annotate data_manager/data_manager_fetch_gff.py @ 0:ac4fde07eaed draft

Uploaded

#!/usr/bin/env python
#Dan Blankenberg

import sys
import os
import tempfile
import shutil
import optparse
from ftplib import FTP
import tarfile
import zipfile
import gzip
import bz2
try:
    # For Python 3.0 and later
    from urllib.request import urlopen
    from io import BytesIO as StringIO
    from io import UnsupportedOperation
except ImportError:
    # Fall back to Python 2's urllib2
    from urllib2 import urlopen
    from StringIO import StringIO
    UnsupportedOperation = AttributeError
from json import loads, dumps


CHUNK_SIZE = 2**20  # 1mb

DATA_TABLE_NAME = 'all_gff'

def cleanup_before_exit( tmp_dir ):
    if tmp_dir and os.path.exists( tmp_dir ):
        shutil.rmtree( tmp_dir )


def stop_err(msg):
    sys.stderr.write(msg)
    sys.exit(1)


def get_dbkey_dbname_id_name( params, dbkey_description=None ):
    dbkey = params['param_dict']['dbkey_source']['dbkey']
    #TODO: ensure sequence_id is unique and does not already appear in location file
    sequence_id = params['param_dict']['sequence_id']
    if not sequence_id:
        sequence_id = dbkey #uuid.uuid4() generate and use an uuid instead?
    
    if params['param_dict']['dbkey_source']['dbkey_source_selector'] == 'new':
        dbkey_name = params['param_dict']['dbkey_source']['dbkey_name']
        if not dbkey_name:
            dbkey_name = dbkey
    else:
        dbkey_name = None
    
    sequence_name = params['param_dict']['sequence_name']
    if not sequence_name:
        sequence_name = dbkey_description
        if not sequence_name:
            sequence_name = dbkey
    return dbkey, dbkey_name, sequence_id, sequence_name


def _get_files_in_ftp_path( ftp, path ):
    path_contents = []
    ftp.retrlines( 'MLSD %s' % ( path ), path_contents.append )
    return [ line.split( ';' )[ -1 ].lstrip() for line in path_contents ]


def _get_stream_readers_for_tar( fh, tmp_dir ):
    fasta_tar = tarfile.open( fileobj=fh, mode='r:*' )
    return [x for x in [fasta_tar.extractfile(member) for member in fasta_tar.getmembers()] if x]


def _get_stream_readers_for_zip( fh, tmp_dir ):
    """
    Unpacks all archived files in a zip file.
    Individual files will be concatenated (in _stream_fasta_to_file)
    """
    fasta_zip = zipfile.ZipFile( fh, 'r' )
    rval = []
    for member in fasta_zip.namelist():
        fasta_zip.extract( member, tmp_dir )
        rval.append( open( os.path.join( tmp_dir, member ), 'rb' ) )
    return rval


def _get_stream_readers_for_gzip( fh, tmp_dir ):
    return [ gzip.GzipFile( fileobj=fh, mode='rb') ]


def _get_stream_readers_for_bz2( fh, tmp_dir ):
    return [ bz2.BZ2File( fh.name, 'rb') ]


def sort_fasta( fasta_filename, sort_method, params ):
    if sort_method is None:
        return
    assert sort_method in SORTING_METHODS, ValueError( "%s is not a valid sorting option." % sort_method )
    return SORTING_METHODS[ sort_method ]( fasta_filename, params )


def _move_and_index_fasta_for_sorting( fasta_filename ):
    unsorted_filename = tempfile.NamedTemporaryFile().name
    shutil.move( fasta_filename, unsorted_filename )
    fasta_offsets = {}
    unsorted_fh = open( unsorted_filename )
    while True:
        offset = unsorted_fh.tell()
        line = unsorted_fh.readline()
        if not line:
            break
        if line.startswith( ">" ):
            line = line.split( None, 1 )[0][1:]
            fasta_offsets[ line ] = offset
    unsorted_fh.close()
    current_order = map( lambda x: x[1], sorted( map( lambda x: ( x[1], x[0] ), fasta_offsets.items() ) ) )
    return ( unsorted_filename, fasta_offsets, current_order )


def _write_sorted_fasta( sorted_names, fasta_offsets, sorted_fasta_filename, unsorted_fasta_filename ):
    unsorted_fh = open( unsorted_fasta_filename )
    sorted_fh = open( sorted_fasta_filename, 'wb+' )
    
    for name in sorted_names:
        offset = fasta_offsets[ name ]
        unsorted_fh.seek( offset )
        sorted_fh.write( unsorted_fh.readline() )
        while True:
            line = unsorted_fh.readline()
            if not line or line.startswith( ">" ):
                break
            sorted_fh.write( line )
    unsorted_fh.close()
    sorted_fh.close()


