scbi_fqbin 0.2.2 → 0.2.3
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/.DS_Store +0 -0
- data/.gitignore +14 -0
- data/Gemfile +4 -0
- data/LICENSE.txt +22 -0
- data/{README.rdoc → README.md} +0 -0
- data/Rakefile +8 -28
- data/lib/scbi_fqbin.rb +3 -5
- data/lib/scbi_fqbin/fastabin.rb +411 -0
- data/lib/scbi_fqbin/fastq_file_c.rb +373 -0
- data/lib/scbi_fqbin/fbin_file.rb +1 -1
- data/lib/scbi_fqbin/t.rb +9 -0
- data/lib/scbi_fqbin/t2.rb +12 -0
- data/lib/scbi_fqbin/version.rb +3 -0
- data/lib_fqbin_src.zip +0 -0
- data/lib_fqbin_src/Makefile +66 -0
- data/lib_fqbin_src/fq +0 -0
- data/lib_fqbin_src/fq.c +165 -0
- data/lib_fqbin_src/hash_fqbin +0 -0
- data/lib_fqbin_src/hash_fqbin.c +212 -0
- data/lib_fqbin_src/idx_fqbin +21 -0
- data/lib_fqbin_src/iterate_fqbin +0 -0
- data/lib_fqbin_src/iterate_fqbin.c +136 -0
- data/lib_fqbin_src/lib_fqbin.c +1748 -0
- data/lib_fqbin_src/lib_fqbin.h +194 -0
- data/lib_fqbin_src/mk_fqbin +0 -0
- data/lib_fqbin_src/mk_fqbin.c +138 -0
- data/lib_fqbin_src/other/bwxform.c +915 -0
- data/lib_fqbin_src/other/bwxform.h +74 -0
- data/lib_fqbin_src/other/find_in_index.c +130 -0
- data/lib_fqbin_src/other/hash_fbin_nogzchunks.c +164 -0
- data/lib_fqbin_src/other/idx_fqbin +0 -0
- data/lib_fqbin_src/other/idx_fqbin.c +67 -0
- data/lib_fqbin_src/other/make_hsh.sh +14 -0
- data/lib_fqbin_src/other/rd_extras_fbin.c +45 -0
- data/lib_fqbin_src/read_fq +0 -0
- data/lib_fqbin_src/read_fq.c +143 -0
- data/lib_fqbin_src/read_fqbin +0 -0
- data/lib_fqbin_src/read_fqbin.c +101 -0
- data/lib_fqbin_src/sort_index +9 -0
- data/lib_fqbin_src/test.rb +13 -0
- data/scbi_fqbin.gemspec +25 -0
- data/test/build.rake +15 -0
- data/test/fbinfile +0 -0
- data/test/fbinfile.index +0 -0
- data/test/no_test_fill_file.rb +66 -0
- data/test/old/app.rb +43 -0
- data/test/old/bin/iterate_fastabin.rb +54 -0
- data/test/old/bin/mk_fastabin.rb +22 -0
- data/test/old/bin/rd_fastabin.rb +36 -0
- data/test/old/bin/rd_fq.rb +20 -0
- data/test/old/bioruby.rb +27 -0
- data/test/old/c/Makefile +34 -0
- data/test/old/c/fbin_lib.zip +0 -0
- data/test/old/c/iterate_fbin.c +54 -0
- data/test/old/c/libreria_gz.c +707 -0
- data/test/old/c/libreria_gz.h +127 -0
- data/test/old/c/main.c +86 -0
- data/test/old/c/mk_fbin.c +24 -0
- data/test/old/c/rd_seq_fbin.c +44 -0
- data/test/old/c/test_ffi/a.out +0 -0
- data/test/old/c/test_ffi/app.c +26 -0
- data/test/old/c/test_ffi/app.rb +19 -0
- data/test/old/c/test_ffi/liblibreria_gz.dylib +0 -0
- data/test/old/c/test_ffi/libmylibrary.dylib +0 -0
- data/test/old/c/test_ffi/my_library.rb +23 -0
- data/test/old/c/test_ffi/mylibrary.c +22 -0
- data/test/old/c/test_ffi/mylibrary.h +6 -0
- data/test/old/c/usage_instructions.txt +62 -0
- data/test/old/ext/Makefile +187 -0
- data/test/old/ext/Makefile.dario +34 -0
- data/test/old/ext/extconf.rb +8 -0
- data/test/old/ext/mk_fbin.c +24 -0
- data/test/old/ext/sample/extras.txt +4 -0
- data/{.gemtest → test/old/ext/sample/extras2.txt} +0 -0
- data/test/old/ext/sample/f1.fasta +10 -0
- data/test/old/ext/sample/f1.fasta.qual +10 -0
