From 0871bf5219c2dff8b7903c5f1ecb441a5544892a Mon Sep 17 00:00:00 2001
From: Mechiel Lukkien <mechiel@ueber.net>
Date: Sat, 7 Dec 2024 13:05:09 +0100
Subject: [PATCH] move checking whether a message needs smtputf8 (has utf8 in
 any of the header sections) to package message

---
 main.go              | 32 +++-----------------------------
 message/part.go      | 33 +++++++++++++++++++++++++++++++++
 smtpserver/server.go | 41 +++++++++++++----------------------------
 3 files changed, 49 insertions(+), 57 deletions(-)

diff --git a/main.go b/main.go
index ff19afa..41b6eb9 100644
--- a/main.go
+++ b/main.go
@@ -1,7 +1,6 @@
 package main
 
 import (
-	"bufio"
 	"bytes"
 	"context"
 	"crypto"
@@ -3519,35 +3518,10 @@ func cmdMessageParse(c *cmd) {
 	err = enc.Encode(part)
 	xcheckf(err, "write")
 
-	hasNonASCII := func(r io.Reader) bool {
-		br := bufio.NewReader(r)
-		for {
-			b, err := br.ReadByte()
-			if err == io.EOF {
-				break
-			}
-			xcheckf(err, "read header")
-			if b > 0x7f {
-				return true
-			}
-		}
-		return false
-	}
-
-	var walk func(p *message.Part) bool
-	walk = func(p *message.Part) bool {
-		if hasNonASCII(p.HeaderReader()) {
-			return true
-		}
-		for _, pp := range p.Parts {
-			if walk(&pp) {
-				return true
-			}
-		}
-		return false
-	}
 	if smtputf8 {
-		fmt.Println("message needs smtputf8:", walk(&part))
+		needs, err := part.NeedsSMTPUTF8()
+		xcheckf(err, "checking if message needs smtputf8")
+		fmt.Println("message needs smtputf8:", needs)
 	}
 }
 
diff --git a/message/part.go b/message/part.go
index bf1b8b7..faba013 100644
--- a/message/part.go
+++ b/message/part.go
@@ -21,6 +21,7 @@ import (
 	"net/textproto"
 	"strings"
 	"time"
+	"unicode"
 
 	"golang.org/x/text/encoding/ianaindex"
 
@@ -598,6 +599,38 @@ func (p *Part) IsDSN() bool {
 		(p.Parts[1].MediaSubType == "DELIVERY-STATUS" || p.Parts[1].MediaSubType == "GLOBAL-DELIVERY-STATUS")
 }
 
+func hasNonASCII(r io.Reader) (bool, error) {
+	br := bufio.NewReader(r)
+	for {
+		b, err := br.ReadByte()
+		if err == io.EOF {
+			break
+		} else if err != nil {
+			return false, err
+		}
+		if b > unicode.MaxASCII {
+			return true, nil
+		}
+	}
+	return false, nil
+}
+
+// NeedsSMTPUTF8 returns whether the part needs the SMTPUTF8 extension to be
+// transported, due to non-ascii in message headers.
+func (p *Part) NeedsSMTPUTF8() (bool, error) {
+	if has, err := hasNonASCII(p.HeaderReader()); err != nil {
+		return false, fmt.Errorf("reading header: %w", err)
+	} else if has {
+		return true, nil
+	}
+	for _, pp := range p.Parts {
+		if has, err := pp.NeedsSMTPUTF8(); err != nil || has {
+			return has, err
+		}
+	}
+	return false, nil
+}
+
 var ErrParamEncoding = errors.New("bad header parameter encoding")
 
 // DispositionFilename tries to parse the disposition header and the "filename"
diff --git a/smtpserver/server.go b/smtpserver/server.go
index e90f0c9..424c88c 100644
--- a/smtpserver/server.go
+++ b/smtpserver/server.go
@@ -1940,47 +1940,32 @@ func (c *conn) cmdRcpt(p *parser) {
 	c.bwritecodeline(smtp.C250Completed, smtp.SeAddr1Other0, "now on the list", nil)
 }
 
-// ../rfc/6531:497
-func (c *conn) isSMTPUTF8Required(part *message.Part) bool {
-	hasNonASCII := func(r io.Reader) bool {
-		br := bufio.NewReader(r)
-		for {
-			b, err := br.ReadByte()
-			if err == io.EOF {
-				break
-			}
-			xcheckf(err, "read header")
-			if b > unicode.MaxASCII {
-				return true
-			}
-		}
-		return false
-	}
-	var hasNonASCIIPartHeader func(p *message.Part) bool
-	hasNonASCIIPartHeader = func(p *message.Part) bool {
-		if hasNonASCII(p.HeaderReader()) {
+func hasNonASCII(s string) bool {
+	for _, c := range []byte(s) {
+		if c > unicode.MaxASCII {
 			return true
 		}
-		for _, pp := range p.Parts {
-			if hasNonASCIIPartHeader(&pp) {
-				return true
-			}
-		}
-		return false
 	}
+	return false
+}
 
+// ../rfc/6531:497
+func (c *conn) isSMTPUTF8Required(part *message.Part) bool {
 	// Check "MAIL FROM".
-	if hasNonASCII(strings.NewReader(string(c.mailFrom.Localpart))) {
+	if hasNonASCII(string(c.mailFrom.Localpart)) {
 		return true
 	}
 	// Check all "RCPT TO".
 	for _, rcpt := range c.recipients {
-		if hasNonASCII(strings.NewReader(string(rcpt.Addr.Localpart))) {
+		if hasNonASCII(string(rcpt.Addr.Localpart)) {
 			return true
 		}
 	}
+
 	// Check header in all message parts.
-	return hasNonASCIIPartHeader(part)
+	smtputf8, err := part.NeedsSMTPUTF8()
+	xcheckf(err, "checking if smtputf8 is required")
+	return smtputf8
 }
 
 // ../rfc/5321:1992 ../rfc/5321:1098