annotate commons/core/parsing/WigParser.py @ 36:44d5973c188c

Uploaded

#
# Copyright INRA-URGI 2009-2010
# 
# This software is governed by the CeCILL license under French law and
# abiding by the rules of distribution of free software. You can use,
# modify and/ or redistribute the software under the terms of the CeCILL
# license as circulated by CEA, CNRS and INRIA at the following URL
# "http://www.cecill.info".
# 
# As a counterpart to the access to the source code and rights to copy,
# modify and redistribute granted by the license, users are provided only
# with a limited warranty and the software's author, the holder of the
# economic rights, and the successive licensors have only limited
# liability.
# 
# In this respect, the user's attention is drawn to the risks associated
# with loading, using, modifying and/or developing or reproducing the
# software by the user in light of its specific status of free software,
# that may mean that it is complicated to manipulate, and that also
# therefore means that it is reserved for developers and experienced
# professionals having in-depth computer knowledge. Users are therefore
# encouraged to load and test the software's suitability as regards their
# requirements in conditions enabling the security of their systems and/or
# data to be ensured and, more generally, to use and operate it in the
# same conditions as regards security.
# 
# The fact that you are presently reading this means that you have had
# knowledge of the CeCILL license and that you accept its terms.
#
import re
import sys
import os.path
import struct
from commons.core.parsing.TranscriptListParser import TranscriptListParser
from SMART.Java.Python.structure.Transcript import Transcript

STRANDTOSTR = {1: "(+)", 0: "(=)", None: "(=)", -1: "(-)"}

nbOpenHandles = 30


class WigParser(TranscriptListParser):
	"""A class that parses a big WIG file, creates an index and make it possible to quickly retrieve some data"""

	def __init__(self, fileName, verbosity = 1):
		self.fileName		   = fileName
		self.filler			   = "\xFF" * struct.calcsize('Q')
		self.strands		   = False
		self.indexFiles	   	   = {}
		self.indexBuilt		   = False
		self.defaultValue	   = 0.0
		self.currentChromosome = None
		self.currentStrand	   = 1
		self.verbosity         = verbosity
		super(WigParser, self).__init__(fileName, verbosity)


	def __def__(self):
		for file in self.indexFiles.values():
			file.close()


	def setStrands(self, strands):
		self.strands = strands


	def setDefaultValue(self, value):
		self.defaultValue = value


	def getFileFormats():
		return ["wig"]
	getFileFormats = staticmethod(getFileFormats)


	def setStrands(self, strands):
		"""
		Consider both strands separately
		"""
		self.strands = strands


	def makeIndexName(self, chromosome, strand = None):
		"""
		Create an index name for a file
		"""
		directoryName = os.path.dirname(self.fileName)
		if strand == None:
			strandName = ""
		else:
			strandName = "+" if strand == 1 else "-"
		indexName = os.path.join(directoryName, ".%s%s.index" % (chromosome, strandName))
		return indexName
	
	
	def findIndexFile(self, chromosome, strand = None):
		"""
		Check if the index of a file exists
		""" 
		indexName = self.makeIndexName(chromosome, strand)
		if os.path.exists(indexName):
			return indexName
		return False
	
	
	def makeIndexFile(self):
		"""
		Create the index for a file
		"""
		if self.indexBuilt:
			return

		inputFile  = open(self.fileName)
		outputFile = None
		index	  = 0
		mark	   = inputFile.tell()
		line	   = inputFile.readline().strip()
		chromosome = None

		while line != "":
			m1 = re.search(r"^\s*-?\d+\.?\d*\s*$", line)
			m2 = re.search(r"^\s*(\d+)\s+-?\d+\.?\d*\s*$", line)
			m3 = re.search(r"^\s*fixedStep\s+chrom=(\S+)\s+start=(\d+)\s+step=1\s*$", line)
			m4 = re.search(r"^\s*fixedStep\s+chrom=\S+\s+start=\d+\s+step=\d+\s+span=\d+\s*$", line)
			m5 = re.search(r"^\s*variableStep\s+chrom=(\S+)\s*$", line)
			m6 = re.search(r"^\s*variableStep\s+chrom=(\S+)\s+span=(\d+)\s*$", line)
						
			if m1 != None:
				outputFile.write(struct.pack("Q", mark))
				index += 1
			elif m2 != None:
				nextIndex = int(m2.group(1))
				if index < nextIndex - 1:
					outputFile.write(self.filler * (nextIndex - index - 1))
				outputFile.write(struct.pack("Q", mark))
				index = nextIndex
			elif m3 != None:
				newChromosome = m3.group(1)
				if newChromosome != chromosome:
					if outputFile != None:
						outputFile.close()
					outputFile = open(self.makeIndexName(newChromosome), "wb")
				chromosome = newChromosome
				nextIndex = int(m3.group(2))
				outputFile.write(self.filler * (nextIndex - index))
				index = nextIndex
			elif m4 != None:
				raise Exception("Error! Cannot parse fixed step WIG files with step > 1 or span > 1")
			elif m5 != None:
				newChromosome = m5.group(1)
				if newChromosome != chromosome:
					if outputFile != None:
						outputFile.close()
					outputFile = open(self.makeIndexName(newChromosome), "wb")
					index = 0
					outputFile.write(self.filler)
				chromosome = newChromosome
			elif m6 != None:
				if m6.group(2) != "1":
					raise Exception("Error! Cannot parse variable step WIG files with step > 1 or span > 1")
				newChromosome = m6.group(1)
				if newChromosome != chromosome:
					if outputFile != None:
						outputFile.close()
					outputFile = open(self.makeIndexName(newChromosome), "wb")
					index = 0
					outputFile.write(self.filler)
				chromosome = newChromosome
			elif (len(line) == 0) or line[0] == "#" or line.startswith("track"):
				pass
			else:
				raise Exception("Error! Cannot understand line '%s' of WIG file while creating index file! Aborting." % (line))
			
