entityextractor 0.0.6 → 0.0.7
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/lib/extractdates.rb +50 -25
- metadata +2 -2
data/lib/extractdates.rb
CHANGED
|
@@ -16,16 +16,8 @@ class ExtractDates
|
|
|
16
16
|
c = @text.chunk
|
|
17
17
|
c.each do |i|
|
|
18
18
|
s = paragraph(i).segment
|
|
19
|
-
dateExtract(s)
|
|
20
19
|
s.each do |j|
|
|
21
|
-
|
|
22
|
-
shash[:date] = dateExtract(j)
|
|
23
|
-
if shash[:date]
|
|
24
|
-
shash[:file] = file
|
|
25
|
-
shash[:title] = j
|
|
26
|
-
shash[:description] = i
|
|
27
|
-
@output.push(shash)
|
|
28
|
-
end
|
|
20
|
+
dateExtract(j, file, j, i)
|
|
29
21
|
end
|
|
30
22
|
end
|
|
31
23
|
rescue
|
|
@@ -34,25 +26,58 @@ class ExtractDates
|
|
|
34
26
|
return @output
|
|
35
27
|
end
|
|
36
28
|
|
|
37
|
-
def dateExtract(blob)
|
|
38
|
-
|
|
39
|
-
|
|
40
|
-
# mm/dd/yyyy
|
|
41
|
-
# Month dd, yyyy
|
|
42
|
-
# Month ddth, yyyy
|
|
43
|
-
# Month yyyy
|
|
44
|
-
|
|
45
|
-
# TOADD:
|
|
46
|
-
# Multiple dates
|
|
47
|
-
# Year detection
|
|
48
|
-
# Conditional American dates
|
|
49
|
-
# Time ranges
|
|
50
|
-
# Filtering
|
|
51
|
-
|
|
29
|
+
def dateExtract(blob, file, title, description)
|
|
30
|
+
blobstring = blob.to_s
|
|
31
|
+
|
|
52
32
|
begin
|
|
53
|
-
|
|
33
|
+
if blobstring.match(/(\d{1,2})\/(\d{1,2})\/(\d{2,4})/)
|
|
34
|
+
save = blobstring.match(/(\d{1,2})\/(\d{1,2})\/(\d{2,4})/)
|
|
35
|
+
addItem(DateTime.parse(blob).to_s, file, title, description, blobstring, save.to_s)
|
|
36
|
+
elsif blobstring.match(/(\d{1,2})-(\d{1,2})-(\d{2,4})/)
|
|
37
|
+
save = blobstring.match(/(\d{1,2})-(\d{1,2})-(\d{2,4})/)
|
|
38
|
+
addItem(DateTime.parse(blob).to_s, file, title, description, blobstring, save.to_s)
|
|
39
|
+
elsif blobstring.match(/(.+?)(\w+ \d{1,2}(st|nd|rd|th|), \d{4})/)
|
|
40
|
+
save = blobstring.match(/(.+?)(\w+ \d{1,2}(st|nd|rd|th|), \d{4})/)
|
|
41
|
+
addItem(DateTime.parse(blob).to_s, file, title, description, blobstring, save.to_s)
|
|
42
|
+
elsif blobstring.match(/(.+?) ((?:Jan(?:uary)?|Feb(?:ruary)?|Mar(?:ch)?|Apr(?:il)?|May|Jun(?:e)?|Jul(?:y)?|Aug(?:ust)?|Sep(?:tember)?|Nov(?:ember)?|Dec(?:ember)?) \d{2}(st|nd|rd|th|)( |\)|\]))/)
|
|
43
|
+
save = blobstring.match(/(.+?) ((?:Jan(?:uary)?|Feb(?:ruary)?|Mar(?:ch)?|Apr(?:il)?|May|Jun(?:e)?|Jul(?:y)?|Aug(?:ust)?|Sep(?:tember)?|Nov(?:ember)?|Dec(?:ember)?) \d{2}(st|nd|rd|th|)( |\)|\]))/)
|
|
44
|
+
addItem(DateTime.parse(blob).to_s, file, title, description, blobstring, save.to_s)
|
|
45
|
+
elsif blobstring.match(/((?:Jan(?:uary)?|Feb(?:ruary)?|Mar(?:ch)?|Apr(?:il)?|May|Jun(?:e)?|Jul(?:y)?|Aug(?:ust)?|Sep(?:tember)?|Nov(?:ember)?|Dec(?:ember)?) [1-2][0,9]\d{2}( |\)|\]))/)
|
|
46
|
+
save = blobstring.match(/((?:Jan(?:uary)?|Feb(?:ruary)?|Mar(?:ch)?|Apr(?:il)?|May|Jun(?:e)?|Jul(?:y)?|Aug(?:ust)?|Sep(?:tember)?|Nov(?:ember)?|Dec(?:ember)?) [1-2][0,9]\d{2}( |\)|\]))/)
|
|
47
|
+
addItem(DateTime.parse(blob).to_s, file, title, description, blobstring, save.to_s)
|
|
48
|
+
elsif blobstring.match(/(\d{4})-(\d{2})-(\d{2})/)
|
|
49
|
+
save = blobstring.match(/(\d{4})-(\d{2})-(\d{2})/)
|
|
50
|
+
addItem(DateTime.parse(blob).to_s, file, title, description, blobstring, save.to_s)
|
|
51
|
+
elsif blobstring.match(/(\d{4})\/(\d{2})\/(\d{2})/)
|
|
52
|
+
save = blobstring.match(/(\d{4})\/(\d{2})\/(\d{2})/)
|
|
53
|
+
addItem(DateTime.parse(blob).to_s, file, title, description, blobstring, save.to_s)
|
|
54
|
+
end
|
|
54
55
|
rescue
|
|
55
56
|
end
|
|
56
57
|
end
|
|
58
|
+
|
|
59
|
+
# Adds and item to the hash
|
|
60
|
+
def addItem(date, file, title, description, blob, regex)
|
|
61
|
+
shash = Hash.new
|
|
62
|
+
shash[:date] = date
|
|
63
|
+
shash[:file] = file
|
|
64
|
+
shash[:title] = title
|
|
65
|
+
shash[:description] = description
|
|
66
|
+
|
|
67
|
+
flag = 0
|
|
68
|
+
@output.each do |o|
|
|
69
|
+
if (o[:date] == shash[:date]) && (o[:file] == shash[:file]) && (o[:title].to_s == shash[:title].to_s)
|
|
70
|
+
flag = 1
|
|
71
|
+
break
|
|
72
|
+
end
|
|
73
|
+
end
|
|
74
|
+
|
|
75
|
+
if flag == 0
|
|
76
|
+
@output.push(shash)
|
|
77
|
+
end
|
|
78
|
+
|
|
79
|
+
blob.slice! regex
|
|
80
|
+
dateExtract(blob, file, title, description)
|
|
81
|
+
end
|
|
57
82
|
end
|
|
58
83
|
|
metadata
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: entityextractor
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
version: 0.0.
|
|
4
|
+
version: 0.0.7
|
|
5
5
|
prerelease:
|
|
6
6
|
platform: ruby
|
|
7
7
|
authors:
|
|
@@ -9,7 +9,7 @@ authors:
|
|
|
9
9
|
autorequire:
|
|
10
10
|
bindir: bin
|
|
11
11
|
cert_chain: []
|
|
12
|
-
date: 2014-03-
|
|
12
|
+
date: 2014-03-18 00:00:00.000000000 Z
|
|
13
13
|
dependencies: []
|
|
14
14
|
description: Extracts entities and terms from any JSON.
|
|
15
15
|
email: shidash@shidash.com
|