Summary
Forwarded messages whose subjects use common FW: or differently-cased Fwd: prefixes are processed as replies instead of forwards when Macro computes body_replyless.
Current implementation
Both crates/email_utils/src/body_replyless/html_parser.rs and plaintext_parser.rs bypass quoted-content removal only when the subject starts with the exact case-sensitive string Fwd:. Providers and clients also emit forms such as FW:, fw:, and fwd:.
When those forms are not recognized, the normal reply splitter can remove the forwarded message content from the parsed and search representations.
Expected result
Fwd: and FW: are recognized case-insensitively.- Leading subject whitespace does not prevent recognition.
- Words that merely begin with
fw or fwd are not treated as forward prefixes.
- HTML and plaintext extraction behave consistently.
Suggested direction
Use one small subject-prefix predicate from both extraction paths and cover it through the public body-replyless behavior.
Non-goals
- Adding localized forward prefixes.
- Changing how new forward subjects are generated in the web composer.
Summary
Forwarded messages whose subjects use common
FW:or differently-casedFwd:prefixes are processed as replies instead of forwards when Macro computesbody_replyless.Current implementation
Both
crates/email_utils/src/body_replyless/html_parser.rsandplaintext_parser.rsbypass quoted-content removal only when the subject starts with the exact case-sensitive stringFwd:. Providers and clients also emit forms such asFW:,fw:, andfwd:.When those forms are not recognized, the normal reply splitter can remove the forwarded message content from the parsed and search representations.
Expected result
Fwd:andFW:are recognized case-insensitively.fworfwdare not treated as forward prefixes.Suggested direction
Use one small subject-prefix predicate from both extraction paths and cover it through the public body-replyless behavior.
Non-goals