def _sort_fasta_as_is( fasta_filename, params ):
    return

def _sort_fasta_lexicographical( fasta_filename, params ):
    ( unsorted_filename, fasta_offsets, current_order ) = _move_and_index_fasta_for_sorting( fasta_filename )
    sorted_names = sorted( fasta_offsets.keys() )
    if sorted_names == current_order:
        shutil.move( unsorted_filename, fasta_filename )
    else:
        _write_sorted_fasta( sorted_names, fasta_offsets, fasta_filename, unsorted_filename )    


def _sort_fasta_gatk( fasta_filename, params ):
    #This method was added by reviewer request.
    ( unsorted_filename, fasta_offsets, current_order ) = _move_and_index_fasta_for_sorting( fasta_filename )
    sorted_names = map( str, range( 1, 23 ) ) + [ 'X', 'Y' ]
    #detect if we have chrN, or just N
    has_chr = False
    for chrom in sorted_names:
        if "chr%s" % chrom in current_order:
            has_chr = True
            break
    
    if has_chr:
        sorted_names = map( lambda x: "chr%s" % x, sorted_names)
        sorted_names.insert( 0, "chrM" )
    else:
        sorted_names.insert( 0, "MT" )
    sorted_names.extend( map( lambda x: "%s_random" % x, sorted_names ) )
    
    existing_sorted_names = []
    for name in sorted_names:
        if name in current_order:
            existing_sorted_names.append( name )
    for name in current_order:
        #TODO: confirm that non-canonical names do not need to be sorted specially
        if name not in existing_sorted_names:
            existing_sorted_names.append( name )
    
    if existing_sorted_names == current_order:
        shutil.move( unsorted_filename, fasta_filename )
    else:
        _write_sorted_fasta( existing_sorted_names, fasta_offsets, fasta_filename, unsorted_filename )


def _sort_fasta_custom( fasta_filename, params ):
    ( unsorted_filename, fasta_offsets, current_order ) = _move_and_index_fasta_for_sorting( fasta_filename )
    sorted_names = []
    for id_repeat in params['param_dict']['sorting']['sequence_identifiers']:
        sorted_names.append( id_repeat[ 'identifier' ] )
    handle_not_listed = params['param_dict']['sorting']['handle_not_listed_selector']
    if handle_not_listed.startswith( 'keep' ):
        add_list = []
        for name in current_order:
            if name not in sorted_names:
                add_list.append( name )
        if add_list:
            if handle_not_listed == 'keep_append':
                sorted_names.extend( add_list )
            else:
                add_list.extend( sorted_names )
                sorted_names = add_list
    if sorted_names == current_order:
        shutil.move( unsorted_filename, fasta_filename )
    else:
        _write_sorted_fasta( sorted_names, fasta_offsets, fasta_filename, unsorted_filename )


def _download_file(start, fh):
    tmp = tempfile.NamedTemporaryFile()
    tmp.write(start)
    tmp.write(fh.read())
    tmp.flush()
    tmp.seek(0)
    return tmp


def get_stream_reader(fh, tmp_dir):
    """
    Check if file is compressed and return correct stream reader.
    If file has to be downloaded, do it now.
    """
    magic_dict = {
        b"\x1f\x8b\x08": _get_stream_readers_for_gzip,
        b"\x42\x5a\x68": _get_stream_readers_for_bz2,
        b"\x50\x4b\x03\x04": _get_stream_readers_for_zip,
    }
    start_of_file = fh.read(CHUNK_SIZE)
    try:
        fh.seek(0)
    except UnsupportedOperation:  # This is if fh has been created by urlopen
        fh = _download_file(start_of_file, fh)
    for k,v in magic_dict.items():
        if start_of_file.startswith(k):
            return v(fh, tmp_dir)
    try:  # Check if file is tar file
        if tarfile.open(fileobj=StringIO(start_of_file)):
            return _get_stream_readers_for_tar(fh, tmp_dir)
    except tarfile.ReadError:
        pass
    return fh


def _get_ucsc_download_address(params, dbkey):
    """
    Check if we can find the correct file for the supplied dbkey on UCSC's FTP server
    """
    UCSC_FTP_SERVER = 'hgdownload.cse.ucsc.edu'
    UCSC_DOWNLOAD_PATH = '/goldenPath/%s/bigZips/'
    COMPRESSED_EXTENSIONS = ['.tar.gz', '.tgz', '.tar.bz2', '.zip', '.fa.gz', '.fa.bz2']

    email = params['param_dict']['__user_email__']
    if not email:
        email = '[e-mail]'

    ucsc_dbkey = params['param_dict']['reference_source']['requested_dbkey'] or dbkey
    UCSC_CHROM_FA_FILENAMES = ['%s.chromFa' % ucsc_dbkey, 'chromFa', ucsc_dbkey]

