Added a new example email and modified tests to include an RFC 5322 format. Refactored the code to handle cases where the Content-Type header is missing by defaulting to plain text (7bit, US-ASCII). Deleted redundant test cases that no longer apply.
Enhanced EML parsing to handle 7bit and 8bit content transfer encodings. Updated related test cases to verify the correct handling of these encodings. This ensures compliance with RFC2045 for defaulting to 7bit when no encoding is specified.
Implemented a new encoding type, `EncodingUSASCII`, representing US-ASCII (7bit) encoding. Updated the test suite to include this new encoding in the test cases, ensuring proper validation and coverage.
The GetBoundary method has been added to the Msg struct in msg.go. Alongside this, a corresponding test method, TestMsg_GetBoundary, was introduced in msg_test.go to ensure the proper working of GetBoundary. This new method returns the boundary string of the Msg, which was previously inaccessible.
This PR introduces an EML parser to go-mail. It allows to read generic EML data from a file, a string or a reader into a go-mail Msg struct. It supports all types of message parts and encodings. It should be able to recognize Mulitpart messages as well as attachments and embeds (inline attachments).
This PR closes#145
Added conditional statements to handle potential failures when writing the EML string to the temporary files during testing. These updates ensure test failures due to write errors are accurately reflected in test results. Also, a minor fix is implemented on file permission in the os.WriteFile function.
The README has been updated to show that go-mail now supports outputting a message as an EML file and parsing an EML file into a go-mail message. This addition enhances the flexibility and control over the email content and format.
The change introduces a new unit test, TestFile_WithContentID, in file_test.go. This test aims to verify the correct function of the WithFileContentID option by using differing scenarios, with assertions for error control and validation of expected content.
In the eml.go and file.go files, the function WithContentID has been renamed to WithFileContentID. This aligns more accurately with the function purpose, which is to set the content ID for a File object.
This commit introduces tests for different email scenarios including emails with no boundaries, emails with attachments, emails with embedded items, and multipart mixed emails. The tests ensure that the code correctly returns all expected parts of the email (e.g., embeds, attachments, parts) and that it correctly processes the email subject.
The debug print statement that outputs the content type of the email has been removed from eml.go. This change improves code cleanliness and avoids unnecessary console output in production.
The comment in the eml.go file was extended to include the possibility of 'Multipart/alternative' parts. Previously, it only mentioned 'Multipart/related' parts. The actual code functionality remains unchanged, this is purely a clarification in the documentation.
The EML parser now includes logic to manage 'multipart/alternative' content types. This adjustment is made within the section handling 'multipart/related' parts, allowing for better handling and parsing of varying content types.
The updated code adds base64 encoding support to email attachments and inline content in eml.go. It does this by introducing a new dataReader which uses a base64 decoder if the content transfer encoding is base64. With this update, attachments with base64 content will be correctly decoded when processed.
The newly added function, WithContentID, allows for setting the Content-ID header for the File. This provides enhanced handling and differentiation of files.
The commit modifies the parseMultiPartHeader function to handle optional fields accurately. The delimiter was changed from "; " to ";" and whitespace is being trimmed from the start of optional fields to ensure correct splitting and mapping.
The error message previously referenced a constant 'HeaderTo' which might not always be the header being parsed. The commit replaces this with 'addrHeader', significantly improving the accuracy of error messages.
