oss/miniflux-v2
oss/miniflux-v2
1
0
Fork 0
mirror of https://github.com/miniflux/v2.git synced 2025-06-27 16:36:00 +00:00
Code Wiki Activity

perf(sanitizer): extract a call to url.Parse and make intensive use of it

Browse source 
Previously, url.Parse(baseUrl) was called on every self-closing tags, and on
most opening tags, accounting for around 15% of the CPU time spent in
processor.ProcessFeedEntries
This commit is contained in:
jvoisin 2025-06-12 15:05:20 +02:00
parent 60ad19c427
commit 2e36522376
1 changed files with 14 additions and 8 deletions
Download patch file Download diff file Expand all files Collapse all files

22
internal/reader/sanitizer/sanitizer.go
View file

@ -128,6 +128,9 @@ func SanitizeHTML(baseURL, rawHTML string, sanitizerOptions *SanitizerOptions) s
var parentTag string
var blockedStack []string
// Errors are a non-issue, so they're handled later in the function.
parsedBaseUrl, _ := url.Parse(baseURL)
tokenizer := html.NewTokenizer(strings.NewReader(rawHTML))
for {
if tokenizer.Next() == html.ErrorToken {
@ -175,7 +178,7 @@ func SanitizeHTML(baseURL, rawHTML string, sanitizerOptions *SanitizerOptions) s
}
if len(blockedStack) == 0 && isValidTag(tagName) {
attrNames, htmlAttributes := sanitizeAttributes(baseURL, tagName, token.Attr, sanitizerOptions)
attrNames, htmlAttributes := sanitizeAttributes(parsedBaseUrl, baseURL, tagName, token.Attr, sanitizerOptions)
if hasRequiredAttributes(tagName, attrNames) {
if len(attrNames) > 0 {
// Rewrite the start tag with allowed attributes.
@ -203,7 +206,7 @@ func SanitizeHTML(baseURL, rawHTML string, sanitizerOptions *SanitizerOptions) s
continue
}
if len(blockedStack) == 0 && isValidTag(tagName) {
attrNames, htmlAttributes := sanitizeAttributes(baseURL, tagName, token.Attr, sanitizerOptions)
attrNames, htmlAttributes := sanitizeAttributes(parsedBaseUrl, baseURL, tagName, token.Attr, sanitizerOptions)
if hasRequiredAttributes(tagName, attrNames) {
if len(attrNames) > 0 {
buffer.WriteString("<" + tagName + " " + htmlAttributes + "/>")
@ -216,7 +219,7 @@ func SanitizeHTML(baseURL, rawHTML string, sanitizerOptions *SanitizerOptions) s
}
}
func sanitizeAttributes(baseURL, tagName string, attributes []html.Attribute, sanitizerOptions *SanitizerOptions) ([]string, string) {
func sanitizeAttributes(parsedBaseUrl *url.URL, baseURL, tagName string, attributes []html.Attribute, sanitizerOptions *SanitizerOptions) ([]string, string) {
var htmlAttrs, attrNames []string
var err error
var isImageLargerThanLayout bool
@ -227,8 +230,6 @@ func sanitizeAttributes(baseURL, tagName string, attributes []html.Attribute, sa
isImageLargerThanLayout = imgWidth > 750
}
parsedBaseUrl, _ := url.Parse(baseURL)
for _, attribute := range attributes {
value := attribute.Val
@ -265,7 +266,7 @@ func sanitizeAttributes(baseURL, tagName string, attributes []html.Attribute, sa
if isExternalResourceAttribute(attribute.Key) {
switch {
case tagName == "iframe":
if !isValidIframeSource(baseURL, attribute.Val) {
if !isValidIframeSource(parsedBaseUrl, baseURL, attribute.Val) {
continue
}
value = rewriteIframeURL(attribute.Val)
@ -447,7 +448,7 @@ func isBlockedResource(src string) bool {
})
}
func isValidIframeSource(baseURL, src string) bool {
func isValidIframeSource(parsedBaseUrl *url.URL, baseURL, src string) bool {
whitelist := []string{
"bandcamp.com",
"cdn.embedly.com",
@ -464,8 +465,13 @@ func isValidIframeSource(baseURL, src string) bool {
}
domain := urllib.Domain(src)
baseDomain := baseURL
if parsedBaseUrl != nil {
baseDomain = parsedBaseUrl.Hostname()
}
// allow iframe from same origin
if urllib.Domain(baseURL) == domain {
if baseDomain == domain {
return true
}