Use UTF-16BE encoding when required

Fix #2
2023-03-10 09:07:27 +01:00
parent 7fd1d3df69
commit 4fb6fafba6
4 changed files with 115 additions and 8 deletions
--- a/go.mod
+++ b/go.mod
@@ -4,8 +4,9 @@ go 1.17

 require (
 	github.com/digitorus/pdf v0.1.2
-	github.com/digitorus/pkcs7 v0.0.0-20221212123742-001c36b64ec3
-	github.com/digitorus/timestamp v0.0.0-20221118121739-0faba6f30e54
+	github.com/digitorus/pkcs7 v0.0.0-20230220124406-51331ccfc40f
+	github.com/digitorus/timestamp v0.0.0-20230220124323-d542479a2425
 	github.com/mattetti/filebuffer v1.0.1
 	golang.org/x/crypto v0.7.0
+	golang.org/x/text v0.8.0
 )
--- a/go.sum
+++ b/go.sum
@@ -1,10 +1,10 @@
 github.com/digitorus/pdf v0.1.2 h1:RjYEJNbiV6Kcn8QzRi6pwHuOaSieUUrg4EZo4b7KuIQ=
 github.com/digitorus/pdf v0.1.2/go.mod h1:05fDDJhPswBRM7GTfqCxNiDyeNcN0f+IobfOAl5pdXw=
 github.com/digitorus/pkcs7 v0.0.0-20221019075359-21b8b40e6bb4/go.mod h1:SKVExuS+vpu2l9IoOc0RwqE7NYnb0JlcFHFnEJkVDzc=
-github.com/digitorus/pkcs7 v0.0.0-20221212123742-001c36b64ec3 h1:rjCXeRWazGsbcBlExMcAW8H1LGdgJ9r619y7+aeKgds=
-github.com/digitorus/pkcs7 v0.0.0-20221212123742-001c36b64ec3/go.mod h1:SKVExuS+vpu2l9IoOc0RwqE7NYnb0JlcFHFnEJkVDzc=
-github.com/digitorus/timestamp v0.0.0-20221118121739-0faba6f30e54 h1:VPfvrlgI2O3mQjBBLywtPxjDMlpWYWobXN3nk26Ht0k=
-github.com/digitorus/timestamp v0.0.0-20221118121739-0faba6f30e54/go.mod h1:6V2ND8Yf8TOJ4h+9pmUlx8kXvNLBB2QplToVVZQ3rF0=
+github.com/digitorus/pkcs7 v0.0.0-20230220124406-51331ccfc40f h1:AoHV/iJ6LjW24bRWrg0zm1xD3Uh83PlNSK0QWH11J0E=
+github.com/digitorus/pkcs7 v0.0.0-20230220124406-51331ccfc40f/go.mod h1:SKVExuS+vpu2l9IoOc0RwqE7NYnb0JlcFHFnEJkVDzc=
+github.com/digitorus/timestamp v0.0.0-20230220124323-d542479a2425 h1:cbnavmdMqZ3b4hcCxizSO/jO+BxyXp/hU9jyzULJ9g8=
+github.com/digitorus/timestamp v0.0.0-20230220124323-d542479a2425/go.mod h1:6V2ND8Yf8TOJ4h+9pmUlx8kXvNLBB2QplToVVZQ3rF0=
 github.com/mattetti/filebuffer v1.0.1 h1:gG7pyfnSIZCxdoKq+cPa8T0hhYtD9NxCdI4D7PTjRLM=
 github.com/mattetti/filebuffer v1.0.1/go.mod h1:YdMURNDOttIiruleeVr6f56OrMc+MydEnTcXwtkxNVs=
 github.com/yuin/goldmark v1.4.13/go.mod h1:6yULJ656Px+3vBD8DxQVa3kxgyrAnzto9xy5taEt/CY=
@@ -37,6 +37,7 @@ golang.org/x/text v0.3.0/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
 golang.org/x/text v0.3.3/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ=
 golang.org/x/text v0.3.7/go.mod h1:u+2+/6zg+i71rQMx5EYifcz6MCKuco9NR6JIITiCfzQ=
 golang.org/x/text v0.7.0/go.mod h1:mrYo+phRRbMaCq/xk9113O4dZlRixOauAjOtrjsXDZ8=
+golang.org/x/text v0.8.0 h1:57P1ETyNKtuIjB4SRd15iJxuhj8Gc416Y78H3qgMh68=
 golang.org/x/text v0.8.0/go.mod h1:e1OnstbJyHTd6l/uOt8jFFHp6TRDWZR/bV3emEE/zU8=
 golang.org/x/tools v0.0.0-20180917221912-90fa682c2a6e/go.mod h1:n7NCudcB/nEzxVGmLbDWY5pfWTLqBcC2KZ6jyYvM4mQ=
 golang.org/x/tools v0.0.0-20191119224855-298f0cb1881e/go.mod h1:b+2E5dAYhXwXZwtnZ6UAqBI28+e2cm9otk0dWdXHAEo=
--- a/sign/helpers.go
+++ b/sign/helpers.go
@@ -11,6 +11,9 @@ import (
 	"time"

