better error handling

Reviewed-on: rannes.dev/sw-jobs-go#1

.gitea/workflows/build.yaml

@@ -0,0 +1,31 @@
+name: Build and Push Docker Image
+
+on:
+  push:
+    branches: [ main ]
+  pull_request:
+    branches: [ main ]
+
+jobs:
+  build-and-push:
+    runs-on: ubuntu-latest
+    permissions:
+      contents: read
+      packages: write
+
+    steps:
+    - uses: actions/checkout@v4
+
+    - name: Log in to Gitea Container Registry
+      uses: docker/login-action@v3
+      with:
+        registry: ${{ secrets.REGISTRY }}
+        username: ${{ secrets.USER }}
+        password: ${{ secrets.TOKEN }}
+
+    - name: Build and push Docker image
+      uses: docker/build-push-action@v5
+      with:
+        context: .
+        push: true
+        tags: ${{ secrets.REGISTRY }}/rannes.dev/sw-jobs-scraper:latest

.gitignore

@@ -1 +1,4 @@
-/lambda-package
+/thehub_cache
+/thehub.json
+/itjobbank_cache
+/it-jobbank.json

Dockerfile

@@ -0,0 +1,11 @@
+FROM golang:1.23 AS builder
+WORKDIR /app
+COPY go.mod go.sum ./
+RUN go mod download
+COPY *.go ./
+RUN CGO_ENABLED=0 GOOS=linux go build -o job-scraper
+
+FROM alpine:3.18
+WORKDIR /app
+COPY --from=builder /app/job-scraper .
+CMD ["./job-scraper"]

build.sh

@@ -1,31 +0,0 @@
-#!/bin/bash
-
-# Set variables
-PACKAGE_DIR="./lambda-package"
-BUILD_FILE="bootstrap"
-ZIP_FILE="lambda-deployment.zip"
-SOURCE_FILE="main.go"
-
-# Delete the content of the lambda-package directory
-rm -rf $PACKAGE_DIR/*
-echo "Deleted the content of $PACKAGE_DIR"
-
-# Set environment variables and build the Go project
-GOOS=linux GOARCH=arm64 go build -o $BUILD_FILE -tags lambda.norpc $SOURCE_FILE
-echo "Built the Go project with GOOS=linux and GOARCH=arm64"
-
-# Move the build file to the lambda-package directory
-mv $BUILD_FILE $PACKAGE_DIR/
-echo "Moved the build file to $PACKAGE_DIR"
-
-# Change directory to lambda-package
-cd $PACKAGE_DIR
-
-# Zip the contents of lambda-package into lambda-deployment.zip
-zip -r $ZIP_FILE *
-echo "Zipped the contents of $PACKAGE_DIR into $ZIP_FILE"
-
-# Return to the original directory
-cd -
-
-echo "Script completed successfully"

go.mod

@@ -8,7 +8,6 @@ require (
 	github.com/antchfx/htmlquery v1.3.1 // indirect
 	github.com/antchfx/xmlquery v1.4.0 // indirect
 	github.com/antchfx/xpath v1.3.0 // indirect
-	github.com/aws/aws-lambda-go v1.47.0 // indirect
 	github.com/gobwas/glob v0.2.3 // indirect
 	github.com/gocolly/colly v1.2.0 // indirect
 	github.com/golang/groupcache v0.0.0-20210331224755-41bb18bfe9da // indirect

go.sum

@@ -8,8 +8,6 @@ github.com/antchfx/xmlquery v1.4.0 h1:xg2HkfcRK2TeTbdb0m1jxCYnvsPaGY/oeZWTGqX/0h
 github.com/antchfx/xmlquery v1.4.0/go.mod h1:Ax2aeaeDjfIw3CwXKDQ0GkwZ6QlxoChlIBP+mGnDFjI=
 github.com/antchfx/xpath v1.3.0 h1:nTMlzGAK3IJ0bPpME2urTuFL76o4A96iYvoKFHRXJgc=
 github.com/antchfx/xpath v1.3.0/go.mod h1:i54GszH55fYfBmoZXapTHN8T8tkcHfRgLyVwwqzXNcs=
-github.com/aws/aws-lambda-go v1.47.0 h1:0H8s0vumYx/YKs4sE7YM0ktwL2eWse+kfopsRI1sXVI=
-github.com/aws/aws-lambda-go v1.47.0/go.mod h1:dpMpZgvWx5vuQJfBt0zqBha60q7Dd7RfgJv23DymV8A=
 github.com/davecgh/go-spew v1.1.0/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/gobwas/glob v0.2.3 h1:A4xDbljILXROh+kObIiy5kIaPYD8e96x1tgBhUI5J+Y=
 github.com/gobwas/glob v0.2.3/go.mod h1:d3Ez4x06l9bZtSvzIay5+Yzi0fmZzPgnTbPcKjJAkT8=

main.go

@@ -1,12 +1,13 @@
 package main
 
 import (
-	"context"
+	"encoding/json"
 	"fmt"
+	"log"
+	"os"
 	"strings"
 	"time"
 
