::handle_N() {
+    _diffNames = true;
+    markUsed(_i - _skipFirstArgs);
+    return true;
+}
+
+bool ContextClosest::handle_t()
+{
+	bool tieError = false;
+    if ((_i+1) < _argc) {
+        _haveTieMode = true;
+        QuickString tieStr(_argv[_i+1]);
+        if (tieStr == "all") {
+        	_tieMode = ALL_TIES;
+        } else if (tieStr == "first") {
+        	_tieMode = FIRST_TIE;
+        } else if (tieStr == "last") {
+        	_tieMode = LAST_TIE;
+        } else {
+        	tieError = true;
+        }
+    } else {
+    	tieError = true;
+    }
+    if (!tieError) {
+        markUsed(_i - _skipFirstArgs);
+        _i++;
+        markUsed(_i - _skipFirstArgs);
+        return true;
+    }
+	_errorMsg = "*****ERROR: Request \"all\", \"first\", \"last\" for Tie Mode (-t)";
+	return false;
+}
+
+bool ContextClosest::handle_mdb()
+{
+	bool mdbError = false;
+    if ((_i+1) < _argc) {
+        QuickString mdbStr(_argv[_i+1]);
+        if (mdbStr == "each") {
+        	_multiDbMode = EACH_DB;
+        } else if (mdbStr == "all") {
+        	_multiDbMode = ALL_DBS;
+        } else {
+        	mdbError = true;
+        }
+    } else {
+    	mdbError = true;
+    }
+    if (!mdbError) {
+        markUsed(_i - _skipFirstArgs);
+        _i++;
+        markUsed(_i - _skipFirstArgs);
+        return true;
+    }
+	_errorMsg = "*****ERROR: Request \"each\" or \"last\" for Multiple Database Mode (-mdb)";
+	return false;
+}

src/utils/Contexts/ContextClosest.h

+/*
+ * ContextClosest.h
+ *
+ *  Created on: Sep 25, 2014
+ *      Author: nek3d
+ */
+
+
+#ifndef CONTEXTCLOSEST_H_
+#define CONTEXTCLOSEST_H_
+
+#include "ContextIntersect.h"
+
+class ContextClosest : public ContextIntersect {
+public:
+	ContextClosest();
+	virtual ~ContextClosest();
+	virtual bool parseCmdArgs(int argc, char **argv, int skipFirstArgs);
+    virtual bool hasIntersectMethods() const { return true; }
+    virtual bool isValidState();
+
+    bool hasTieMode() const { return _haveTieMode; }
+    bool ignoreOverlaps() const { return _ignoreOverlaps; }
+    bool ignoreUpstream() const { return _ignoreUpstream; }
+    bool ignoreDownstream() const { return _ignoreDownstream; }
+    bool reportDistance() const { return _reportDistance; }
+    bool signDistance() const { return _signDistance; }
+    bool hasStrandedDistMode() const { return _haveStrandedDistMode; }
+    bool diffNames() const { return _diffNames; }
+
+    typedef enum { FIRST_TIE, LAST_TIE, ALL_TIES} tieModeType;
+    tieModeType getTieMode() const { return _tieMode; }
+
+    typedef enum { REF_DIST, A_DIST, B_DIST} strandedDistanceModeType;
+    strandedDistanceModeType getStrandedDistMode() const { return _strandedDistMode; }
+
+    typedef enum { EACH_DB, ALL_DBS } multiDbModeType;
+    multiDbModeType getMultiDbMode() const { return _multiDbMode; }
+
+private:
+    bool _haveTieMode;
+    bool _ignoreOverlaps;
+    bool _ignoreUpstream;
+    bool _ignoreDownstream;
+    bool _reportDistance;
+    bool _signDistance;
+    bool _haveStrandedDistMode;
+    bool _diffNames;
+    tieModeType _tieMode;
+    strandedDistanceModeType _strandedDistMode;
+    multiDbModeType _multiDbMode;
+
+    bool handle_d();
+    bool handle_D();
+    bool handle_io();
+    bool handle_iu();
+    bool handle_id();
+    bool handle_N();
+    bool handle_t();
+    bool handle_mdb();
+
+};
+
+
+#endif /* CONTEXTCLOSEST_H_ */

