check return of ReadString()

Recursively create ad dir including output dir. The output dir itself
is not being created separately anymore. That way, no directory will
be created if no ads could be downloaded.

Makefile

@@ -63,6 +63,7 @@ lint:
 
 lint-full:
 	golangci-lint run --enable-all --exclude-use-default --disable exhaustivestruct,exhaustruct,depguard,interfacer,deadcode,golint,structcheck,scopelint,varcheck,ifshort,maligned,nosnakecase,godot,funlen,gofumpt,cyclop,noctx,gochecknoglobals,paralleltest
+	gocritic check -enableAll *.go
 
 testfuzzy: clean
 	go test -fuzz ./... $(ARGS)

README-de.md

@@ -222,6 +222,49 @@ Sowie alle Bilder.
 Das Format kann man mit der Variable `template` in der Konfiguration
 ändern. Die `example.conf` enthält ein Beispiel für das Standard Template.
 
+## Verhalten des Tools
+
+Es gibt einige Dinge über das Verhalten von kleingebäck, über die Du
+Bescheid wissen solltest:
+
+- alle HTML Seiten und Bilder werden immer heruntergeladen
+- es wird ein (konfigurierbarer) Useragent verwendet
+- HTTP Cookies werden beachtet
+- bei Fehlern wird dreimal mit unterschiedlichem Abstand erneut
+  versucht
+- Bilder Downloads laufen parallelisiert mit leicht unterschiedlichen
+  zeitlichen Abständen ab
+- Gleich aussehende Bilder werden nicht überschrieben
+
+Der letzte Punkt muss genauer erläutert werden:
+
+Wenn man bei Kleinanzeigen.de eine Anzeige einstellt und Bilder
+postet, werden diese dort in ihrer Grösse reduziert (durch Kompression
+und Verkleinerung der Bilder usw.). Diese reduzierten Bilder werden
+dann von kleingebäck heruntergeladen. Falls Du Deine original Bilder
+behalten hast, kannst Du diese danach in das Backupverzeichnis
+kopieren. Bei einem erneuten kleingebäck-Lauf werden diese Bilder dann
+nicht überschrieben.
+
+Wir verwenden dafür einen Algorythmus namens [distance
+hashing](https://github.com/corona10/goimagehash). Dieser Algorithmus
+prüft die Ähnlichkeit von Bildern. Diese können in ihrer Auflösung,
+Kompression, Farbtiefe und vielem mehr manipuliert worden sein und
+trotzdem als das "gleiche Bild" erkannt werden (wohlgemerkt nicht "das
+selbe": die Dateien sind durchaus unterschiedlich!). Bis zu einer
+Distance von 5 überschreiben wir keine Bilder, weil wir dann davon
+ausgehen, dass das lokal Vorhandene das Original ist.
+
+Bitte beachte aber, dass dies KEIN Cachingmechanismus ist: die Bilder
+werden trotzdem immer alle heruntergeladen. Das muss so sein, da wir
+uns nicht die Dateinamen anschauen können, da kleinanzeigen.de diese
+nämlich zu Zahlen umbenennt. Und die Dateinamen können sich auch
+ändern, wenn der User in der Anzeige die Bilder umarrangiert hat.
+
+Du kannst dieses Verhalten mit der Option **--force** ausschalten. Du
+kannst ausserdem mit der Option **--ignoreerrors** auch alle Fehler
+ignorieren, die beim Bilderdownload auftreten könnten.
+
 ## Documentation
 
 Die Dokumentation kann man

README.md

@@ -207,6 +207,48 @@ variable. The supplied sample config contains the default template.
 
 All images will be stored in the same directory.
 