-	"github.com/aws/aws-lambda-go/lambda"
 	"github.com/gocolly/colly"
 )
 
@@ -19,6 +20,8 @@ type job struct {
 	Description string `json:"description"`
 	Link        string `json:"link"`
 	Skills      skills `json:"skills"`
+	Scraped     string `json:"scraped"`
+	Source      string `json:"source"`
 }
 
 type skills struct {
@@ -28,53 +31,85 @@ type skills struct {
 	Svelte     bool `json:"svelte"`
 	Nextjs     bool `json:"nextjs"`
 	Typescript bool `json:"typescript"`
+	Tailwind   bool `json:"tailwind"`
 }
 
-var (
-	jobs      []job
-	lastFetch time.Time
-	cacheTTL  = time.Minute * 5
-	jobLimit  = 20
-)
+// Utility functions
 
+// Checks if a string contains any of the given keywords
 func skillChecker(description string) skills {
 	return skills{
 		React:      strings.Contains(description, "React"),
 		Python:     strings.Contains(description, "Python"),
-		Golang:     strings.Contains(description, "Go"),
+		Golang:     strings.Contains(description, "Golang"),
 		Svelte:     strings.Contains(description, "Svelte"),
 		Nextjs:     strings.Contains(description, "Next.js"),
 		Typescript: strings.Contains(description, "TypeScript"),
+		Tailwind:   strings.Contains(description, "Tailwind"),
 	}
 }
 
-func fetchData() error {
+// Converts job struct to json
+func jobsToJson(file *os.File, jobs []job, fName string) {
+	// Encode jobs slice to JSON
+	encoder := json.NewEncoder(file)
+	encoder.SetIndent("", "  ") // Pretty-print with indentation
+	if err := encoder.Encode(jobs); err != nil {
+		log.Fatalf("Cannot write to file %q: %s", fName, err)
+	}
 
-	baseUrl := "https://thehub.io"
-	// Instantiate default collector
-	c := colly.NewCollector(
-		// visit only the hub
-		colly.AllowedDomains("www.thehub.io", "thehub.io"),
+	fmt.Println("Job details successfully written to", fName)
+}
 
-		// Cache responses to prevent multiple requests
-		colly.CacheDir("./tmp"),
-	)
-
-	// Slice of excluded words in the job titles
-	excluded := []string{"senior", "lead"}
-	// Instantiate a new collector to visit the job details page
-	detailsCollector := c.Clone()
-
-	// Limit the number of jobs to fetch
-	jobCount := 0
-
-	// On every <div> element with class "card__content attribute call callback
-	c.OnHTML("div[class=card__content]", func(e *colly.HTMLElement) {
-		// Return if the job limit has been reached
-		if jobCount >= jobLimit {
+func checkIfPaid(description string) {
+	for _, keyword := range unpaidKeywords {
+		if strings.Contains(strings.ToLower(description), keyword) {
 			return
 		}
-		// Get the title and ensure it doesn't contain any excluded words
+	}
+}
+
+func checkIfStudent(description string) string {
+	for _, keyword := range studentKeywords {
+		if strings.Contains(strings.ToLower(description), keyword) {
+			return "student"
+		}
+	}
+	return "full time"
+}
+
+// Slice to store job details
+var (
+	excluded        = []string{"senior", "lead", "founder", "cto", "vp of", "erfaren", "arkitekt", "architect", "manager", "ulønnet", "unpaid", "praktik", "cyber", "leder", "sikkerhed", "supporter", "sr."}
+	unpaidKeywords  = []string{"unpaid", "praktik", "ulønnet"}
+	studentKeywords = []string{"studerende", "studenter", "student", "medhjælper"}
+)
+
+func scrapeHub() {
+	var (
+		jobs         []job
+		jobCount     int
+		fName        = "/app/data/thehub.json"
+		maxJobs      = 20
+		baseUrl      = "https://thehub.io"
+		searchString = "https://thehub.io/jobs?roles=frontenddeveloper&roles=fullstackdeveloper&roles=backenddeveloper&roles=devops&paid=true&countryCode=DK&sorting=newJobs"
+	)
+
+	// Create file after scraping is complete
+	c := colly.NewCollector(
+		colly.AllowedDomains("www.thehub.io", "thehub.io"),
+	)
+
+	detailsCollector := colly.NewCollector(
+		colly.AllowedDomains("www.thehub.io", "thehub.io"),
+		colly.CacheDir("/app/data/thehub_cache"),
+	)
+
+	c.OnHTML("div[class=card__content]", func(e *colly.HTMLElement) {
+		if jobCount >= maxJobs {
+			return
+		}
+
 		title := e.ChildText("span.card-job-find-list__position")
 		for _, excludedWord := range excluded {
 			if strings.Contains(strings.ToLower(title), excludedWord) {
@@ -91,66 +126,189 @@ func fetchData() error {
 		fmt.Println("Visiting", r.URL.String())
 	})
 