    ftp = FTP(UCSC_FTP_SERVER)
    ftp.login('anonymous', email)

    ucsc_path = UCSC_DOWNLOAD_PATH % ucsc_dbkey
    path_contents = _get_files_in_ftp_path(ftp, ucsc_path)
    ftp.quit()

    for ucsc_chrom_fa_filename in UCSC_CHROM_FA_FILENAMES:
        for ext in COMPRESSED_EXTENSIONS:
            if "%s%s" % (ucsc_chrom_fa_filename, ext) in path_contents:
                ucsc_file_name = "%s%s%s" % (ucsc_path, ucsc_chrom_fa_filename, ext)
                return "ftp://%s%s" % (UCSC_FTP_SERVER, ucsc_file_name)

    raise Exception('Unable to determine filename for UCSC Genome for %s: %s' % (ucsc_dbkey, path_contents))

def add_fasta_to_table(data_manager_dict, fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params):
    for data_table_name, data_table_entry in _stream_fasta_to_file( fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params ):
        if data_table_entry:
            _add_data_table_entry( data_manager_dict, data_table_entry, data_table_name )


def download_from_ucsc( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir ):
    url = _get_ucsc_download_address(params, dbkey)
    fasta_readers = get_stream_reader(urlopen(url), tmp_dir)
    add_fasta_to_table(data_manager_dict, fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params)


def download_from_ncbi( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir ):
    NCBI_DOWNLOAD_URL = 'http://togows.dbcls.jp/entry/ncbi-nucleotide/%s.fasta' #FIXME: taken from dave's genome manager...why some japan site?
    requested_identifier = params['param_dict']['reference_source']['requested_identifier']
    url = NCBI_DOWNLOAD_URL % requested_identifier
    fasta_readers = get_stream_reader(urlopen(url), tmp_dir)
    add_fasta_to_table(data_manager_dict, fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params)


def download_from_url( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir ):
    urls = filter( bool, map( lambda x: x.strip(), params['param_dict']['reference_source']['user_url'].split( '\n' ) ) )
    fasta_readers = [ get_stream_reader(urlopen( url ), tmp_dir) for url in urls ]
    add_fasta_to_table(data_manager_dict, fasta_readers, target_directory, dbkey, dbkey_name, sequence_id,sequence_name, params)


def download_from_history( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir ):
    #TODO: allow multiple FASTA input files
    input_filename = params['param_dict']['reference_source']['input_fasta']
    if isinstance( input_filename, list ):
        fasta_readers = [ get_stream_reader(open(filename, 'rb'), tmp_dir) for filename in input_filename ]
    else:
        fasta_readers = get_stream_reader(open(input_filename), tmp_dir)
    add_fasta_to_table(data_manager_dict, fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params)


def copy_from_directory( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir ):
    input_filename = params['param_dict']['reference_source']['fasta_filename']
    create_symlink = params['param_dict']['reference_source']['create_symlink'] == 'create_symlink'
    if create_symlink:
        data_table_entries = _create_symlink( input_filename, target_directory, dbkey, dbkey_name, sequence_id, sequence_name )
    else:
        if isinstance( input_filename, list ):
            fasta_readers = [ get_stream_reader(open(filename, 'rb'), tmp_dir) for filename in input_filename ]
        else:
            fasta_readers = get_stream_reader(open(input_filename), tmp_dir)
        data_table_entries = _stream_fasta_to_file( fasta_readers, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params )
    for data_table_name, data_table_entry in data_table_entries:
        if data_table_entry:
            _add_data_table_entry( data_manager_dict, data_table_entry, data_table_name )


def _add_data_table_entry( data_manager_dict, data_table_entry, data_table_name ):
    data_manager_dict['data_tables'] = data_manager_dict.get( 'data_tables', {} )
    data_manager_dict['data_tables'][data_table_name] = data_manager_dict['data_tables'].get( DATA_TABLE_NAME, [] )
    data_manager_dict['data_tables'][data_table_name].append( data_table_entry )
    return data_manager_dict


def _stream_fasta_to_file( fasta_stream, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, params, close_stream=True ):
    fasta_base_filename = "%s.gff" % sequence_id
    fasta_filename = os.path.join( target_directory, fasta_base_filename )
    with open( fasta_filename, 'wb+' ) as fasta_writer:

        if isinstance( fasta_stream, list ) and len( fasta_stream ) == 1:
            fasta_stream = fasta_stream[0]

        if isinstance( fasta_stream, list ):
            last_char = None
            for fh in fasta_stream:
                if last_char not in [ None, '\n', '\r', b'\n', b'\r' ]:
                    fasta_writer.write( b'\n' )
                while True:
                    data = fh.read( CHUNK_SIZE )
                    if data:
                        fasta_writer.write( data )
                        last_char = data[-1]
                    else:
                        break
                if close_stream:
                    fh.close()
        else:
            while True:
                data = fasta_stream.read( CHUNK_SIZE )
                if data:
                    fasta_writer.write( data )
                else:
                    break
            if close_stream:
                fasta_stream.close()

