refactor: highlight code blocks function

.air.toml

@@ -23,7 +23,7 @@ tmp_dir = "tmp"
   rerun = false
   rerun_delay = 500
   send_interrupt = false
-  stop_on_error = false
+  stop_on_error = true
 
 [color]
   app = ""

README.md

@@ -37,6 +37,7 @@ This project is super lightweight by design. The UI is simple and the frontend i
 | [anonymousoverflow.privacyfucking.rocks](https://anonymousoverflow.privacyfucking.rocks/) | Germany                         | Operated by [privacyfucking.rocks](https://privacyfucking.rocks)                                 |
 | [exchange.seitan-ayoub.lol](https://exchange.seitan-ayoub.lol)                            | Germany                         | Operated by [Seitan Ayoub](https://seitan-ayoub.lol)                                             |
 | [overflow.r4fo.com](https://overflow.r4fo.com)                                            | The Netherlands                 | Operated by [r4fo.com](https://r4fo.com)                                                         |
+| [overflow.ducks.party](https://overflow.ducks.party)                                      | The Netherlands                 | Operated by [ducks.party](https://ducks.party)                                                   |
 
 ## Other Instances
 

src/routes/question.go

@@ -29,7 +29,6 @@ var soSortValues = map[string]string{
 }
 
 func ViewQuestion(c *gin.Context) {
-	client := resty.New()
 
 	questionId := c.Param("id")
 	if _, err := strconv.Atoi(questionId); err != nil {
@@ -41,42 +40,24 @@ func ViewQuestion(c *gin.Context) {
 		return
 	}
 
-	questionTitle := c.Param("title")
-
-	sortValue := c.Query("sort_by")
-	if sortValue == "" {
-		sortValue = "votes"
+	params, err := parseAndValidateParameters(c)
+	if err != nil {
+		return
 	}
 
-	soSortValue, ok := soSortValues[sortValue]
-	if !ok {
-		soSortValue = soSortValues["votes"]
-	}
-
-	sub := c.Param("sub")
-
 	domain := "stackoverflow.com"
 
-	if sub != "" {
-		domain = fmt.Sprintf("%s.stackexchange.com", sub)
+	if params.Sub != "" {
+		domain = fmt.Sprintf("%s.stackexchange.com", params.Sub)
 	}
 
-	soLink := fmt.Sprintf("https://%s/questions/%s/%s?answertab=%s", domain, questionId, questionTitle, soSortValue)
+	soLink := fmt.Sprintf("https://%s/questions/%s/%s?answertab=%s", domain, questionId, params.QuestionTitle, params.SoSortValue)
 
-	resp, err := client.R().Get(soLink)
-	if err != nil {
-		c.HTML(500, "home.html", gin.H{
-			"errorMessage": "Unable to fetch question data",
-			"theme":        c.MustGet("theme").(string),
-			"version":      config.Version,
-		})
-		return
-	}
-	defer resp.RawResponse.Body.Close()
+	resp, err := fetchQuestionData(soLink)
 
 	if resp.StatusCode() != 200 {
 		c.HTML(500, "home.html", gin.H{
-			"errorMessage": "Received a non-OK status code",
+			"errorMessage": fmt.Sprintf("Received a non-OK status code %d", resp.StatusCode()),
 			"theme":        c.MustGet("theme").(string),
 			"version":      config.Version,
 		})
@@ -97,181 +78,25 @@ func ViewQuestion(c *gin.Context) {
 		return
 	}
 
-	newFilteredQuestion := types.FilteredQuestion{}
-
-	questionTextParent := doc.Find("h1.fs-headline1")
-
-	questionText := questionTextParent.Children().First().Text()
-
-	newFilteredQuestion.Title = questionText
-
-	questionPostLayout := doc.Find("div.post-layout").First()
-
-	questionTags := utils.GetPostTags(questionPostLayout)
-	newFilteredQuestion.Tags = questionTags
-
-	questionBodyParent := doc.Find("div.s-prose")
-
-	questionBodyParentHTML, err := questionBodyParent.Html()
+	newFilteredQuestion, err := extractQuestionData(doc, domain)
 	if err != nil {
 		c.HTML(500, "home.html", gin.H{
-			"errorMessage": "Unable to parse question body",
+			"errorMessage": "Failed to extract question data",
 			"theme":        c.MustGet("theme").(string),
 			"version":      config.Version,
 		})
 		return
 	}
 
