#!/usr/bin/python
'''
Called by ucsc_view.sh
Tidy and arrange the result of miRNA binding sites. Generate the format for the UCSC genome browser's custom tracks.
Convert the transcript coordinates to chromosome coordinates. If a binding sites crossed the junction, use two blocks to present it, just like junction reads.
'''

import sys

f_exon_bed = open(sys.argv[1])

lCoords = []
# m = 1
exon_counter = 0
dExonMap = {}
for row in f_exon_bed.readlines():
	exon_counter += 1
	cols = row.rstrip('\n').split('\t')
	chr = cols[0]
	strand = cols[5]
	start = int(cols[1]) + 1
	end = int(cols[2])

	for k in range(start, end + 1):
		# print k
		dExonMap[k] = exon_counter
		# m += 1

	lCoords.extend([start, end])

if strand == '-':
	lCoords = lCoords[::-1]

# print lCoords

dLocMap = {}
n = 1
for i in range(0, len(lCoords), 2):
	# print i
	start = lCoords[i]
	# print 'start', start
	t = i + 1
	end = lCoords[t]	
	# print 'end', end

	if strand == '+':
		r = range(start, end + 1)
	else:
		r = range(start, end - 1, -1)
	for j in r:
		dLocMap[n] = j
		n += 1

# print dExonMap
# print m
if strand == '-':
	lCoords = lCoords[::-1]

for row in sys.stdin.readlines():
	cols = row.rstrip('\n').split('\t')
	mir = cols[0]
	trans_start = int(cols[1])
	trans_end = int(cols[2])
	clip_support = cols[6]

	color = '3,112,192'
	if clip_support == 'Yes': 
		color = '192,0,1'

	chr_start = dLocMap[trans_start]
	chr_end = dLocMap[trans_end]
	if strand == '-':
		chr_start, chr_end = chr_end, chr_start

	if dExonMap[chr_start] == dExonMap[chr_end]:
		block_count = 1
		block_size = trans_end - trans_start + 1
		block_starts = 0
	else:
		block_count = 2
		if strand == '-':
			trans_start, trans_end = trans_end, trans_start
		index_of_first_exon_end = 2 * dExonMap[dLocMap[trans_start]] - 1
		index_of_second_exon_start = index_of_first_exon_end + 1
		first_exon_end = lCoords[index_of_first_exon_end]
		second_exon_start = lCoords[index_of_second_exon_start]
		
		# print 
		first_block_size = first_exon_end - chr_start + 1
		second_block_size = chr_end - second_exon_start + 1
		block_size = '%s,%s' % (first_block_size, second_block_size)

		second_start = second_exon_start - chr_start
		block_starts = '0,%s' % second_start

	chr_start -= 1
	l_output_bed = [chr, chr_start, chr_end, mir, 0, strand, chr_start, chr_end, color, block_count, block_size, block_starts]
	l_output_bed = [str(s) for s in l_output_bed]
	print '\t'.join(l_output_bed)