initial commit

.gitignore

@@ -0,0 +1,2 @@
+*.env
+news.json

Makefile

@@ -0,0 +1,8 @@
+.PHONY: fmt
+fmt: 
+	go mod tidy
+	go fmt ./...
+
+.PHONY: refresh
+refresh:
+	go run cmd/refresh_news/main.go

cmd/refresh_news/main.go

@@ -0,0 +1,58 @@
+package main
+
+import (
+	"encoding/json"
+	"log"
+	"net/http"
+	"os"
+
+	"sarcastick/pkg/news"
+
+	"golang.org/x/net/html"
+)
+
+func main() {
+	response, err := http.Get("https://www.expats.cz")
+	if err != nil {
+		log.Fatal(err)
+	}
+	defer response.Body.Close()
+
+	if response.StatusCode != http.StatusOK {
+		log.Fatalf("HTTP request failed with status code %d", response.StatusCode)
+	}
+
+	htmlBody, err := html.Parse(response.Body)
+	if err != nil {
+		log.Fatal(err)
+	}
+	link := news.FindNewsLink(htmlBody)
+	if link == "" {
+		log.Fatal("No link found")
+	}
+
+	response, err = http.Get("https://www.expats.cz" + link)
+	if err != nil {
+		log.Fatal(err)
+	}
+	defer response.Body.Close()
+
+	if response.StatusCode != http.StatusOK {
+		log.Fatalf("HTTP request failed with status code %d", response.StatusCode)
+	}
+
+	htmlBody, err = html.Parse(response.Body)
+	if err != nil {
+		log.Fatal(err)
+	}
+
+	ct := news.ExtractContentWithTitle(htmlBody)
+	file, err := os.Create("news.json")
+	if err != nil {
+		log.Fatal(err)
+	}
+	err = json.NewEncoder(file).Encode(ct)
+	if err != nil {
+		log.Fatal(err)
+	}
+}

go.mod

@@ -0,0 +1,9 @@
+module sarcastick
+
+go 1.21
+
+require golang.org/x/net v0.24.0
+
+require github.com/go-telegram/bot v1.2.1
+
+require github.com/sashabaranov/go-openai v1.20.5

go.sum

@@ -0,0 +1,6 @@
+github.com/go-telegram/bot v1.2.1 h1:FkrixLCtMtPUQAN4plXdNElbhkdXkx2p68YPXKBruDg=
+github.com/go-telegram/bot v1.2.1/go.mod h1:i2TRs7fXWIeaceF3z7KzsMt/he0TwkVC680mvdTFYeM=
+github.com/sashabaranov/go-openai v1.20.5 h1:Sab4nzBLtoyxm4jRqH9G9pkIh50WpBFacPTEpruPB6o=
+github.com/sashabaranov/go-openai v1.20.5/go.mod h1:lj5b/K+zjTSFxVLijLSTDZuP7adOgerWeFyZLUhAKRg=
+golang.org/x/net v0.24.0 h1:1PcaxkF854Fu3+lvBIx5SYn9wRlBzzcnHZSiaFFAb0w=
+golang.org/x/net v0.24.0/go.mod h1:2Q7sJY5mzlzWjKtYUEXSlBWCdyaioyXzRB2RtU8KVE8=

main.go

@@ -0,0 +1,43 @@
+package main
+
+import (
+	"context"
+	"log"
+	"os"
+	"regexp"
+	"sarcastick/pkg/news"
+	"sarcastick/pkg/openai"
+	"strings"
+
+	"github.com/go-telegram/bot"
+	"github.com/go-telegram/bot/models"
+)
+
+var detailsRegex = regexp.MustCompile("^/details [0-9]*$")
+var sarcasmRegex = regexp.MustCompile("^/sarcasm [0-9]*$")
+
+func main() {
+	b, err := bot.New(os.Getenv("TELEGRAM_BOT_KEY"))
+	if err != nil {
+		log.Fatal(err)
+	}
+	client := openai.NewClient()
+	b.RegisterHandler(bot.HandlerTypeMessageText, "/start", bot.MatchTypeExact, helpHandler)
+	b.RegisterHandler(bot.HandlerTypeMessageText, "/help", bot.MatchTypeExact, helpHandler)
+	b.RegisterHandler(bot.HandlerTypeMessageText, "/news", bot.MatchTypeExact, news.SummaryHandler)
+	b.RegisterHandlerRegexp(bot.HandlerTypeMessageText, detailsRegex, news.DetailsHandler)
+	b.RegisterHandlerRegexp(bot.HandlerTypeMessageText, sarcasmRegex, client.SarcasmHandler)
+	b.Start(context.Background())
+}
+
+func helpHandler(ctx context.Context, b *bot.Bot, update *models.Update) {
+	var sb strings.Builder
+	sb.WriteString("- use /news to get the main titles of the day (given with their ID)\n")
+	sb.WriteString("- use /details with the ID of the news to know more\n")
+	sb.WriteString("- use /sarcasm with the ID of the news to generate a sarcastic comment on the news\n")
+	sb.WriteString("- use /help to show this message \n")
+	b.SendMessage(ctx, &bot.SendMessageParams{
+		ChatID: update.Message.Chat.ID,
+		Text:   sb.String(),
+	})
+}

