/tools/expression/upload.py
https://bitbucket.org/cistrome/cistrome-harvard/ · Python · 363 lines · 303 code · 18 blank · 42 comment · 107 complexity · 93825743bfbd1a2179749ffdd7a4cd16 MD5 · raw file
- #!/usr/bin/env python
- #Processes uploads from the user.
- # WARNING: Changes in this tool (particularly as related to parsing) may need
- # to be reflected in galaxy.web.controllers.tool_runner and galaxy.tools
- import urllib, sys, os, gzip, tempfile, shutil, re, gzip, zipfile, codecs, binascii
- from galaxy import eggs
- # need to import model before sniff to resolve a circular import dependency
- import galaxy.model
- from galaxy.datatypes import sniff
- from galaxy.datatypes.binary import *
- from galaxy.datatypes.registry import Registry
- from galaxy import util
- from galaxy.util.json import *
- assert sys.version_info[:2] >= ( 2, 4 )
- def stop_err( msg, ret=1 ):
- sys.stderr.write( msg )
- sys.exit( ret )
- def file_err( msg, dataset, json_file ):
- json_file.write( to_json_string( dict( type = 'dataset',
- ext = 'data',
- dataset_id = dataset.dataset_id,
- stderr = msg ) ) + "\n" )
- try:
- os.remove( dataset.path )
- except:
- pass
- def safe_dict(d):
- """
- Recursively clone json structure with UTF-8 dictionary keys
- http://mellowmachines.com/blog/2009/06/exploding-dictionary-with-unicode-keys-as-python-arguments/
- """
- if isinstance(d, dict):
- return dict([(k.encode('utf-8'), safe_dict(v)) for k,v in d.iteritems()])
- elif isinstance(d, list):
- return [safe_dict(x) for x in d]
- else:
- return d
- def check_html( temp_name, chunk=None ):
- if chunk is None:
- temp = open(temp_name, "U")
- else:
- temp = chunk
- regexp1 = re.compile( "<A\s+[^>]*HREF[^>]+>", re.I )
- regexp2 = re.compile( "<IFRAME[^>]*>", re.I )
- regexp3 = re.compile( "<FRAMESET[^>]*>", re.I )
- regexp4 = re.compile( "<META[^>]*>", re.I )
- regexp5 = re.compile( "<SCRIPT[^>]*>", re.I )
- lineno = 0
- for line in temp:
- lineno += 1
- matches = regexp1.search( line ) or regexp2.search( line ) or regexp3.search( line ) or regexp4.search( line ) or regexp5.search( line )
- if matches:
- if chunk is None:
- temp.close()
- return True
- if lineno > 100:
- break
- if chunk is None:
- temp.close()
- return False
- def check_binary( temp_name ):
- is_binary = False
- temp = open( temp_name, "U" )
- chars_read = 0
- for chars in temp:
- for char in chars:
- chars_read += 1
- if ord( char ) > 128:
- is_binary = True
- break
- if chars_read > 100:
- break
- if chars_read > 100:
- break
- temp.close()
- return is_binary
- def check_bam( temp_name ):
- return Bam().sniff( temp_name )
- def check_sff( temp_name ):
- return Sff().sniff( temp_name )
- def check_gzip( temp_name ):
- # This method returns a tuple of booleans representing ( is_gzipped, is_valid )
- # Make sure we have a gzipped file
- try:
- temp = open( temp_name, "U" )
- magic_check = temp.read( 2 )
- temp.close()
- if magic_check != util.gzip_magic:
- return ( False, False )
- except:
- return ( False, False )
- # We support some binary data types, so check if the compressed binary file is valid
- # If the file is Bam, it should already have been detected as such, so we'll just check
- # for sff format.