+## Tool Behavior
+
+There are a bunch of things you might want to know about the behavior
+of the kleingebäck tool:
+
+- all HTML pages and IMAGEs are always being downloaded
+- we use a (customizable) user agent
+- we respect HTTP cookies
+- in the case of an error, the tool does 3 retries, the time it waits
+  between tries is longer for each retry
+- image download is parallized using small time differences to look
+  more natural
+- same images are not being overwritten on subsequent download
+
+
+The latter needs to be elaborated a bit more:
+
+If you publish an ad on kleinanzeigen.de and post images, those images
+will be reduced in size by the site (by compressing and down sizing
+them). This reduced images will be downloaded by kleingebäck. However,
+you may still own the original images and may want to put them into
+that backup directory so that you have all things for one ad together.
+
+You can easily do that, because kleingebäck won't overwrite those
+original images. It uses something called a distance hash using
+[goimagehash](https://github.com/corona10/goimagehash). This
+algorithmus checks the similarity of images. If an image has been
+resized it is still very similar to the original one. We accept a
+maximum of a distance of 5, everything above leads to overwrite.
+
+This works with resizes, cropped and otherwise manipulated images as
+long as the image still shows the original contents good enough.
+
+Also note, that this is NOT a caching mechanism: the images will be
+downloaded anyway during each run. We also can't look at the file
+names because kleinanzeigen.de renames all images to numbers. And
+those might even change if the user re-arranges the images.
+
+You can override this behavior using the **--force** option. Another
+option, **--ignoreerrors**, can be used to ignore all kinds of image
+errors. 
+
 ## Documentation
 
 You can read the documentation [online](https://github.com/TLINDEN/kleingebaeck/blob/main/kleingebaeck.pod) or locally once you have installed kleingebaeck with: `kleingebaeck --manual`.

SECURITY.md

@@ -0,0 +1,17 @@
+# Security Policy
+
+## Supported Versions
+
+Only  the latest  release  is supported.  If you  find  an issue  (any
+issue!), please check with the latest release first.
+
+## Reporting a Vulnerability
+
+I don't agree with the "responsible disclosure" process most projects
+(and companies) work these days.
+
+So, if you find a vulnerability of any kind, please just open an
+[issue](https://github.com/TLINDEN/kleingebaeck/issues). Please add
+all details required to reproduce the vulnerability. You won't be chased.
+
+That's just all about it.

ad.go

@@ -73,7 +73,7 @@ func (ad *Ad) Incomplete() bool {
 }
 
 func (ad *Ad) CalculateExpire() {
-	if len(ad.Created) > 0 {
+	if ad.Created != "" {
 		ts, err := time.Parse("02.01.2006", ad.Created)
 		if err == nil {
 			ad.Expire = ts.AddDate(0, ExpireMonths, ExpireDays).Format("02.01.2006")

config.go

@@ -34,7 +34,7 @@ import (
 )
 
 const (
-	VERSION    string = "0.3.4"
+	VERSION    string = "0.3.6"
 	Baseuri    string = "https://www.kleinanzeigen.de"
 	Listuri    string = "/s-bestandsliste.html"
 	Defaultdir string = "."

fetch.go

@@ -52,7 +52,7 @@ func NewFetcher(conf *Config) (*Fetcher, error) {
 }
 
 func (f *Fetcher) Get(uri string) (io.ReadCloser, error) {
-	req, err := http.NewRequest(http.MethodGet, uri, nil)
+	req, err := http.NewRequest(http.MethodGet, uri, http.NoBody)
 	if err != nil {
 		return nil, fmt.Errorf("failed to create a new HTTP request obj: %w", err)
 	}

go.mod

@@ -24,6 +24,7 @@ require (
 	github.com/corona10/goimagehash v1.1.0 // indirect
 	github.com/fatih/color v1.16.0 // indirect
 	github.com/fsnotify/fsnotify v1.6.0 // indirect
+	github.com/inconshreveable/mousetrap v1.1.0 // indirect
 	github.com/knadh/koanf/maps v0.1.1 // indirect
 	github.com/mattn/go-colorable v0.1.13 // indirect
 	github.com/mitchellh/copystructure v1.2.0 // indirect