- data/test/old/ext/sample/f1.fbin +0 -0
- data/test/old/ext/sample/f1.fbin.index +0 -0
- data/test/old/ext/sample/main.c +86 -0
- data/test/old/ext/usage_instructions.txt +62 -0
- data/test/old/t_scbi_fastabin.rb +140 -0
- data/test/read_tests/10-original_sizes.sh +16 -0
- data/test/read_tests/20-fq_time.sh +23 -0
- data/test/read_tests/30-fbin_read_time.sh +23 -0
- data/test/read_tests/40-bsc_read_time.sh +21 -0
- data/test/read_tests/50-fq_time_x4.sh +25 -0
- data/test/read_tests/60-fbin_read_time_x4.sh +24 -0
- data/test/read_tests/70-bsc_read_time_x4.sh +32 -0
- data/test/results_bio_scbi_fasta.txt +11 -0
- data/test/{test_scbi_fbin_file.rb → scbi_fbin_file_test.rb} +0 -0
- data/test/speed.txt +81 -0
- data/test/t_scbi_fasta.rb +12 -0
- data/test/write_tests/10-original_sizes.sh +16 -0
- data/test/write_tests/20-zip_time.sh +17 -0
- data/test/write_tests/30-mk_fbin_time.sh +23 -0
- data/test/write_tests/31-mk_fbin_time_f30.sh +21 -0
- data/test/write_tests/40-gzip_time.sh +16 -0
- data/test/write_tests/41-bsc_time.sh +16 -0
- data/test/write_tests/50-zip_sizes.sh +16 -0
- data/test/write_tests/60-fbin_sizes.sh +17 -0
- data/test/write_tests/61-fbin_sizes_f30.sh +16 -0
- data/test/write_tests/70-gzip_sizes.sh +17 -0
- data/test/write_tests/80-bsc_sizes.sh +17 -0
- data/website/index.html +87 -0
- data/website/index.txt +81 -0
- data/website/javascripts/rounded_corners_lite.inc.js +285 -0
- data/website/stylesheets/screen.css +159 -0
- data/website/template.html.erb +50 -0
- metadata +208 -95
- data/History.txt +0 -19
- data/Manifest.txt +0 -12
- data/PostInstall.txt +0 -7
- data/script/console +0 -10
- data/script/destroy +0 -14
- data/script/generate +0 -14
checksums.yaml
ADDED
@@ -0,0 +1,7 @@
|
|
1
|
+
---
|
2
|
+
SHA1:
|
3
|
+
metadata.gz: dcb4c18dbccdd6482584cb3802ed537b58c3786e
|
4
|
+
data.tar.gz: 7b545c529fd75b6411d4d611d48f2ceb8b81f845
|
5
|
+
SHA512:
|
6
|
+
metadata.gz: 5948652a1740474b8f03589bbfe3a6761d6fab7a048ceef0d29ba29203793f21e233b53f78c79e867d7cb6f672ca835dd44468b5d6bab2968dfe9fc297fa7015
|
7
|
+
data.tar.gz: e0dd01ce455d5789caa5bc7a888b77d295871e759f5509ce4f1b002fffab5f7e68f3ae01ff5a2d8bcb2f55f6fcbc33c5b3ccd63922a0a597aff3de19393f4e63
|
data/.DS_Store
ADDED
Binary file
|
data/.gitignore
ADDED
data/Gemfile
ADDED
data/LICENSE.txt
ADDED
@@ -0,0 +1,22 @@
|
|
1
|
+
Copyright (c) 2014 dariogf
|
2
|
+
|
3
|
+
MIT License
|
4
|
+
|
5
|
+
Permission is hereby granted, free of charge, to any person obtaining
|
6
|
+
a copy of this software and associated documentation files (the
|
7
|
+
"Software"), to deal in the Software without restriction, including
|
8
|
+
without limitation the rights to use, copy, modify, merge, publish,
|
9
|
+
distribute, sublicense, and/or sell copies of the Software, and to
|
10
|
+
permit persons to whom the Software is furnished to do so, subject to
|
11
|
+
the following conditions:
|
12
|
+
|
13
|
+
The above copyright notice and this permission notice shall be
|
14
|
+
included in all copies or substantial portions of the Software.
