xml_row_finder 0.3.0 → 0.3.1
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.
- checksums.yaml +4 -4
- checksums.yaml.gz.sig +0 -0
- data/lib/xml_row_finder.rb +8 -8
- data.tar.gz.sig +0 -0
- metadata +1 -1
- metadata.gz.sig +0 -0
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA256:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 805ba05e167f14ca224f83611d91ada484c31b7c95b329468f4e2fce47302e0e
|
4
|
+
data.tar.gz: f70f9dd37c9e9f4be41d2d0c5d4408c70049d810273c84eeec1b451c40ad0df4
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 2edc584e01ae048ffb4d7b6ee88d4b08a1dd79ce9b593cecf0b24f9fe606809eb1ae50f6e85d5dee70f75a172dcce4eced9e2d3934f90bb23b77321c00fa04e6
|
7
|
+
data.tar.gz: 00ddeb4c7b968e5e31acf8e3accd7d466d9d9941e0d53d7892a9ffb6b82965b2b97f48d7d3cb32c6bb5013466888a12e384a3bd27c52486bddf79b75178758dc
|
checksums.yaml.gz.sig
CHANGED
Binary file
|
data/lib/xml_row_finder.rb
CHANGED
@@ -13,7 +13,7 @@ class XMLRowFinder
|
|
13
13
|
|
14
14
|
@debug = debug
|
15
15
|
|
16
|
-
|
16
|
+
doc = if raws =~ /^http/ then
|
17
17
|
|
18
18
|
nki = Nokorexi.new(url=raws) do |doc1|
|
19
19
|
doc1.xpath('//*[@onclick]').each do |e|
|
@@ -32,9 +32,11 @@ class XMLRowFinder
|
|
32
32
|
Rexle.new(raws)
|
33
33
|
end
|
34
34
|
|
35
|
+
@doc = Rexle.new(doc.xml)
|
36
|
+
|
35
37
|
a = []
|
36
38
|
|
37
|
-
|
39
|
+
doc.root.each_recursive do |e|
|
38
40
|
e.attributes.delete
|
39
41
|
a << e.backtrack.to_xpath
|
40
42
|
end
|
@@ -52,7 +54,7 @@ class XMLRowFinder
|
|
52
54
|
|
53
55
|
# using Nokogiri since Rexle has a bug with xpath predicates
|
54
56
|
#
|
55
|
-
@doc2 = Nokogiri::XML(
|
57
|
+
@doc2 = Nokogiri::XML(doc.root.xml)
|
56
58
|
|
57
59
|
a5 = a4[0..-2].map do |xpath2|
|
58
60
|
[@doc2.xpath(xpath2).length, xpath2]
|
@@ -65,11 +67,11 @@ class XMLRowFinder
|
|
65
67
|
# find the container element
|
66
68
|
xpath = @xpath[/^[^\[]+/]
|
67
69
|
axpath = xpath.split('/')
|
68
|
-
e =
|
70
|
+
e = doc.element xpath
|
69
71
|
|
70
72
|
until (e.xml.include? last_row) do
|
71
73
|
axpath.pop
|
72
|
-
e =
|
74
|
+
e = doc.element axpath.join('/')
|
73
75
|
end
|
74
76
|
|
75
77
|
@cont_xpath = axpath.join('/')
|
@@ -77,10 +79,9 @@ class XMLRowFinder
|
|
77
79
|
end
|
78
80
|
|
79
81
|
# returns the container element for all rows
|
80
|
-
# object returned: Rexle::Element
|
81
82
|
#
|
82
83
|
def body()
|
83
|
-
@doc.element
|
84
|
+
Rexle.new(@doc.element(@cont_xpath).xml)
|
84
85
|
end
|
85
86
|
|
86
87
|
# returns the xpath pointing to the container element for all rows
|
@@ -89,7 +90,6 @@ class XMLRowFinder
|
|
89
90
|
@cont_xpath
|
90
91
|
end
|
91
92
|
|
92
|
-
|
93
93
|
# returns rows
|
94
94
|
# object returned: An array of Nokogiri XML Element object
|
95
95
|
#
|
data.tar.gz.sig
CHANGED
Binary file
|
metadata
CHANGED
metadata.gz.sig
CHANGED
Binary file
|