From 2d5a1949f90374b3746f9c30b83764fdb3fc0e6e Mon Sep 17 00:00:00 2001
From: sloum <sloum@colorfield.space>
Date: Sat, 12 Sep 2020 23:16:34 -0700
Subject: [PATCH] Adds the worls worst html parser to the gemini module

---
 gemini/gemini.go      |  19 +++---
 gemini/html_parser.go | 136 ++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 148 insertions(+), 7 deletions(-)
 create mode 100644 gemini/html_parser.go

diff --git a/gemini/gemini.go b/gemini/gemini.go
index db02ca5..099e27d 100644
--- a/gemini/gemini.go
+++ b/gemini/gemini.go
@@ -78,7 +78,7 @@ func (t *TofuDigest) Match(host, localCert string, cState *tls.ConnectionState)
 			return fmt.Errorf("EXP")
 		}
 
-		if err := cert.VerifyHostname(host); err != nil {
+		if err := cert.VerifyHostname(host); err != nil && cert.Subject.CommonName != host {
 			return fmt.Errorf("Certificate error: %s", err)
 		}
 
@@ -328,14 +328,18 @@ func Visit(host, port, resource string, td *TofuDigest) (Capsule, error) {
 		}
 		capsule.MimeMaj = minMajMime[0]
 		capsule.MimeMin = minMajMime[1]
+
+		if len(resource) > 0 && resource[0] != '/' {
+			resource = fmt.Sprintf("/%s", resource)
+		} else if resource == "" {
+			resource = "/"
+		}
+		currentUrl := fmt.Sprintf("gemini://%s:%s%s", host, port, resource)
+
 		if capsule.MimeMaj == "text" && capsule.MimeMin == "gemini" {
-			if len(resource) > 0 && resource[0] != '/' {
-				resource = fmt.Sprintf("/%s", resource)
-			} else if resource == "" {
-				resource = "/"
-			}
-			currentUrl := fmt.Sprintf("gemini://%s:%s%s", host, port, resource)
 			capsule.Content, capsule.Links = parseGemini(body, currentUrl)
+		} else if capsule.MimeMaj == "text" && capsule.MimeMin == "html" {
+			capsule.Content, capsule.Links = ParseHTML(body, currentUrl)
 		} else {
 			capsule.Content = body
 		}
@@ -408,6 +412,7 @@ func parseGemini(b, currentUrl string) (string, []string) {
 	return strings.Join(splitContent[:outputIndex], "\n"), links
 }
 
+
 // handleRelativeUrl provides link completion
 func HandleRelativeUrl(relLink, current string) (string, error) {
 	base, err := url.Parse(current)
diff --git a/gemini/html_parser.go b/gemini/html_parser.go
new file mode 100644
index 0000000..c1ff5f4
--- /dev/null
+++ b/gemini/html_parser.go
@@ -0,0 +1,136 @@
+package gemini
+
+import (
+	"fmt"
+	"strings"
+)
+
+type tag struct {
+	name       string
+	attributes map[string]string
+	text       string
+}
+
+var fields []string
+var out strings.Builder
+var links []string = make([]string, 0, 10)
+var inHead bool = false
+
+func ParseHTML(body, currentURL string) (string, []string) {
+	out.Reset()
+	links = make([]string, 0, 10)
+	inHead = false
+
+	body = strings.Replace(body, "<", " <", -1)
+	body = strings.Replace(body, ">", "> ", -1)
+	fields = strings.Fields(body)
+
+	var text string
+	for i := 0; i < len(fields); i++ {
+		if strings.HasPrefix(fields[i], "</") {
+			text = parseClose(i)
+	  } else if fields[i][0] == '<' {
+			var newIndex int
+			text, newIndex = parseTag(i, currentURL)
+			i = newIndex
+		} else {
+			text = fields[i] + " "
+		}
+		if text != "" {
+			out.WriteString(text)
+		}
+		text = ""
+	}
+
+	return out.String(), links
+}
+
+func skipToClose(tag string, i int) int {
+	for ;i < len(fields) && fields[i] != tag; i++ {
+		continue
+	}
+	return i
+}
+
+func parseTag(i int, currentURL string) (string, int) {
+	out := ""
+	hitClose := false
+	tag := strings.Replace(fields[i], "<", "", 1)
+	if len(tag) < 1 {
+		i++
+		tag = fields[i]
+	}
+	if strings.HasSuffix(tag, ">") && len(tag) >= 2 {
+		tag = tag[:len(tag)-1]
+		hitClose = true
+	}
+	tag = strings.ToLower(tag)
+MainSwitch:
+	switch tag {
+	case "head":
+		i = skipToClose("</head>", i)
+	case "script":
+		i = skipToClose("</script>", i)
+	case "h1":
+		out = "\n# "
+	case "h2":
+		out = "\n## "
+	case "h3":
+		out = "\n### "
+	case "li", "dt":
+		out = "* "
+	case "hr":
+		out = "\n\n-------------------------\n\n"
+	case "blockquote":
+		out = "\n> "
+	case "p", "header", "nav", "footer", "aside", "div", "main", "article", "details", "summary", "ul", "ol", "dl":
+		out = "\n"
+	case "a", "img":
+		target := "href="
+		if tag == "img" {
+			target = "src="
+		}
+		for ;i < len(fields) && !strings.HasPrefix(fields[i], target) ; i++ {
+			if strings.HasSuffix(fields[i], ">") {
+				hitClose = true
+				break MainSwitch
+			}
+		}
+		fields[i] = strings.Replace(fields[i], target, "", 1)
+		fields[i] = strings.Replace(fields[i], "\"", "", -1)
+		fields[i] = strings.Replace(fields[i], "'", "", -1)
+		if strings.HasSuffix(fields[i], ">") {
+			fields[i] = fields[i][:len(fields[i])-1]
+			hitClose = true
+		}
+		link := fields[i]
+		if strings.Index(link, "://") < 0 {
+			link, _ = HandleRelativeUrl(link, currentURL)
+		}
+
+		links = append(links, link)
+		if tag == "img" {
+			out = fmt.Sprintf("\n[%d]IMG\n", len(links))
+		} else {
+			out = fmt.Sprintf("[%d]", len(links))
+		}
+	}
+	if !hitClose {
+		for ;!strings.HasSuffix(fields[i], ">"); i++ {
+			continue
+		}
+	}
+
+	return out, i
+}
+
+
+func parseClose(i int) string {
+	tag := strings.Trim(fields[i], " \n\r\t<>/")
+	switch tag {
+	case "p", "header", "nav", "footer", "aside", "div", "main", "article", "details", "summary", "ul", "ol", "dl", "li", "dd", "blockquote", "h1", "h2", "h3", "h4", "h5", "h6":
+		return "\n"
+	default:
+		return ""
+	}
+}