go.sum

@@ -15,6 +15,8 @@ github.com/fatih/color v1.16.0 h1:zmkK9Ngbjj+K0yRhTVONQh1p/HknKYSlNT+vZCzyokM=
 github.com/fatih/color v1.16.0/go.mod h1:fL2Sau1YI5c0pdGEVCbKQbLXB6edEj1ZgiY4NijnWvE=
 github.com/fsnotify/fsnotify v1.6.0 h1:n+5WquG0fcWoWp6xPWfHdbskMCQaFnG6PfBrh1Ky4HY=
 github.com/fsnotify/fsnotify v1.6.0/go.mod h1:sl3t1tCWJFWoRz9R8WJCbQihKKwmorjAbSClcnxKAGw=
+github.com/inconshreveable/mousetrap v1.1.0 h1:wN+x4NVGpMsO7ErUn/mUI3vEoE6Jt13X2s0bqwp9tc8=
+github.com/inconshreveable/mousetrap v1.1.0/go.mod h1:vpF70FUmC8bwa3OWnCshd2FqLfsEA9PFc4w1p2J65bw=
 github.com/jarcoal/httpmock v1.3.1 h1:iUx3whfZWVf3jT01hQTO/Eo5sAYtB2/rqaUuOtpInww=
 github.com/jarcoal/httpmock v1.3.1/go.mod h1:3yb8rc4BI7TCBhFY8ng0gjuLKJNquuDNiPaZjnENuYg=
 github.com/knadh/koanf/maps v0.1.1 h1:G5TjmUh2D7G2YWf5SQQqSiHRJEjaicvU0KpypqB3NIs=

image.go

@@ -49,7 +49,7 @@ func (img *Image) LogValue() slog.Value {
 // holds all images of an ad
 type Cache []*goimagehash.ImageHash
 
-func NewImage(buf *bytes.Reader, filename string, uri string) *Image {
+func NewImage(buf *bytes.Reader, filename, uri string) *Image {
 	img := &Image{
 		Filename: filename,
 		URI:      uri,
@@ -134,7 +134,7 @@ func ReadImages(addir string, dont bool) (Cache, error) {
 			reader := bytes.NewReader(data.Bytes())
 
 			img := NewImage(reader, filename, "")
-			if err = img.CalcHash(); err != nil {
+			if err := img.CalcHash(); err != nil {
 				return nil, err
 			}
 

main.go

@@ -18,13 +18,16 @@ along with this program. If not, see <http://www.gnu.org/licenses/>.
 package main
 
 import (
+	"bufio"
 	"errors"
 	"fmt"
 	"io"
 	"log/slog"
 	"os"
+	"runtime"
 	"runtime/debug"
 
+	"github.com/inconshreveable/mousetrap"
 	"github.com/lmittmann/tint"
 	"github.com/tlinden/yadu"
 )
@@ -35,6 +38,25 @@ func main() {
 	os.Exit(Main(os.Stdout))
 }
 
+func init() {
+	// if we're running on Windows  AND if the user double clicked the
+	// exe  file from explorer, we  tell them and then  wait until any
+	// key has been hit, which  will make the cmd window disappear and
+	// thus give the user time to read it.
+	if runtime.GOOS == "windows" {
+		if mousetrap.StartedByExplorer() {
+			fmt.Println("Do no double click kleingebaeck.exe!")
+			fmt.Println("Please open a command shell and run it from there.")
+			fmt.Println()
+			fmt.Print("Press any key to quit: ")
+			_, err := bufio.NewReader(os.Stdin).ReadString('\n')
+			if err != nil {
+				panic(err)
+			}
+		}
+	}
+}
+
 func Main(output io.Writer) int {
 	logLevel := &slog.LevelVar{}
 	opts := &tint.Options{
@@ -116,11 +138,6 @@ func Main(output io.Writer) int {
 	if err != nil {
 		return Die(err)
 	}
-
-	err = Mkdir(outdir)
-	if err != nil {
-		return Die(err)
-	}
 	conf.Outdir = outdir
 
