diff --git a/parsemail.go b/parsemail.go
index bc19e93..6289f70 100644
--- a/parsemail.go
+++ b/parsemail.go
@@ -8,6 +8,7 @@ import (
"io/ioutil"
"mime"
"mime/multipart"
+ "mime/quotedprintable"
"net/mail"
"strings"
"time"
@@ -38,6 +39,8 @@ func Parse(r io.Reader) (email Email, err error) {
return
}
+ cte := msg.Header.Get("Content-Transfer-Encoding")
+
switch contentType {
case contentTypeMultipartMixed:
email.TextBody, email.HTMLBody, email.Attachments, email.EmbeddedFiles, err = parseMultipartMixed(msg.Body, params["boundary"])
@@ -47,14 +50,36 @@ func Parse(r io.Reader) (email Email, err error) {
email.TextBody, email.HTMLBody, email.EmbeddedFiles, err = parseMultipartRelated(msg.Body, params["boundary"])
case contentTypeTextPlain:
message, _ := ioutil.ReadAll(msg.Body)
+ var reader io.Reader
+ reader, err = decodeContent(strings.NewReader(string(message[:])), cte)
+ if err != nil {
+ return
+ }
+
+ message, err = ioutil.ReadAll(reader)
+ if err != nil {
+ return
+ }
+
email.TextBody = strings.TrimSuffix(string(message[:]), "\n")
case contentTypeTextHtml:
message, _ := ioutil.ReadAll(msg.Body)
+ var reader io.Reader
+ reader, err = decodeContent(strings.NewReader(string(message[:])), cte)
+ if err != nil {
+ return
+ }
+
+ message, err = ioutil.ReadAll(reader)
+ if err != nil {
+ return
+ }
+
email.HTMLBody = strings.TrimSuffix(string(message[:]), "\n")
case contentTypeOctetStream:
email.Attachments, err = parseAttachmentOnlyEmail(msg.Body, msg.Header)
default:
- email.Content, err = decodeContent(msg.Body, msg.Header.Get("Content-Transfer-Encoding"))
+ email.Content, err = decodeContent(msg.Body, cte)
}
return
@@ -134,7 +159,7 @@ func parseAttachmentOnlyEmail(body io.Reader, header mail.Header) (attachments [
func parseMultipartRelated(msg io.Reader, boundary string) (textBody, htmlBody string, embeddedFiles []EmbeddedFile, err error) {
pmr := multipart.NewReader(msg, boundary)
for {
- part, err := pmr.NextPart()
+ part, err := pmr.NextRawPart()
if err == io.EOF {
break
@@ -142,6 +167,8 @@ func parseMultipartRelated(msg io.Reader, boundary string) (textBody, htmlBody s
return textBody, htmlBody, embeddedFiles, err
}
+ cte := part.Header.Get("Content-Transfer-Encoding")
+
contentType, params, err := mime.ParseMediaType(part.Header.Get("Content-Type"))
if err != nil {
return textBody, htmlBody, embeddedFiles, err
@@ -149,14 +176,22 @@ func parseMultipartRelated(msg io.Reader, boundary string) (textBody, htmlBody s
switch contentType {
case contentTypeTextPlain:
- ppContent, err := ioutil.ReadAll(part)
+ decoded, err := decodeContent(part, cte)
+ if err != nil {
+ return textBody, htmlBody, embeddedFiles, err
+ }
+ ppContent, err := ioutil.ReadAll(decoded)
if err != nil {
return textBody, htmlBody, embeddedFiles, err
}
textBody += strings.TrimSuffix(string(ppContent[:]), "\n")
case contentTypeTextHtml:
- ppContent, err := ioutil.ReadAll(part)
+ decoded, err := decodeContent(part, cte)
+ if err != nil {
+ return textBody, htmlBody, embeddedFiles, err
+ }
+ ppContent, err := ioutil.ReadAll(decoded)
if err != nil {
return textBody, htmlBody, embeddedFiles, err
}
@@ -191,7 +226,7 @@ func parseMultipartRelated(msg io.Reader, boundary string) (textBody, htmlBody s
func parseMultipartAlternative(msg io.Reader, boundary string) (textBody, htmlBody string, embeddedFiles []EmbeddedFile, err error) {
pmr := multipart.NewReader(msg, boundary)
for {
- part, err := pmr.NextPart()
+ part, err := pmr.NextRawPart()
if err == io.EOF {
break
@@ -199,6 +234,8 @@ func parseMultipartAlternative(msg io.Reader, boundary string) (textBody, htmlBo