|
15
|
+
|
16
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
|
17
|
+
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
|
18
|
+
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
|
19
|
+
NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
|
20
|
+
LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
|
21
|
+
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
|
22
|
+
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
|
data/{README.rdoc → README.md}
RENAMED
File without changes
|
data/Rakefile
CHANGED
@@ -1,28 +1,8 @@
|
|
1
|
-
require
|
2
|
-
|
3
|
-
require '
|
4
|
-
|
5
|
-
|
6
|
-
|
7
|
-
|
8
|
-
|
9
|
-
# Hoe.plugin :cucumberfeatures
|
10
|
-
|
11
|
-
# Generate all the Rake tasks
|
12
|
-
# Run 'rake -T' to see list of generated tasks (from gem root directory)
|
13
|
-
$hoe = Hoe.spec 'scbi_fqbin' do
|
14
|
-
self.developer 'Dario Guerrero', 'dariogf@gmail.com'
|
15
|
-
self.post_install_message = 'PostInstall.txt' # TODO remove if post-install message not required
|
16
|
-
self.rubyforge_name = self.name # TODO this is default value
|
17
|
-
self.extra_deps = [['scbi_fasta','>= 0.0.1'],['ffi','>= 0.0.1']]
|
18
|
-
|
19
|
-
end
|
20
|
-
|
21
|
-
require 'newgem/tasks'
|
22
|
-
Dir['tasks/**/*.rake'].each { |t| load t; puts t}
|
23
|
-
|
24
|
-
# TODO - want other tests/tasks run by default? Add them to the list
|
25
|
-
|
26
|
-
# remove_task :default
|
27
|
-
task :default => [:build, :install,:spec, :features]
|
28
|
-
|
1
|
+
require "bundler/gem_tasks"
|
2
|
+
|
3
|
+
require 'rake/testtask'
|
4
|
+
|
5
|
+
Rake::TestTask.new do |t|
|
6
|
+
t.libs << 'test'
|
7
|
+
t.pattern = "test/*_test.rb"
|
8
|
+
end
|
data/lib/scbi_fqbin.rb
CHANGED
@@ -1,8 +1,6 @@
|
|
1
|
-
|
2
|
-
|
3
|
-
# require 'fastabin'
|
4
|
-
require 'fbin_file'
|
1
|
+
require "scbi_fqbin/version"
|
2
|
+
require 'scbi_fqbin/fbin_file'
|
5
3
|
|
6
4
|
module ScbiFqbin
|
7
|
-
|
5
|
+
# Your code goes here...
|
8
6
|
end
|
@@ -0,0 +1,411 @@
|
|
1
|
+
########################################################
|
2
|
+
# Author: Dario Guerrero & Rafael Larrosa
|
3
|
+
#
|
4
|
+
# Encapsulates FastaBIN format routines
|
5
|
+
#
|
6
|
+
#
|
7
|
+
#
|
8
|
+
########################################################
|
9
|
+
|
10
|
+
require 'zlib'
|
11
|
+
require 'scbi_fasta'
|
12
|
+
|
13
|
+
class Fastabin
|
14
|
+
|
15
|
+
#--------------------------------------
|
16
|
+
# CONSTANT DEFINITIONS
|
17
|
+
#--------------------------------------
|
18
|
+
|
19
|
+
#Compression type. Valid values are Zlib::NO_COMPRESSION, Zlib::BEST_SPEED, Zlib::BEST_COMPRESSION, Zlib::DEFAULT_COMPRESSION, and an integer from 0 to 9.
