Actualiser Makefile

itself.

.gitmodules

@@ -1,3 +1,3 @@
 [submodule "sumalibs"]
 	path = sumalibs
-	url = https://git.metabarcoding.org/obitools/sumalibs.git
+	url = https://forge.metabarcoding.org/obitools/sumalibs.git

Makefile

@@ -1,3 +1,7 @@
+PREFIX=/usr/local
+
+CFLAGS=-I$(PREFIX)/include
+
 EXEC = sumatra
 
 SUMATRA_SRC = sumatra.c \
@@ -5,10 +9,9 @@ SUMATRA_SRC= sumatra.c \
 
 SUMATRA_OBJ = $(patsubst %.c,%.o,$(SUMATRA_SRC))
 
-
 SRCS = $(SUMATRA_SRC)
 
-LIB= -lfasta -llcs -lfile -lutils -lz -lm
+LIB = -lsuma -lz -lm
 
 include ./global.mk
 
@@ -23,8 +26,8 @@ all: $(EXEC)
 
 # executable compilation and link
 
-sumatra: $(SUMATRA_OBJ) $(LIBFASTA) $(LIBLCS) $(LIBFILE) $(LIBUTILS)
-	$(CC) $(LDFLAGS) -o $@ -pthread $(SUMATRA_OBJ) $(LIBFASTAPATH) $(LIBLCSPATH) $(LIBFILEPATH) $(LIBUTILSPATH) $(LIB)
+sumatra: $(SUMATRA_OBJ) ./sumalibs/libsuma.a
+	$(CC) $(LDFLAGS) -o $@ -pthread $(SUMATRA_OBJ) $(LIBSUMAPATH) $(LIB)
 
 ########
 #
@@ -33,12 +36,10 @@ sumatra: $(SUMATRA_OBJ) $(LIBFASTA) $(LIBLCS) $(LIBFILE) $(LIBUTILS)
 ########
 
 clean:
-	rm -f *.o	
-	rm -f *.P
+	rm -f $(SUMATRA_OBJ)
 	rm -f $(EXEC)
-	$(MAKE) -C ./sumalibs/libfasta clean
-	$(MAKE) -C ./sumalibs/liblcs clean
-	$(MAKE) -C ./sumalibs/libfile clean
-	$(MAKE) -C ./sumalibs/libutils clean
-
+	$(MAKE) -C ./sumalibs clean
 
+install: all
+	install -d $(DESTDIR)$(PREFIX)/bin/
+	install -m 755 $(EXEC) $(DESTDIR)$(PREFIX)/bin/

README.md

@@ -0,0 +1 @@
+[See the wiki](https://git.metabarcoding.org/obitools/sumatra/wikis/home)

global.mk

@@ -1,22 +1,15 @@
+LIBSUMAPATH = -L./sumalibs
 
-LIBFASTAPATH = -L./sumalibs/libfasta
-LIBLCSPATH   = -L./sumalibs/liblcs
-LIBFILEPATH  = -L./sumalibs/libfile
-LIBUTILSPATH = -L./sumalibs/libutils
-
-LIBFASTA = ./sumalibs/libfasta/libfasta.a
-LIBLCS   = ./sumalibs/liblcs/liblcs.a
-LIBFILE  = ./sumalibs/libfile/libfile.a
-LIBUTILS = ./sumalibs/libutils/libutils.a
+LIBSUMA = ./sumalibs/libsuma.a
 
 CC=gcc
 LDFLAGS=
 
 
 ifeq ($(CC),gcc)
-        CFLAGS = -O3 -s -DOMP_SUPPORT -fopenmp -w
+        CFLAGS = -I sumalibs -O3 -s -DOMP_SUPPORT -w
 else
-        CFLAGS = -O3 -w
+        CFLAGS = -I sumalibs -O3 -w
 endif
 
