jekyll-artisync 0.2 → 0.8

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: bdd1c95dc86f0e265d5e748dd9559cff17968ad55570e7d610ec2441a7339382
-  data.tar.gz: 184d76a6fe6174d9c0888dbe8126b2bf16a1e6306f28dcd836b347dd2698527a
+  metadata.gz: 01afb17eccd2097c95f4071c0dce0074ec1156b70e455b78ec8c0c961e6aaf3e
+  data.tar.gz: 72720da82e6a39b40b0682f427c4d75397559cd7ecc23e802e9088675db60616
 SHA512:
-  metadata.gz: 6eead02cfc58bfb5bb24d3860e222a6e04ba3883a9a40a8a9c8e673efc94b29b22f851631e103dbc93ca1ff0a9e43ccefeca1eec1bbf0839c395baa0d2c86a94
-  data.tar.gz: 7c34719be25ff7b9ab5501c2a2ef5512d71678a4c3ea918ff7b6d53ae8f874adc7683790d8468e1afaf177e167584ae7fe111d8d35f9ea1f432665aa1244cb97
+  metadata.gz: fd53ace05990ec40ebe945eebc5981cbddf8c8270417f7885297e8a827dada3a1fd8fb7c7f74dae13efa5a2a4beb0b704a795c752041df16da816e3c249bbd0e
+  data.tar.gz: e9df8042e29b3301787bb4247e826f0b8b6ae8d247a61e31e2a9978383b0e5f39a89d66981ea558140fe2e166f03721f19faf99757c9604d43544fae9d16899b

data/lib/jekyll-artisync.rb

@@ -1,14 +1,6 @@
-require 'net/http'
-require 'nokogiri'
-
+require 'syncers/syncer_factory'
 require "jekyll"
 
-# user agent is necessary otherwise certain sites such as Zhihu throws 400
-USER_AGENT = 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.86 Safari/537.36'
-
-SITE_TO_ARTICLE_XPATH = {
-  'zhihu' => '//div[contains(@class, "Post-RichText") and contains(@class, "ztext")]',
-}
 
 class ArticleSyncEmbed < Liquid::Tag
 
@@ -17,40 +9,11 @@ class ArticleSyncEmbed < Liquid::Tag
     @content = content
   end
 
-  def _fetch_html(url)
-    uri = URI(url)
-    res = Net::HTTP.start(uri.hostname, uri.port, :use_ssl => true) do |http|
-      # :use_ssl => true for the uri is https
-      http.request(Net::HTTP::Get.new(uri, {'User-Agent' => USER_AGENT}))
-    end
-
-    res.body
-  end
-
-  def _handle_node(node)
-    case node.name
-    when 'figure'
-      img_node = node.children[1]
-
-      img_url = img_node['data-actualsrc']
-      img_url['_b.jpg'] = '_720w.jpg'
-      img_url['/v2'] = "/80/v2"
-
-      img_node['src'] = img_url
-    end
-    node.to_html
-  end
-
   def render(context)
-    url, site = @content.strip.split
-    page_html = self._fetch_html(url)
-    article = Nokogiri::HTML(page_html).xpath(SITE_TO_ARTICLE_XPATH[site])
-    content = []
-    article.children.each do |node|
-      content.append self._handle_node(node)
-    end
-
-    content.join("\n")
+    url = Liquid::Template.parse(@content).render(context).strip
+    uri = URI(url)
+    syncer = SyncerFactory.get_syncer(uri)
+    return syncer.gen_html
   end
 
   Liquid::Template.register_tag "artisync", self

data/lib/syncers/abstract_syncer.rb

@@ -0,0 +1,35 @@
+require 'net/http'
+
+
+# user agent is necessary otherwise certain sites such as Zhihu throws 400
+USER_AGENT = 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.86 Safari/537.36'
+
+MESSAGE = "AbstractSyncer::[%s]: override required."
+
+class AbstractSyncer
+  def initialize(uri)
+    @uri = uri
+  end
+
+  def _fetch_html
+    uri = @uri
+    res = Net::HTTP.start(uri.hostname, uri.port, :use_ssl => true) do |http|
+      # :use_ssl => true for the uri is https
+      http.request(Net::HTTP::Get.new(uri, {'User-Agent' => USER_AGENT}))
+    end
+
+    res.body
+  end
+
+  def get_article_xpath
+    raise MESSAGE % [__method__]
+  end
+
+  def get_article_nodes
+    raise MESSAGE % [__method__]
+  end
+
+  def gen_html
+    raise MESSAGE % [__method__]
+  end
+end

data/lib/syncers/csdn_syncer.rb

@@ -0,0 +1,14 @@
+require_relative 'per_node_syncer'
+require 'nokogiri'
+
+
+class CSDNSyncer < PerNodeSyncer
+  def get_article_xpath
+    '//div[contains(@id, "content_views")]'
+  end
+
+  def _handle_node(node)
+    node.to_html
+  end
+
+end