return textBody, htmlBody, embeddedFiles, err
}
+ cte := part.Header.Get("Content-Transfer-Encoding")
+
contentType, params, err := mime.ParseMediaType(part.Header.Get("Content-Type"))
if err != nil {
return textBody, htmlBody, embeddedFiles, err
@@ -206,14 +243,22 @@ func parseMultipartAlternative(msg io.Reader, boundary string) (textBody, htmlBo
switch contentType {
case contentTypeTextPlain:
- ppContent, err := ioutil.ReadAll(part)
+ decoded, err := decodeContent(part, cte)
+ if err != nil {
+ return textBody, htmlBody, embeddedFiles, err
+ }
+ ppContent, err := ioutil.ReadAll(decoded)
if err != nil {
return textBody, htmlBody, embeddedFiles, err
}
textBody += strings.TrimSuffix(string(ppContent[:]), "\n")
case contentTypeTextHtml:
- ppContent, err := ioutil.ReadAll(part)
+ decoded, err := decodeContent(part, cte)
+ if err != nil {
+ return textBody, htmlBody, embeddedFiles, err
+ }
+ ppContent, err := ioutil.ReadAll(decoded)
if err != nil {
return textBody, htmlBody, embeddedFiles, err
}
@@ -248,7 +293,7 @@ func parseMultipartAlternative(msg io.Reader, boundary string) (textBody, htmlBo
func parseMultipartMixed(msg io.Reader, boundary string) (textBody, htmlBody string, attachments []Attachment, embeddedFiles []EmbeddedFile, err error) {
mr := multipart.NewReader(msg, boundary)
for {
- part, err := mr.NextPart()
+ part, err := mr.NextRawPart()
if err == io.EOF {
break
} else if err != nil {
@@ -265,11 +310,21 @@ func parseMultipartMixed(msg io.Reader, boundary string) (textBody, htmlBody str
continue
}
+ cte := part.Header.Get("Content-Transfer-Encoding")
+
contentType, params, err := mime.ParseMediaType(part.Header.Get("Content-Type"))
if err != nil {
return textBody, htmlBody, attachments, embeddedFiles, err
}
+ if isAttachment(part) {
+ at, err := decodeAttachment(part)
+ if err != nil {
+ return textBody, htmlBody, attachments, embeddedFiles, err
+ }
+ attachments = append(attachments, at)
+ }
+
if contentType == contentTypeMultipartAlternative {
textBody, htmlBody, embeddedFiles, err = parseMultipartAlternative(part, params["boundary"])
if err != nil {
@@ -281,14 +336,22 @@ func parseMultipartMixed(msg io.Reader, boundary string) (textBody, htmlBody str
return textBody, htmlBody, attachments, embeddedFiles, err
}
} else if contentType == contentTypeTextPlain {
- ppContent, err := ioutil.ReadAll(part)
+ decoded, err := decodeContent(part, cte)
+ if err != nil {
+ return textBody, htmlBody, attachments, embeddedFiles, err
+ }
+ ppContent, err := ioutil.ReadAll(decoded)
if err != nil {
return textBody, htmlBody, attachments, embeddedFiles, err
}
textBody += strings.TrimSuffix(string(ppContent[:]), "\n")
} else if contentType == contentTypeTextHtml {
- ppContent, err := ioutil.ReadAll(part)
+ decoded, err := decodeContent(part, cte)
+ if err != nil {
+ return textBody, htmlBody, attachments, embeddedFiles, err
+ }
+ ppContent, err := ioutil.ReadAll(decoded)
if err != nil {
return textBody, htmlBody, attachments, embeddedFiles, err
}
@@ -383,17 +446,25 @@ func decodeContent(content io.Reader, encoding string) (io.Reader, error) {
if err != nil {
return nil, err
}
-
return bytes.NewReader(b), nil
- case "7bit", "8bit", "binary:
- dd, err := ioutil.ReadAll(content)
+ case "quoted-printable":
+ decoded := quotedprintable.NewReader(content)
+ b, err := ioutil.ReadAll(decoded)
if err != nil {
return nil, err
}
-
- return bytes.NewReader(dd), nil
- case "":
- return content, nil
+ return bytes.NewReader(b), nil
+ // The values "8bit", "7bit", and "binary" all imply that NO encoding has been performed and data need to be read as bytes.