|
20
|
+
COMPRESSION=Zlib::BEST_COMPRESSION
|
21
|
+
SEQUENCES_PER_BLOCK=10
|
22
|
+
READ_BIN_REG_EXP= /^([^\s]+)\s+([^\s]+)\s+([^\s]+)\s+([^\s]+)/
|
23
|
+
|
24
|
+
READ_REG_EXP= /^([^\s]+)\s+([^\s]+)\s+([^\s]+)\s+([^\s]+)\s+([^\s]+)/
|
25
|
+
|
26
|
+
#--------------------------------------
|
27
|
+
# Constructor
|
28
|
+
#--------------------------------------
|
29
|
+
def initialize(filename, mode = 'r', index_filename=nil)
|
30
|
+
|
31
|
+
@filename=filename
|
32
|
+
@index_filename = index_filename ||= filename+'.index';
|
33
|
+
@bin_file = nil
|
34
|
+
@need_to_regenerate_index=false
|
35
|
+
|
36
|
+
@added_sequence_count=0
|
37
|
+
|
38
|
+
# check open mode read or write
|
39
|
+
if mode.upcase.index('W')
|
40
|
+
bin_mode='wb'
|
41
|
+
if File.exists?(@index_filename)
|
42
|
+
File.delete(@index_filename)
|
43
|
+
end
|
44
|
+
else
|
45
|
+
|
46
|
+
# if read mode, check if file exits
|
47
|
+
if !File.exists?(filename)
|
48
|
+
raise "File #{filename} doesn't exists'"
|
49
|
+
end
|
50
|
+
|
51
|
+
# if index doesn't exits, recreate it
|
52
|
+
if !File.exists?(@index_filename)
|
53
|
+
regenerate_index
|
54
|
+
end
|
55
|
+
|
56
|
+
# check open mode
|
57
|
+
bin_mode = 'rb'
|
58
|
+
|
59
|
+
end
|
60
|
+
|
61
|
+
#open files
|
62
|
+
@bin_io = File.open(filename,bin_mode)
|
63
|
+
@bin_file = Zlib::GzipWriter.new(@bin_io)
|
64
|
+
|
65
|
+
end
|
66
|
+
|
67
|
+
|
68
|
+
#--------------------------------------
|
69
|
+
# Add a pair of fasta and qual files to the Fastabin file
|
70
|
+
#--------------------------------------
|
71
|
+
def add_fasta_qual(fasta_file_name,qual_file_name)
|
72
|
+
|
73
|
+
# use FastaQualFile to read fasta
|
74
|
+
qf = FastaQualFile.new(fasta_file_name,qual_file_name)
|
75
|
+
|
76
|
+
|
77
|
+
# iterate over sequences
|
78
|
+
qf.each do |name,fasta,qual|
|
79
|
+
|
80
|
+
if (qf.num_seqs % 10000) == 0
|
81
|
+
puts Time.now.to_s + ',' + qf.num_seqs.to_s + ':' + name
|
82
|
+
end
|
83
|
+
#
|
84
|
+
#add them to fastabin
|
85
|
+
add_seq(name,fasta,qual,nil)
|
86
|
+
|
87
|
+
end
|
88
|
+
|
89
|
+
qf.close
|
90
|
+
|
91
|
+
end
|
92
|
+
|
93
|
+
#--------------------------------------
|
94
|
+
# Add one seq to the fastabin file
|
95
|
+
#--------------------------------------
|
96
|
+
def add_seq(seq_name, seq_fasta, seq_qual, seq_extras=nil)
|
97
|
+
|
98
|
+
if (@added_sequence_count % SEQUENCES_PER_BLOCK==0)
|
99
|
+
#@bin_file.flush
|
100
|
+
@bin_file.close
|
101
|
+
#@bin_io.close
|
102
|
+
|
103
|
+
@bin_io = File.open(@filename,'ab')
|
104
|
+
@bin_file = Zlib::GzipWriter.new(@bin_io)
|
105
|
+
puts "NEW BLOCK"
|
106
|
+
|
107
|
+
end
|
108
|
+
@added_sequence_count += 1
|
109
|
+
|
110
|
+
@need_to_regenerate_index = true
|
111
|
+
|
112
|
+
zipped_fasta = ''
|
113
|
+
zipped_qual = ''
|
114
|
+
zipped_extras = ''
|
115
|
+
|
116
|
+
ini = @bin_file.pos
|
117
|
+
|
118
|
+
# get current pos and write deflated data to fastabin format
|
119
|
+
#zipped_fasta = deflate_fasta(seq_fasta.strip)
|
120
|
+
zipped_fasta = seq_fasta.strip
|
121
|
+
|
122
|
+
if !seq_qual.empty?
