fix: word count calculation

metadata/epub.go

@@ -53,10 +53,12 @@ func countEPUBWords(filepath string) (int64, error) {
 	rf := rc.Rootfiles[0]
 
 	var completeCount int64
-	for _, item := range rf.Spine.Itemrefs {
+	for _, item := range rf.Itemrefs {
 		f, _ := item.Open()
 		doc, _ := goquery.NewDocumentFromReader(f)
-		completeCount = completeCount + int64(len(strings.Fields(doc.Text())))
+		doc.Find("script, style, noscript, iframe").Remove()
+		words := len(strings.Fields(doc.Text()))
+		completeCount = completeCount + int64(words)
 	}
 
 	return completeCount, nil

metadata/metadata.go

@@ -87,7 +87,7 @@ func GetWordCount(filepath string) (*int64, error) {
 		}
 		return &totalWords, nil
 	} else {
-		return nil, fmt.Errorf("Invalid extension: %s", fileExtension)
+		return nil, fmt.Errorf("invalid extension: %s", fileExtension)
 	}
 }
 

search/anna.go

@@ -4,14 +4,11 @@ import (
 	"fmt"
 	"io"
 	"net/url"
-	"regexp"
 	"strings"
 
 	"github.com/PuerkitoBio/goquery"
 )
 
-var commentRE = regexp.MustCompile(`(?s)<!--(.*?)-->`)
-
 func searchAnnasArchive(query string) ([]SearchItem, error) {
 	searchURL := "https://annas-archive.org/search?index=&q=%s&ext=epub&sort=&lang=en"
 	url := fmt.Sprintf(searchURL, url.QueryEscape(query))
@@ -32,62 +29,34 @@ func parseAnnasArchive(body io.ReadCloser) ([]SearchItem, error) {
 
 	// Normalize Results
 	var allEntries []SearchItem
-	doc.Find("#aarecord-list > div.justify-center").Each(func(ix int, rawBook *goquery.Selection) {
+	doc.Find(".js-aarecord-list-outer > div > div").Each(func(ix int, rawBook *goquery.Selection) {
-		rawBook = getAnnasArchiveBookSelection(rawBook)
 
 		// Parse Details
-		details := rawBook.Find("div:nth-child(2) > div:nth-child(1)").Text()
+		details := rawBook.Find("div:nth-child(3)").Text()
-		detailsSplit := strings.Split(details, ", ")
+		detailsSplit := strings.Split(details, " · ")
 
 		// Invalid Details
-		if len(detailsSplit) < 4 {
+		if len(detailsSplit) < 3 {
 			return
 		}
 
 		// Parse MD5
-		itemHref, _ := rawBook.Find("a").Attr("href")
+		titleAuthorDetails := rawBook.Find("div a")
+		titleEl := titleAuthorDetails.Eq(0)
+		itemHref, _ := titleEl.Attr("href")
 		hrefArray := strings.Split(itemHref, "/")
 		id := hrefArray[len(hrefArray)-1]
 
 		allEntries = append(allEntries, SearchItem{
 			ID:       id,
-			Title:    rawBook.Find("h3").First().Text(),
+			Title:    titleEl.Text(),
-			Author:   rawBook.Find("div:nth-child(2) > div:nth-child(4)").First().Text(),
+			Author:   titleAuthorDetails.Eq(1).Text(),
 			Language: detailsSplit[0],
 			FileType: detailsSplit[1],
-			FileSize: detailsSplit[3],
+			FileSize: detailsSplit[2],
 		})
 	})
 
 	// Return Results
 	return allEntries, nil
 }
-
-// getAnnasArchiveBookSelection parses potentially commented out HTML. For some reason
-// Annas Archive comments out blocks "below the fold". They aren't rendered until you
-// scroll. This attempts to parse the commented out HTML.
-func getAnnasArchiveBookSelection(rawBook *goquery.Selection) *goquery.Selection {
-	rawHTML, err := rawBook.Html()
-	if err != nil {
-		return rawBook
-	}
-
-	strippedHTML := strings.TrimSpace(rawHTML)
-	if !strings.HasPrefix(strippedHTML, "<!--") || !strings.HasSuffix(strippedHTML, "-->") {
-		return rawBook
-	}
-
-	allMatches := commentRE.FindAllStringSubmatch(strippedHTML, -1)
-	if len(allMatches) != 1 || len(allMatches[0]) != 2 {
-		return rawBook
-	}
-
-	captureGroup := allMatches[0][1]
-	docReader := strings.NewReader(captureGroup)
-	doc, err := goquery.NewDocumentFromReader(docReader)
-	if err != nil {
-		return rawBook
-	}
-
-	return doc.Selection
-}