cassiopee 0.1.3 → 0.1.4
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- data/Changelog +1 -0
- data/lib/cassiopee-mt.rb +44 -15
- metadata +3 -3
data/Changelog
CHANGED
data/lib/cassiopee-mt.rb
CHANGED
|
@@ -15,6 +15,8 @@ module CassiopeeMt
|
|
|
15
15
|
# Filtering is used to split the input data according to maxthread
|
|
16
16
|
# Matches of each thread are merge to matches of CrawlerMT
|
|
17
17
|
class CrawlerMt < Crawler
|
|
18
|
+
|
|
19
|
+
MINSEQSIZE=10
|
|
18
20
|
|
|
19
21
|
# Max number fo threads to use
|
|
20
22
|
attr_accessor :maxthread
|
|
@@ -33,42 +35,69 @@ module CassiopeeMt
|
|
|
33
35
|
crawler.file_suffix = @file_suffix
|
|
34
36
|
crawler.loadIndex()
|
|
35
37
|
#crawler.file_suffix = @file_suffix+"."+threadId.to_s
|
|
36
|
-
crawler.indexString(@sequence)
|
|
37
38
|
end
|
|
38
39
|
|
|
39
40
|
def searchExact(pattern)
|
|
40
|
-
|
|
41
|
-
|
|
41
|
+
len = @sequence.length
|
|
42
|
+
if(@min_position>0)
|
|
43
|
+
min = @min_position
|
|
44
|
+
else
|
|
45
|
+
min = 0
|
|
46
|
+
end
|
|
47
|
+
if(@max_position>0)
|
|
48
|
+
max = @max_position
|
|
49
|
+
else
|
|
50
|
+
max= @sequence.length
|
|
51
|
+
end
|
|
52
|
+
len = max - min
|
|
53
|
+
if(len<MINSEQSIZE)
|
|
54
|
+
@maxthread=1
|
|
55
|
+
end
|
|
56
|
+
nb = len.div(maxthread)
|
|
42
57
|
(1..maxthread).each do |i|
|
|
43
58
|
crawler = Crawler.new
|
|
44
59
|
setParams(crawler,i)
|
|
45
|
-
|
|
60
|
+
curmax = min + nb
|
|
46
61
|
if(i==maxthread)
|
|
47
|
-
|
|
62
|
+
curmax = max
|
|
48
63
|
end
|
|
49
|
-
crawler.filter_position(min,
|
|
50
|
-
$log.debug("Start new Thread between " << min.to_s << " and " <<
|
|
64
|
+
crawler.filter_position(min,curmax)
|
|
65
|
+
$log.debug("Start new Thread between " << min.to_s << " and " << curmax.to_s)
|
|
51
66
|
@th[i-1] = Thread.new{ Thread.current["matches"] = crawler.searchExact(pattern) }
|
|
52
|
-
min =
|
|
67
|
+
min = curmax + 1
|
|
53
68
|
end
|
|
54
69
|
@th.each {|t| t.join; t["matches"].each { |m| @matches << m }}
|
|
55
70
|
return @matches
|
|
56
71
|
end
|
|
57
72
|
|
|
58
73
|
def searchApproximate(s,edit)
|
|
59
|
-
|
|
60
|
-
|
|
74
|
+
len = @sequence.length
|
|
75
|
+
if(@min_position>0)
|
|
76
|
+
min = @min_position
|
|
77
|
+
else
|
|
78
|
+
min = 0
|
|
79
|
+
end
|
|
80
|
+
if(@max_position>0)
|
|
81
|
+
max = @max_position
|
|
82
|
+
else
|
|
83
|
+
max = @sequence.length
|
|
84
|
+
end
|
|
85
|
+
len = max - min
|
|
86
|
+
if(len<MINSEQSIZE)
|
|
87
|
+
@maxthread=1
|
|
88
|
+
end
|
|
89
|
+
nb = len.div(maxthread)
|
|
61
90
|
(1..maxthread).each do |i|
|
|
62
91
|
crawler = Crawler.new
|
|
63
92
|
setParams(crawler,i)
|
|
64
|
-
|
|
93
|
+
curmax = min + nb
|
|
65
94
|
if(i==maxthread)
|
|
66
|
-
|
|
95
|
+
curmax = max
|
|
67
96
|
end
|
|
68
|
-
crawler.filter_position(min,
|
|
69
|
-
$log.debug("Start new Thread between " << min.to_s << " and " <<
|
|
97
|
+
crawler.filter_position(min,curmax)
|
|
98
|
+
$log.debug("Start new Thread between " << min.to_s << " and " << curmax.to_s)
|
|
70
99
|
@th[i-1] = Thread.new{ Thread.current["matches"] = crawler.searchApproximate(s,edit) }
|
|
71
|
-
min =
|
|
100
|
+
min = curmax + 1
|
|
72
101
|
end
|
|
73
102
|
@th.each {|t| t.join; t["matches"].each { |m| @matches << m }}
|
|
74
103
|
return @matches
|
metadata
CHANGED
|
@@ -1,13 +1,13 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: cassiopee
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
hash:
|
|
4
|
+
hash: 19
|
|
5
5
|
prerelease: false
|
|
6
6
|
segments:
|
|
7
7
|
- 0
|
|
8
8
|
- 1
|
|
9
|
-
-
|
|
10
|
-
version: 0.1.
|
|
9
|
+
- 4
|
|
10
|
+
version: 0.1.4
|
|
11
11
|
platform: ruby
|
|
12
12
|
authors:
|
|
13
13
|
- Olivier Sallou
|