html2doc 1.4.2 → 1.4.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 834f2d69887ce656cd78e78822c78c98a19017e01ff1fbaffb5bcf9951fb1419
-  data.tar.gz: b7c12ae5d35ae84ac8de00d3cde42dc27549f5567623a5d88a50c3757a4c6dcd
+  metadata.gz: bb9828e9edbe4e0dabb525267ac5274793a2691d72eb837f46de3191c3159031
+  data.tar.gz: 5ad4f6a3856ef6ab784db79db8e4f69fdd2b6aeeb5b179e90f0b7d06c1c55eea
 SHA512:
-  metadata.gz: 02a2b5f53337616b673d43f20a3f90594bcf5355288235d4eafe2622cac8e6af01d24ea2b05a8d2fea0fb9256b4d43686f3c81cb075ea562f6fcb616791abfeb
-  data.tar.gz: 565a6554f2de5e99bbaa9cc5de81e3c7d9789e36d9251641518cf82f44c2183a36a9305d2054535396813b21e475a31af1ca9577c9cb7c4bd7a9dbc0b872cf6b
+  metadata.gz: 4d904e3cc7747cd0167c440b41d4123a6a46d00709f114149b4784bd9c6824361c26169569a44e0f3837c8d102c3eefea2fa256533fa6dbc492f6d31a13e5234
+  data.tar.gz: 167943e0582c2fcae1724e9d6c9f38fdefdb2641cd2e5ccd6dfc47ebea7b29a6748e353f58383be314390ea306895da781ab85f9a0ebb070672d17f84d30001b

data/.github/workflows/rake.yml

@@ -10,6 +10,6 @@ on:
 
 jobs:
   rake:
-    uses: metanorma/metanorma-build-scripts/.github/workflows/generic-rake.yml@main
+    uses: metanorma/ci/.github/workflows/generic-rake.yml@main
     secrets:
       pat_token: ${{ secrets.METANORMA_CI_PAT_TOKEN }}

data/.github/workflows/release.yml

@@ -0,0 +1,24 @@
+# Auto-generated by Cimas: Do not edit it manually!
+# See https://github.com/metanorma/cimas
+name: release
+
+on:
+  workflow_dispatch:
+    inputs:
+      next_version:
+        description: |
+          Next release version. Possible values: x.y.z, major, minor, patch or pre|rc|etc
+        required: true
+        default: 'skip'
+  push:
+    tags: [ v* ]
+
+jobs:
+  release:
+    uses: metanorma/ci/.github/workflows/rubygems-release.yml@main
+    with:
+      next_version: ${{ github.event.inputs.next_version }}
+    secrets:
+      rubygems-api-key: ${{ secrets.METANORMA_CI_RUBYGEMS_API_KEY }}
+      pat_token: ${{ secrets.METANORMA_CI_PAT_TOKEN }}
+

data/lib/html2doc/lists.rb

@@ -58,6 +58,7 @@ class Html2Doc
       l["class"] ||= "MsoListParagraphCxSpMiddle"
       next unless l.first_element_child&.name == "p"
 
+      l["style"] ||= ""
       l["style"] += (l.first_element_child["style"]&.sub(/mso-list[^;]+;/, "") || "")
       l.first_element_child.replace(l.first_element_child.children)
     end

data/lib/html2doc/mime.rb

@@ -67,22 +67,23 @@ class Html2Doc
   end
 
   def contentid(mhtml)
-    mhtml.gsub %r{(<img[^>]*?src=")([^\"']+)(['"])}m do |m|
+    mhtml.gsub %r{(<img[^>]*?src=")([^"']+)(['"])}m do |m|
       repl = "#{$1}cid:#{File.basename($2)}#{$3}"
-      /^data:|^https?:/.match($2) ? m : repl
-    end.gsub %r{(<v:imagedata[^>]*?src=")([^\"']+)(['"])}m do |m|
+      /^data:|^https?:/ =~ $2 ? m : repl
+    end.gsub %r{(<v:imagedata[^>]*?src=")([^"']+)(['"])}m do |m|
       repl = "#{$1}cid:#{File.basename($2)}#{$3}"
-      /^data:|^https?:/.match($2) ? m : repl
+      /^data:|^https?:/ =~ $2 ? m : repl
     end
   end
 