 
@@ -32,14 +25,5 @@ default: all
 #
 ########
 
-./sumalibs/libfasta/libfasta.a:
-	$(MAKE) -C ./sumalibs/libfasta
-
-./sumalibs/liblcs/liblcs.a:
-	$(MAKE) -C ./sumalibs/liblcs
-
-./sumalibs/libfile/libfile.a:
-	$(MAKE) -C ./sumalibs/libfile
-
-./sumalibs/libutils/libutils.a:
-	$(MAKE) -C ./sumalibs/libutils
+./sumalibs/libsuma.a:
+	$(MAKE) -C ./sumalibs

mtcompare_sumatra.c

@@ -10,10 +10,10 @@
 #include <stdio.h>
 #include <string.h>
 #include "sumatra.h"
-#include "./sumalibs/libfasta/sequence.h"
-#include "./sumalibs/libutils/utilities.h"
-#include "./sumalibs/liblcs/upperband.h"
-#include "./sumalibs/liblcs/sse_banded_LCS_alignment.h"
+#include "libfasta/sequence.h"
+#include "libutils/utilities.h"
+#include "liblcs/upperband.h"
+#include "liblcs/sse_banded_LCS_alignment.h"
 
 
 typedef struct {

sumatra.c

@@ -14,13 +14,13 @@
 
 #include <sys/time.h>
 
-#include "./sumalibs/libfasta/sequence.h"
-#include "./sumalibs/liblcs/upperband.h"
-#include "./sumalibs/liblcs/sse_banded_LCS_alignment.h"
-#include "./sumalibs/libutils/utilities.h"
+#include "libfasta/sequence.h"
+#include "liblcs/upperband.h"
+#include "liblcs/sse_banded_LCS_alignment.h"
+#include "libutils/utilities.h"
 #include "mtcompare_sumatra.h"
 
-#define VERSION "1.0.10"
+#define VERSION "1.0.36"
 
 
 /* ----------------------------------------------- */
@@ -57,7 +57,9 @@ static void PrintHelp()
         PP      " -g       : n's are replaced with a's (default: sequences with n's are discarded).\n");
         PP      " -x       : Adds four extra columns with the count and length of both sequences.\n");
         PP      "-----------------------------------------------------------------------------------------------------------------------------\n");
-        PP      " First argument  : the nucleotide dataset to analyze\n\n");
+        PP      " First argument  : the nucleotide dataset to analyze (or nothing   \n");
+        PP      "                   if there is only one dataset and the standard   \n");
+        PP		"                   input should be used).                        \n\n");
         PP      " Second argument : optionally the second nucleotide dataset\n");
         PP      "-----------------------------------------------------------------------------------------------------------------------------\n");
         PP      " Results table description : \n");
@@ -84,7 +86,7 @@ static void PrintHelp()
 static void ExitUsage(stat)
         int stat;
 {
-        PP      "usage: sumatra [-l|L|a|n|r|d|g|x] [-t threshold_value] [-p number of threads] dataset1 [dataset2]\n");
+        PP      "usage: sumatra [-l|L|a|n|r|d|g|x] [-t threshold_value] [-p number of threads] [dataset1] [dataset2]\n");
         PP      "type \"sumatra -h\" for help\n");
 
         if (stat)
@@ -212,7 +214,7 @@ int compare1(fastaSeqCount db1, double threshold, BOOL normalize, int reference,
 	BOOL     always = TRUE;
 	int64_t  pairs = (int64_t)(db1.count - 1) * (int64_t)db1.count /2;
 	BOOL     print;
-	double   score;
+	double   score, scoreG;
 	int32_t  i,j;
 	char* s1;
 	char* s2;
@@ -237,7 +239,7 @@ int compare1(fastaSeqCount db1, double threshold, BOOL normalize, int reference,
 	calculateMaxAndMinLenDB(db1, &lmax, &lmin);
 	sizeForSeqs = prepareTablesForSumathings(lmax, lmin, threshold, normalize, reference, lcsmode, &address, &iseq1, &iseq2);
 
