mirror of
https://github.com/gohugoio/hugo.git
synced 2024-07-04 16:21:57 +00:00
98ee69bce2
This commit adds the interface ContentReWriter in the tranformer chain. This is backed by two pooled byte buffers, alternating between being the reader or the writer. This keeps the performance characteristic of the old implementation, but in a thread safe way. Fixes #911 Benchmark old vs new: benchmark old ns/op new ns/op delta BenchmarkAbsURL 17614 17384 -1.31% BenchmarkXMLAbsURL 9431 9248 -1.94% benchmark old allocs new allocs delta BenchmarkAbsURL 24 28 +16.67% BenchmarkXMLAbsURL 12 14 +16.67% benchmark old bytes new bytes delta BenchmarkAbsURL 3295 3424 +3.92% BenchmarkXMLAbsURL 1954 1987 +1.69%
140 lines
6.8 KiB
Go
140 lines
6.8 KiB
Go
package transform
|
||
|
||
import (
|
||
"bytes"
|
||
"github.com/spf13/hugo/helpers"
|
||
"strings"
|
||
"testing"
|
||
)
|
||
|
||
const H5_JS_CONTENT_DOUBLE_QUOTE = "<!DOCTYPE html><html><head><script src=\"foobar.js\"></script><script src=\"/barfoo.js\"></script></head><body><nav><h1>title</h1></nav><article>content <a href=\"foobar\">foobar</a>. <a href=\"/foobar\">Follow up</a></article></body></html>"
|
||
const H5_JS_CONTENT_SINGLE_QUOTE = "<!DOCTYPE html><html><head><script src='foobar.js'></script><script src='/barfoo.js'></script></head><body><nav><h1>title</h1></nav><article>content <a href='foobar'>foobar</a>. <a href='/foobar'>Follow up</a></article></body></html>"
|
||
const H5_JS_CONTENT_ABS_URL = "<!DOCTYPE html><html><head><script src=\"http://user@host:10234/foobar.js\"></script></head><body><nav><h1>title</h1></nav><article>content <a href=\"https://host/foobar\">foobar</a>. Follow up</article></body></html>"
|
||
const H5_JS_CONTENT_ABS_URL_SCHEMALESS = "<!DOCTYPE html><html><head><script src=\"//host/foobar.js\"></script><script src='//host2/barfoo.js'></head><body><nav><h1>title</h1></nav><article>content <a href=\"//host/foobar\">foobar</a>. <a href='//host2/foobar'>Follow up</a></article></body></html>"
|
||
const CORRECT_OUTPUT_SRC_HREF_DQ = "<!DOCTYPE html><html><head><script src=\"foobar.js\"></script><script src=\"http://base/barfoo.js\"></script></head><body><nav><h1>title</h1></nav><article>content <a href=\"foobar\">foobar</a>. <a href=\"http://base/foobar\">Follow up</a></article></body></html>"
|
||
const CORRECT_OUTPUT_SRC_HREF_SQ = "<!DOCTYPE html><html><head><script src='foobar.js'></script><script src='http://base/barfoo.js'></script></head><body><nav><h1>title</h1></nav><article>content <a href='foobar'>foobar</a>. <a href='http://base/foobar'>Follow up</a></article></body></html>"
|
||
|
||
const H5_XML_CONTENT_ABS_URL = "<?xml version=\"1.0\" encoding=\"utf-8\" standalone=\"yes\" ?><feed xmlns=\"http://www.w3.org/2005/Atom\"><entry><content type=\"html\"><p><a href="/foobar">foobar</a></p> <p>A video: <iframe src='/foo'></iframe></p></content></entry></feed>"
|
||
const CORRECT_OUTPUT_SRC_HREF_IN_XML = "<?xml version=\"1.0\" encoding=\"utf-8\" standalone=\"yes\" ?><feed xmlns=\"http://www.w3.org/2005/Atom\"><entry><content type=\"html\"><p><a href="http://base/foobar">foobar</a></p> <p>A video: <iframe src='http://base/foo'></iframe></p></content></entry></feed>"
|
||
const H5_XML_CONTENT_GUARDED = "<?xml version=\"1.0\" encoding=\"utf-8\" standalone=\"yes\" ?><feed xmlns=\"http://www.w3.org/2005/Atom\"><entry><content type=\"html\"><p><a href="//foobar">foobar</a></p> <p>A video: <iframe src='//foo'></iframe></p></content></entry></feed>"
|
||
|
||
// additional sanity tests for replacements testing
|
||
const REPLACE_1 = "No replacements."