- try:
- header = gzip.open( temp_name ).read(4)
- if binascii.b2a_hex( header ) == binascii.hexlify( '.sff' ):
- return ( True, True )
- except:
- return( False, False )
- CHUNK_SIZE = 2**15 # 32Kb
- gzipped_file = gzip.GzipFile( temp_name, mode='rb' )
- chunk = gzipped_file.read( CHUNK_SIZE )
- gzipped_file.close()
- # See if we have a compressed HTML file
- if check_html( temp_name, chunk=chunk ):
- return ( True, False )
- return ( True, True )
- def check_zip( temp_name ):
- # Return: (is_zip, known_ext, exactly_one_pheno, gt_one, homogeneous, ext)
- if not zipfile.is_zipfile( temp_name ):
- return (False, False, False, False, False, None)
- zip_file = zipfile.ZipFile( temp_name, "r" )
- # Make sure the archive consists of valid files. The current rules are:
- # 1. The file type in the zip is homegeneous, except that there is exactly one .txt pheno file
- # 2. The rest of the files must be either .cel or .xys
- # 3. There must be at least two .cel or .xys
- hasPheno = False
- count = 0
- test_ext = None
- for name in zip_file.namelist():
- #Reason:modification to support folder in zip file
- #ext = name.split(".")[1].strip().lower()
- ext = os.path.splitext( name )[1].strip().lower().replace(".","")
- if(ext==""): #ignore folder
- continue
- count += 1
- if (not (ext == "txt" or ext == "cel" or ext == "xys")):
- return (True, False, False, False, False, ext)
- if (ext == "txt"):
- if (hasPheno):
- return (True, True, False, False, False, None)
- else:
- hasPheno = True
- elif (test_ext == None):
- test_ext = ext
- elif (ext != test_ext):
- return (True, True, True, True, False, None)
- zip_file.close()
- return ( True, True, hasPheno, (count >= 3), True, test_ext )
- def parse_outputs( args ):
- rval = {}
- for arg in args:
- id, files_path, path = arg.split( ':', 2 )
- rval[int( id )] = ( path, files_path )
- return rval
- def add_file( dataset, json_file, output_path ):
- data_type = None
- line_count = None
- converted_path = None
- stdout = None
- link_data_only = dataset.get( 'link_data_only', 'copy_files' )
-
- if dataset.type == 'url':
- try:
- temp_name, dataset.is_multi_byte = sniff.stream_to_file( urllib.urlopen( dataset.path ), prefix='url_paste' )
- except Exception, e:
- file_err( 'Unable to fetch %s\n%s' % ( dataset.path, str( e ) ), dataset, json_file )
- return
- dataset.path = temp_name
- # See if we have an empty file
- if not os.path.exists( dataset.path ):
- file_err( 'Uploaded temporary file (%s) does not exist.' % dataset.path, dataset, json_file )
- return
- if not os.path.getsize( dataset.path ) > 0:
- file_err( 'The uploaded file is empty', dataset, json_file )
- return
- if not dataset.type == 'url':
- # Already set is_multi_byte above if type == 'url'
- try:
- dataset.is_multi_byte = util.is_multi_byte( codecs.open( dataset.path, 'r', 'utf-8' ).read( 100 ) )
- except UnicodeDecodeError, e:
- dataset.is_multi_byte = False
- # Is dataset content multi-byte?
- if dataset.is_multi_byte:
- data_type = 'multi-byte char'
- ext = sniff.guess_ext( dataset.path, is_multi_byte=True )
- # Is dataset content supported sniffable binary?
- elif check_bam( dataset.path ):
- ext = 'bam'
- data_type = 'bam'
- elif check_sff( dataset.path ):
- ext = 'sff'
- data_type = 'sff'
- else:
- # See if we have a gzipped file, which, if it passes our restrictions, we'll uncompress
- is_gzipped, is_valid = check_gzip( dataset.path )
- if is_gzipped and not is_valid:
- file_err( 'The uploaded file contains inappropriate content', dataset, json_file )
- return
- elif is_gzipped and is_valid:
- # We need to uncompress the temp_name file, but BAM files must remain compressed in the BGZF format
- CHUNK_SIZE = 2**20 # 1Mb
- fd, uncompressed = tempfile.mkstemp( prefix='data_id_%s_upload_gunzip_' % dataset.dataset_id, dir=os.path.dirname( dataset.path ), text=False )
- gzipped_file = gzip.GzipFile( dataset.path, 'rb' )
- while 1:
- try:
- chunk = gzipped_file.read( CHUNK_SIZE )
- except IOError:
- os.close( fd )
- os.remove( uncompressed )
- file_err( 'Problem decompressing gzipped data', dataset, json_file )
- return
- if not chunk:
- break
- os.write( fd, chunk )
- os.close( fd )
- gzipped_file.close()
- # Replace the gzipped file with the decompressed file
- shutil.move( uncompressed, dataset.path )
- dataset.name = dataset.name.rstrip( '.gz' )
- data_type = 'gzip'
- if not data_type:
- # See if we have a zip archive
- is_zipped, known_ext, one_pheno, gt_one, homogeneous, test_ext = check_zip( dataset.path )
- if (not is_zipped):
- file_err("CEL or NimbleGen files must be zipped.", dataset, json_file)
- if (not known_ext):
- file_err("Unknown file type in zip: " + test_ext, dataset, json_file)
- if (not one_pheno):
- file_err("There must be exactly one .txt pheno file in the zip.", dataset, json_file)
