markitdown 0.1.0 → 0.1.1
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.
- data/lib/markitdown.rb +5 -1
- data/lib/markitdown/version.rb +1 -1
- data/spec/table2.html +5 -5
- data/spec/table2.markdown +3 -3
- metadata +3 -3
data/lib/markitdown.rb
CHANGED
|
@@ -23,7 +23,8 @@ module Markitdown
|
|
|
23
23
|
states.unshift node.name.downcase
|
|
24
24
|
pre = prefix(states)
|
|
25
25
|
recurse = true
|
|
26
|
-
|
|
26
|
+
strip_content = false
|
|
27
|
+
flatten_content = false
|
|
27
28
|
case node.name
|
|
28
29
|
when "head"
|
|
29
30
|
return []
|
|
@@ -135,8 +136,10 @@ module Markitdown
|
|
|
135
136
|
after = "\n\n"
|
|
136
137
|
when "th"
|
|
137
138
|
results << "|"
|
|
139
|
+
flatten_content = true
|
|
138
140
|
when "td"
|
|
139
141
|
results << "|"
|
|
142
|
+
flatten_content = true
|
|
140
143
|
when "tr"
|
|
141
144
|
after = "|\n"
|
|
142
145
|
table = find_parent(node.parent, "table")
|
|
@@ -153,6 +156,7 @@ module Markitdown
|
|
|
153
156
|
node.children.each do |child|
|
|
154
157
|
contents = self.parse_node(child, states)
|
|
155
158
|
contents = contents.flatten.compact.join.strip if strip_content
|
|
159
|
+
contents = contents.flatten.compact.join.gsub("\n", " ") if flatten_content
|
|
156
160
|
results << contents
|
|
157
161
|
end
|
|
158
162
|
end
|
data/lib/markitdown/version.rb
CHANGED
data/spec/table2.html
CHANGED
|
@@ -10,8 +10,8 @@
|
|
|
10
10
|
|
|
11
11
|
<table>
|
|
12
12
|
<tr>
|
|
13
|
-
<th>This
|
|
14
|
-
<th>is</th>
|
|
13
|
+
<th>This <br/></th>
|
|
14
|
+
<th><p>is</p></th>
|
|
15
15
|
<th>a</th>
|
|
16
16
|
<th>table</th>
|
|
17
17
|
</tr>
|
|
@@ -42,12 +42,12 @@
|
|
|
42
42
|
<td>This</td>
|
|
43
43
|
<td>is</td>
|
|
44
44
|
<td>a</td>
|
|
45
|
-
<td>table
|
|
45
|
+
<td>table <br/></td>
|
|
46
46
|
</tr>
|
|
47
47
|
<tr>
|
|
48
48
|
<td>This</td>
|
|
49
|
-
<td>is</td>
|
|
50
|
-
<td>a
|
|
49
|
+
<td><p>is</p></td>
|
|
50
|
+
<td><p>a<br/></p><br/></td>
|
|
51
51
|
<td>table</td>
|
|
52
52
|
</tr>
|
|
53
53
|
</body>
|
data/spec/table2.markdown
CHANGED
|
@@ -8,12 +8,12 @@ This is only a test
|
|
|
8
8
|
|
|
9
9
|
This [announcement](http://www.google.com) would be followed by instructions.
|
|
10
10
|
|
|
11
|
-
|This|is|a|table|
|
|
11
|
+
|This | is |a|table|
|
|
12
12
|
|---|---|---|---|
|
|
13
13
|
|This|is|a|
|
|
14
14
|
|This|is|a| [announcement](http://www.google.com) |
|
|
15
15
|
|This|is|a| _table_ |
|
|
16
16
|
|This|is|a| **table** |
|
|
17
|
-
|This|is|a|table|
|
|
18
|
-
|This|is|a|table|
|
|
17
|
+
|This|is|a|table |
|
|
18
|
+
|This| is | a |table|
|
|
19
19
|
|
metadata
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: markitdown
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
version: 0.1.
|
|
4
|
+
version: 0.1.1
|
|
5
5
|
prerelease:
|
|
6
6
|
platform: ruby
|
|
7
7
|
authors:
|
|
@@ -103,7 +103,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
|
|
|
103
103
|
version: '0'
|
|
104
104
|
segments:
|
|
105
105
|
- 0
|
|
106
|
-
hash:
|
|
106
|
+
hash: -667411799113990729
|
|
107
107
|
required_rubygems_version: !ruby/object:Gem::Requirement
|
|
108
108
|
none: false
|
|
109
109
|
requirements:
|
|
@@ -112,7 +112,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
|
|
|
112
112
|
version: '0'
|
|
113
113
|
segments:
|
|
114
114
|
- 0
|
|
115
|
-
hash:
|
|
115
|
+
hash: -667411799113990729
|
|
116
116
|
requirements: []
|
|
117
117
|
rubyforge_project:
|
|
118
118
|
rubygems_version: 1.8.23
|