src/utils/Contexts/Makefile

@@ -19,9 +19,10 @@ INCLUDES =  -I$(UTILITIES_DIR)/general/ \
 # ----------------------------------
 # define our source and object files
 # ----------------------------------
-SOURCES= ContextBase.cpp ContextBase.h ContextIntersect.cpp ContextIntersect.h ContextMap.cpp \
-	ContextMap.h ContextSample.cpp ContextSample.h ContextMerge.h ContextMerge.cpp ContextJaccard.h ContextJaccard.cpp ContextFisher.h ContextFisher.cpp
-OBJECTS= ContextBase.o ContextIntersect.o ContextMap.o ContextSample.o ContextMerge.o ContextJaccard.o ContextFisher.o
+SOURCES= ContextBase.cpp ContextBase.h ContextIntersect.cpp ContextIntersect.h ContextFisher.cpp ContextFisher.h ContextMap.cpp \
+	ContextMap.h ContextSample.cpp ContextSample.h ContextMerge.h ContextMerge.cpp ContextJaccard.h ContextJaccard.cpp \
+	ContextClosest.cpp ContextClosest.h
+OBJECTS= ContextBase.o ContextIntersect.o ContextFisher.o ContextMap.o ContextSample.o ContextMerge.o ContextJaccard.o ContextClosest.o
 _EXT_OBJECTS=ParseTools.o QuickString.o
 EXT_OBJECTS=$(patsubst %,$(OBJ_DIR)/%,$(_EXT_OBJECTS))
 BUILT_OBJECTS= $(patsubst %,$(OBJ_DIR)/%,$(OBJECTS))
@@ -38,10 +39,11 @@ clean:
 	@echo "Cleaning up."
 	@rm -f $(OBJ_DIR)/ContextBase.o \
 	       $(OBJ_DIR)/ContextIntersect.o \
+	       $(OBJ_DIR)/ContextFisher.o \
 	       $(OBJ_DIR)/ContextMap.o \
 	       $(OBJ_DIR)/ContextSample.o \
 	       $(OBJ_DIR)/ContextMerge.o \
 	       $(OBJ_DIR)/ContextJaccard.o \
-	       $(OBJ_DIR)/ContextFisher.o \
+	       $(OBJ_DIR)/ContextClosest.o \
 
 .PHONY: clean

src/utils/FileRecordTools/Records/Record.cpp

@@ -122,7 +122,7 @@ bool Record::chromAfter(const Record *other) const
 
 bool Record::after(const Record *other) const
 {
-	return (_chrId == other->_chrId && _startPos >= other->_endPos);
+	return (sameChrom(other) && _startPos >= other->_endPos);
 }
 
 bool Record::intersects(const Record *record,

src/utils/NewChromsweep/CloseSweep.h

+/*
+ * CloseSweep.h
+ *
+ *  Created on: Sep 25, 2014
+ *      Author: nek3d
+ */
+
+#ifndef CLOSESWEEP_H_
+#define CLOSESWEEP_H_
+
+#include "NewChromsweep.h"
+
+class ContextClosest;
+
+class CloseSweep : public NewChromSweep {
+public:
+	CloseSweep(ContextClosest *context);
+	~CloseSweep(void);
+
+	const vector<int> &getDistances() { return _finalDistances; }
+
+private:
+   ContextClosest *_context;
+
+	typedef vector<const Record * > distRecVecType;
+	vector<distRecVecType *> _minUpstreamRecs;
+	vector<int> _minUpstreamDist;
+	vector<distRecVecType *> _minDownstreamRecs;
+	vector<int> _minDownstreamDist;
+	vector<distRecVecType *> _overlapRecs;
+	vector<int> _maxPrevLeftClosestEndPos;
+
+	vector<int> _finalDistances;
+
+
+	//structs to help with finding closest among all of multiple dbs.
+	RecordKeyVector _copyRetList;
+	vector<int> _copyDists;
+
+	//override these two methods from chromsweep
+	void masterScan(RecordKeyVector &retList);
+    void scanCache(int dbIdx, RecordKeyVector &retList);
+    bool chromChange(int dbIdx, RecordKeyVector &retList);
+
+
+	typedef enum { IGNORE, DELETE } rateOvlpType;
+    rateOvlpType considerRecord(const Record *cacheRec, int dbIdx, bool &stopScanning);
+    void finalizeSelections(int dbIdx, RecordKeyVector &retList);
+    void checkMultiDbs(RecordKeyVector &retList);
+};
+
+
+#endif /* CLOSESWEEP_H_ */

src/utils/NewChromsweep/Makefile

-OBJ_DIR = ../../../obj/
-BIN_DIR = ../../../bin/
-UTILITIES_DIR = ../../utils/
-# -------------------
-# define our includes
-# -------------------