Rework markup link rendering (#26745)

Fixes #26548

This PR refactors the rendering of markup links. The old code uses
`strings.Replace` to change some urls while the new code uses more
context to decide which link should be generated.

The added tests should ensure the same output for the old and new
behaviour (besides the bug).

We may need to refactor the rendering a bit more to make it clear how
the different helper methods render the input string. There are lots of
options (resolve links / images / mentions / git hashes / emojis / ...)
but you don't really know what helper uses which options. For example,
we currently support images in the user description which should not be
allowed I think:

<details>
  <summary>Profile</summary> 

https://try.gitea.io/KN4CK3R

![grafik](https://github.com/go-gitea/gitea/assets/1666336/109ae422-496d-4200-b52e-b3a528f553e5)

</details>

---------

Co-authored-by: wxiaoguang <[email protected]>

Author: KN4CK3R

models/issues/comment_code.go

@@ -109,9 +109,11 @@ func findCodeComments(ctx context.Context, opts FindCommentsOptions, issue *Issu
 
 		var err error
 		if comment.RenderedContent, err = markdown.RenderString(&markup.RenderContext{
-			Ctx:       ctx,
-			URLPrefix: issue.Repo.Link(),
-			Metas:     issue.Repo.ComposeMetas(ctx),
+			Ctx: ctx,
+			Links: markup.Links{
+				Base: issue.Repo.Link(),
+			},
+			Metas: issue.Repo.ComposeMetas(ctx),
 		}, comment.Content); err != nil {
 			return nil, err
 		}

models/repo/repo.go

@@ -584,8 +584,7 @@ func (repo *Repository) CanEnableEditor() bool {
 // DescriptionHTML does special handles to description and return HTML string.
 func (repo *Repository) DescriptionHTML(ctx context.Context) template.HTML {
 	desc, err := markup.RenderDescriptionHTML(&markup.RenderContext{
-		Ctx:       ctx,
-		URLPrefix: repo.HTMLURL(),
+		Ctx: ctx,
 		// Don't use Metas to speedup requests
 	}, repo.Description)
 	if err != nil {

modules/markup/external/external.go

@@ -79,9 +79,10 @@ func envMark(envName string) string {
 // Render renders the data of the document to HTML via the external tool.
 func (p *Renderer) Render(ctx *markup.RenderContext, input io.Reader, output io.Writer) error {
 	var (
-		urlRawPrefix = strings.Replace(ctx.URLPrefix, "/src/", "/raw/", 1)
-		command      = strings.NewReplacer(envMark("GITEA_PREFIX_SRC"), ctx.URLPrefix,
-			envMark("GITEA_PREFIX_RAW"), urlRawPrefix).Replace(p.Command)
+		command = strings.NewReplacer(
+			envMark("GITEA_PREFIX_SRC"), ctx.Links.SrcLink(),
+			envMark("GITEA_PREFIX_RAW"), ctx.Links.RawLink(),
+		).Replace(p.Command)
 		commands = strings.Fields(command)
 		args     = commands[1:]
 	)
@@ -121,14 +122,14 @@ func (p *Renderer) Render(ctx *markup.RenderContext, input io.Reader, output io.
 		ctx.Ctx = graceful.GetManager().ShutdownContext()
 	}
 
-	processCtx, _, finished := process.GetManager().AddContext(ctx.Ctx, fmt.Sprintf("Render [%s] for %s", commands[0], ctx.URLPrefix))
+	processCtx, _, finished := process.GetManager().AddContext(ctx.Ctx, fmt.Sprintf("Render [%s] for %s", commands[0], ctx.Links.SrcLink()))
 	defer finished()
 
 	cmd := exec.CommandContext(processCtx, commands[0], args...)
 	cmd.Env = append(
 		os.Environ(),
-		"GITEA_PREFIX_SRC="+ctx.URLPrefix,
-		"GITEA_PREFIX_RAW="+urlRawPrefix,
+		"GITEA_PREFIX_SRC="+ctx.Links.SrcLink(),
+		"GITEA_PREFIX_RAW="+ctx.Links.RawLink(),
 	)
 	if !p.IsInputFile {
 		cmd.Stdin = input

modules/markup/html.go

@@ -80,15 +80,10 @@ const keywordClass = "issue-keyword"
 
