From 062b36fe5725d1267c66db2e506b4131d78ce772 Mon Sep 17 00:00:00 2001
From: Philipp Tanlak <philipp.tanlak@gmail.com>
Date: Fri, 11 Aug 2023 18:31:20 +0200
Subject: simplify project structure

---
 cmd/flyscrape/main.go | 31 +++++++++++--------
 cmd/watch/main.go     | 83 +++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 102 insertions(+), 12 deletions(-)
 create mode 100644 cmd/watch/main.go

(limited to 'cmd')

diff --git a/cmd/flyscrape/main.go b/cmd/flyscrape/main.go
index fb31056..85e40a9 100644
--- a/cmd/flyscrape/main.go
+++ b/cmd/flyscrape/main.go
@@ -7,9 +7,9 @@ import (
 	"io"
 	"net/http"
 	"os"
+	"time"
 
-	"flyscrape/flyscrape"
-	"flyscrape/js"
+	"flyscrape"
 )
 
 func main() {
@@ -17,14 +17,20 @@ func main() {
 		exit("Please provide a file to run.")
 	}
 
-	opts, scrape, err := js.Compile(os.Args[1])
+	src, err := os.ReadFile(os.Args[1])
+	if err != nil {
+		exit(fmt.Sprintf("Error reading file: %v", err))
+	}
+
+	opts, scrape, err := flyscrape.Compile(string(src))
 	if err != nil {
 		exit(fmt.Sprintf("Error compiling JavaScript file: %v", err))
 	}
 
-	svc := flyscrape.Service{
-		ScrapeOptions: *opts,
+	svc := flyscrape.Scraper{
+		ScrapeOptions: opts,
 		ScrapeFunc:    scrape,
+		Concurrency:   5,
 		FetchFunc: func(url string) (string, error) {
 			resp, err := http.Get(url)
 			if err != nil {
@@ -39,14 +45,15 @@ func main() {
 			return string(data), nil
 		},
 	}
-	results := svc.Scrape()
-	if err != nil {
-	}
-	fmt.Printf("%T\n", results[0])
 
-	data, _ := json.MarshalIndent(results, "", "  ")
-	fmt.Println(string(data))
-	return
+	count := 0
+	start := time.Now()
+	for result := range svc.Scrape() {
+		data, _ := json.MarshalIndent(result, "", "  ")
+		fmt.Println(string(data))
+		count++
+	}
+	fmt.Printf("Scraped %d websites in %v\n", count, time.Since(start))
 }
 
 func exit(msg string) {
diff --git a/cmd/watch/main.go b/cmd/watch/main.go
new file mode 100644
index 0000000..5065d8b
--- /dev/null
+++ b/cmd/watch/main.go
@@ -0,0 +1,83 @@
+package main
+
+import (
+	"encoding/json"
+	"fmt"
+	"io"
+	"log"
+	"net/http"
+	"os"
+
+	"flyscrape"
+
+	"github.com/cornelk/hashmap"
+	"github.com/inancgumus/screen"
+)
+
+func main() {
+	if len(os.Args) != 2 {
+		fmt.Println("Please provide a file to run.")
+		os.Exit(1)
+	}
+
+	cache := hashmap.New[string, string]()
+
+	err := flyscrape.Watch(os.Args[1], func(s string) error {
+		opts, scrape, err := flyscrape.Compile(s)
+		if err == nil {
+			run(cache, opts, scrape)
+		}
+		return nil
+	})
+	if err != nil {
+		log.Fatal(err)
+	}
+}
+
+func run(cache *hashmap.Map[string, string], opts flyscrape.ScrapeOptions, fn flyscrape.ScrapeFunc) {
+	opts.Depth = 0
+
+	svc := flyscrape.Scraper{
+		Concurrency:   20,
+		ScrapeOptions: opts,
+		ScrapeFunc:    fn,
+		FetchFunc: func(url string) (string, error) {
+			if html, ok := cache.Get(url); ok {
+				return html, nil
+			}
+			html, err := fetch(url)
+			if err != nil {
+				return "", err
+			}
+			cache.Set(url, html)
+			return html, nil
+		},
+	}
+
+	result := <-svc.Scrape()
+	if result.Error != nil {
+		fmt.Println(result.Error)
+	}
+
+	screen.Clear()
+	screen.MoveTopLeft()
+
+	enc := json.NewEncoder(os.Stdout)
+	enc.SetEscapeHTML(false)
+	enc.SetIndent("", "   ")
+	enc.Encode(result)
+}
+
+func fetch(url string) (string, error) {
+	resp, err := http.Get(url)
+	if err != nil {
+		return "", err
+	}
+	defer resp.Body.Close()
+	body, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return "", err
+	}
+
+	return string(body), nil
+}
-- 
cgit v1.2.3