    #sort_fasta( fasta_filename, params['param_dict']['sorting']['sort_selector'], params )
    
    dbkey_dict = None
    if dbkey_name:
        #do len calc here
        #len_base_name = "%s.len" % ( dbkey )
        #compute_fasta_length( fasta_filename, os.path.join( target_directory, len_base_name ), keep_first_word=True )
        dbkey_dict = dict( value=dbkey, name=dbkey_name, len_path='' )
    
    return [ ( '__dbkeys__', dbkey_dict ), ( DATA_TABLE_NAME, dict( value=sequence_id, dbkey=dbkey, name=sequence_name, path=fasta_base_filename ) ) ]


def compute_fasta_length( fasta_file, out_file, keep_first_word=False ):

    infile = fasta_file
    out = open( out_file, 'w')

    fasta_title = ''
    seq_len = 0

    first_entry = True

    for line in open( infile ):
        line = line.strip()
        if not line or line.startswith( '#' ):
            continue
        if line[0] == '>':
            if first_entry == False:
                if keep_first_word:
                    fasta_title = fasta_title.split()[0]
                out.write( "%s\t%d\n" % ( fasta_title[ 1: ], seq_len ) )
            else:
                first_entry = False
            fasta_title = line
            seq_len = 0
        else:
            seq_len += len(line)

    # last fasta-entry
    if keep_first_word:
        fasta_title = fasta_title.split()[0]
    out.write( "%s\t%d\n" % ( fasta_title[ 1: ], seq_len ) )
    out.close()


def _create_symlink( input_filename, target_directory, dbkey, dbkey_name, sequence_id, sequence_name ):
    fasta_base_filename = "%s.fa" % sequence_id
    fasta_filename = os.path.join( target_directory, fasta_base_filename )
    os.symlink( input_filename, fasta_filename )
    
    dbkey_dict = None
    if dbkey_name:
        #do len calc here
        len_base_name = "%s.len" % ( dbkey )
        compute_fasta_length( fasta_filename, os.path.join( target_directory, len_base_name ), keep_first_word=True )
        dbkey_dict = dict( value=dbkey, name=dbkey_name, len_path=len_base_name )
    
    return [ ( '__dbkeys__', dbkey_dict ), ( DATA_TABLE_NAME, dict( value=sequence_id, dbkey=dbkey, name=sequence_name, path=fasta_base_filename ) ) ]


REFERENCE_SOURCE_TO_DOWNLOAD = dict( ucsc=download_from_ucsc, ncbi=download_from_ncbi, url=download_from_url, history=download_from_history, directory=copy_from_directory )

SORTING_METHODS = dict( as_is=_sort_fasta_as_is, lexicographical=_sort_fasta_lexicographical, gatk=_sort_fasta_gatk, custom=_sort_fasta_custom )


def main():
    #Parse Command Line
    parser = optparse.OptionParser()
    parser.add_option( '-d', '--dbkey_description', dest='dbkey_description', action='store', type="string", default=None, help='dbkey_description' )
    parser.add_option( '-t', '--type', dest='file_type', action='store', type='string', default=None, help='file_type')
    (options, args) = parser.parse_args()
    
    filename = args[0]
    global DATA_TABLE_NAME
    if options.file_type == 'representative':
       DATA_TABLE_NAME= 'representative_gff'
    params = loads( open( filename ).read() )
    target_directory = params[ 'output_data' ][0]['extra_files_path']
    os.mkdir( target_directory )
    data_manager_dict = {}
    
    dbkey, dbkey_name, sequence_id, sequence_name = get_dbkey_dbname_id_name( params, dbkey_description=options.dbkey_description ) 
    
    if dbkey in [ None, '', '?' ]:
        raise Exception( '"%s" is not a valid dbkey. You must specify a valid dbkey.' % ( dbkey ) )

    # Create a tmp_dir, in case a zip file needs to be uncompressed
    tmp_dir = tempfile.mkdtemp()
    #Fetch the FASTA
    try:
        REFERENCE_SOURCE_TO_DOWNLOAD[ params['param_dict']['reference_source']['reference_source_selector'] ]( data_manager_dict, params, target_directory, dbkey, dbkey_name, sequence_id, sequence_name, tmp_dir )
    finally:
        cleanup_before_exit(tmp_dir)
    #save info to json file
    open( filename, 'wb' ).write( dumps( data_manager_dict ).encode() )
        
if __name__ == "__main__":
    main()