- if (not gt_one):
- file_err("There must be more than one CEL or XYS file in the zip.", dataset, json_file)
- if (not homogeneous):
- file_err("Except the .txt pheno file, other files must be all CEL or XYS.", dataset, json_file)
- data_type = 'zip'
- if (test_ext == 'cel'):
- ext = 'cel.zip'
- file_type = 'cel.zip'
- else:
- ext = 'xys.zip'
- file_type = 'xys.zip'
- if not data_type:
- if check_binary( dataset.path ):
- # We have a binary dataset, but it is not Bam or Sff
- data_type = 'binary'
- #binary_ok = False
- parts = dataset.name.split( "." )
- if len( parts ) > 1:
- ext = parts[1].strip().lower()
- if ext not in unsniffable_binary_formats:
- file_err( 'The uploaded file contains inappropriate content', dataset, json_file )
- return
- elif ext in unsniffable_binary_formats and dataset.file_type != ext:
- err_msg = "You must manually set the 'File Format' to '%s' when uploading %s files." % ( ext.capitalize(), ext )
- file_err( err_msg, dataset, json_file )
- return
- if not data_type:
- # We must have a text file
- if check_html( dataset.path ):
- file_err( 'The uploaded file contains inappropriate content', dataset, json_file )
- return
- if data_type != 'binary' and data_type != 'zip':
- # don't convert newlines on data we're only going to symlink
- if link_data_only == 'copy_files':
- in_place = True
- if dataset.type in ( 'server_dir', 'path_paste' ):
- in_place = False
- if dataset.space_to_tab:
- line_count, converted_path = sniff.convert_newlines_sep2tabs( dataset.path, in_place=in_place )
- else:
- line_count, converted_path = sniff.convert_newlines( dataset.path, in_place=in_place )
- if dataset.file_type == 'auto':
- ext = sniff.guess_ext( dataset.path )
- else:
- ext = dataset.file_type
- data_type = ext
- # Save job info for the framework
- if ext == 'auto' and dataset.ext:
- ext = dataset.ext
- if ext == 'auto':
- ext = 'data'
- # Move the dataset to its "real" path
- if link_data_only == 'link_to_files':
- pass # data will remain in place
- elif dataset.type in ( 'server_dir', 'path_paste' ):
- if converted_path is not None:
- shutil.copy( converted_path, output_path )
- try:
- os.remove( converted_path )
- except:
- pass
- else:
- # this should not happen, but it's here just in case
- shutil.copy( dataset.path, output_path )
- else:
- shutil.move( dataset.path, output_path )
- # Write the job info
- info = dict( type = 'dataset',
- dataset_id = dataset.dataset_id,
- ext = ext,
- stdout = 'uploaded %s file' % data_type,
- name = dataset.name,
- line_count = line_count )
- json_file.write( to_json_string( info ) + "\n" )
- # Groom the dataset content if necessary
- datatype = Registry().get_datatype_by_extension( ext )
- datatype.groom_dataset_content( output_path )
- def add_composite_file( dataset, json_file, output_path, files_path ):
- if dataset.composite_files:
- os.mkdir( files_path )
- for name, value in dataset.composite_files.iteritems():
- value = util.bunch.Bunch( **value )
- if dataset.composite_file_paths[ value.name ] is None and not value.optional:
- file_err( 'A required composite data file was not provided (%s)' % name, dataset, json_file )
- break
- elif dataset.composite_file_paths[value.name] is not None:
- if not value.is_binary:
- if uploaded_dataset.composite_files[ value.name ].space_to_tab:
- sniff.convert_newlines_sep2tabs( dataset.composite_file_paths[ value.name ][ 'path' ] )
- else:
- sniff.convert_newlines( dataset.composite_file_paths[ value.name ][ 'path' ] )
- shutil.move( dataset.composite_file_paths[ value.name ][ 'path' ], os.path.join( files_path, name ) )
- # Move the dataset to its "real" path
- shutil.move( dataset.primary_file, output_path )
- # Write the job info
- info = dict( type = 'dataset',
- dataset_id = dataset.dataset_id,
- stdout = 'uploaded %s file' % dataset.file_type )
- json_file.write( to_json_string( info ) + "\n" )
- def __main__():
- if len( sys.argv ) < 2:
- print >>sys.stderr, 'usage: upload.py <json paramfile> <output spec> ...'
- sys.exit( 1 )
- output_paths = parse_outputs( sys.argv[2:] )
- json_file = open( 'galaxy.json', 'w' )
- for line in open( sys.argv[1], 'r' ):
- dataset = from_json_string( line )
- dataset = util.bunch.Bunch( **safe_dict( dataset ) )
- try:
- output_path = output_paths[int( dataset.dataset_id )][0]
- except:
- print >>sys.stderr, 'Output path for dataset %s not found on command line' % dataset.dataset_id
- sys.exit( 1 )
- if dataset.type == 'composite':
- files_path = output_paths[int( dataset.dataset_id )][1]
- add_composite_file( dataset, json_file, output_path, files_path )
- else:
- add_file( dataset, json_file, output_path )
- # clean up paramfile
- try:
- os.remove( sys.argv[1] )
- except:
- pass
- if __name__ == '__main__':
- __main__()