-	newFilteredQuestion.Body = template.HTML(utils.ReplaceImgTags(questionBodyParentHTML))
-
-	questionBodyText := questionBodyParent.Text()
-
-	// remove all whitespace to create the shortened body desc
-	shortenedBody := strings.TrimSpace(questionBodyText)
-
-	// remove all newlines
-	shortenedBody = strings.ReplaceAll(shortenedBody, "\n", " ")
-
-	// get the first 50 chars
-	shortenedBody = shortenedBody[:50]
-
-	newFilteredQuestion.ShortenedBody = shortenedBody
-
-	comments := utils.FindAndReturnComments(questionBodyParentHTML, domain, questionPostLayout)
-	newFilteredQuestion.Comments = comments
-
-	// parse any code blocks and highlight them
-	answerCodeBlocks := questionCodeBlockRegex.FindAllString(questionBodyParentHTML, -1)
-	for _, codeBlock := range answerCodeBlocks {
-		codeBlock = utils.StripBlockTags(codeBlock)
-
-		// syntax highlight
-		highlightedCodeBlock := utils.HighlightSyntaxViaContent(codeBlock)
-
-		// replace the code block with the highlighted code block
-		questionBodyParentHTML = strings.Replace(questionBodyParentHTML, codeBlock, highlightedCodeBlock, 1)
-	}
-
-	questionCard := doc.Find("div.postcell")
-
-	questionMetadata := questionCard.Find("div.user-info")
-	questionTimestamp := ""
-	questionMetadata.Find("span.relativetime").Each(func(i int, s *goquery.Selection) {
-		// get the second
-		if i == 0 {
-			if s.Text() != "" {
-				// if it's not been edited, it means it's the first
-				questionTimestamp = s.Text()
-				return
-			}
-		}
-
-		// otherwise it's the second element
-		if i == 1 {
-			questionTimestamp = s.Text()
-			return
-		}
-	})
-
-	newFilteredQuestion.Timestamp = questionTimestamp
-
-	userDetails := questionMetadata.Find("div.user-details")
-
-	questionAuthor := ""
-	questionAuthorURL := fmt.Sprintf("https://%s", domain)
-
-	// check if the question has been edited
-	isQuestionEdited := false
-	questionMetadata.Find("a.js-gps-track").Each(func(i int, s *goquery.Selection) {
-		if strings.Contains(s.Text(), "edited") {
-			isQuestionEdited = true
-			return
-		}
-	})
-
-	userDetails.Find("a").Each(func(i int, s *goquery.Selection) {
-		// if question has been edited, the author is the second element.
-
-		if isQuestionEdited {
-			if i == 1 {
-				questionAuthor = s.Text()
-				questionAuthorURL += s.AttrOr("href", "")
-				return
-			}
-		} else {
-			// otherwise it's the first element
-			if i == 0 {
-				questionAuthor = s.Text()
-				questionAuthorURL += s.AttrOr("href", "")
-				return
-			}
-		}
-	})
-
-	newFilteredQuestion.AuthorName = questionAuthor
-	newFilteredQuestion.AuthorURL = questionAuthorURL
-
-	answers := []types.FilteredAnswer{}
-
-	doc.Find("div.answer").Each(func(i int, s *goquery.Selection) {
-
-		newFilteredAnswer := types.FilteredAnswer{}
-
-		postLayout := s.Find("div.post-layout")
-		voteCell := postLayout.Find("div.votecell")
-		answerCell := postLayout.Find("div.answercell")
-		answerBody := answerCell.Find("div.s-prose")
-		answerBodyHTML, _ := answerBody.Html()
-
-		voteCount := html.EscapeString(voteCell.Find("div.js-vote-count").Text())
-
-		newFilteredAnswer.Upvotes = voteCount
-		newFilteredAnswer.IsAccepted = s.HasClass("accepted-answer")
-
-		answerFooter := s.Find("div.mt24")
-
-		answerAuthorURL := fmt.Sprintf("https://%s", domain)
-		answerAuthorName := ""
-		answerTimestamp := ""
-
-		answerFooter.Find("div.post-signature").Each(func(i int, s *goquery.Selection) {
-			answerAuthorDetails := s.Find("div.user-details")
-
-			if answerAuthorDetails.Length() > 0 {
-				questionAuthor := answerAuthorDetails.Find("a").First()
-				answerAuthorName = html.EscapeString(questionAuthor.Text())
-				answerAuthorURL += html.EscapeString(questionAuthor.AttrOr("href", ""))
-			}
-
-			answerTimestamp = html.EscapeString(s.Find("span.relativetime").Text())
+	answers, err := extractAnswersData(doc, domain)
+	if err != nil {
+		c.HTML(500, "home.html", gin.H{
+			"errorMessage": "Failed to extract answer data",
+			"theme":        c.MustGet("theme").(string),
+			"version":      config.Version,
 		})
-
-		answerId, _ := s.Attr("data-answerid")
-		newFilteredAnswer.ID = answerId
-
-		newFilteredAnswer.AuthorName = answerAuthorName
-		newFilteredAnswer.AuthorURL = answerAuthorURL
-		newFilteredAnswer.Timestamp = answerTimestamp
-
-		// parse any code blocks and highlight them
-		answerCodeBlocks := codeBlockRegex.FindAllString(answerBodyHTML, -1)
-		for _, codeBlock := range answerCodeBlocks {
-			codeBlock = utils.StripBlockTags(codeBlock)
-
-			// syntax highlight
-			highlightedCodeBlock := utils.HighlightSyntaxViaContent(codeBlock)
-
-			// replace the code block with the highlighted code block
-			answerBodyHTML = strings.Replace(answerBodyHTML, codeBlock, highlightedCodeBlock, 1)
-		}
-
-		comments = utils.FindAndReturnComments(answerBodyHTML, domain, postLayout)
-
-		newFilteredAnswer.Comments = comments
-		newFilteredAnswer.Body = template.HTML(utils.ReplaceImgTags(answerBodyHTML))
-
-		answers = append(answers, newFilteredAnswer)
-	})
+		return
+	}
 
