Apply external markup rendering to non-text files

[causal-agent]

• Gitea version (or commit ref): 1.4.3
• Git version: 2.18.0
• Operating system: FreeBSD 11.2-RELEASE
• Database (use [x]):
  • PostgreSQL
  • MySQL
  • MSSQL
  • SQLite
• Can you reproduce the bug at https://try.gitea.io:
  • Yes (provide example URL)
  • No
  • Not relevant
• Log gist:

Description

Currently, files are only checked against external markup extensions when the file is a text file:

gitea/routers/repo/view.go

Lines 201 to 214 in fc0001c

	case isTextFile:
	if blob.Size() >= setting.UI.MaxDisplayFileSize {
	ctx.Data["IsFileTooLarge"] = true
	break
	}
	d, _ := ioutil.ReadAll(dataRc)
	buf = templates.ToUTF8WithFallback(append(buf, d...))
	readmeExist := markup.IsReadmeFile(blob.Name())
	ctx.Data["ReadmeExist"] = readmeExist
	if markup.Type(blob.Name()) != "" {
	ctx.Data["IsMarkup"] = true
	ctx.Data["FileContent"] = string(markup.Render(blob.Name(), buf, path.Dir(treeLink), ctx.Repo.Repository.ComposeMetas()))

I would like to be able to use external markup renderers to render previews of some binary formats, similar to how PDFs can be displayed inline. Instead, I always get the "view raw" page for binary files.

[HarvsG]

I have the same issue trying to render docx files using pandoc. I created an issue #7614.

[HarvsG]

As you identified the type of rendering is determined by this switch

gitea/routers/repo/view.go

Lines 200 to 269 in fc0001c

	switch {
	case isTextFile:
	if blob.Size() >= setting.UI.MaxDisplayFileSize {
	ctx.Data["IsFileTooLarge"] = true
	break
	}
	d, _ := ioutil.ReadAll(dataRc)
	buf = templates.ToUTF8WithFallback(append(buf, d...))
	readmeExist := markup.IsReadmeFile(blob.Name())
	ctx.Data["ReadmeExist"] = readmeExist
	if markup.Type(blob.Name()) != "" {
	ctx.Data["IsMarkup"] = true
	ctx.Data["FileContent"] = string(markup.Render(blob.Name(), buf, path.Dir(treeLink), ctx.Repo.Repository.ComposeMetas()))
	} else if readmeExist {
	ctx.Data["IsRenderedHTML"] = true
	ctx.Data["FileContent"] = strings.Replace(
	gotemplate.HTMLEscapeString(string(buf)), "\n", `<br>`, -1,
	)
	} else {
	// Building code view blocks with line number on server side.
	var fileContent string
	if content, err := templates.ToUTF8WithErr(buf); err != nil {
	if err != nil {
	log.Error(4, "ToUTF8WithErr: %v", err)
	}
	fileContent = string(buf)
	} else {
	fileContent = content
	}
	var output bytes.Buffer
	lines := strings.Split(fileContent, "\n")
	//Remove blank line at the end of file
	if len(lines) > 0 && lines[len(lines)-1] == "" {
	lines = lines[:len(lines)-1]
	}
	for index, line := range lines {
	line = gotemplate.HTMLEscapeString(line)
	if index != len(lines)-1 {
	line += "\n"
	}
	output.WriteString(fmt.Sprintf(`<li class="L%d" rel="L%d">%s</li>`, index+1, index+1, line))
	}
	ctx.Data["FileContent"] = gotemplate.HTML(output.String())
	output.Reset()
	for i := 0; i < len(lines); i++ {
	output.WriteString(fmt.Sprintf(`<span id="L%d">%d</span>`, i+1, i+1))
	}
	ctx.Data["LineNums"] = gotemplate.HTML(output.String())
	}
	if ctx.Repo.CanEnableEditor() {
	ctx.Data["CanEditFile"] = true
	ctx.Data["EditFileTooltip"] = ctx.Tr("repo.editor.edit_this_file")
	} else if !ctx.Repo.IsViewBranch {
	ctx.Data["EditFileTooltip"] = ctx.Tr("repo.editor.must_be_on_a_branch")
	} else if !ctx.Repo.IsWriter() {
	ctx.Data["EditFileTooltip"] = ctx.Tr("repo.editor.fork_before_edit")
	}
	case base.IsPDFFile(buf):
	ctx.Data["IsPDFFile"] = true
	case base.IsVideoFile(buf):
	ctx.Data["IsVideoFile"] = true
	case base.IsImageFile(buf):
	ctx.Data["IsImageFile"] = true
	}

But interestingly I cant see a case base.IsBinaryFile(buf): or a default: case that would manage binary files. So they must either be handled before the switch or after it.

Edit:

In the case of the render type not being known then this is run which provides the 'view raw' link.

gitea/templates/repo/view_file.tmpl

Lines 67 to 68 in bcbc9f3

	{{else}}
	<a href="{{EscapePound $.RawFileLink}}" rel="nofollow" class="btn btn-gray btn-radius">{{.i18n.Tr "repo.file_view_raw"}}</a>

So a change must be made that makes either of .IsMarkup or .IsRenderedHTML options true:

gitea/templates/repo/view_file.tmpl

Lines 49 to 52 in bcbc9f3

	{{if .IsMarkup}}
	{{if .FileContent}}{{.FileContent | Safe}}{{end}}
	{{else if .IsRenderedHTML}}
	<pre>{{if .FileContent}}{{.FileContent | Str2html}}{{end}}</pre>

Edit2, Solution:

Perhaps a solution would be to add

	default:
		if blob.Size() >= setting.UI.MaxDisplayFileSize {
			ctx.Data["IsFileTooLarge"] = true
			break
		}

		d, _ := ioutil.ReadAll(dataRc)
		buf = templates.ToUTF8WithFallback(append(buf, d...))
		if markup.Type(blob.Name()) != "" {
			ctx.Data["IsMarkup"] = true // or ctx.Data["IsRenderedHTML"] = true
			ctx.Data["FileContent"] = string(markup.Render(blob.Name(), buf, path.Dir(treeLink), ctx.Repo.Repository.ComposeMetas()))
		}

at line

gitea/routers/repo/view.go

Line 278 in fc0001c

}