Merge pull request #131 from buildkite-plugins/fix/handle_emojis_SUP-5786

Fix unicode handling in git diff output

Author: scadu

pipeline.go

@@ -4,6 +4,7 @@ import (
 	"fmt"
 	"os"
 	"reflect"
+	"strconv"
 	"strings"
 
 	"github.com/bmatcuk/doublestar/v4"
@@ -104,7 +105,25 @@ func diff(command string) ([]string, error) {
 		return nil, fmt.Errorf("diff command failed: %v", err)
 	}
 
-	return strings.Fields(strings.TrimSpace(output)), nil
+	fields := strings.Fields(strings.TrimSpace(output))
+	paths := make([]string, 0, len(fields))
+
+	for _, field := range fields {
+		// Git quotes paths with special characters using C-style quoting
+		if strings.HasPrefix(field, "\"") && strings.HasSuffix(field, "\"") {
+			// Unquote to decode escape sequences (e.g., \360\237\252\201 -> 🪁)
+			if unquoted, err := strconv.Unquote(field); err == nil {
+				paths = append(paths, unquoted)
+			} else {
+				// If unquoting fails, fall back to removing quotes
+				paths = append(paths, strings.Trim(field, "\""))
+			}
+		} else {
+			paths = append(paths, field)
+		}
+	}
+
+	return paths, nil
 }
 
 func stepsToTrigger(files []string, watch []WatchConfig) ([]Step, error) {

pipeline_test.go

@@ -119,6 +119,34 @@ func TestDiffWithSubshell(t *testing.T) {
 	assert.Equal(t, want, got)
 }
 
+func TestDiffWithQuotedPaths(t *testing.T) {
+	want := []string{
+		"projects/test/pages/17_🪁_testfile.py",
+		"normal/file.txt",
+	}
+	got, err := diff(`printf '"projects/test/pages/17_\360\237\252\201_testfile.py" normal/file.txt'`)
+	assert.NoError(t, err)
+	assert.Equal(t, want, got)
+}
+
+func TestStepsToTriggerWithEmojiPaths(t *testing.T) {
+	watch := []WatchConfig{
+		{
+			Paths: []string{"projects/**"},
+			Step:  Step{Trigger: "test-pipeline"},
+		},
+	}
+
+	changedFiles := []string{
+		"projects/test/pages/17_🪁_testfile.py",
+		"other/file.txt",
+	}
+
+	steps, err := stepsToTrigger(changedFiles, watch)
+	assert.NoError(t, err)
+	assert.Equal(t, []Step{{Trigger: "test-pipeline"}}, steps)
+}
+
 func TestPipelinesToTriggerGetsListOfPipelines(t *testing.T) {
 	want := []string{"service-1", "service-2", "service-4"}