 	imagePolicy := "'self' https:"
 
@@ -284,9 +109,171 @@ func ViewQuestion(c *gin.Context) {
 		"answers":     answers,
 		"imagePolicy": imagePolicy,
 		"theme":       c.MustGet("theme").(string),
-		"currentUrl":  fmt.Sprintf("%s/questions/%s/%s", os.Getenv("APP_URL"), questionId, questionTitle),
-		"sortValue":   sortValue,
+		"currentUrl":  fmt.Sprintf("%s/questions/%s/%s", os.Getenv("APP_URL"), questionId, params.QuestionTitle),
+		"sortValue":   params.SoSortValue,
 		"domain":      domain,
 	})
 
 }
+
+type viewQuestionInputs struct {
+	QuestionID    string
+	QuestionTitle string
+	SoSortValue   string
+	Sub           string
+}
+
+// parseAndValidateParameters consolidates the URL and query parameters into an easily-accessible struct.
+func parseAndValidateParameters(c *gin.Context) (inputs viewQuestionInputs, err error) {
+
+	questionId := c.Param("id")
+	if _, err = strconv.Atoi(questionId); err != nil {
+		c.HTML(400, "home.html", gin.H{
+			"errorMessage": "Invalid question ID",
+			"theme":        c.MustGet("theme").(string),
+			"version":      config.Version,
+		})
+		return
+	}
+
+	inputs.QuestionID = questionId
+
+	sortValue := c.Query("sort_by")
+	if sortValue == "" {
+		sortValue = "votes"
+	}
+
+	soSortValue, ok := soSortValues[sortValue]
+	if !ok {
+		soSortValue = soSortValues["votes"]
+	}
+
+	inputs.SoSortValue = soSortValue
+
+	sub := c.Param("sub")
+
+	inputs.Sub = sub
+
+	return
+}
+
+// fetchQuestionData sends the request to StackOverflow.
+func fetchQuestionData(soLink string) (resp *resty.Response, err error) {
+	client := resty.New()
+	resp, err = client.R().Get(soLink)
+	return
+}
+
+// extractQuestionData parses the HTML document and extracts question data.
+func extractQuestionData(doc *goquery.Document, domain string) (question types.FilteredQuestion, err error) {
+	// Extract the question title.
+	questionTextParent := doc.Find("h1.fs-headline1").First()
+	question.Title = strings.TrimSpace(questionTextParent.Children().First().Text())
+
+	// Extract question tags.
+	questionTags := utils.GetPostTags(doc.Find("div.post-layout").First())
+	question.Tags = questionTags
+
+	// Extract and process the question body.
+	questionBodyParent := doc.Find("div.s-prose").First()
+	questionBodyParentHTML, err := questionBodyParent.Html()
+	if err != nil {
+		return question, err
+	}
+	question.Body = template.HTML(processHTMLBody(questionBodyParentHTML))
+
+	// Extract the shortened body description.
+	shortenedBody := strings.TrimSpace(questionBodyParent.Text())
+	shortenedBody = strings.ReplaceAll(shortenedBody, "\n", " ")
+	if len(shortenedBody) > 50 {
+		shortenedBody = shortenedBody[:50]
+	}
+	question.ShortenedBody = shortenedBody
+
+	// Extract question comments.
+	comments := utils.FindAndReturnComments(questionBodyParentHTML, domain, doc.Find("div.post-layout").First())
+	question.Comments = comments
+
+	// Extract question timestamp and author information.
+	questionCard := doc.Find("div.postcell").First()
+	extractMetadata(questionCard, &question, domain)
+
+	return
+}
+
+// extractMetadata extracts author and timestamp information from a given selection.
