4 from optparse import OptionParser
6 from subprocess import Popen, PIPE
9 from htsworkflow.util.opener import autoopen
10 from htsworkflow.version import version
12 LOGGER = logging.getLogger(__name__)
14 # constants for our fastq finite state machine
17 FASTQ_SEQUENCE_HEADER = 2
20 def main(cmdline=None):
21 parser = make_parser()
22 opts, args = parser.parse_args(cmdline)
25 logging.basicConfig(level=logging.INFO)
27 logging.basicConfig(level=logging.WARN)
34 parser.error("Requires one argument, got: %s" % (str(args)))
36 if opts.flowcell is not None:
37 header = "%s_" % (opts.flowcell,)
42 left = open_write(opts.single, opts.force)
44 left = open_write(opts.left, opts.force)
45 right = open_write(opts.right, opts.force)
47 # open the srf, fastq, or compressed fastq
49 source = srf_open(args[0], opts.srf2fastq, opts.cnf1)
51 source = autoopen(args[0])
54 convert_single_to_fastq(source, left, header)
56 convert_single_to_two_fastq(source, left, right, opts.mid, header)
61 parser = OptionParser("""%prog: [options] file
63 file can be either a fastq file or a srf file.
64 You can also force the flowcell ID to be added to the header.""")
66 parser.add_option('--force', default=False, action="store_true",
67 help="overwrite existing files.")
68 parser.add_option('--flowcell', default=None,
69 help="add flowcell id header to sequence")
70 parser.add_option('-l','--left', default="r1.fastq",
71 help='left side filename')
72 parser.add_option('-m','--mid', default=None,
73 help='actual sequence mid point')
74 parser.add_option('-r','--right', default="r2.fastq",
75 help='right side filename')
76 parser.add_option('-s','--single', default=None,
77 help="single fastq target name")
78 parser.add_option('-v', '--verbose', default=False, action="store_true",
79 help="show information about what we're doing.")
80 parser.add_option('--version', default=False, action="store_true",
81 help="Report software version")
82 parser.add_option('--cnf1', default=False, action="store_true",
83 help="Force cnf1 mode in srf2fastq")
84 parser.add_option('--srf2fastq', default='srf2fastq',
85 help='specify srf2fastq command')
89 def srf_open(filename, srf2fastq_cmd, cnf1=False):
91 Make a stream from srf file using srf2fastq
93 if not os.path.exists(srf2fastq_cmd):
94 LOGGER.error("srf command: %s doesn't exist" % (srf2fastq_cmd,))
96 if cnf1 or is_cnf1(filename):
100 LOGGER.info('srf command: %s' % (" ".join(cmd),))
101 p = Popen(cmd, stdout=PIPE)
105 def convert_single_to_fastq(instream, target1, header=''):
108 for line in instream:
111 if state == FASTQ_HEADER:
112 write_header(target1, header, line)
113 state = FASTQ_SEQUENCE
115 elif state == FASTQ_SEQUENCE:
116 write_sequence(target1, line)
117 state = FASTQ_SEQUENCE_HEADER
119 elif state == FASTQ_SEQUENCE_HEADER:
120 # the sequence header isn't really sequence, but
121 # we're just passing it through
122 write_sequence(target1, line)
123 state = FASTQ_QUALITY
124 # sequence or quality data
125 elif state == FASTQ_QUALITY:
126 write_sequence(target1, line)
129 raise RuntimeError("Unrecognized STATE in fastq split")
133 def convert_single_to_two_fastq(instream, target1, target2, mid=None, header=''):
135 read a fastq file where two paired ends have been run together into
138 instream is the source stream
139 target1 and target2 are the destination streams
145 for line in instream:
148 if state == FASTQ_HEADER:
149 write_header(target1, header, line, "/1")
150 write_header(target2, header, line, "/2")
151 state = FASTQ_SEQUENCE
153 elif state == FASTQ_SEQUENCE:
156 write_split_sequence(target1, target2, line, mid)
157 state = FASTQ_SEQUENCE_HEADER
159 elif state == FASTQ_SEQUENCE_HEADER:
160 # the sequence header isn't really sequence, but
161 # we're just passing it through
162 write_sequence(target1, line)
163 write_sequence(target2, line)
165 state = FASTQ_QUALITY
166 # sequence or quality data
167 elif state == FASTQ_QUALITY:
168 write_split_sequence(target1, target2, line, mid)
171 raise RuntimeError("Unrecognized STATE in fastq split")
173 def write_header(target, prefix, line, suffix=''):
176 target.write(line[1:])
178 target.write(os.linesep)
180 def write_sequence(target, line):
182 target.write(os.linesep)
184 def write_split_sequence(target1, target2, line, mid):
185 target1.write(line[:mid])
186 target1.write(os.linesep)
188 target2.write(line[mid:])
189 target2.write(os.linesep)
191 def is_srf(filename):
193 Check filename to see if it is likely to be a SRF file
195 f = open(filename, 'r')
198 return header == "SSRF"
200 def is_cnf1(filename):
202 Brute force detection if a SRF file is using CNF1/CNF4 records
204 max_header = 1024 ** 2
205 PROGRAM_ID = 'PROGRAM_ID\000'
206 cnf4_apps = set(("solexa2srf v1.4",
207 "illumina2srf v1.11.5.Illumina.1.3"))
209 if not is_srf(filename):
210 raise ValueError("%s must be a srf file" % (filename,))
212 fd = os.open(filename, os.O_RDONLY)
213 f = mmap.mmap(fd, 0, access=mmap.ACCESS_READ)
214 # alas the max search length requires python 2.6+
215 program_id_location = f.find(PROGRAM_ID, 0) #, max_header)
216 program_header_start = program_id_location+len(PROGRAM_ID)
217 next_null = f.find('\000', program_header_start) #, max_header)
218 program_id_header = f[program_header_start:next_null]
222 if program_id_header in cnf4_apps:
227 def open_write(filename, force=False):
229 Open a file, but throw an exception if it already exists
232 if os.path.exists(filename):
233 raise RuntimeError("%s exists" % (filename,))
235 return open(filename, 'w')
238 path, name = os.path.split(filename)
239 base, ext = os.path.splitext(name)
241 target1_name = base + '_r1.fastq'
242 target2_name = base + '_r2.fastq'
244 for target_name in [target1_name, target2_name]:
245 print 'target name', target_name
246 if os.path.exists(target_name):
247 raise RuntimeError("%s exists" % (target_name,))
249 instream = open(filename,'r')
250 target1 = open(target1_name,'w')
251 target2 = open(target2_name,'w')
254 if __name__ == "__main__":
255 sys.exit(main(sys.argv[1:]))