data/lib/syncers/medium_syncer.rb

@@ -0,0 +1,31 @@
+require_relative 'per_node_syncer'
+require 'nokogiri'
+
+
+class MediumSyncer < PerNodeSyncer
+  def get_article_xpath
+    '//article/div/section/div/div'
+  end
+
+  # override
+  def get_article_nodes
+    page_html = self._fetch_html
+    article = Nokogiri::HTML(page_html).xpath(self.get_article_xpath)
+
+    # Medium embeds Author section, which is not needed for article
+    article.children[1..]
+  end
+
+  def _handle_node(node)
+    case node.name
+    when "figure"
+      img_nodes = node.css('img')
+      node = img_nodes[-1] if img_nodes
+      # to make sure image scales right
+      node.remove_attribute('width')
+      node.remove_attribute('height')
+    end
+    node.to_html
+  end
+
+end

data/lib/syncers/per_node_syncer.rb

@@ -0,0 +1,34 @@
+require_relative 'abstract_syncer'
+require 'nokogiri'
+
+module NodeAttrModule
+  ATTRS = ['class', 'id']
+  def NodeAttrModule.remove_common_attr(node)
+    ATTRS.each do |attr|
+      node.remove_attribute(attr) if node[attr]
+    end
+  end
+end
+
+
+class PerNodeSyncer < AbstractSyncer
+  include NodeAttrModule
+
+  def get_article_nodes
+    page_html = self._fetch_html
+    article = Nokogiri::HTML(page_html).xpath(self.get_article_xpath)
+    article.children
+  end
+
+  def gen_html
+    article_nodes = self.get_article_nodes
+    content = []
+    article_nodes.each do |node|
+      NodeAttrModule.remove_common_attr(node)
+      content.append self._handle_node(node)
+    end
+
+    content.join("\n")
+  end
+
+end

data/lib/syncers/syncer_factory.rb

@@ -0,0 +1,30 @@
+require_relative 'zhihu_syncer'
+require_relative 'weixin_syncer'
+require_relative 'medium_syncer'
+require_relative 'csdn_syncer'
+
+
+ZHIHU_ZHUANLAN = 'zhuanlan.zhihu.com'
+WEIXIN = 'mp.weixin.qq.com'
+MEDIUM = 'medium.com'
+CSDN = 'blog.csdn.net'
+
+class SyncerFactory
+  def self.get_syncer(uri)
+    host_name = uri.host
+
+    case host_name
+    when ZHIHU_ZHUANLAN
+      return ZhihuSyncer.new(uri)
+    when WEIXIN
+      return WeixinSyncer.new(uri)
+    when MEDIUM
+      return MediumSyncer.new(uri)
+    when CSDN
+      return CSDNSyncer.new(uri)
+    else
+      raise "Not supported website for host: #{host_name}"
+    end
+  end
+end
+

data/lib/syncers/weixin_syncer.rb

@@ -0,0 +1,21 @@
+require_relative 'per_node_syncer'
+require 'nokogiri'
+
+
+class WeixinSyncer < PerNodeSyncer
+  def get_article_xpath
+    '//div[contains(@class, "rich_media_content") and contains(@id, "js_content")]'
+  end
+
+  def _handle_node(node)
+    node.children.each do |child_node|
+      case child_node.name
+      when 'img'
+        child_node['src'] = child_node['data-src']
+        child_node['data-src'] = nil
+      end
+    end
+    node.to_html
+  end
+
+end

data/lib/syncers/zhihu_syncer.rb

@@ -0,0 +1,23 @@
+require_relative 'per_node_syncer'
+require 'nokogiri'
+
+
+class ZhihuSyncer < PerNodeSyncer
+  def get_article_xpath
+    '//div[contains(@class, "Post-RichText") and contains(@class, "ztext")]'
+  end
+
+  def _handle_node(node)
+    case node.name
+    when 'figure'
+      img_node = node.css('img')[-1]
+      if img_node
+        node = img_node
+        node['src'] = node['data-actualsrc']
+      end
+    end
+
+    node.to_html
+  end
+
+end

metadata

@@ -1,15 +1,29 @@
 --- !ruby/object:Gem::Specification
 name: jekyll-artisync
 version: !ruby/object:Gem::Version
-  version: '0.2'
+  version: '0.8'
 platform: ruby
 authors:
-- Junhan
+- Junhan Zhu
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2020-05-24 00:00:00.000000000 Z
-dependencies: []
+date: 2020-06-06 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: nokogiri
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.10'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.10'
 description: Take an article from a given site and display on current Jekyll page.
 email:
 - junhanoct@gmail.com
@@ -18,6 +32,13 @@ extensions: []
 extra_rdoc_files: []
 files:
 - lib/jekyll-artisync.rb
+- lib/syncers/abstract_syncer.rb
+- lib/syncers/csdn_syncer.rb
+- lib/syncers/medium_syncer.rb
+- lib/syncers/per_node_syncer.rb
+- lib/syncers/syncer_factory.rb
+- lib/syncers/weixin_syncer.rb
+- lib/syncers/zhihu_syncer.rb
 homepage: https://github.com/junhan-z/jekyll-artisync
 licenses:
 - MIT