+func extractMetadata(selection *goquery.Selection, question *types.FilteredQuestion, domain string) {
+	questionMetadata := selection.Find("div.user-info").First()
+	question.Timestamp = questionMetadata.Find("span.relativetime").First().Text()
+
+	questionAuthorURL := "https://" + domain
+	questionAuthor := selection.Find("div.post-signature.owner div.user-info div.user-details a").First()
+	question.AuthorName = questionAuthor.Text()
+	questionAuthorURL += questionAuthor.AttrOr("href", "")
+	question.AuthorURL = questionAuthorURL
+
+	// Determine if the question has been edited and update author details accordingly.
+	isQuestionEdited := selection.Find("a.js-gps-track").Text() == "edited"
+	if isQuestionEdited {
+		editedAuthor := questionMetadata.Find("a").Last()
+		question.AuthorName = editedAuthor.Text()
+		question.AuthorURL = "https://" + domain + editedAuthor.AttrOr("href", "")
+	}
+}
+
+// extractAnswersData parses the HTML document and extracts answers data.
+func extractAnswersData(doc *goquery.Document, domain string) ([]types.FilteredAnswer, error) {
+	var answers []types.FilteredAnswer
+
+	// Iterate over each answer block.
+	doc.Find("div.answer").Each(func(i int, s *goquery.Selection) {
+		var answer types.FilteredAnswer
+
+		postLayout := s.Find("div.post-layout").First()
+
+		// Extract upvotes.
+		voteCell := postLayout.Find("div.votecell").First()
+		voteCount := html.EscapeString(voteCell.Find("div.js-vote-count").Text())
+		answer.Upvotes = voteCount
+
+		// Check if the answer is accepted.
+		answer.IsAccepted = s.HasClass("accepted-answer")
+
+		// Extract answer body and process it.
+		answerCell := postLayout.Find("div.answercell").First()
+		answerBody := answerCell.Find("div.s-prose").First()
+		answerBodyHTML, _ := answerBody.Html()
+
+		// Process code blocks within the answer.
+		processedAnswerBody := processHTMLBody(answerBodyHTML)
+		answer.Body = template.HTML(html.UnescapeString(processedAnswerBody))
+
+		// Extract author information and timestamp.
+		extractAnswerAuthorInfo(s, &answer, domain)
+
+		answers = append(answers, answer)
+	})
+
+	return answers, nil
+}
+
+// processHTMLBody highlights syntax and replaces images with proxied versions.
+func processHTMLBody(bodyHTML string) string {
+	highlightedBody := utils.HighlightCodeBlocks(bodyHTML)
+	imageProxiedBody := utils.ReplaceImgTags(highlightedBody)
+	return imageProxiedBody
+}
+
+// extractAnswerAuthorInfo extracts the author name, URL, and timestamp from an answer block.
+// It directly mutates the answer.
+func extractAnswerAuthorInfo(selection *goquery.Selection, answer *types.FilteredAnswer, domain string) {
+	authorDetails := selection.Find("div.post-signature").Last()
+
+	authorName := html.EscapeString(authorDetails.Find("div.user-details a").First().Text())
+	authorURL := "https://" + domain + authorDetails.Find("div.user-details a").AttrOr("href", "")
+	timestamp := html.EscapeString(authorDetails.Find("span.relativetime").Text())
+
+	answer.AuthorName = authorName
+	answer.AuthorURL = authorURL
+	answer.Timestamp = timestamp
+}

