Add <base> support

Thibaut Courouble committed 9 years ago

Commit 4e41ed91 parent 6f0214e6 files · +99 −3Browse files

Unified · Split

6 files

lib/docs/core/filter.rb +3 −1
lib/docs/core/scraper.rb +1 −1
lib/docs/core/scrapers/url_scraper.rb +1 −1
lib/docs/filters/core/apply_base_url.rb +21 −0
test/lib/docs/core/filter_test.rb +4 −0
test/lib/docs/filters/core/apply_base_url_test.rb +69 −0

@@ -1,3 +1,5 @@
	1	+# frozen_string_literal: true
	2	+
1	3	module Docs
2	4	class Filter < ::HTML::Pipeline::Filter
3	5	def css(*args)
@@ -73,7 +75,7 @@ module Docs · 69 unmodified lines
73	75	end
74	76
75	77	def relative_url_string?(str)
76		− !fragment_url_string?(str) && str !~ SCHEME_RGX
	78	+ str !~ SCHEME_RGX && !fragment_url_string?(str) && !data_url_string?(str)
77	79	end
78	80
79	81	def absolute_url_string?(str)

lib/docs/core/scraper.rb+1−1

@@ -41,7 +41,7 @@ module Docs · 40 unmodified lines
41	41	self.html_filters = FilterStack.new
42	42	self.text_filters = FilterStack.new
43	43
44		− html_filters.push 'container', 'clean_html', 'normalize_urls', 'internal_urls', 'normalize_paths'
	44	+ html_filters.push 'apply_base_url', 'container', 'clean_html', 'normalize_urls', 'internal_urls', 'normalize_paths'
45	45	text_filters.push 'inner_html', 'clean_text', 'attribution'
46	46
47	47	def initialize

lib/docs/core/scrapers/url_scraper.rb+1−1

@@ -129,7 +129,7 @@ module Docs · 128 unmodified lines
129	129
130	130	def fetch_redirections
131	131	result = {}
132		− with_filters 'container', 'normalize_urls', 'internal_urls' do
	132	+ with_filters 'apply_base_url', 'container', 'normalize_urls', 'internal_urls' do
133	133	build_pages do \|page\|
134	134	next if page[:response_effective_path] == page[:response_path]
135	135	result[page[:response_path].downcase] = page[:response_effective_path]

lib/docs/filters/core/apply_base_url.rb+21−0

@@ -0,0 +1,21 @@
	1	+module Docs
	2	+ class ApplyBaseUrlFilter < Filter
	3	+ URL_ATTRIBUTES = { 'a': 'href', 'img': 'src', 'iframe': 'src' }
	4	+ SCHEME_RGX = /\A[^:\/?#]+:/
	5	+
	6	+ def call
	7	+ base_url = at_css('base').try(:[], 'href')
	8	+ return doc unless base_url
	9	+
	10	+ URL_ATTRIBUTES.each_pair do \|tag, attribute\|
	11	+ css(tag).each do \|node\|
	12	+ next unless value = node[attribute]
	13	+ next if !relative_url_string?(value) \|\| value[0] == '/'.freeze
	14	+ node[attribute] = "#{base_url}#{node[attribute]}"
	15	+ end
	16	+ end
	17	+
	18	+ doc
	19	+ end
	20	+ end
	21	+end

test/lib/docs/core/filter_test.rb+4−0

@@ -149,6 +149,10 @@ class DocsFilterTest < MiniTest::Spec · 148 unmodified lines
149	149	it "returns false with 'mailto:test@example.com'" do
150	150	refute filter.relative_url_string?('mailto:test@example.com')
151	151	end
	152	+
	153	+ it "returns false with 'data:image/gif;base64,foo'" do
	154	+ refute filter.relative_url_string?('data:image/gif;base64,foo')
	155	+ end
152	156	end
153	157
154	158	describe "#absolute_url_string?" do

test/lib/docs/filters/core/apply_base_url_test.rb+69−0

@@ -0,0 +1,69 @@
	1	+require 'test_helper'
	2	+require 'docs'
	3	+
	4	+class ApplyBaseUrlFilterTest < MiniTest::Spec
	5	+ include FilterTestHelper
	6	+ self.filter_class = Docs::ApplyBaseUrlFilter
	7	+ self.filter_type = 'html'
	8	+
	9	+ context "when there is no <base>" do
	10	+ it "does nothing" do
	11	+ @body = make_body nil, link_to('test')
	12	+ assert_equal link_to('test'), filter_output.at_css('body').inner_html
	13	+ end
	14	+ end
	15	+
	16	+ context "when <base> is '/base/'" do
	17	+ it "rewrites relative urls" do
	18	+ @body = make_body '/base/', link_to('path#frag')
	19	+ assert_equal link_to('/base/path#frag'), filter_output.at_css('body').inner_html
	20	+ end
	21	+
	22	+ it "rewrites relative image urls" do
	23	+ @body = make_body '/base/', '<img src="../img.png">'
	24	+ assert_equal '<img src="/base/../img.png">', filter_output.at_css('body').inner_html
	25	+ end
	26	+
	27	+ it "rewrites relative iframe urls" do
	28	+ @body = make_body '/base/', '<iframe src="./test"></iframe>'
	29	+ assert_equal '<iframe src="/base/./test"></iframe>', filter_output.at_css('body').inner_html
	30	+ end
	31	+
	32	+ it "doesn't rewrite absolute urls" do
	33	+ @body = make_body '/base/', link_to('http://example.com')
	34	+ assert_equal link_to('http://example.com'), filter_output.at_css('body').inner_html
	35	+ end
	36	+
	37	+ it "doesn't rewrite protocol-less urls" do
	38	+ @body = make_body '/base/', link_to('//example.com')
	39	+ assert_equal link_to('//example.com'), filter_output.at_css('body').inner_html
	40	+ end
	41	+
	42	+ it "doesn't rewrite root-relative urls" do
	43	+ @body = make_body '/base/', link_to('/path')
	44	+ assert_equal link_to('/path'), filter_output.at_css('body').inner_html
	45	+ end
	46	+
	47	+ it "doesn't rewrite fragment-only urls" do
	48	+ @body = make_body '/base/', link_to('#test')
	49	+ assert_equal link_to('#test'), filter_output.at_css('body').inner_html
	50	+ end
	51	+
	52	+ it "doesn't rewrite email urls" do
	53	+ @body = make_body '/base/', link_to('mailto:test@example.com')
	54	+ assert_equal link_to('mailto:test@example.com'), filter_output.at_css('body').inner_html
	55	+ end
	56	+
	57	+ it "doesn't rewrite data urls" do
	58	+ @body = make_body '/base/', '<img src="data:image/gif;base64,aaaa">'
	59	+ assert_equal '<img src="data:image/gif;base64,aaaa">', filter_output.at_css('body').inner_html
	60	+ end
	61	+ end
	62	+
	63	+ private
	64	+
	65	+ def make_body(base, body)
	66	+ base = %(<base href="#{base}">) if base
	67	+ "<html><meta charset=utf-8><title></title>#{base}#{body}</html>"
	68	+ end
	69	+end