Make string sorting (e.g. ByTitle, ByLinkTitle and ByParam) language aware

Fixes #2180

2 files changed, 96 insertions, 2 deletions

diff --git a/langs/language.go b/langs/language.go
index 0df2914a1..244f6a743 100644
--- a/langs/language.go
+++ b/langs/language.go
@@ -19,6 +19,9 @@ import (
 	"sync"
 	"time"
 
+	"golang.org/x/text/collate"
+	"golang.org/x/text/language"
+
 	"github.com/pkg/errors"
 
 	"github.com/gohugoio/hugo/common/htime"
@@ -80,8 +83,9 @@ type Language struct {
 	// TODO(bep) do the same for some of the others.
 	translator    locales.Translator
 	timeFormatter htime.TimeFormatter
-
-	location *time.Location
+	tag           language.Tag
+	collator      *Collator
+	location      *time.Location
 
 	// Error during initialization. Will fail the buld.
 	initErr error
@@ -111,6 +115,18 @@ func NewLanguage(lang string, cfg config.Provider) *Language {
 		}
 	}
 
+	var coll *Collator
+	tag, err := language.Parse(lang)
+	if err == nil {
+		coll = &Collator{
+			c: collate.New(tag),
+		}
+	} else {
+		coll = &Collator{
+			c: collate.New(language.English),
+		}
+	}
+
 	l := &Language{
 		Lang:       lang,
 		ContentDir: cfg.GetString("contentDir"),
@@ -119,6 +135,8 @@ func NewLanguage(lang string, cfg config.Provider) *Language {
 		params:        params,
 		translator:    translator,
 		timeFormatter: htime.NewTimeFormatter(translator),
+		tag:           tag,
+		collator:      coll,
 	}
 
 	if err := l.loadLocation(cfg.GetString("timeZone")); err != nil {
@@ -275,6 +293,10 @@ func GetLocation(l *Language) *time.Location {
 	return l.location
 }
 
+func GetCollator(l *Language) *Collator {
+	return l.collator
+}
+
 func (l *Language) loadLocation(tzStr string) error {
 	location, err := time.LoadLocation(tzStr)
 	if err != nil {
@@ -284,3 +306,16 @@ func (l *Language) loadLocation(tzStr string) error {
 
 	return nil
 }
+
+type Collator struct {
+	sync.Mutex
+	c *collate.Collator
+}
+
+// CompareStrings compares a and b.
+// It returns -1 if a < b, 1 if a > b and 0 if a == b.
+// Note that the Collator is not thread safe, so you may want
+// to aquire a lock on it before calling this method.
+func (c *Collator) CompareStrings(a, b string) int {
+	return c.c.CompareString(a, b)
+}
diff --git a/langs/language_test.go b/langs/language_test.go
index e6ef94824..264e813a0 100644
--- a/langs/language_test.go
+++ b/langs/language_test.go
@@ -14,10 +14,13 @@
 package langs
 
 import (
+	"sync"
 	"testing"
 
 	qt "github.com/frankban/quicktest"
 	"github.com/gohugoio/hugo/config"
+	"golang.org/x/text/collate"
+	"golang.org/x/text/language"
 )
 
 func TestGetGlobalOnlySetting(t *testing.T) {
@@ -47,3 +50,59 @@ func TestLanguageParams(t *testing.T) {
 	c.Assert(lang.Params()["p1"], qt.Equals, "p1p")
 	c.Assert(lang.Get("p1"), qt.Equals, "p1cfg")
 }
+
+func TestCollator(t *testing.T) {
+
+	c := qt.New(t)
+
+	var wg sync.WaitGroup
+
+	coll := &Collator{c: collate.New(language.English, collate.Loose)}
+
+	for i := 0; i < 10; i++ {
+		wg.Add(1)
+		go func() {
+			coll.Lock()
+			defer coll.Unlock()
+			defer wg.Done()
+			for j := 0; j < 10; j++ {
+				k := coll.CompareStrings("abc", "def")
+				c.Assert(k, qt.Equals, -1)
+			}
+		}()
+	}
+	wg.Wait()
+
+}
+
+func BenchmarkCollator(b *testing.B) {
+	s := []string{"foo", "bar", "éntre", "baz", "qux", "quux", "corge", "grault", "garply", "waldo", "fred", "plugh", "xyzzy", "thud"}
+
+	doWork := func(coll *Collator) {
+		for i := 0; i < len(s); i++ {
+			for j := i + 1; j < len(s); j++ {
+				_ = coll.CompareStrings(s[i], s[j])
+			}
+		}
+	}
+
+	b.Run("Single", func(b *testing.B) {
+		coll := &Collator{c: collate.New(language.English, collate.Loose)}
+		for i := 0; i < b.N; i++ {
+			doWork(coll)
+		}
+	})
+
+	b.Run("Para", func(b *testing.B) {
+		b.RunParallel(func(pb *testing.PB) {
+			coll := &Collator{c: collate.New(language.English, collate.Loose)}
+
+			for pb.Next() {
+				coll.Lock()
+				doWork(coll)
+				coll.Unlock()
+			}
+		})
+	})
+
+}