-	detailsCollector.OnHTML("div.view-job-details", func(e *colly.HTMLElement) {
-		if jobCount >= jobLimit {
+	detailsCollector.OnHTML("div[class='view-job-details']", func(e *colly.HTMLElement) {
+		if jobCount >= maxJobs {
 			return
 		}
-		// Get logo and trim the url
 
 		logo := e.ChildAttr("div.media-item__image", "style")
 		cutLeft := "background-image:url("
 		cutRight := ");"
 		trimmedLogo := strings.Trim(logo, cutLeft+cutRight)
 
-		// Get company name
+		descriptionHTML, err := e.DOM.Find("content.text-block__content > span").Html()
+		if err != nil {
+			log.Printf("Error getting HTML of description: %s", err)
+			return
+		}
+
 		jobDetails := job{
 			Title:       e.ChildText("h2[class=view-job-details__title]"),
 			Logo:        trimmedLogo,
 			Company:     e.ChildText(".bullet-inline-list > a:first-child"),
 			Location:    e.ChildText(".bullet-inline-list > a:nth-child(2)"),
 			Type:        e.ChildText(".bullet-inline-list > a:nth-child(3)"),
-			Description: e.ChildText("content.text-block__content > span"),
+			Description: descriptionHTML,
 			Link:        e.Request.URL.String(),
 			Skills:      skillChecker(e.ChildText("content.text-block__content > span")),
+			Scraped:     time.Now().String(),
+			Source:      baseUrl,
 		}
 		jobs = append(jobs, jobDetails)
 		jobCount++
+		fmt.Printf("Scraped job %d from TheHub\n", jobCount)
 	})
-	// Handle pagination
+
 	c.OnHTML("a.page-link", func(e *colly.HTMLElement) {
+		if jobCount >= maxJobs {
+			return
+		}
 		nextPage := e.Attr("href")
 		if nextPage != "" {
 			fullNextPage := baseUrl + nextPage
-			fmt.Println("Visiting next page:", fullNextPage)
 			e.Request.Visit(fullNextPage)
 		}
 	})
-	// Visit the initial URL to start scraping
-	err := c.Visit("https://thehub.io/jobs?roles=frontenddeveloper&roles=fullstackdeveloper&roles=backenddeveloper&search=developer&paid=true&countryCode=DK&sorting=newJobs")
+
+	// Add error handling for the initial visit
+	err := c.Visit(searchString)
 	if err != nil {
-		return err
+		log.Printf("Error visiting TheHub: %s", err)
+		return
+	}
+
+	// Wait for all collectors to finish
+	c.Wait()
+	detailsCollector.Wait()
+
+	// Write jobs to file after scraping is complete
+	if len(jobs) > 0 {
+		file, err := os.Create(fName)
+		if err != nil {
+			log.Printf("Cannot create file %q: %s", fName, err)
+			return
+		}
+		defer file.Close()
+
+		jobsToJson(file, jobs, fName)
+		fmt.Printf("Successfully scraped %d jobs from TheHub\n", len(jobs))
+	} else {
+		log.Println("No jobs were scraped from TheHub")
 	}
-	return nil
 }
 
-func handler(ctx context.Context) ([]job, error) {
-	// Check if cache is valid
-	if time.Since(lastFetch) < cacheTTL && len(jobs) > 0 {
-		return jobs, nil
-	}
+func scrapeItJobBank() {
+	var (
+		jobs         []job
+		jobCount     int
+		fName        = "/app/data/it-jobbank.json"
+		maxJobs      = 20
+		baseUrl      = "https://www.it-jobbank.dk"
+		searchString = "https://www.it-jobbank.dk/jobsoegning/udvikling"
+	)
 