			mark = inputFile.tell()
			line = inputFile.readline().strip()

		inputFile.close
		if outputFile != None:
			outputFile.close()
		self.indexBuilt = True
	
	
	def getIndexFileHandle(self, chromosome, strand = None):
		"""
		Get the handle of an index file
		"""
		indexFileKey = chromosome
		if strand != None:
			indexFileKey += "+" if strand == 1 else "-"
		if indexFileKey in self.indexFiles:
			return self.indexFiles[indexFileKey]

		indexFileName = self.makeIndexName(chromosome, strand)
		if not self.findIndexFile(chromosome, strand):
			self.makeIndexFile()
	
		if not os.path.exists(indexFileName):
			print "Warning! Index for chromosome %s, strand %s does not exist." % (chromosome, STRANDTOSTR[strand])
			return False
		indexFile = open(indexFileName, "rb")

		if len(self.indexFiles.keys()) > nbOpenHandles:
			removedKey = set(self.indexFiles.keys()).pop()
			self.indexFiles[removedKey].close()
			del self.indexFiles[removedKey]
		self.indexFiles[indexFileKey] = indexFile
		return indexFile
		

	
	def findIndex(self, chromosome, start, strand = None):
		"""
		Find the point where to start reading file
		"""

		sizeOfLong = struct.calcsize("Q")
		empty	  = int(struct.unpack("Q", self.filler)[0])
		offset	 = empty
		indexFile  = self.getIndexFileHandle(chromosome, strand)
	
		if not indexFile:
			return (None, None)
		
		while offset == empty:
			address = start * sizeOfLong
			indexFile.seek(address, os.SEEK_SET)
			
			buffer = indexFile.read(sizeOfLong)
			if len(buffer) != sizeOfLong:
				if buffer == "":
					print "Warning! Index position %d of chromosome %s on strand %s seems out of range!" % (start, chromosome, STRANDTOSTR[strand])
					return (None, None)
				else:
					raise Exception("Problem fetching position %d of chromosome %s on strand %s seems out of range!" % (start, chromosome, STRANDTOSTR[strand]))
			
			offset = int(struct.unpack("Q", buffer)[0])
			start += 1
			
		start -= 1
		return (offset, start)
	
	

	def getRange(self, chromosome, start, end):
		"""
		Parse a wig file and output a range
		"""
		arrays  = {}
		strands = {1: "+", -1: "-"} if self.strands else {0: ""}

		for strand in strands:

			array = [self.defaultValue] * (end - start + 1)
			file  = open(self.fileName)
			offset, index = self.findIndex(chromosome, start, strand if self.strands else None)
			if offset == None:
				arrays[strand] = array
				continue
			file.seek(offset, os.SEEK_SET)

			for line in file:
				line = line.strip()

				m1 = re.search(r"^\s*(-?\d+\.?\d*)\s*$", line)
				m2 = re.search(r"^\s*(\d+)\s+(-?\d+\.?\d*)\s*$", line)
				m3 = re.search(r"^\s*fixedStep\s+chrom=(\S+)\s+start=(\d+)\s+step=\d+\s*$", line)
				m4 = re.search(r"^\s*variableStep\s+chrom=(\S+)\s*$", line)

				if m1 != None:
					if index > end:
						break
					if index >= start:
						array[index - start] = float(m1.group(1))
					index += 1
				elif m2 != None:
					index = int(m2.group(1))
					if index > end:
						break
					if index >= start:
						array[index - start] = float(m2.group(2))
					index += 1
				elif m3 != None:
					if m3.group(1) != "%s%s" % (chromosome, strands[strand]):
						break
					index = int(m3.group(2))
				elif m4 != None:
					if m4.group(1) != "%s%s" % (chromosome, strands[strand]):
						break
				elif (len(line) == 0) or (line[0] == "#") or line.startswith("track"):
					pass
				else:
					raise Exception("Error! Cannot read line '%s' of wig file" % (line))

			file.close()
	
			arrays[strand] = array
			
		if self.strands:
			return arrays
		return array
	

	def skipFirstLines(self):
		return

	
	def parseLine(self, line):
		if line.startswith("track"):
			return None
		m = re.search(r"^\s*variableStep\s+chrom=(\S+)", line)
		if m != None:
			chromosome = m.group(1)
			if chromosome.endswith("+"):
				self.currentStrand = 1
				self.currentChromosome = chromosome[:-1]
			elif chromosome.endswith("-"):
				self.currentStrand = -1
				self.currentChromosome = chromosome[:-1]
			else:
				self.currentStrand = 1
				self.currentChromosome = chromosome
			return None
		position, value = line.split()
		position = int(position)
		value	= float(value)
		transcript = Transcript()
		transcript.setChromosome(self.currentChromosome)
		transcript.setStart(position)
		transcript.setEnd(position)
		transcript.setDirection(self.currentStrand)
		transcript.setTagValue("ID", "wig_%s_%d_%d" % (self.currentChromosome, self.currentStrand, position))
		transcript.setTagValue("nbElements", value)
		return transcript