From d6b7a1f774a65c37b7b98a8b6c69b7397b059241 Mon Sep 17 00:00:00 2001
From: "Brian P. Walenz" <thebri@gmail.com>
Date: Fri, 18 Sep 2020 12:42:39 -0400
Subject: [PATCH] Add -minlength option to discard short overlaps.

---
 src/mhap/mhapConvert.C | 16 ++++++++++++++--
 1 file changed, 14 insertions(+), 2 deletions(-)

diff --git a/src/mhap/mhapConvert.C b/src/mhap/mhapConvert.C
index 87da59f0c..ee1db19d8 100644
--- a/src/mhap/mhapConvert.C
+++ b/src/mhap/mhapConvert.C
@@ -28,6 +28,7 @@ int
 main(int argc, char **argv) {
   char           *outName     = NULL;
   char           *seqName     = NULL;
+  int32           minLength   = 0;
 
   vector<char *>  files;
 
@@ -41,6 +42,9 @@ main(int argc, char **argv) {
     } else if (strcmp(argv[arg], "-S") == 0) {
       seqName = argv[++arg];
 
+    } else if (strcmp(argv[arg], "-minlength") == 0) {
+      minLength = strtoint32(argv[++arg]);
+
     } else if (fileExists(argv[arg])) {
       files.push_back(argv[arg]);
 
@@ -55,6 +59,7 @@ main(int argc, char **argv) {
   if ((err) || (seqName == NULL) || (outName == NULL) || (files.size() == 0)) {
     fprintf(stderr, "usage: %s -S seqStore -o output.ovb input.mhap[.gz]\n", argv[0]);
     fprintf(stderr, "  Converts mhap native output to ovb\n");
+    fprintf(stderr, "    -minlength X    discards overlaps below X bp long.\n");
 
     if (seqName == NULL)
       fprintf(stderr, "ERROR:  no seqStore (-S) supplied\n");
@@ -147,9 +152,16 @@ main(int argc, char **argv) {
                 ov.dat.ovl.bhg5, ov.dat.ovl.bhg3,
                 (ov.dat.ovl.flipped) ? " flipped" : ""), exit(1);
 
-      //  Overlap looks good, write it!
+      //  Overlap looks good, write it if its long enough.  Bogart is
+      //  computing overlap length as the max number of bases covered on
+      //  either read.
+
+      int32  oalen = alen - ov.dat.ovl.ahg5 - ov.dat.ovl.ahg3;
+      int32  oblen = blen - ov.dat.ovl.bhg5 - ov.dat.ovl.bhg3;
 
-      of->writeOverlap(&ov);
+      if ((minLength <= oalen) ||
+          (minLength <= oblen))
+        of->writeOverlap(&ov);
     }
 
     delete in;