 	// used for all HTTP requests

main_test.go

@@ -334,14 +334,14 @@ type Adsource struct {
 }
 
 // Render a HTML template for an adlisting or an ad
-func GetTemplate(adconfigs []AdConfig, adconfig AdConfig, htmltemplate string) string {
+func GetTemplate(adconfigs []AdConfig, adconfig *AdConfig, htmltemplate string) string {
 	tmpl, err := tpl.New("template").Parse(htmltemplate)
 	if err != nil {
 		panic(err)
 	}
 
 	var out bytes.Buffer
-	if len(adconfig.ID) == 0 {
+	if adconfig.ID == "" {
 		err = tmpl.Execute(&out, adconfigs)
 	} else {
 		err = tmpl.Execute(&out, adconfig)
@@ -376,15 +376,15 @@ func InitValidSources() []Adsource {
 	ads := []Adsource{
 		{
 			uri:     fmt.Sprintf("%s%s?userId=1", Baseuri, Listuri),
-			content: GetTemplate(list1, empty, LISTTPL),
+			content: GetTemplate(list1, &empty, LISTTPL),
 		},
 		{
 			uri:     fmt.Sprintf("%s%s?userId=1&pageNum=2", Baseuri, Listuri),
-			content: GetTemplate(list2, empty, LISTTPL),
+			content: GetTemplate(list2, &empty, LISTTPL),
 		},
 		{
 			uri:     fmt.Sprintf("%s%s?userId=1&pageNum=3", Baseuri, Listuri),
-			content: GetTemplate(list3, empty, LISTTPL),
+			content: GetTemplate(list3, &empty, LISTTPL),
 		},
 	}
 
@@ -392,7 +392,7 @@ func InitValidSources() []Adsource {
 	for _, ad := range adsrc {
 		ads = append(ads, Adsource{
 			uri:     fmt.Sprintf("%s/s-anzeige/%s/%s", Baseuri, ad.Slug, ad.ID),
-			content: GetTemplate(nil, ad, ADTPL),
+			content: GetTemplate(nil, &ad, ADTPL),
 		})
 	}
 
@@ -405,28 +405,28 @@ func InitInvalidSources() []Adsource {
 		{
 			// valid ad page but without content
 			uri:     fmt.Sprintf("%s/s-anzeige/empty/1", Baseuri),
-			content: GetTemplate(nil, empty, EMPTYPAGE),
+			content: GetTemplate(nil, &empty, EMPTYPAGE),
 		},
 		{
 			// some random foreign webpage
 			uri:     INVALIDURI,
-			content: GetTemplate(nil, empty, "<html>foo</html>"),
+			content: GetTemplate(nil, &empty, "<html>foo</html>"),
 		},
 		{
 			// some invalid page path
 			uri:     fmt.Sprintf("%s/anzeige/name/1", Baseuri),
-			content: GetTemplate(nil, empty, "<html></html>"),
+			content: GetTemplate(nil, &empty, "<html></html>"),
 		},
 		{
 			// some none-ad page
 			uri:     fmt.Sprintf("%s/anzeige/name/1/foo/bar", Baseuri),
-			content: GetTemplate(nil, empty, "<html>HTTP 404: /eine-anzeige/ does not exist!</html>"),
+			content: GetTemplate(nil, &empty, "<html>HTTP 404: /eine-anzeige/ does not exist!</html>"),
 			status:  404,
 		},
 		{
 			// valid ad page but 503
 			uri:     fmt.Sprintf("%s/s-anzeige/503/1", Baseuri),
-			content: GetTemplate(nil, empty, "<html>HTTP 503: service unavailable</html>"),
+			content: GetTemplate(nil, &empty, "<html>HTTP 503: service unavailable</html>"),
 			status:  503,
 		},
 	}
@@ -465,7 +465,7 @@ func SetIntercept(ads []Adsource) {
 	}
 }
 
-func VerifyAd(advertisement AdConfig) error {
+func VerifyAd(advertisement *AdConfig) error {
 	body := advertisement.Title + advertisement.Price + advertisement.ID + "Kleinanzeigen => " +
 		advertisement.Category + advertisement.Condition + advertisement.Created
 
