feat: SEO backend output (2/4)

internal/infrastructure/useCases/seoBody.go

@@ -0,0 +1,101 @@
+package useCases
+
+import (
+	"strings"
+	"unicode/utf8"
+
+	"golang.org/x/net/html"
+)
+
+type bodyFacts struct {
+	spacedText       string
+	textRunes        int
+	hasSubheading    bool
+	imagesWithoutAlt int
+}
+
+var blockElements = map[string]bool{
+	"p": true, "div": true, "br": true, "li": true, "ul": true, "ol": true, "blockquote": true, "pre": true,
+	"h1": true, "h2": true, "h3": true, "h4": true, "h5": true, "h6": true, "tr": true, "section": true, "article": true,
+	"figure": true, "figcaption": true, "table": true, "hr": true,
+}
+
+var subheadings = map[string]bool{"h2": true, "h3": true, "h4": true, "h5": true, "h6": true}
+
+var skippedElements = map[string]bool{"script": true, "style": true}
+
+const sentenceTerminators = ".!?"
+
+func analyzeBody(body string) bodyFacts {
+	tokenizer := html.NewTokenizer(strings.NewReader(body))
+	var plain, spaced strings.Builder
+	facts := bodyFacts{}
+	skipping := false
+	for {
+		kind := tokenizer.Next()
+		if kind == html.ErrorToken {
+			break
+		}
+		switch kind {
+		case html.TextToken:
+			if !skipping {
+				text := string(tokenizer.Text())
+				plain.WriteString(text)
+				spaced.WriteString(text)
+			}
+		case html.StartTagToken, html.SelfClosingTagToken, html.EndTagToken:
+			token := tokenizer.Token()
+			skipping = nextSkipping(skipping, kind, token)
+			if blockElements[token.Data] {
+				spaced.WriteString(" ")
+			}
+			if kind == html.EndTagToken {
+				continue
+			}
+			if subheadings[token.Data] {
+				facts.hasSubheading = true
+			}
+			if token.Data == "img" && !hasAlt(token) {
+				facts.imagesWithoutAlt++
+			}
+		}
+	}
+	facts.textRunes = utf8.RuneCountInString(strings.TrimSpace(plain.String()))
+	facts.spacedText = strings.Join(strings.Fields(spaced.String()), " ")
+	return facts
+}
+
+func nextSkipping(skipping bool, kind html.TokenType, token html.Token) bool {
+	if !skippedElements[token.Data] {
+		return skipping
+	}
+	switch kind {
+	case html.StartTagToken:
+		return true
+	case html.EndTagToken:
+		return false
+	}
+	return skipping
+}
+
+func hasAlt(token html.Token) bool {
+	for _, attribute := range token.Attr {
+		if attribute.Key == "alt" {
+			return strings.TrimSpace(attribute.Val) != ""
+		}
+	}
+	return false
+}
+
+func firstSentence(text string) string {
+	runes := []rune(text)
+	for index, character := range runes {
+		if !strings.ContainsRune(sentenceTerminators, character) {
+			continue
+		}
+		if index+1 == len(runes) || runes[index+1] == ' ' {
+			return string(runes[:index+1])
+		}
+	}
+	return text
+}