|
123
|
+
q = seq_qual.strip.split(' ')
|
124
|
+
q.map!{|e| (e.to_i+33).chr}
|
125
|
+
#zipped_qual = deflate_qual(q.join)
|
126
|
+
zipped_qual = q.join
|
127
|
+
|
128
|
+
#puts q.join
|
129
|
+
#raise
|
130
|
+
#zipped_qual = deflate_qual(seq_qual.strip)
|
131
|
+
|
132
|
+
end
|
133
|
+
|
134
|
+
if !seq_extras.nil?
|
135
|
+
#zipped_extras = deflate_extras(seq_extras)
|
136
|
+
zipped_extras = seq_extras
|
137
|
+
end
|
138
|
+
|
139
|
+
# write data to index file and bin (to retrieve it later if index file gets lost)
|
140
|
+
head = "#{seq_name} #{zipped_fasta.size} #{zipped_qual.size} #{zipped_extras.size}"
|
141
|
+
bin_index_line ="#{head.size.to_s.rjust(4)}#{head}"
|
142
|
+
|
143
|
+
#index_line ="#{seq_name} #{ini+bin_index_line.size} #{zipped_fasta.size} #{zipped_qual.size} #{zipped_extras.size}"
|
144
|
+
#puts index_line
|
145
|
+
#index_file.puts index_line
|
146
|
+
|
147
|
+
@bin_file.write bin_index_line
|
148
|
+
|
149
|
+
# puts zipped_fasta
|
150
|
+
# puts zipped_qual
|
151
|
+
# puts zipped_extras
|
152
|
+
|
153
|
+
# puts "1F:#{zipped_fasta}@#{zipped_fasta.size.to_s}@#{zipped_fasta.length.to_s}"
|
154
|
+
# puts "1Q:#{zipped_qual}@#{zipped_qual.size.to_s}@#{zipped_qual.length.to_s}"
|
155
|
+
# puts "1E:#{zipped_extras}@#{zipped_extras.size.to_s}@#{zipped_extras.length.to_s}"
|
156
|
+
#
|
157
|
+
|
158
|
+
@bin_file.puts zipped_fasta
|
159
|
+
@bin_file.puts zipped_qual
|
160
|
+
@bin_file.puts zipped_extras
|
161
|
+
|
162
|
+
|
163
|
+
|
164
|
+
end
|
165
|
+
|
166
|
+
#--------------------------------------
|
167
|
+
# Iterate over all sequences of a fastabin file
|
168
|
+
#--------------------------------------
|
169
|
+
def each(get_fasta=true,get_qual=true,get_extras=true)
|
170
|
+
|
171
|
+
#bin = ''
|
172
|
+
@bin_file.pos=0
|
173
|
+
|
174
|
+
|
175
|
+
while !@bin_file.eof?