-	for (i=0; i < db1.count; i++) // ...??
+	for (i=0; i < db1.count; i++) // ...?? db1.count - 1 probably
 		for (j=i+1; j < db1.count; j++)
 		{
 			print = FALSE;
@@ -250,7 +252,23 @@ int compare1(fastaSeqCount db1, double threshold, BOOL normalize, int reference,
 				l1 = (db1.fastaSeqs+i)->length;
 				s2 = (db1.fastaSeqs+j)->sequence;
 				l2 = (db1.fastaSeqs+j)->length;
+
+/*				fprintf(stderr, "\n%s", s1);
+				fprintf(stderr, "\n%s", s2);
+				fprintf(stderr, "\n%f", threshold);
+				fprintf(stderr, "\n%d", normalize);
+				fprintf(stderr, "\n%d", reference);
+				fprintf(stderr, "\n%d\n", lcsmode);
+*/
+//				score = generic_sse_banded_lcs_align(s1, s2, threshold, normalize, reference, lcsmode);
+//				fprintf(stderr, "\nscore generic = %f", scoreG);
 				score = alignForSumathings(s1, iseq1, s2, iseq2, l1, l2, normalize, reference, lcsmode, address, sizeForSeqs, LCSmin);
+//				fprintf(stderr, "\nscore = %f\n", score);
+//				if (scoreG != score)
+//				{
+//					fprintf(stderr, "\nscores differents\n");
+//					exit(1);
+//				}
 				print = always || (((normalize || lcsmode) && (score >= threshold)) || ((!lcsmode && !normalize) && (score <= threshold)));
 				if (print && !lcsmode && normalize)
 					score = 1.0 - score;
@@ -320,7 +338,7 @@ int compare2(fastaSeqCount db1, fastaSeqCount db2, double threshold, BOOL normal
 				score = alignForSumathings(s1, iseq1, s2, iseq2, l1, l2, normalize, reference, lcsmode, address, sizeForSeqs, LCSmin);
 				print = always || (((normalize || lcsmode) && (score >= threshold)) || ((!lcsmode && !normalize) && (score <= threshold)));
 				if (print && !lcsmode && normalize)
-					score = 1.0 - score;
+					score = 1.0 - score;	// TODO isn't that already done?
 			}
 			printResults(db1.fastaSeqs+i, db2.fastaSeqs+j, score, extradata, pairs, print);
 		}
@@ -425,8 +443,6 @@ int main(int argc, char **argv)
 	}
 
 	ndb = argc - optind;
-	if (ndb < 1)
-        errflag++;
 
 	if (errflag)
 		ExitUsage(errflag);

sumatra.h

@@ -8,7 +8,7 @@
 #ifndef SUMATRA_H_
 #define SUMATRA_H_
 
-#include "./sumalibs/libfasta/sequence.h"
+#include "libfasta/sequence.h"
 
 void printResults(fastaSeqPtr seq1, fastaSeqPtr seq2, double score, BOOL extradata, int64_t pairs, BOOL print);
 

sumatra_user_manual.md

@@ -29,7 +29,8 @@ Untar the archive, go into the newly created directory and compile:
 ```
 tar –zxvf sumatra_v[x.x.xx].tar.gz
 cd sumatra_v[x.x.xx]
-make
+make -C sumalibs install
+make install
 ```
 
 ## Documentation
@@ -40,12 +41,12 @@ Sumatra computes the pairwise alignment scores from one dataset or between two d
 
 #### Input
 
-Files must be in FASTA format.
+If there is one dataset, the input can be either the standard input (stdin), or a file in FASTA format. If there are two datasets to compare, the input must be two files in FASTA format.
 
 #### Usage
 
 ```
-sumatra [-l|L|a|n|r|d|g|x] [-t threshold_value] [-p number of threads] dataset1 [dataset2] 
+sumatra [-l|L|a|n|r|d|g|x] [-t threshold_value] [-p number of threads] [dataset1] [dataset2]
 ```
 First argument: the sequence dataset in fasta format to analyse.