upmark 1.0.0 → 2.0.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: e6e70467bec74fa706947ab9b0aea3c6cf626c6cf201a07aa918b5fba980a1a2
-  data.tar.gz: a237b227e1c64e116dad035548ac4d364f9fca96956b26ad7b90f0b5919b29fc
+  metadata.gz: 30ee9a20ac459c48e9a74d6b2ce2df2328366421c2193b0b81b7d60504dbb171
+  data.tar.gz: 5cd1167cf96a88018f9a34cd1ea7b330b21c7732ffaf1525f110694f83f90edc
 SHA512:
-  metadata.gz: 8c132b5f1a0fdc78cd7c5a4e3c7b3a9bad1a7aacfc7ad3e59f9d2a88bdb03bdd610bd5a907bddf565e2183c0c70dc0558834b80e7fddac8ad9421fe245f2e99c
-  data.tar.gz: 7a4997aea6c74b5a2ee3adcf377e963663ca1ce6e4aec1cb48dc7d3066f0f8ec46710235885df300e6bd0fe15923825b130f0ab1de22bfc15d649bacbaae98b8
+  metadata.gz: d3e08e299cc8405a1c2833637b85ffd04aa078f7af031f8a52745fd742d316aae118efbfa0c32db8a6411ea94be684c488902a9864b9bc664eab0f2c39a84247
+  data.tar.gz: 858361e80fa96cd4001c317b3af32dfc36bf30d6ca70375c1fac81f24b3d96cd9c02f0ad1ee17f9059c4141a6e9b2916a587f456bb91004cc6e8c44c15e9d0c3

data/README.md

@@ -42,7 +42,7 @@ It will also pass through block and span-level HTML elements (e.g. `table`, `div
 
 ## How it works
 
-Upmark defines a parsing expression grammar (PEG) using the very awesome [Parslet](http://kschiess.github.com/parslet/) gem. This PEG is then used to convert HTML into Markdown in 4 steps:
+Upmark defines a parsing expression grammar (PEG) using the very awesome [Parslet](https://github.com/kschiess/parslet/) gem. This PEG is then used to convert HTML into Markdown in 4 steps:
 
 1. Parse the XHTML into an abstract syntax tree (AST).
 2. Normalize the AST into a nested hash of HTML elements.

data/lib/upmark/transform/markdown.rb

@@ -48,12 +48,12 @@ module Upmark
       element(:li) {|element| "#{text(element)}" }
 
       element(:ul) do |element|
-        children = element[:children].map {|value| value.strip != "" ? value : nil }.compact
+        children = element[:children].flatten.map {|value| value.strip != "" ? value : nil }.compact
         children.map {|value| "* #{value.gsub(/^\s*•\s*/,'')}\n" }
       end
 
       element(:ol) do |element|
-        children = element[:children].map {|value| value.strip != "" ? value : nil }.compact
+        children = element[:children].flatten.map {|value| value.strip != "" ? value : nil }.compact
         children.map.with_index {|value, i| "#{i + 1}. #{value}\n" }
       end
 
@@ -71,7 +71,7 @@ module Upmark
 
       element(:img) do |element|
         attributes = map_attributes_subtree(element[:attributes])
-        href       = attributes[:src]
+        href       = attributes[:src].to_s
         title      = attributes[:title]
         alt_text   = attributes[:alt]
 

data/lib/upmark.rb

@@ -14,7 +14,11 @@ module Upmark
     preprocess   = Transform::Preprocess.new
     markdown     = Transform::Markdown.new
 
-    ast = xml.parse(html.strip)
+    # Remove base64 data URLs that cause parser issues
+    html = html.gsub(/(data:image\/[^;]*;base64,)[A-Za-z0-9+\/=]+/, '').strip
+
+    ast = xml.parse(html)
+
     ast = normalise.apply(ast)
     ast = preprocess.apply(ast)
     ast = markdown.apply(ast)

data/spec/acceptance/upmark_spec.rb

@@ -86,6 +86,17 @@ Labor MP calls to end dogs
 ![messenger bag skateboard](http://helvetica.com/image.gif "art party organic")
       MD
     end
+
+    specify "removes base64 data URLs" do
+      expect(<<~HTML).to convert_to("")
+        <img src="data:image/png;base64,abc" />
+      HTML
+
+      src = "abc" * 10000
+      expect(<<~HTML).to convert_to("")
+        <img src="data:image/png;base64,#{src}" />
+      HTML
+    end
   end
 
   context "<p>" do
@@ -341,4 +352,36 @@ messenger **bag** skateboard
       expect(html).to convert_to("Hi\nthere")
     end
   end
-end
+
+  context "nested unordered lists" do
+  let(:html) do
+      <<-HTML
+      <ul>
+        <ul>
+          <li>List item</li>
+        </ul>
+      </ul>
+      HTML
+    end
+
+    it "generates readable output" do
+      expect(html).to convert_to("* * List item")
+    end
+  end
+
+  context "nested ordered lists" do
+    let(:html) do
+        <<-HTML
+        <ol>
+          <ol>
+            <li>List item</li>
+          </ol>
+        </ol>
+        HTML
+      end
+
+      it "generates readable output" do
+        expect(html).to convert_to("1. 1. List item")
+      end
+    end
+  end

metadata

@@ -1,16 +1,16 @@
 --- !ruby/object:Gem::Specification
 name: upmark
 version: !ruby/object:Gem::Version
-  version: 1.0.0
+  version: 2.0.0
 platform: ruby
 authors:
 - Josh Bassett
 - Gus Gollings
 - James Healy
-autorequire: 
+autorequire:
 bindir: bin
 cert_chain: []
-date: 2018-03-26 00:00:00.000000000 Z
+date: 2025-09-25 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rspec
@@ -95,7 +95,7 @@ files:
 homepage: http://github.com/conversation/upmark
 licenses: []
 metadata: {}
-post_install_message: 
+post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -110,14 +110,13 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubyforge_project: upmark
-rubygems_version: 2.7.0
-signing_key: 
+rubygems_version: 3.4.10
+signing_key:
 specification_version: 4
 summary: A HTML to Markdown converter.
 test_files:
+- spec/acceptance/upmark_spec.rb
+- spec/errors_spec.rb
 - spec/spec_helper.rb
 - spec/unit/lib/upmark/parser/xml_spec.rb
 - spec/unit/lib/upmark/transform/markdown_spec.rb
-- spec/acceptance/upmark_spec.rb
-- spec/errors_spec.rb