jaculabilis/danbooru
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

hentai foundry: fix 'Document tree depth exceeded' when parsing commentaries.

Browse Source 
Fix a regression in 38c9559fe that caused #4657 to fail again.
This commit is contained in:
evazion
2021-09-01 01:36:33 -05:00
parent d00aa847ae
commit 19c0027d1f
1 changed files with 12 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
app/logical/d_text.rb
View File

@@ -49,7 +49,7 @@ class DText
# @param artists [Array<Artist>] # @param artists [Array<Artist>]
# @return [String] the HTML output # @return [String] the HTML output
def self.postprocess(html, wiki_pages, tags, artists) def self.postprocess(html, wiki_pages, tags, artists)
fragment = Nokogiri::HTML5.fragment(html) fragment = parse_html(html)
fragment.css("a.dtext-wiki-link").each do |node| fragment.css("a.dtext-wiki-link").each do |node|
path = Addressable::URI.parse(node["href"]).path path = Addressable::URI.parse(node["href"]).path
@@ -174,7 +174,7 @@ class DText
# @return [Array<String>] the list of wiki page names # @return [Array<String>] the list of wiki page names
def self.parse_wiki_titles(text) def self.parse_wiki_titles(text)
html = DTextRagel.parse(text) html = DTextRagel.parse(text)
fragment = Nokogiri::HTML5.fragment(html) fragment = parse_html(html)
titles = fragment.css("a.dtext-wiki-link").map do |node| titles = fragment.css("a.dtext-wiki-link").map do |node|
title = node["href"][%r{\A/wiki_pages/(.*)\z}i, 1] title = node["href"][%r{\A/wiki_pages/(.*)\z}i, 1]
@@ -191,7 +191,7 @@ class DText
# @return [Array<String>] the list of external URLs # @return [Array<String>] the list of external URLs
def self.parse_external_links(text) def self.parse_external_links(text)
html = DTextRagel.parse(text) html = DTextRagel.parse(text)
fragment = Nokogiri::HTML5.fragment(html) fragment = parse_html(html)
links = fragment.css("a.dtext-external-link").map { |node| node["href"] } links = fragment.css("a.dtext-external-link").map { |node| node["href"] }
links.uniq links.uniq
@@ -326,7 +326,7 @@ class DText
# @param html [String] the HTML input # @param html [String] the HTML input
# @return [String] the Markdown output # @return [String] the Markdown output
def self.html_to_markdown(html) def self.html_to_markdown(html)
html = Nokogiri::HTML5.fragment(html) html = parse_html(html)
html.children.map do |node| html.children.map do |node|
case node.name case node.name
@@ -349,7 +349,7 @@ class DText
# @param inline [Boolean] if true, convert <img> tags to plaintext # @param inline [Boolean] if true, convert <img> tags to plaintext
# @return [String] the DText output # @return [String] the DText output
def self.from_html(text, inline: false, &block) def self.from_html(text, inline: false, &block)
html = Nokogiri::HTML5.fragment(text) html = parse_html(text)
dtext = html.children.map do |element| dtext = html.children.map do |element|
block.call(element) if block.present? block.call(element) if block.present?
@@ -427,4 +427,11 @@ class DText
def self.excerpt(text, length: 160) def self.excerpt(text, length: 160)
strip_dtext(text).split(/\r\n|\r|\n/).first.to_s.truncate(length) strip_dtext(text).split(/\r\n|\r|\n/).first.to_s.truncate(length)
end end
# Parse a string of HTML to a document object.
# @param html [String]
# @return [Nokogiri::HTML5::DocumentFragment]
def self.parse_html(html)
Nokogiri::HTML5.fragment(html, max_tree_depth: -1)
end
end end