56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
|
# File 'lib/jekyll-import/importers/wordpressdotcom.rb', line 56
def self.process(options)
source = options.fetch('source', "wordpress.xml")
fetch = !options.fetch('no_fetch_images', false)
assets_folder = options.fetch('assets_folder', 'assets')
FileUtils.mkdir_p(assets_folder)
import_count = Hash.new(0)
doc = Hpricot::XML(File.read(source))
authors = Hash[
(doc/:channel/'wp:author').map do |author|
[author.at("wp:author_login").inner_text.strip, {
"login" => author.at("wp:author_login").inner_text.strip,
"email" => author.at("wp:author_email").inner_text,
"display_name" => author.at("wp:author_display_name").inner_text,
"first_name" => author.at("wp:author_first_name").inner_text,
"last_name" => author.at("wp:author_last_name").inner_text
}]
end
] rescue {}
(doc/:channel/:item).each do |item|
title = item.at(:title).inner_text.strip
permalink_title = item.at('wp:post_name').inner_text
if permalink_title == ""
permalink_title = sluggify(title)
end
date = Time.parse(item.at('wp:post_date').inner_text)
status = item.at('wp:status').inner_text
if status == "publish"
published = true
else
published = false
end
type = item.at('wp:post_type').inner_text
categories = item.search('category[@domain="category"]').map{|c| c.inner_text}.reject{|c| c == 'Uncategorized'}.uniq
tags = item.search('category[@domain="post_tag"]').map{|t| t.inner_text}.uniq
metas = Hash.new
item.search("wp:postmeta").each do |meta|
key = meta.at('wp:meta_key').inner_text
value = meta.at('wp:meta_value').inner_text
metas[key] = value
end
author_login = item.at('dc:creator').inner_text.strip
name = "#{date.strftime('%Y-%m-%d')}-#{permalink_title}.html"
= {
'layout' => type,
'title' => title,
'date' => date,
'categories' => categories,
'tags' => tags,
'status' => status,
'type' => type,
'published' => published,
'meta' => metas,
'author' => authors[author_login]
}
begin
content = Hpricot(item.at('content:encoded').inner_text)
if fetch
download_images(title, content, assets_folder)
end
FileUtils.mkdir_p "_#{type}s"
File.open("_#{type}s/#{name}", "w") do |f|
f.puts .to_yaml
f.puts '---'
f.puts Util.wpautop(content.to_html)
end
rescue => e
puts "Couldn't import post!"
puts "Title: #{title}"
puts "Name/Slug: #{name}\n"
puts "Error: #{e.message}"
next
end
import_count[type] += 1
end
import_count.each do |key, value|
puts "Imported #{value} #{key}s"
end
end
|