github-mirror/discourse
2
0
Fork 0
mirror of https://github.com/discourse/discourse.git synced 2025-05-21 18:12:32 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

Add 'keep_cid_imgs' option to HTML to Markdown converter to improve incoming email parsing

Browse Source
This commit is contained in:
Régis Hanol
2017-05-03 22:53:47 +02:00
parent 9d7917f79d
commit 768c63c103
2 changed files with 22 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
lib/html_to_markdown.rb
View File

@ -136,7 +136,7 @@ class HtmlToMarkdown
end
def visit_img(node)
if is_valid_url?(node["src"]) && is_visible_img?(node)
if is_valid_src?(node["src"]) && is_visible_img?(node)
if @opts[:keep_img_tags]
@stack[-1].markdown << node.to_html
else
@ -147,7 +147,7 @@ class HtmlToMarkdown
end
def visit_a(node)
if is_valid_url?(node["href"])
if is_valid_href?(node["href"])
@stack[-1].markdown << "["
traverse(node)
@stack[-1].markdown << "](#{node["href"]})"
@ -206,14 +206,20 @@ class HtmlToMarkdown
(lines + [""]).join("\n")
end
def is_valid_url?(url)
url.present? && (url.start_with?("http") || url.start_with?("www."))
def is_valid_href?(href)
href.present? && (href.start_with?("http") || href.start_with?("www."))
end
def is_valid_src?(src)
return false if src.blank?
return true if @opts[:keep_cid_imgs] && src.start_with?("cid:")
src.start_with?("http") || src.start_with?("www.")
end
def is_visible_img?(img)
return false if img["width"].present? && img["width"].to_i == 0
return false if img["width"].present? && img["width"].to_i == 0
return false if img["height"].present? && img["height"].to_i == 0
return false if img["style"].present? && img["style"][/(width|height)\s*:\s*0/]
return false if img["style"].present? && img["style"][/(width|height)\s*:\s*0/]
true
end