 // IsLink reports whether link fits valid format.
 func IsLink(link []byte) bool {
-	return isLink(link)
-}
-
-// isLink reports whether link fits valid format.
-func isLink(link []byte) bool {
 	return validLinksPattern.Match(link)
 }
 
-func isLinkStr(link string) bool {
+func IsLinkStr(link string) bool {
 	return validLinksPattern.MatchString(link)
 }
 
@@ -344,7 +339,7 @@ func postProcess(ctx *RenderContext, procs []processor, input io.Reader, output
 		node = node.FirstChild
 	}
 
-	visitNode(ctx, procs, procs, node)
+	visitNode(ctx, procs, node)
 
 	newNodes := make([]*html.Node, 0, 5)
 
@@ -375,7 +370,7 @@ func postProcess(ctx *RenderContext, procs []processor, input io.Reader, output
 	return nil
 }
 
-func visitNode(ctx *RenderContext, procs, textProcs []processor, node *html.Node) {
+func visitNode(ctx *RenderContext, procs []processor, node *html.Node) {
 	// Add user-content- to IDs and "#" links if they don't already have them
 	for idx, attr := range node.Attr {
 		val := strings.TrimPrefix(attr.Val, "#")
@@ -390,35 +385,29 @@ func visitNode(ctx *RenderContext, procs, textProcs []processor, node *html.Node
 		}
 
 		if attr.Key == "class" && attr.Val == "emoji" {
-			textProcs = nil
+			procs = nil
 		}
 	}
 
 	// We ignore code and pre.
 	switch node.Type {
 	case html.TextNode:
-		textNode(ctx, textProcs, node)
+		textNode(ctx, procs, node)
 	case html.ElementNode:
 		if node.Data == "img" {
 			for i, attr := range node.Attr {
 				if attr.Key != "src" {
 					continue
 				}
-				if len(attr.Val) > 0 && !isLinkStr(attr.Val) && !strings.HasPrefix(attr.Val, "data:image/") {
-					prefix := ctx.URLPrefix
-					if ctx.IsWiki {
-						prefix = util.URLJoin(prefix, "wiki", "raw")
-					}
-					prefix = strings.Replace(prefix, "/src/", "/media/", 1)
-
-					attr.Val = util.URLJoin(prefix, attr.Val)
+				if len(attr.Val) > 0 && !IsLinkStr(attr.Val) && !strings.HasPrefix(attr.Val, "data:image/") {
+					attr.Val = util.URLJoin(ctx.Links.ResolveMediaLink(ctx.IsWiki), attr.Val)
 				}
 				attr.Val = camoHandleLink(attr.Val)
 				node.Attr[i] = attr
 			}
 		} else if node.Data == "a" {
 			// Restrict text in links to emojis
-			textProcs = emojiProcessors
+			procs = emojiProcessors
 		} else if node.Data == "code" || node.Data == "pre" {
 			return
 		} else if node.Data == "i" {
@@ -444,7 +433,7 @@ func visitNode(ctx *RenderContext, procs, textProcs []processor, node *html.Node
 			}
 		}
 		for n := node.FirstChild; n != nil; n = n.NextSibling {
-			visitNode(ctx, procs, textProcs, n)
+			visitNode(ctx, procs, n)
 		}
 	}
 	// ignore everything else
@@ -641,10 +630,6 @@ func mentionProcessor(ctx *RenderContext, node *html.Node) {
 }
 
 func shortLinkProcessor(ctx *RenderContext, node *html.Node) {
-	shortLinkProcessorFull(ctx, node, false)
-}
-
-func shortLinkProcessorFull(ctx *RenderContext, node *html.Node, noLink bool) {
 	next := node.NextSibling
 	for node != nil && node != next {
 		m := shortLinkPattern.FindStringSubmatchIndex(node.Data)
@@ -665,7 +650,7 @@ func shortLinkProcessorFull(ctx *RenderContext, node *html.Node, noLink bool) {
 			if equalPos := strings.IndexByte(v, '='); equalPos == -1 {
 				// There is no equal in this argument; this is a mandatory arg
 				if props["name"] == "" {
-					if isLinkStr(v) {
+					if IsLinkStr(v) {
 						// If we clearly see it is a link, we save it so
 