 	"github.com/digitorus/pdf"
+
+	"golang.org/x/text/encoding/unicode"
+	"golang.org/x/text/transform"
 )

 func findFirstPage(parent pdf.Value) (pdf.Value, error) {
@@ -35,11 +38,28 @@ func findFirstPage(parent pdf.Value) (pdf.Value, error) {
 }

 func pdfString(text string) string {
+	if !isASCII(text) {
+		// UTF-16BE
+		enc := unicode.UTF16(unicode.BigEndian, unicode.UseBOM).NewEncoder()
+		res, _, err := transform.String(enc, text)
+		if err != nil {
+			panic(err)
+		}
+		return "(" + res + ")"
+	}
+
+	// UTF-8
+	// (\357\273\277Layer 1)               % UTF-8 Layer 1 Name
+	// <EF BB BF DA AF DA 86 D9 BE DA 98>  % UTF-8 Layer 2 Name
+	// text = "\357\273\277" + text
+	// text = hex.EncodeToString([]byte(text))
+	// text = "<" + text + ">"
+
+	// PDFDocEncoded
 	text = strings.Replace(text, "\\", "\\\\", -1)
 	text = strings.Replace(text, ")", "\\)", -1)
 	text = strings.Replace(text, "(", "\\(", -1)
 	text = strings.Replace(text, "\r", "\\r", -1)
-
 	text = "(" + text + ")"

 	return text
@@ -167,3 +187,12 @@ func getOIDFromHashAlgorithm(target crypto.Hash) asn1.ObjectIdentifier {
 	}
 	return nil
 }
+
+func isASCII(s string) bool {
+	for _, r := range s {
+		if r > '\u007F' {
+			return false
+		}
+	}
+	return true
+}
--- a/sign/sign_test.go
+++ b/sign/sign_test.go
@@ -68,7 +68,7 @@ func TestReaderCanReadPDF(t *testing.T) {
 		if ext != ".pdf" {
 			continue
 		}
-		
+
 		fileName := f.Name()
 		t.Run(fileName, func(st *testing.T) {
 			st.Parallel()
@@ -285,6 +285,82 @@ func TestSignPDFFile(t *testing.T) {
 	}
 }

+func TestSignPDFFileUTF8(t *testing.T) {
+	certificate_data_block, _ := pem.Decode([]byte(signCertPem))
+	if certificate_data_block == nil {
+		t.Errorf("failed to parse PEM block containing the certificate")
+		return
+	}
+
+	cert, err := x509.ParseCertificate(certificate_data_block.Bytes)
+	if err != nil {
+		t.Errorf("%s", err.Error())
+		return
+	}
+
+	key_data_block, _ := pem.Decode([]byte(signKeyPem))
+	if key_data_block == nil {
+		t.Errorf("failed to parse PEM block containing the private key")
+		return
+	}
+
+	pkey, err := x509.ParsePKCS1PrivateKey(key_data_block.Bytes)
+	if err != nil {
+		t.Errorf("%s", err.Error())
+		return
+	}
+
+	tmpfile, err := os.CreateTemp("", "pdfsign_test")
+	if err != nil {
+		t.Errorf("%s", err.Error())
+		return
+	}
+
+	signerName := "姓名"
+	signerLocation := "位置"
+	err = SignFile("../testfiles/testfile20.pdf", tmpfile.Name(), SignData{
+		Signature: SignDataSignature{
+			Info: SignDataSignatureInfo{
+				Name:        signerName,
+				Location:    signerLocation,
+				Reason:      "Test with UTF-8",
+				ContactInfo: "None",
+				Date:        time.Now().Local(),
+			},
+			CertType:   CertificationSignature,
+			DocMDPPerm: AllowFillingExistingFormFieldsAndSignaturesPerms,
+		},
+		DigestAlgorithm: crypto.SHA512,
+		Signer:          pkey,
+		Certificate:     cert,
+	})
+
+	if err != nil {
+		os.Remove(tmpfile.Name())
+		t.Errorf("%s: %s", "testfile20.pdf", err.Error())
+		return
+	}
+
+	info, err := verify.File(tmpfile)
+	if err != nil {
+		t.Errorf("%s: %s", tmpfile.Name(), err.Error())
+
+		err2 := os.Rename(tmpfile.Name(), "../testfiles/failed/testfile20.pdf")
+		if err2 != nil {
+			t.Error(err2)
+		}
+	} else {
+		if info.Signers[0].Name != signerName {
+			t.Errorf("expected %q, got %q", signerName, info.Signers[0].Name)
+		}
+		if info.Signers[0].Location != signerLocation {
+			t.Errorf("expected %q, got %q", signerLocation, info.Signers[0].Location)
+		}
+
+		os.Remove(tmpfile.Name())
+	}
+}
+
 func BenchmarkSignPDF(b *testing.B) {
 	certificate_data_block, _ := pem.Decode([]byte(signCertPem))
 	if certificate_data_block == nil {