|
||
const REPLACE_2 = "ᚠᛇᚻ ᛒᛦᚦ ᚠᚱᚩᚠᚢᚱ\nᚠᛁᚱᚪ ᚷᛖᚻᚹᛦᛚᚳᚢᛗ"
|
||
|
||
// Issue: 816, schemaless links combined with others
|
||
const REPLACE_SCHEMALESS_HTML = `Pre. src='//schemaless' src='/normal' <a href="//schemaless">Schemaless</a>. <a href="/normal">normal</a>. Post.`
|
||
const REPLACE_SCHEMALESS_HTML_CORRECT = `Pre. src='//schemaless' src='http://base/normal' <a href="//schemaless">Schemaless</a>. <a href="http://base/normal">normal</a>. Post.`
|
||
const REPLACE_SCHEMALESS_XML = `Pre. src="//schemaless" src="/normal" <a href='//schemaless'>Schemaless</a>. <a href='/normal'>normal</a>. Post.`
|
||
const REPLACE_SCHEMALESS_XML_CORRECT = `Pre. src="//schemaless" src="http://base/normal" <a href='//schemaless'>Schemaless</a>. <a href='http://base/normal'>normal</a>. Post.`
|
||
|
||
var abs_url_bench_tests = []test{
|
||
{H5_JS_CONTENT_DOUBLE_QUOTE, CORRECT_OUTPUT_SRC_HREF_DQ},
|
||
{H5_JS_CONTENT_SINGLE_QUOTE, CORRECT_OUTPUT_SRC_HREF_SQ},
|
||
{H5_JS_CONTENT_ABS_URL, H5_JS_CONTENT_ABS_URL},
|
||
{H5_JS_CONTENT_ABS_URL_SCHEMALESS, H5_JS_CONTENT_ABS_URL_SCHEMALESS},
|
||
}
|
||
|
||
var xml_abs_url_bench_tests = []test{
|
||
{H5_XML_CONTENT_ABS_URL, CORRECT_OUTPUT_SRC_HREF_IN_XML},
|
||
{H5_XML_CONTENT_GUARDED, H5_XML_CONTENT_GUARDED},
|
||
}
|
||
|
||
var sanity_tests = []test{{REPLACE_1, REPLACE_1}, {REPLACE_2, REPLACE_2}}
|
||
var extra_tests_html = []test{{REPLACE_SCHEMALESS_HTML, REPLACE_SCHEMALESS_HTML_CORRECT}}
|
||
var abs_url_tests = append(abs_url_bench_tests, append(sanity_tests, extra_tests_html...)...)
|
||
var extra_tests_xml = []test{{REPLACE_SCHEMALESS_XML, REPLACE_SCHEMALESS_XML_CORRECT}}
|
||
var xml_abs_url_tests = append(xml_abs_url_bench_tests, append(sanity_tests, extra_tests_xml...)...)
|
||
|
||
func TestChainZeroTransformers(t *testing.T) {
|
||
tr := NewChain()
|
||
in := new(bytes.Buffer)
|
||
out := new(bytes.Buffer)
|
||
if err := tr.Apply(in, out); err != nil {
|
||
t.Errorf("A zero transformer chain returned an error.")
|
||
}
|
||
}
|
||
|
||
func TestChaingMultipleTransformers(t *testing.T) {
|
||
f1 := func(rw ContentReWriter) {
|
||
rw.Write(bytes.Replace(rw.Content(), []byte("f1"), []byte("f1r"), -1))
|
||
}
|
||
f2 := func(rw ContentReWriter) {
|
||
rw.Write(bytes.Replace(rw.Content(), []byte("f2"), []byte("f2r"), -1))
|
||
}
|
||
f3 := func(rw ContentReWriter) {
|
||
rw.Write(bytes.Replace(rw.Content(), []byte("f3"), []byte("f3r"), -1))
|
||
}
|
||
|
||
f4 := func(rw ContentReWriter) {
|
||
rw.Write(bytes.Replace(rw.Content(), []byte("f4"), []byte("f4r"), -1))
|
||
}
|
||
|
||
tr := NewChain(f1, f2, f3, f4)
|
||
|
||
out := new(bytes.Buffer)
|
||
if err := tr.Apply(out, helpers.StringToReader("Test: f4 f3 f1 f2 f1 The End.")); err != nil {
|
||
t.Errorf("Multi transformer chain returned an error: %s", err)
|
||
}
|
||
|
||
expected := "Test: f4r f3r f1r f2r f1r The End."
|
||
|
||
if string(out.Bytes()) != expected {
|
||
t.Errorf("Expected %s got %s", expected, string(out.Bytes()))
|
||
}
|
||
}
|
||
|
||
func BenchmarkAbsURL(b *testing.B) {
|
||
absURL, _ := AbsURL("http://base")
|
||
tr := NewChain(absURL...)
|
||
|
||
b.ResetTimer()
|
||
for i := 0; i < b.N; i++ {
|
||
apply(b.Errorf, tr, abs_url_bench_tests)
|
||
}
|
||
}
|
||
|
||
func TestAbsURL(t *testing.T) {
|
||
absURL, _ := AbsURL("http://base")
|
||
tr := NewChain(absURL...)
|
||
|
||
apply(t.Errorf, tr, abs_url_tests)
|
||
|
||
}
|
||
|
||
func BenchmarkXMLAbsURL(b *testing.B) {
|
||
absURLInXML, _ := AbsURLInXML("http://base")
|
||
tr := NewChain(absURLInXML...)
|
||
|
||
b.ResetTimer()
|
||
for i := 0; i < b.N; i++ {
|
||
apply(b.Errorf, tr, xml_abs_url_bench_tests)
|
||
}
|
||
}
|
||
|
||
func TestXMLAbsURL(t *testing.T) {
|
||
absURLInXML, _ := AbsURLInXML("http://base")
|
||
tr := NewChain(absURLInXML...)
|
||
apply(t.Errorf, tr, xml_abs_url_tests)
|
||
}
|
||
|
||
type errorf func(string, ...interface{})
|
||
|
||
func apply(ef errorf, tr chain, tests []test) {
|
||
for _, test := range tests {
|
||
out := new(bytes.Buffer)
|
||
err := tr.Apply(out, strings.NewReader(test.content))
|
||
if err != nil {
|
||
ef("Unexpected error: %s", err)
|
||
}
|
||
if test.expected != string(out.Bytes()) {
|
||
ef("Expected:\n%s\nGot:\n%s", test.expected, string(out.Bytes()))
|
||
}
|
||
}
|
||
}
|
||
|
||
type test struct {
|
||
content string
|
||
expected string
|
||
}
|