From b3e65978b4a60bede5d5c9dc68daa1e5b5888a7a Mon Sep 17 00:00:00 2001
From: Darius Kazemi <darius.kazemi@gmail.com>
Date: Tue, 30 Apr 2019 15:29:28 -0700
Subject: [PATCH] Support locally cached inline images in Articles

---
 app/lib/activitypub/activity.rb        |  4 +-
 app/lib/activitypub/activity/create.rb | 53 ++++++++++++++++++++++++++
 app/lib/formatter.rb                   |  5 +++
 app/lib/sanitize_config.rb             |  3 +-
 4 files changed, 62 insertions(+), 3 deletions(-)

diff --git a/app/lib/activitypub/activity.rb b/app/lib/activitypub/activity.rb
index 66b5763a9..9b4a88382 100644
--- a/app/lib/activitypub/activity.rb
+++ b/app/lib/activitypub/activity.rb
@@ -4,8 +4,8 @@ class ActivityPub::Activity
   include JsonLdHelper
   include Redisable
 
-  SUPPORTED_TYPES = %w(Note Question).freeze
-  CONVERTED_TYPES = %w(Image Video Article Page).freeze
+  SUPPORTED_TYPES = %w(Note Question Article).freeze
+  CONVERTED_TYPES = %w(Image Video Page).freeze
 
   def initialize(json, account, **options)
     @json    = json
diff --git a/app/lib/activitypub/activity/create.rb b/app/lib/activitypub/activity/create.rb
index 00f0dd42d..1e47de2d9 100644
--- a/app/lib/activitypub/activity/create.rb
+++ b/app/lib/activitypub/activity/create.rb
@@ -30,6 +30,7 @@ class ActivityPub::Activity::Create < ActivityPub::Activity
     @mentions = []
     @params   = {}
 
+    process_inline_images if @object['content'].present? && @object['type'] == 'Article'
     process_status_params
     process_tags
     process_audience
@@ -73,6 +74,56 @@ class ActivityPub::Activity::Create < ActivityPub::Activity
     end
   end
 
+  class Handler < ::Ox::Sax
+    attr_reader :srcs
+    attr_reader :alts
+    def initialize(block)
+      @stack = []
+      @srcs = []
+      @alts = {}
+    end
+
+    def start_element(element_name)
+      @stack << [element_name, {}]
+    end
+
+    def end_element(element_name)
+      self_name, self_attributes = @stack[-1]
+      if self_name == :img && !self_attributes[:src].nil?
+        @srcs << self_attributes[:src]
+        @alts[self_attributes[:src]] = self_attributes[:alt]
+      end
+      @stack.pop
+    end
+
+    def attr(attribute_name, attribute_value)
+      _name, attributes = @stack.last
+      attributes[attribute_name] = attribute_value
+    end
+  end
+
+  def process_inline_images
+    proc = Proc.new { |name| puts name }
+    handler = Handler.new(proc)
+    Ox.sax_parse(handler, @object['content'])
+    handler.srcs.each do |src|
+      if skip_download?
+        @object['content'].gsub!(src, '')
+        next
+      end
+
+      media_attachment = MediaAttachment.create(account: @account, remote_url: src, description: handler.alts[src], focus: nil)
+      media_attachment.file_remote_url = src
+      media_attachment.save
+      if unsupported_media_type?(media_attachment.file.content_type)
+        @object['content'].gsub!(src, '')
+        media_attachment.delete
+      else
+        @object['content'].gsub!(src, media_attachment.file.url(:small))
+      end
+    end
+  end
+
   def process_audience
     (as_array(@object['to']) + as_array(@object['cc'])).uniq.each do |audience|
       next if audience == ActivityPub::TagManager::COLLECTIONS[:public]
@@ -310,6 +361,8 @@ class ActivityPub::Activity::Create < ActivityPub::Activity
   def text_from_content
     return Formatter.instance.linkify([[text_from_name, text_from_summary.presence].compact.join("\n\n"), object_url || @object['id']].join(' ')) if converted_object_type?
 
+    return Formatter.instance.format_article(@object['content']) if @object['content'].present? && @object['type'] == 'Article'
+
     if @object['content'].present?
       @object['content']
     elsif content_language_map?
diff --git a/app/lib/formatter.rb b/app/lib/formatter.rb
index 8a1aad41a..9c2c0cd29 100644
--- a/app/lib/formatter.rb
+++ b/app/lib/formatter.rb
@@ -90,6 +90,11 @@ class Formatter
     html.html_safe # rubocop:disable Rails/OutputSafety
   end
 
+  def format_article(text)
+    text = text.gsub(/>\n+</, "><")
+    text.html_safe # rubocop:disable Rails/OutputSafety
+  end
+
   def linkify(text)
     html = encode_and_link_urls(text)
     html = simple_format(html, {}, sanitize: false)
diff --git a/app/lib/sanitize_config.rb b/app/lib/sanitize_config.rb
index fa2d0784b..c157fc16a 100644
--- a/app/lib/sanitize_config.rb
+++ b/app/lib/sanitize_config.rb
@@ -34,13 +34,14 @@ class Sanitize
     end
 
     MASTODON_STRICT ||= freeze_config(
-      elements: %w(p br span a abbr del pre blockquote code b strong i em h1 h2 h3 h4 h5 ul ol li),
+      elements: %w(p br span a abbr del pre blockquote code b strong i em h1 h2 h3 h4 h5 ul ol li img),
 
       attributes: {
         'a'          => %w(href rel class title),
         'span'       => %w(class),
         'abbr'       => %w(title),
         'blockquote' => %w(cite),
+        'img'        => %w(src alt),
       },
 
       add_attributes: {