Welcome to mirror list, hosted at ThFree Co, Russian Federation.

github.com/moses-smt/salm.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat (limited to 'Src/SuffixArrayApplications/SuffixArrayScan/Applications/CalcCountOfCounts.cpp')
-rwxr-xr-xSrc/SuffixArrayApplications/SuffixArrayScan/Applications/CalcCountOfCounts.cpp34
1 files changed, 34 insertions, 0 deletions
diff --git a/Src/SuffixArrayApplications/SuffixArrayScan/Applications/CalcCountOfCounts.cpp b/Src/SuffixArrayApplications/SuffixArrayScan/Applications/CalcCountOfCounts.cpp
new file mode 100755
index 0000000..d7c96a2
--- /dev/null
+++ b/Src/SuffixArrayApplications/SuffixArrayScan/Applications/CalcCountOfCounts.cpp
@@ -0,0 +1,34 @@
+
+#include "_SuffixArrayScanningBase.h"
+#include "stdio.h"
+#include "stdlib.h"
+#include <iostream>
+#include <fstream>
+#include <map>
+
+using namespace std;
+
+/**
+* Given a corpus indexed by its suffix array, output the count-of-count information
+* Revision $Rev: 3665 $
+* Last Modified $LastChangedDate: 2007-06-16 15:40:59 -0400 (Sat, 16 Jun 2007) $
+**/
+int main(int argc, char * argv[]){
+ //-----------------------------------------------------------------------------
+ //check parameter
+ //-----------------------------------------------------------------------------
+ if(argc<4){
+ fprintf(stderr,"\nGiven an indexed corpus, output the count of counts for n-grams.\n");
+ fprintf(stderr,"\nUsage:\n");
+ fprintf(stderr,"\n%s fileNameStem maxN maxFreq\n\n",argv[0]);
+ exit(0);
+ }
+
+ unsigned int maxN = atoi(argv[2]);
+ unsigned int maxFreq = atoi(argv[3]);
+
+ C_SuffixArrayScanningBase saObj(argv[1], maxN);
+ saObj.scanSuffixArrayForCountofCounts(maxFreq);
+
+ return 1;
+}