src/utils/cf.go

@@ -1,56 +0,0 @@
-package utils
-
-import (
-	"encoding/hex"
-	"net/url"
-	"strings"
-)
-
-// decodeEmail decodes an obfuscated email string.
-func DecodeEmail(encoded string) (string, error) {
-	// remove the prefix if present
-	prefix := "/cdn-cgi/l/email-protection#"
-	encoded = strings.TrimPrefix(encoded, prefix)
-
-	// get the key for XOR operation
-	key, err := hex.DecodeString(encoded[:2])
-	if err != nil {
-		return "", err
-	}
-
-	// decode the rest of the string using the key
-	var decodedChars []byte
-	// start a loop from the third character of the encoded string (index 2), as the first two characters are used as the key for XOR operation.
-	// the loop increments by 2 in each iteration because the encoded string is in hexadecimal format, and each hexadecimal character represents 4 bits.
-	// so, two hexadecimal characters represent one byte (8 bits), which is the size of a character in ASCII.
-	for i := 2; i < len(encoded); i += 2 {
-		// get the current pair of hexadecimal characters from the encoded string.
-		encodedPair := encoded[i : i+2]
-
-		// decode the hexadecimal pair to get the original byte. this is done by converting the hexadecimal pair to a byte using the DecodeString function from the encoding/hex package.
-		// if the hexadecimal pair is not a valid hexadecimal number, DecodeString will return an error.
-		pairBytes, err := hex.DecodeString(encodedPair)
-		if err != nil {
-			// if there is an error in decoding, return the error and an empty string.
-			return "", err
-		}
-
-		// perform an XOR operation on the decoded byte with the key. this is the reverse operation of the original encoding process.
-		// the XOR operation is a bitwise operation that returns 0 if the two bits are the same, and 1 otherwise.
-		// by performing the XOR operation with the same key used in the encoding process, we can retrieve the original byte.
-		decodedChar := pairBytes[0] ^ key[0]
-
-		// append the decoded character to the decodedChars slice. this slice will hold all the decoded characters of the encoded string.
-		decodedChars = append(decodedChars, decodedChar)
-	}
-	// convert the decoded bytes to a string
-	decoded := string(decodedChars)
-
-	// unescape any URL-encoded characters
-	unescaped, err := url.QueryUnescape(decoded)
-	if err != nil {
-		return "", err
-	}
-
-	return unescaped, nil
-}

src/utils/syntax.go

@@ -12,7 +12,9 @@ import (
 	"github.com/alecthomas/chroma/styles"
 )
 
-func HighlightSyntaxViaContent(content string) (htmlOut string) {
+// highlightSyntaxViaContent uses Chroma to lex code content and apply the appropriate tokenizer engine.
+// If it can't find one, it defaults to JavaScript syntax highlighting.
+func highlightSyntaxViaContent(content string) (htmlOut string) {
 	content = html.UnescapeString(content)
 
 	fallbackOut := html.EscapeString(content)
@@ -20,9 +22,7 @@ func HighlightSyntaxViaContent(content string) (htmlOut string) {
 	// identify the language
 	lexer := lexers.Analyse(content)
 	if lexer == nil {
-		// unable to identify, so just return the wrapped content
-		htmlOut = fallbackOut
-		return
+		lexer = lexers.Get(".js")
 	}
 
 	style := styles.Get("xcode")
@@ -54,7 +54,9 @@ func HighlightSyntaxViaContent(content string) (htmlOut string) {
 
 var preClassRegex = regexp.MustCompile(`(?s)<pre class=".+">`)
 
-func StripBlockTags(content string) (result string) {
+// stripBlockTags takes an extracted code block from HTML and strips it of its pre and code tags.
+// What's returned is just the code.
+func stripBlockTags(content string) (result string) {
 	// strip all "<code>" tags
 	content = strings.Replace(content, "<code>", "", -1)
 	content = strings.Replace(content, "</code>", "", -1)
@@ -68,3 +70,26 @@ func StripBlockTags(content string) (result string) {
 
 	return
 }
+
+var codeBlockRegex = regexp.MustCompile(`(?s)<pre><code>(.*?)<\/code><\/pre>`)
+
+// HighlightCodeBlocks uses both highlightSyntaxViaContent stripCodeBlocks and returns the newly highlighted code HTML.
+func HighlightCodeBlocks(html string) string {
+	// Replace each code block with the highlighted version
+	highlightedHTML := codeBlockRegex.ReplaceAllStringFunc(html, func(codeBlock string) string {
+		// Extract the code content from the code block
+		codeContent := codeBlockRegex.FindStringSubmatch(codeBlock)[1]
+
+		codeContent = stripBlockTags(codeContent)
+
+		// Highlight the code content
+		highlightedCode := highlightSyntaxViaContent(codeContent)
+
+		// Replace the original code block with the highlighted version
+		highlightedCodeBlock := "<pre>" + highlightedCode + "</pre>"
+
+		return highlightedCodeBlock
+	})
+
+	return highlightedHTML
+}