From 0a9476fe2aa09610a23b2a54974bbb733b478a92 Mon Sep 17 00:00:00 2001
From: Calvin Morrison <mutantturkey@gmail.com>
Date: Fri, 30 Aug 2013 14:33:50 -0400
Subject: check for fasta headers

---
 count_nucleobases   | Bin 0 -> 5496 bytes
 count_nucleobases.c |  19 +++++++++++++------
 2 files changed, 13 insertions(+), 6 deletions(-)
 create mode 100755 count_nucleobases

diff --git a/count_nucleobases b/count_nucleobases
new file mode 100755
index 0000000..8c1b122
Binary files /dev/null and b/count_nucleobases differ
diff --git a/count_nucleobases.c b/count_nucleobases.c
index d1a4f1c..c2aafe5 100755
--- a/count_nucleobases.c
+++ b/count_nucleobases.c
@@ -5,27 +5,34 @@
 #include <stdint.h>
 int main(int argc, char **argv) {
 
+  char *line = NULL;
+  size_t len = 0;
+  ssize_t read;
 
   if(argc != 2) {
     printf("Please supply a filename, and only a filename\n");
     exit(EXIT_FAILURE);
   }
+
   FILE *fh = fopen(argv[1], "r" );
   if(fh == NULL) {
     fprintf(stderr, "Couldn't open: %s\n", argv[1]);
     exit(EXIT_FAILURE);
   }
 
-  char line[8192];
-  long unsigned long counts[256];
-  while (fgets(line, 8192, fh) != NULL) {
+  unsigned long long counts[256] = {0};
 
-    unsigned int i = 0;
-    for(i = 0; i < strlen(line); i++) {
-     counts[line[i]]++;
+  while ((read = getline(&line, &len, fh)) != -1) {
+    if(line[0] != '>')   {
+      unsigned int i = 0;
+      for(i = 0; i < strlen(line); i++) {
+        counts[line[i]]++;
+      }
     }
   }
 
+  free(line);
+
   printf("A:%llu\nC:%llu\nG:%llu\nT:%llu\n",
          counts['a'] + counts['A'],
          counts['c'] + counts['C'],
-- 
cgit v1.2.3