   # max width for Word document is 400, max height is 680
   def image_resize(img, path, maxheight, maxwidth)
-    realsize = ImageSize.path(path).size
-    s = [img["width"].to_i, img["height"].to_i]
-    s = realsize if s[0].zero? && s[1].zero?
-    return [nil, nil] if realsize.nil? || realsize[0].nil? || realsize[1].nil?
+    s, realsize = get_image_size(img, path)
+    return s if s[0] == nil && s[1] == nil
 
+    if img.name == "svg" && !img["viewBox"]
+      img["viewBox"] = "0 0 #{s[0]} #{s[1]}"
+    end
     s[1] = s[0] * realsize[1] / realsize[0] if s[1].zero? && !s[0].zero?
     s[0] = s[1] * realsize[0] / realsize[1] if s[0].zero? && !s[1].zero?
     s = [(s[0] * maxheight / s[1]).ceil, maxheight] if s[1] > maxheight
@@ -90,6 +91,16 @@ class Html2Doc
     s
   end
 
+  def get_image_size(img, path)
+    realsize = ImageSize.path(path).size # does not support emf
+    s = [img["width"].to_i, img["height"].to_i]
+    return [s, s] if /\.emf$/.match?(path)
+
+    s = realsize if s[0].zero? && s[1].zero?
+    s = [nil, nil]  if realsize.nil? || realsize[0].nil? || realsize[1].nil?
+    [s, realsize]
+  end
+
   IMAGE_PATH = "//*[local-name() = 'img' or local-name() = 'imagedata']".freeze
 
   def mkuuid

data/lib/html2doc/version.rb

@@ -1,3 +1,3 @@
 class Html2Doc
-  VERSION = "1.4.2".freeze
+  VERSION = "1.4.3".freeze
 end

data/spec/html2doc_spec.rb

@@ -626,7 +626,7 @@ RSpec.describe Html2Doc do
   end
 
   it "resizes images with missing or auto sizes" do
-    image = { "src" => "spec/19160-8.jpg" }
+    image = Nokogiri::XML("<img src='spec/19160-8.jpg'/>").root
     expect(Html2Doc.new({}).image_resize(image, "spec/19160-8.jpg", 100, 100))
       .to eq [30, 100]
     image["width"] = "20"
@@ -666,6 +666,12 @@ RSpec.describe Html2Doc do
       .to eq [30, 100]
   end
 
+  it "resizes SVG with missing or auto sizes" do
+    image = Nokogiri::XML(File.read("spec/odf.svg")).root
+    Html2Doc.new({}).image_resize(image, "spec/odf.svg", 100, 100)
+    expect(image.to_xml).to match_fuzzy '<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 100 100"/>'
+  end
+
   it "does not move images if they are external URLs" do
     simple_body = '<img src="https://example.com/19160-6.png">'
     Html2Doc.new(filename: "test", imagedir: ".")

data/spec/odf.svg

@@ -0,0 +1 @@
+<svg xmlns="http://www.w3.org/2000/svg" width=100 height=100><circle fill="#009" r="45" cx="50" cy="50"/><path d="M33,26H78A37,37,0,0,1,33,83V57H59V43H33Z" fill="#FFF"/></svg>

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: html2doc
 version: !ruby/object:Gem::Version
-  version: 1.4.2
+  version: 1.4.3
 platform: ruby
 authors:
 - Ribose Inc.
-autorequire:
+autorequire: 
 bindir: bin
 cert_chain: []
-date: 2022-07-04 00:00:00.000000000 Z
+date: 2022-10-21 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: asciimath
@@ -274,6 +274,7 @@ extra_rdoc_files: []
 files:
 - ".gitattributes"
 - ".github/workflows/rake.yml"
+- ".github/workflows/release.yml"
 - ".gitignore"
 - ".hound.yml"
 - ".oss-guides.rubocop.yml"
@@ -313,13 +314,14 @@ files:
 - spec/header.html
 - spec/header_img.html
 - spec/html2doc_spec.rb
+- spec/odf.svg
 - spec/spec_helper.rb
 homepage: https://github.com/metanorma/html2doc
 licenses:
 - CC-BY-SA-3.0
 - BSD-2-Clause
 metadata: {}
-post_install_message:
+post_install_message: 
 rdoc_options: []
 require_paths:
 - lib
@@ -334,8 +336,8 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.3.16
-signing_key:
+rubygems_version: 3.1.6
+signing_key: 
 specification_version: 4
 summary: Convert HTML document to Microsoft Word document
 test_files: []