|
176
|
+
|
177
|
+
head_size = @bin_file.read(4).to_i
|
178
|
+
line = @bin_file.read(head_size)
|
179
|
+
|
180
|
+
if line =~ READ_BIN_REG_EXP
|
181
|
+
|
182
|
+
name = $1
|
183
|
+
i = @bin_file.pos
|
184
|
+
fz = $2.to_i
|
185
|
+
qz=$3.to_i
|
186
|
+
ez=$4.to_i
|
187
|
+
|
188
|
+
@bin_file.pos = i+fz+qz+ez
|
189
|
+
|
190
|
+
name,fasta,qual,extras = extract_seq(name,i,fz,qz,ez)
|
191
|
+
|
192
|
+
yield(name,fasta,qual,extras)
|
193
|
+
|
194
|
+
else
|
195
|
+
raise "Invalid index line found at each"
|
196
|
+
end
|
197
|
+
|
198
|
+
end
|
199
|
+
|
200
|
+
end
|
201
|
+
|
202
|
+
#--------------------------------------
|
203
|
+
# Iterate over all sequences of a fastabin file
|
204
|
+
#--------------------------------------
|
205
|
+
def each_by_index(get_fasta=true,get_qual=true)
|
206
|
+
index_file = Zlib::GzipReader.open(@index_filename)
|
207
|
+
|
208
|
+
# iterate over each line of index_file
|
209
|
+
index_file.each_line do |e|
|
210
|
+
|
211
|
+
# parse params
|
212
|
+
if e=~ READ_REG_EXP
|
213
|
+
|
214
|
+
name = $1
|
215
|
+
i = $2.to_i
|
216
|
+
fz = $3.to_i
|
217
|
+
qz=$4.to_i
|
218
|
+
ez=$5.to_i
|
219
|
+
|
220
|
+
name,fasta,qual,extras = extract_seq(name,i,fz,qz,ez)
|
221
|
+
|
222
|
+
yield(name,fasta,qual,extras)
|
223
|
+
end
|
224
|
+
|
225
|
+
end
|
226
|
+
index_file.close
|
227
|
+
|
228
|
+
end
|
229
|
+
|
230
|
+
#--------------------------------------
|
231
|
+
# Read one seq from the fastabin file by name
|
232
|
+
#--------------------------------------
|
233
|
+
def read_seq(seq_name)
|
234
|
+
|
235
|
+
index_file = Zlib::GzipReader.open(@index_filename)
|
236
|
+
|
237
|
+
res = nil
|
238
|
+
e=nil
|
239
|
+
|
240
|
+
index_file.grep(/^#{seq_name}\s/) do |line|
|
241
|
+
|
242
|
+
e=line.chomp
|
243
|
+
|
244
|
+
# parse params
|
245
|
+
if e=~ READ_REG_EXP
|
246
|
+
name = $1
|
247
|
+
i = $2.to_i
|
248
|
+
fz = $3.to_i
|
249
|
+
qz=$4.to_i
|
250
|
+
ez=$5.to_i
|
251
|
+
|
252
|
+
res = extract_seq(name,i,fz,qz,ez)
|
253
|
+
end
|
254
|
+
break
|
255
|
+
end
|
256
|
+
|
257
|
+
index_file.close
|
258
|
+
|
259
|
+
return res
|
260
|
+
end
|
261
|
+
|
262
|
+
|
263
|
+
#--------------------------------------
|
264
|
+
# Count lines in index file. This is sequence count
|
265
|
+
#--------------------------------------
|
266
|
+
def count
|
267
|
+
index_file = Zlib::GzipReader.open(@index_filename)
|
268
|
+
|
269
|
+
res=index_file.readlines.count
|
270
|
+
|
271
|
+
index_file.close
|
272
|
+
|
273
|
+
return res
|
274
|
+
end
|
275
|
+
|
276
|
+
#--------------------------------------
|
277
|
+
# Close files
|
278
|
+
#--------------------------------------
|
279
|
+
def close
|
280
|
+
|
281
|
+
@bin_file.close if !@bin_file.closed?
|
282
|
+
#@bin_io.close if !@bin_io.closed?
|
283
|
+
|
284
|
+
regenerate_index if @need_to_regenerate_index
|
285
|
+
end
|
286
|
+
|
287
|
+
|
288
|
+
private
|
289
|
+
|
290
|
+
#--------------------------------------
|
291
|
+
# Extract/regenerate index from bin file
|
292
|
+
#--------------------------------------
|
293
|
+
|
294
|
+
def regenerate_index
|
295
|
+
#puts Time.now.to_s + " - extracting index"
|
296
|
+
|
297
|
+
bin_io = File.open(@filename,'rb')
|
298
|
+
bin_file = Zlib::GzipReader.new(bin_io).to_io
|
299
|
+
|
300
|
+
index_file = Zlib::GzipWriter.open(@index_filename)
|
301
|
+
#@bin_file.pos=0
|
302
|
+
i=0
|
303
|
+
while !bin_file.eof?
