Integrate all previous PRs

.circleci/config.yml

@@ -1,12 +1,14 @@
-# Golang CircleCI 2.0 configuration file
+# Golang CircleCI 2.1 configuration file
 #
 # Check https://circleci.com/docs/2.0/language-go/ for more details
-version: 2
+version: 2.1
 jobs:
   build:
     docker:
       - image: circleci/golang:1.14
     working_directory: /go/src/github.com/{{ORG_NAME}}/{{REPO_NAME}}
     steps:
       - checkout
-      - run: go test -v ./...
+      - run: GO111MODULE=off go get github.com/mattn/goveralls
+      - run: go test -v -cover -race -coverprofile=coverage.out ./...
+      - run: $GOPATH/bin/goveralls -coverprofile=coverage.out -service=circle-ci -repotoken=$COVERALLS_TOKEN

CONTRIBUTING.md

@@ -16,4 +16,4 @@ Every change merges to master. No development is done in other branches.
 - Wait for CI to verify you didn't break anything
     - If you did, rewrite it
 - If CI passes, wait for manual review by repo's owner
-- Your pull request will be merged into master
+- Your pull request will be merged into master

README.md

@@ -1,6 +1,8 @@
 # Parsemail - simple email parsing Go library
 
-[![Build Status](https://circleci.com/gh/DusanKasan/parsemail.svg?style=shield&circle-token=:circle-token)](https://circleci.com/gh/DusanKasan/parsemail) [![Coverage Status](https://coveralls.io/repos/github/DusanKasan/Parsemail/badge.svg?branch=master)](https://coveralls.io/github/DusanKasan/Parsemail?branch=master) [![Go Report Card](https://goreportcard.com/badge/github.com/DusanKasan/parsemail)](https://goreportcard.com/report/github.com/DusanKasan/parsemail)
+[![Build Status](https://circleci.com/gh/k3a/parsemail.svg?style=shield)](https://circleci.com/gh/k3a/parsemail) [![Coverage Status](https://coveralls.io/repos/github/k3a/parsemail/badge.svg?branch=master)](https://coveralls.io/github/k3a/parsemail?branch=master) [![Go Report Card](https://goreportcard.com/badge/github.com/k3a/parsemail)](https://goreportcard.com/report/github.com/k3a/parsemail)
+
+> The [original repo](https://github.com/DusanKasan/parsemail) is unmaintained. This is an active fork replacing the original repo. Contributions are welcome.
 
 This library allows for parsing an email message into a more convenient form than the `net/mail` provides. Where the `net/mail` just gives you a map of header fields and a `io.Reader` of its body, Parsemail allows access to all the standard header fields set in [RFC5322](https://tools.ietf.org/html/rfc5322), html/text body as well as attachements/embedded content as binary streams with metadata.
 
@@ -55,4 +57,4 @@ for _, a := range(email.EmbeddedFiles) {
     fmt.Println(a.ContentType)
     //and read a.Data
 }
-```
+```

go.mod

@@ -1,3 +1,3 @@
-module github.com/DusanKasan/parsemail
+module github.com/k3a/parsemail
 
-go 1.12
+go 1.12

parsemail.go

@@ -8,6 +8,7 @@ import (
 	"io/ioutil"
 	"mime"
 	"mime/multipart"
+	"mime/quotedprintable"
 	"net/mail"
 	"strings"
 	"time"
@@ -18,6 +19,7 @@ const contentTypeMultipartAlternative = "multipart/alternative"
 const contentTypeMultipartRelated = "multipart/related"
 const contentTypeTextHtml = "text/html"
 const contentTypeTextPlain = "text/plain"
+const contentTypeOctetStream = "application/octet-stream"
 
 // Parse an email message read from io.Reader into parsemail.Email struct
 func Parse(r io.Reader) (email Email, err error) {
@@ -37,6 +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"])
@@ -46,12 +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
@@ -103,32 +131,67 @@ func parseContentType(contentTypeHeader string) (contentType string, params map[
 	return mime.ParseMediaType(contentTypeHeader)
 }
 
+func parseAttachmentOnlyEmail(body io.Reader, header mail.Header) (attachments []Attachment, err error) {
+	contentDisposition := header.Get("Content-Disposition")
+
+	if len(contentDisposition) > 0 && strings.Contains(contentDisposition, "attachment;") {
+
+		attachmentData, err := decodeContent(body, header.Get("Content-Transfer-Encoding"))
+		if err != nil {
+			return attachments, err
+		}
+
+		fileName := strings.Replace(contentDisposition, "attachment; filename=\"", "", -1)
+		fileName = strings.TrimRight(fileName, "\"")
+
+		at := Attachment{
+			Filename:    fileName,
+			ContentType: "application/octet-stream",
+			Data:        attachmentData,
+		}
+
+		attachments = append(attachments, at)
+	}
+
+	return attachments, nil
+}
+
 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
 		} else if err != nil {
 			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
 		}
 
 		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
 			}
@@ -163,29 +226,39 @@ 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
 		} else if err != nil {
 			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
 		}
 
 		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
 			}
@@ -220,18 +293,38 @@ 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 {
 			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)
+			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 {
@@ -243,26 +336,27 @@ 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
 			}
-
-			htmlBody += strings.TrimSuffix(string(ppContent[:]), "\n")
-		} else if isAttachment(part) {
-			at, err := decodeAttachment(part)
+			ppContent, err := ioutil.ReadAll(decoded)
 			if err != nil {
 				return textBody, htmlBody, attachments, embeddedFiles, err
 			}
 
-			attachments = append(attachments, at)
+			htmlBody += strings.TrimSuffix(string(ppContent[:]), "\n")
 		} else {
 			return textBody, htmlBody, attachments, embeddedFiles, fmt.Errorf("Unknown multipart/mixed nested mime type: %s", contentType)
 		}
@@ -345,24 +439,32 @@ func decodeAttachment(part *multipart.Part) (at Attachment, err error) {
 }
 
 func decodeContent(content io.Reader, encoding string) (io.Reader, error) {
-	switch encoding {
+	switch strings.ToLower(encoding) {
 	case "base64":
 		decoded := base64.NewDecoder(base64.StdEncoding, content)
 		b, err := ioutil.ReadAll(decoded)
 		if err != nil {
 			return nil, err
 		}
-
 		return bytes.NewReader(b), nil
-	case "7bit":
-		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)
 	}
@@ -483,11 +585,11 @@ type Email struct {
 	ResentMessageID string
 
 	ContentType string
-	Content io.Reader
+	Content     io.Reader
 
 	HTMLBody string
 	TextBody string
 
 	Attachments   []Attachment
 	EmbeddedFiles []EmbeddedFile
-}
+}