@@ -525,7 +525,7 @@ func TestMain(t *testing.T) {
 
 	// verify if downloaded ads match
 	for _, ad := range adsrc {
-		if err := VerifyAd(ad); err != nil {
+		if err := VerifyAd(&ad); err != nil {
 			t.Errorf(err.Error())
 		}
 	}

mkrel.sh

@@ -22,7 +22,12 @@ freebsd/amd64
 linux/amd64
 netbsd/amd64
 openbsd/amd64
-windows/amd64"
+windows/amd64
+freebsd/arm64
+linux/arm64
+netbsd/arm64
+openbsd/arm64
+windows/arm64"
 
 tool="$1"
 version="$2"

scrape.go

@@ -126,21 +126,32 @@ func ScrapeAd(fetch *Fetcher, uri string) error {
 
 	advertisement.CalculateExpire()
 
-	proceed := CheckAdVisited(fetch.Config, advertisement.Slug)
+	// prepare ad dir name
+	addir, err := AdDirName(fetch.Config, advertisement)
+	if err != nil {
+		return err
+	}
+
+	proceed := CheckAdVisited(fetch.Config, addir)
 	if !proceed {
 		return nil
 	}
 
 	// write listing
-	addir, err := WriteAd(fetch.Config, advertisement)
+	err = WriteAd(fetch.Config, advertisement, addir)
 	if err != nil {
 		return err
 	}
 
+	// tell the user
 	slog.Debug("extracted ad listing", "ad", advertisement)
 
+	// stats
 	fetch.Config.IncrAds()
 
+	// register for later checks
+	DirsVisited[addir] = 1
+
 	return ScrapeImages(fetch, advertisement, addir)
 }
 

store.go

@@ -72,19 +72,13 @@ func AdDirName(conf *Config, advertisement *Ad) (string, error) {
 	return buf.String(), nil
 }
 
-func WriteAd(conf *Config, advertisement *Ad) (string, error) {
+func WriteAd(conf *Config, advertisement *Ad, addir string) error {
-	// prepare ad dir name
-	addir, err := AdDirName(conf, advertisement)
-	if err != nil {
-		return "", err
-	}
-
 	// prepare output dir
 	dir := filepath.Join(conf.Outdir, addir)
 
-	err = Mkdir(dir)
+	err := Mkdir(dir)
 	if err != nil {
-		return "", err
+		return err
 	}
 
 	// write ad file
@@ -92,7 +86,7 @@ func WriteAd(conf *Config, advertisement *Ad) (string, error) {
 
 	listingfd, err := os.Create(listingfile)
 	if err != nil {
-		return "", fmt.Errorf("failed to create Adlisting.txt: %w", err)
+		return fmt.Errorf("failed to create Adlisting.txt: %w", err)
 	}
 	defer listingfd.Close()
 
@@ -104,17 +98,17 @@ func WriteAd(conf *Config, advertisement *Ad) (string, error) {
 
 	tmpl, err := tpl.New("adlisting").Parse(conf.Template)
 	if err != nil {
-		return "", fmt.Errorf("failed to parse adlisting template: %w", err)
+		return fmt.Errorf("failed to parse adlisting template: %w", err)
 	}
 
 	err = tmpl.Execute(listingfd, advertisement)
 	if err != nil {
-		return "", fmt.Errorf("failed to execute adlisting template: %w", err)
+		return fmt.Errorf("failed to execute adlisting template: %w", err)
 	}
 
 	slog.Info("wrote ad listing", "listingfile", listingfile)
 
-	return addir, nil
+	return nil
 }
 
 func WriteImage(filename string, reader *bytes.Reader) error {
@@ -176,9 +170,6 @@ func CheckAdVisited(conf *Config, adname string) bool {
 		return false
 	}
 
-	// register
-	DirsVisited[adname] = 1
-
 	// overwrite
 	return true
 }

util.go

@@ -32,7 +32,7 @@ import (
 
 func Mkdir(dir string) error {
 	if _, err := os.Stat(dir); errors.Is(err, os.ErrNotExist) {
-		err := os.Mkdir(dir, os.ModePerm)
+		err := os.MkdirAll(dir, os.ModePerm)
 		if err != nil {
 			return fmt.Errorf("failed to create directory %s: %w", dir, err)
 		}