|
304
|
+
|
305
|
+
head_size = bin_file.read(4).to_i
|
306
|
+
line = bin_file.read(head_size)
|
307
|
+
|
308
|
+
if line =~ READ_BIN_REG_EXP
|
309
|
+
|
310
|
+
name = $1
|
311
|
+
i = bin_file.pos
|
312
|
+
fz = $2.to_i
|
313
|
+
qz=$3.to_i
|
314
|
+
ez=$4.to_i
|
315
|
+
|
316
|
+
index_line ="#{name} #{i} #{fz} #{qz} #{ez}"
|
317
|
+
#puts index_line
|
318
|
+
i += 1
|
319
|
+
|
320
|
+
puts i
|
321
|
+
index_file.puts index_line
|
322
|
+
|
323
|
+
bin_file.pos = i+fz+qz+ez
|
324
|
+
|
325
|
+
else
|
326
|
+
raise "Invalid index line found at each"
|
327
|
+
end
|
328
|
+
|
329
|
+
end
|
330
|
+
|
331
|
+
#puts Time.now.to_s + ' - done'
|
332
|
+
|
333
|
+
index_file.close
|
334
|
+
bin_file.close
|
335
|
+
#bin_io.close
|
336
|
+
|
337
|
+
|
338
|
+
end
|
339
|
+
|
340
|
+
def extract_seq(name,i,fz,qz,ez)
|
341
|
+
|
342
|
+
#annotate current pos in files to revert changes later
|
343
|
+
current_bin_pos = @bin_file.pos
|
344
|
+
|
345
|
+
res = nil
|
346
|
+
|
347
|
+
# read bin data and inflate it
|
348
|
+
@bin_file.pos = i
|
349
|
+
|
350
|
+
fasta_d = @bin_file.read(fz)
|
351
|
+
# puts "2F:#{fasta_d}@#{fasta_d.size}"
|
352
|
+
fasta = inflate(fasta_d)
|
353
|
+
fasta_d=nil
|
354
|
+
|
355
|
+
qual_d = @bin_file.read(qz)
|
356
|
+
qual = inflate(qual_d)
|
357
|
+
# puts "2Q:#{qual_d}@#{qual_d.size}"
|
358
|
+
qual_d=nil
|
359
|
+
|
360
|
+
extras_d = @bin_file.read(ez)
|
361
|
+
extras = inflate(extras_d)
|
362
|
+
# puts "2E:#{extras_d}@#{extras_d.size}"
|
363
|
+
extras_d=nil
|
364
|
+
|
365
|
+
res =[name, fasta, qual, extras]
|
366
|
+
|
367
|
+
@bin_file.pos = current_bin_pos
|
368
|
+
|
369
|
+
return res
|
370
|
+
end
|
371
|
+
|
372
|
+
def deflate_fasta(input)
|
373
|
+
|
374
|
+
zipper = Zlib::Deflate.new(Zlib::BEST_COMPRESSION,15,9,Zlib::FILTERED)
|
375
|
+
res= zipper.deflate(input, Zlib::FINISH)
|
376
|
+
zipper.close
|
377
|
+
|
378
|
+
return res
|
379
|
+
end
|
380
|
+
|
381
|
+
def deflate_qual(input)
|
382
|
+
|
383
|
+
# zipper = Zlib::Deflate.new(Zlib::BEST_COMPRESSION,15,9,Zlib::FILTERED)
|
384
|
+
zipper = Zlib::Deflate.new(Zlib::BEST_COMPRESSION,15,9)
|
385
|
+
res= zipper.deflate(input, Zlib::FINISH)
|
386
|
+
zipper.close
|
387
|
+
|
388
|
+
return res
|
389
|
+
end
|
390
|
+
|
391
|
+
def deflate_extras(input)
|
392
|
+
|
393
|
+
zipper = Zlib::Deflate.new(Zlib::BEST_COMPRESSION,15,9,Zlib::HUFFMAN_ONLY)
|
394
|
+
res= zipper.deflate(input, Zlib::FINISH)
|
395
|
+
zipper.close
|
396
|
+
|
397
|
+
return res
|
398
|
+
end
|
399
|
+
|
400
|
+
def inflate(input)
|
401
|
+
|
402
|
+
unzipper = Zlib::Inflate.new(15)
|
403
|
+
res= unzipper.inflate(input)
|
404
|
+
#unzipper.finish
|
405
|
+
unzipper.close
|
406
|
+
|
407
|
+
return res
|
408
|
+
end
|
409
|
+
|
410
|
+
end
|
411
|
+
|