-	// Fetch new data
-	err := fetchData()
+	c := colly.NewCollector(
+		colly.AllowedDomains("www.it-jobbank.dk", "it-jobbank.dk"),
+	)
+
+	detailsCollector := colly.NewCollector(
+		colly.AllowedDomains("www.it-jobbank.dk", "it-jobbank.dk"),
+		colly.CacheDir("/app/data/itjobbank_cache"),
+	)
+
+	c.OnHTML("div[class=result]", func(e *colly.HTMLElement) {
+		if jobCount >= maxJobs {
+			return
+		}
+
+		title := e.ChildText("h3.job-title > a")
+		for _, excludedWord := range excluded {
+			if strings.Contains(strings.ToLower(title), excludedWord) {
+				return
+			}
+		}
+		fullLink := e.ChildAttr("h3.job-title > a", "href")
+
+		detailsCollector.Visit(fullLink)
+	})
+
+	detailsCollector.OnRequest(func(r *colly.Request) {
+		fmt.Println("Visiting", r.URL.String())
+	})
+
+	detailsCollector.OnHTML("section > div", func(e *colly.HTMLElement) {
+		if jobCount >= maxJobs {
+			return
+		}
+
+		descriptionHTML, err := e.DOM.Find("div[id=job_ad]").Html()
+		if err != nil {
+			log.Printf("Error getting HTML of description: %s", err)
+			return
+		}
+
+		checkIfPaid(descriptionHTML)
+
+		title := e.ChildText("h1.title")
+		if title == "" {
+			title = e.ChildText("h1[id=jobtitle]")
+		}
+
+		jobDetails := job{
+			Title:       title,
+			Logo:        baseUrl + e.ChildAttr("div.company-logo > img", "src"),
+			Company:     e.ChildText("p.published"),
+			Location:    e.ChildText("div.job-location > p.caption"),
+			Type:        checkIfStudent(descriptionHTML),
+			Description: descriptionHTML,
+			Link:        e.Request.URL.String(),
+			Skills:      skillChecker(descriptionHTML),
+			Scraped:     time.Now().String(),
+			Source:      baseUrl,
+		}
+		jobs = append(jobs, jobDetails)
+		jobCount++
+		fmt.Printf("Scraped job %d from IT JobBank\n", jobCount)
+	})
+
+	c.OnHTML("a.page-link", func(e *colly.HTMLElement) {
+		if jobCount >= maxJobs {
+			return
+		}
+		nextPage := e.Attr("href")
+		if nextPage != "" {
+			e.Request.Visit(nextPage)
+		}
+	})
+
+	// Add error handling for the initial visit
+	err := c.Visit(searchString)
 	if err != nil {
-		return nil, err
+		log.Printf("Error visiting IT JobBank: %s", err)
+		return
 	}
 
-	// Update cache timestamp
-	lastFetch = time.Now()
+	// Wait for all collectors to finish
+	c.Wait()
+	detailsCollector.Wait()
 
-	return jobs, nil
+	// Write jobs to file after scraping is complete
+	if len(jobs) > 0 {
+		file, err := os.Create(fName)
+		if err != nil {
+			log.Printf("Cannot create file %q: %s", fName, err)
+			return
+		}
+		defer file.Close()
+
+		jobsToJson(file, jobs, fName)
+		fmt.Printf("Successfully scraped %d jobs from IT JobBank\n", len(jobs))
+	} else {
+		log.Println("No jobs were scraped from IT JobBank")
+	}
 }
 
 func main() {
-	lambda.Start(handler)
+	scrapeHub()
+	scrapeItJobBank()
 }

readme.md

@@ -1,13 +1,18 @@
-# The Hub Scraper
+# IT jobs scraper
 
-deprecated as lambda was a bad solution for this, without setting up dynamodb, api etc. This will go live in a ec2 so it can write to local storage instead of running on demand.
+This is a simple scraper that extracts job details from the [The Hub](https://thehub.io) website and itjobbank.
 
-Go is fast but free tier lambda is not and I am not yet a smart man.
 
-<del>This is  a simple scraper that extracts job details from the [The Hub](https://thehub.io) website.</del>
+## Filtering
 
-<del>It's a fork of the original [The Hub Scraper](https://gitea.rannes.dev/rannes.dev/sw-jobs-go) by [Rannes](https://gitea.rannes.dev/rannes.dev).</del>
+The scraper filters out a list of keywords like senior, architect etc. as I wrote it for entry and mid level roles. It also filters out unpaid form the hub, and keyword based from itjobbank.
 
-<del>## Usage</del>
+## Usage
 
-<del>To run the scraper zip it deploy it to AWS Lambda and then call the function. </del>
+To run the scraper, simply execute the following command:
+
+```bash
+go run scraper.go
+```
+
+The scraper will create a `thehub.json` and `itjobbank.json` file in the current directory, which contains a list of job details in JSON format. It caches the pages, so very light on resources and requests.

run-scrapers.sh

@@ -0,0 +1,3 @@
+#!/bin/bash
+cd /home/admin/sw-jobs-go
+go run go-scraper >> scraper.log 2>&1