pkg/news/extract.go

@@ -0,0 +1,85 @@
+package news
+
+import (
+	"slices"
+	"strings"
+
+	"golang.org/x/net/html"
+)
+
+func FindNewsLink(n *html.Node) string {
+	if n.Type == html.ElementNode && n.Data == "a" {
+		for _, a := range n.Attr {
+			if a.Key == "href" && strings.Contains(a.Val, "czech-news-in-brief") {
+				return a.Val
+			}
+		}
+	}
+	for c := n.FirstChild; c != nil; c = c.NextSibling {
+		if l := FindNewsLink(c); l != "" {
+			return l
+		}
+	}
+	return ""
+}
+
+func ExtractContentWithTitle(n *html.Node) []News {
+	if n.Type == html.ElementNode && n.Data == "div" && slices.ContainsFunc(n.Attr, containContentClass) {
+		ct := []News{}
+		for c := n.FirstChild; c != nil; c = c.NextSibling {
+			if c.Type == html.ElementNode && c.Data == "div" && slices.ContainsFunc(c.Attr, containTitleClass) {
+				ct = append(ct, News{
+					Title:   extractText(c),
+					Content: extractContent(c.NextSibling),
+				})
+			}
+		}
+		if len(ct) > 0 {
+			return ct
+		}
+	}
+	for c := n.FirstChild; c != nil; c = c.NextSibling {
+		if ct := ExtractContentWithTitle(c); len(ct) > 0 {
+			return ct
+		}
+	}
+	return nil
+}
+
+func containTextWrapperClass(a html.Attribute) bool {
+	return a.Key == "class" && strings.Contains(a.Val, "widget text")
+}
+
+func containTitleClass(a html.Attribute) bool {
+	return a.Key == "class" && strings.Contains(a.Val, "headinglevel2")
+}
+
+func containContentClass(a html.Attribute) bool {
+	return a.Key == "class" && strings.Contains(a.Val, "content")
+}
+
+func extractContent(n *html.Node) string {
+	r := ""
+	for c := n; c != nil; c = c.NextSibling {
+		if c.Type == html.ElementNode && c.Data == "div" && slices.ContainsFunc(c.Attr, containTextWrapperClass) {
+			r += extractText(c.FirstChild)
+		}
+		if c.Type == html.ElementNode && c.Data == "div" && slices.ContainsFunc(c.Attr, containTitleClass) {
+			break
+		}
+	}
+	return r
+}
+
+func extractText(n *html.Node) string {
+	if n.Type == html.TextNode {
+		return n.Data
+	}
+	r := ""
+	if n.Type == html.ElementNode {
+		for c := n.FirstChild; c != nil; c = c.NextSibling {
+			r += extractText(c)
+		}
+	}
+	return r
+}

pkg/news/handlers.go

@@ -0,0 +1,49 @@
+package news
+
+import (
+	"context"
+	"fmt"
+	"strconv"
+	"strings"
+
+	"github.com/go-telegram/bot"
+	"github.com/go-telegram/bot/models"
+)
+
+func SummaryHandler(ctx context.Context, b *bot.Bot, update *models.Update) {
+	t, err := getSummary()
+	if err != nil {
+		fmt.Println(err.Error())
+		return
+	}
+	b.SendMessage(ctx, &bot.SendMessageParams{
+		ChatID: update.Message.Chat.ID,
+		Text:   t,
+	})
+}
+
+func DetailsHandler(ctx context.Context, b *bot.Bot, update *models.Update) {
+	id, err := ParseID(update)
+	if err != nil {
+		fmt.Println("Error", err.Error())
+		return
+	}
+	n, err := GetNewsById(id)
+	if err != nil {
+		fmt.Println("Error", err.Error())
+		return
+	}
+	var qb strings.Builder
+	qb.WriteString(n.Title + "\n\n")
+	qb.WriteString(n.Content)
+	b.SendMessage(ctx, &bot.SendMessageParams{
+		ChatID: update.Message.Chat.ID,
+		Text:   qb.String(),
+	})
+}
+
+func ParseID(update *models.Update) (int, error) {
+	text := update.Message.Text
+	id := strings.Split(text, " ")[1]
+	return strconv.Atoi(id)
+}

pkg/news/news.go

@@ -0,0 +1,51 @@
+package news
+
+import (
+	"encoding/json"
+	"fmt"
+	"os"
+	"strings"
+)
+
+type News struct {
+	Title   string
+	Content string
+}
+
+func getSummary() (string, error) {
+	allNews, err := getAllNews()
+	if err != nil {
+		return "", err
+	}
+	var sb strings.Builder
+	sb.WriteString("TODAY'S NEWS (based on expat.cz)\n")
+	for i, n := range allNews {
+		sb.WriteString(fmt.Sprintf("%d. %s\n", i+1, n.Title))
+	}
+	sb.WriteString("Use /details {news} or /sarcasm {news} for more")
+	return sb.String(), nil
+}
+
+func getAllNews() ([]*News, error) {
+	file, err := os.Open("news.json")
+	if err != nil {
+		return nil, err
+	}
+	var allNews []*News
+	err = json.NewDecoder(file).Decode(&allNews)
+	if err != nil {
+		return nil, err
+	}
+	return allNews, nil
+}
+
+func GetNewsById(id int) (*News, error) {
+	allNews, err := getAllNews()
+	if err != nil {
+		return nil, err
+	}
+	if len(allNews) < id {
+		return nil, nil
+	}
+	return allNews[id-1], nil
+}