+ // "7bit" means that the data is all represented as short lines of US-ASCII data.
+ // "8bit" means that the lines are short, but there may be non-ASCII characters (octets with the high-order bit set).
+ // "Binary" means that not only may non-ASCII characters be present, but also that the lines are not necessarily short enough for SMTP transport.
+ case "", "7bit", "8bit", "binary":
+ decoded := quotedprintable.NewReader(content)
+ b, err := ioutil.ReadAll(decoded)
+ if err != nil {
+ return nil, err
+ }
+ return bytes.NewReader(b), nil
default:
return nil, fmt.Errorf("unknown encoding: %s", encoding)
}
diff --git a/parsemail_test.go b/parsemail_test.go
index cb088aa..973efa7 100644
--- a/parsemail_test.go
+++ b/parsemail_test.go
@@ -372,15 +372,15 @@ So, "Hello".`,
htmlBody: "
",
attachments: []attachmentData{
{
- filename: "unencoded.csv",
- contentType: "application/csv",
- data: fmt.Sprintf("\n"+`"%s", "%s", "%s", "%s", "%s"`+"\n"+`"%s", "%s", "%s", "%s", "%s"`+"\n", "Some", "Data", "In", "Csv", "Format", "Foo", "Bar", "Baz", "Bum", "Poo"),
+ filename: "unencoded.csv",
+ contentType: "application/csv",
+ data: fmt.Sprintf("\n"+`"%s", "%s", "%s", "%s", "%s"`+"\n"+`"%s", "%s", "%s", "%s", "%s"`+"\n", "Some", "Data", "In", "Csv", "Format", "Foo", "Bar", "Baz", "Bum", "Poo"),
},
},
},
13: {
contentType: "multipart/related; boundary=\"000000000000ab2e2205a26de587\"",
- mailData: multipartRelatedExample,
+ mailData: multipartRelatedExample,
subject: "Saying Hello",
from: []mail.Address{
{
@@ -389,7 +389,7 @@ So, "Hello".`,
},
},
sender: mail.Address{
- Name: "Michael Jones",
+ Name: "Michael Jones",
Address: "mjones@machine.example",
},
to: []mail.Address{
@@ -401,7 +401,7 @@ So, "Hello".`,
messageID: "1234@local.machine.example",
date: parseDate("Fri, 21 Nov 1997 09:55:06 -0600"),
htmlBody: "",
- textBody: "Time for the egg.",
+ textBody: "Time for the egg.",
},
14: {
mailData: data3,
@@ -563,10 +563,14 @@ So, "Hello".`,
t.Error(err)
}
- if ra.Filename == ad.filename && string(b) == ad.data && ra.ContentType == ad.contentType {
+ if ra.Filename == ad.filename && ra.ContentType == ad.contentType {
found = true
attachs = append(attachs[:i], attachs[i+1:]...)
}
+
+ if string(b) != ad.data {
+ t.Errorf("[Test Case %v] Bad data for attachment: \nEXPECTED:\n%s\nHAVE:\n%s", index, ad.data, string(b))
+ }
}
if !found {
@@ -623,9 +627,9 @@ func parseDate(in string) time.Time {
}
type attachmentData struct {
- filename string
- contentType string
- data string
+ filename string
+ contentType string
+ data string
}
type embeddedFileData struct {
@@ -869,8 +873,8 @@ Message-ID: <5678.21-Nov-1997@example.com>
Hi everyone.
`
-//todo: not yet implemented in net/mail
-//once there is support for this, add it
+// todo: not yet implemented in net/mail
+// once there is support for this, add it
var rfc5322exampleA13 = `From: Pete
To: A Group:Ed Jones ,joe@where.test,John ;
Cc: Undisclosed recipients:;
@@ -880,7 +884,7 @@ Message-ID:
Testing.
`
-//we skipped the first message bcause it's the same as A 1.1
+// we skipped the first message bcause it's the same as A 1.1
var rfc5322exampleA2a = `From: Mary Smith
To: John Doe
Reply-To: "Mary Smith: Personal Account"