 						// But first we need to ensure, that if both mandatory args provided
@@ -740,7 +725,7 @@ func shortLinkProcessorFull(ctx *RenderContext, node *html.Node, noLink bool) {
 			DataAtom:   atom.A,
 		}
 		childNode.Parent = linkNode
-		absoluteLink := isLinkStr(link)
+		absoluteLink := IsLinkStr(link)
 		if !absoluteLink {
 			if image {
 				link = strings.ReplaceAll(link, " ", "+")
@@ -751,16 +736,9 @@ func shortLinkProcessorFull(ctx *RenderContext, node *html.Node, noLink bool) {
 				link = url.PathEscape(link)
 			}
 		}
-		urlPrefix := ctx.URLPrefix
 		if image {
 			if !absoluteLink {
-				if IsSameDomain(urlPrefix) {
-					urlPrefix = strings.Replace(urlPrefix, "/src/", "/raw/", 1)
-				}
-				if ctx.IsWiki {
-					link = util.URLJoin("wiki", "raw", link)
-				}
-				link = util.URLJoin(urlPrefix, link)
+				link = util.URLJoin(ctx.Links.ResolveMediaLink(ctx.IsWiki), link)
 			}
 			title := props["title"]
 			if title == "" {
@@ -789,18 +767,15 @@ func shortLinkProcessorFull(ctx *RenderContext, node *html.Node, noLink bool) {
 		} else {
 			if !absoluteLink {
 				if ctx.IsWiki {
-					link = util.URLJoin("wiki", link)
+					link = util.URLJoin(ctx.Links.WikiLink(), link)
+				} else {
+					link = util.URLJoin(ctx.Links.SrcLink(), link)
 				}
-				link = util.URLJoin(urlPrefix, link)
 			}
 			childNode.Type = html.TextNode
 			childNode.Data = name
 		}
-		if noLink {
-			linkNode = childNode
-		} else {
-			linkNode.Attr = []html.Attribute{{Key: "href", Val: link}}
-		}
+		linkNode.Attr = []html.Attribute{{Key: "href", Val: link}}
 		replaceContent(node, m[0], m[1], linkNode)
 		node = node.NextSibling.NextSibling
 	}

modules/markup/html_internal_test.go

@@ -287,8 +287,8 @@ func TestRender_IssueIndexPattern_Document(t *testing.T) {
 }
 
 func testRenderIssueIndexPattern(t *testing.T, input, expected string, ctx *RenderContext) {
-	if ctx.URLPrefix == "" {
-		ctx.URLPrefix = TestAppURL
+	if ctx.Links.Base == "" {
+		ctx.Links.Base = TestRepoURL
 	}
 
 	var buf strings.Builder
@@ -303,19 +303,23 @@ func TestRender_AutoLink(t *testing.T) {
 	test := func(input, expected string) {
 		var buffer strings.Builder
 		err := PostProcess(&RenderContext{
-			Ctx:       git.DefaultContext,
-			URLPrefix: TestRepoURL,
-			Metas:     localMetas,
+			Ctx: git.DefaultContext,
+			Links: Links{
+				Base: TestRepoURL,
+			},
+			Metas: localMetas,
 		}, strings.NewReader(input), &buffer)
 		assert.Equal(t, err, nil)
 		assert.Equal(t, strings.TrimSpace(expected), strings.TrimSpace(buffer.String()))
 
 		buffer.Reset()
 		err = PostProcess(&RenderContext{
-			Ctx:       git.DefaultContext,
-			URLPrefix: TestRepoURL,
-			Metas:     localMetas,
-			IsWiki:    true,
+			Ctx: git.DefaultContext,
+			Links: Links{
+				Base: TestRepoURL,
+			},
+			Metas:  localMetas,
+			IsWiki: true,
 		}, strings.NewReader(input), &buffer)
 		assert.Equal(t, err, nil)
 		assert.Equal(t, strings.TrimSpace(expected), strings.TrimSpace(buffer.String()))
@@ -342,9 +346,11 @@ func TestRender_FullIssueURLs(t *testing.T) {
 	test := func(input, expected string) {
 		var result strings.Builder
 		err := postProcess(&RenderContext{
-			Ctx:       git.DefaultContext,
-			URLPrefix: TestRepoURL,
-			Metas:     localMetas,
+			Ctx: git.DefaultContext,
+			Links: Links{
+				Base: TestRepoURL,
+			},
+			Metas: localMetas,
 		}, []processor{fullIssuePatternProcessor}, strings.NewReader(input), &result)
 		assert.NoError(t, err)
 		assert.Equal(t, expected, result.String())