Merge covBam with covBamExons.

With the new implementation of RNA-seq coverage calculation, it has become
possible to calculate coverage for ChIP-seq using the same code. This
change reduces our burden of maintaining two functions that often need to
be modified together.

The merge introduces paired-end functionality into ChIP-seq too. However, I
noticed that the speed for ChIP-seq at TSS reduced by two folds. This is NOT
due to the extra computation on paired-end functionality. It remains a future
improvement to enhance the speed of the new covBamExons function.

Author: lishen

TODO

@@ -1,5 +1,6 @@
 - May 20, 2014
-* Need to fix: bamFileList returns a bam file list that may not be unique.
+* Need to fix: bamFileList returns a bam file list that may not be unique. 
+  (done)
 
 - Sep 5, 2013
 * Add sanity check for BED file. (done)

bin/ngs.plot.r

@@ -47,7 +47,7 @@ cmd.help <- function(){
 ###########################################################################
 #################### Deal with program input arguments ####################
 args <- commandArgs(T)
-# args <- unlist(strsplit('-G hg19 -R genebody -F rnaseq -C hesc.RNAseq.1M.bam -O rna_paired_1M_new_refac -Debug 1', ' '))
+# args <- unlist(strsplit('-G hg19 -R tss -C hesc.H3k4me3.Rep1.sort.bam -O tmp2 -Debug 1', ' '))
 
 # Input argument parser.
 args.tbl <- parseArgs(args, c('-G', '-C', '-R', '-O'))