Keywords: Don't split words with a dash

Signed-off-by: Michael Mayer <michael@liquidbytes.net>

internal/entity/location.go

@@ -87,14 +87,11 @@ func (m *Location) Find(db *gorm.DB, api string) error {
 
 // Keywords computes keyword based on a Location
 func (m *Location) Keywords() (result []string) {
-	result = append(result, txt.Keywords(m.City())...)
+	result = append(result, txt.Keywords(txt.ReplaceSpaces(m.City(), "-"))...)
-	result = append(result, txt.Keywords(m.State())...)
+	result = append(result, txt.Keywords(txt.ReplaceSpaces(m.State(), "-"))...)
-	result = append(result, txt.Keywords(m.CountryName())...)
+	result = append(result, txt.Keywords(txt.ReplaceSpaces(m.CountryName(), "-"))...)
 	result = append(result, txt.Keywords(m.Category())...)
-
 	result = append(result, txt.Keywords(m.Name())...)
-	result = append(result, txt.Keywords(m.Label())...)
-	result = append(result, txt.Keywords(m.Notes())...)
 
 	result = txt.UniqueWords(result)
 

internal/photoprism/index_mediafile.go

@@ -308,7 +308,7 @@ func (ind *Index) MediaFile(m *MediaFile, o IndexOptions, originalName string) (
 		}
 
 		w = append(w, locKeywords...)
-		w = append(w, txt.Keywords(file.OriginalName)...)
+		w = append(w, txt.FilenameWords(file.OriginalName)...)
 		w = append(w, file.FileMainColor)
 		w = append(w, labels.Keywords()...)
 

internal/remote/discover_test.go

@@ -14,7 +14,7 @@ func TestDiscover(t *testing.T) {
 			t.Fatal(err)
 		}
 
-		assert.Equal(t, "Webdav", r.AccName)
+		assert.Equal(t, "Webdav-Dummy", r.AccName)
 		assert.Equal(t, "webdav", r.AccType)
 		assert.Equal(t, "http://webdav-dummy/", r.AccURL)
 		assert.Equal(t, "admin", r.AccUser)
@@ -28,7 +28,7 @@ func TestDiscover(t *testing.T) {
 			t.Fatal(err)
 		}
 
-		assert.Equal(t, "Webdav", r.AccName)
+		assert.Equal(t, "Webdav-Dummy", r.AccName)
 		assert.Equal(t, "webdav", r.AccType)
 		assert.Equal(t, "http://webdav-dummy/", r.AccURL)
 		assert.Equal(t, "admin", r.AccUser)

pkg/fs/base_test.go

@@ -61,5 +61,4 @@ func TestBaseAbs(t *testing.T) {
 		assert.Equal(t, "/testdata/Test (4)", result)
 	})
 
-
 }

pkg/txt/words.go

@@ -6,13 +6,25 @@ import (
 	"strings"
 )
 
-var KeywordsRegexp = regexp.MustCompile("[\\p{L}]{3,}")
+var KeywordsRegexp = regexp.MustCompile("[\\p{L}\\-]{3,}")
 
-// Words returns a slice of words with at least 3 characters from a string.
+// Words returns a slice of words with at least 3 characters from a string, dashes count as character ("ile-de-france").
 func Words(s string) (results []string) {
 	return KeywordsRegexp.FindAllString(s, -1)
 }
 
+// ReplaceSpaces replaces all spaces with another string.
+func ReplaceSpaces(s string, char string) string {
+	return strings.Replace(s, " ", char, -1)
+}
+
+var FilenameKeywordsRegexp = regexp.MustCompile("[\\p{L}]{3,}")
+
+// FilenameWords returns a slice of words with at least 3 characters from a string ("ile", "france").
+func FilenameWords(s string) (results []string) {
+	return FilenameKeywordsRegexp.FindAllString(s, -1)
+}
+
 // Keywords returns a slice of keywords without stopwords.
 func Keywords(s string) (results []string) {
 	for _, w := range Words(s) {

pkg/txt/words_test.go

@@ -7,9 +7,9 @@ import (
 )
 
 func TestWords(t *testing.T) {
-	t.Run("I'm a lazy brown fox!", func(t *testing.T) {
+	t.Run("I'm a lazy-brown fox!", func(t *testing.T) {
-		result := Words("I'm a lazy BRoWN fox!")
+		result := Words("I'm a lazy-BRoWN fox!")
-		assert.Equal(t, []string{"lazy", "BRoWN", "fox"}, result)
+		assert.Equal(t, []string{"lazy-BRoWN", "fox"}, result)
 	})
 	t.Run("no result", func(t *testing.T) {
 		result := Words("x")
@@ -17,6 +17,17 @@ func TestWords(t *testing.T) {
 	})
 }
 
+func TestFilenameWords(t *testing.T) {
+	t.Run("I'm a lazy-brown fox!", func(t *testing.T) {
+		result := FilenameWords("I'm a lazy-BRoWN fox!")
+		assert.Equal(t, []string{"lazy", "BRoWN", "fox"}, result)
+	})
+	t.Run("no result", func(t *testing.T) {
+		result := FilenameWords("x")
+		assert.Equal(t, []string(nil), result)
+	})
+}
+
 func TestKeywords(t *testing.T) {
 	t.Run("I'm a lazy brown fox!", func(t *testing.T) {
 		result := Keywords("I'm a lazy BRoWN img!")
@@ -30,8 +41,8 @@ func TestKeywords(t *testing.T) {
 
 func TestUniqueWords(t *testing.T) {
 	t.Run("many", func(t *testing.T) {
-		result := UniqueWords([]string{"lazy", "brown", "apple", "brown"})
+		result := UniqueWords([]string{"lazy", "brown", "apple", "brown", "new-york"})
-		assert.Equal(t, []string{"apple", "brown", "lazy"}, result)
+		assert.Equal(t, []string{"apple", "brown", "lazy", "new-york"}, result)
 	})
 	t.Run("one", func(t *testing.T) {
 		result := UniqueWords([]string{"lazy"})
@@ -41,8 +52,8 @@ func TestUniqueWords(t *testing.T) {
 
 func TestUniqueKeywords(t *testing.T) {
 	t.Run("many", func(t *testing.T) {
-		result := UniqueKeywords("lazy, brown, apple, brown, ...")
+		result := UniqueKeywords("lazy, brown, apple, new-york, brown, ...")
-		assert.Equal(t, []string{"apple", "brown", "lazy"}, result)
+		assert.Equal(t, []string{"apple", "brown", "lazy", "new-york"}, result)
 	})
 	t.Run("one", func(t *testing.T) {
 		result := UniqueKeywords("")