Improved Email Grammar

book/src/generated/lang-support.md

@@ -178,7 +178,7 @@
 | lua-format-string | ✓ |  |  |  |  |  |
 | luap | ✓ |  |  |  |  |  |
 | luau | ✓ | ✓ | ✓ |  |  | `luau-lsp` |
-| mail | ✓ | ✓ |  |  | ✓ |  |
+| mail | ✓ | ✓ | ✓ | ✓ | ✓ |  |
 | make | ✓ |  | ✓ |  |  |  |
 | markdoc | ✓ |  |  |  |  | `markdoc-ls` |
 | markdown | ✓ |  |  | ✓ |  | `marksman`, `markdown-oxide`, `rumdl` |

languages.toml

@@ -2094,13 +2094,13 @@ source = { git = "https://github.com/Flakebi/tree-sitter-tablegen", rev = "3e9c4
 
 [[language]]
 name = "mail"
-scope = "text.mail"
-file-types = ["eml"]
+scope = "source.mail"
+file-types = ["eml", "email"]
 injection-regex = "mail|eml|email"
 
 [[grammar]]
 name = "mail"
-source = { git = "https://codeberg.org/ficd/tree-sitter-mail", rev = "5eddbfdbec4c893182c79047499901c196332e78" }
+source = { git = "https://codeberg.org/philocalyst/MailGrammar", rev = "631ce96da3fcf6a817f599d8667ef985438f5d36" }
 
 [[language]]
 name = "markdown"

runtime/queries/mail/folds.scm

@@ -0,0 +1,18 @@
+; Fold the entire header block.
+(header) @fold
+
+; Fold long recipient lists and individual Received: trace headers.
+(address_list) @fold
+(received_field) @fold
+
+; Fold MIME sections independently.
+(mime_part) @fold
+(part_headers) @fold
+(part_body) @fold
+(multipart_body) @fold
+
+; Fold body paragraphs, quote groups, and signatures.
+(body_block) @fold
+(quote_group) @fold
+(quoted_block) @fold
+(signature) @fold

runtime/queries/mail/highlights.scm

@@ -1,28 +1,151 @@
-; header fields
-[
-  (header_field_email)
-  (header_field_subject)
-  (header_field)
-] @keyword
-
-; delimited punctuation
-(header_separator) @punctuation.delimiter
-(email_delimiter) @punctuation.delimiter
-
-; email subject contents
-(header_subject
-  (subject) @markup.bold)
-; extra metadata headers
-(header_other
-  (header_unstructured) @comment)
-
-; Addressee Name (Firstname, Lastname, etc.)
-(atom) @variable
-
-; Email Address
-(email) @string
-
-; Quoted Reply
-(quote_marker) @punctuation.special
+(ERROR)          @error
+(malformed_line) @error
+
+(address_field                    name: (field_name) @constant)
+(date_field                       name: (field_name) @constant)
+(message_id_field                 name: (field_name) @constant)
+(in_reply_to_field                name: (field_name) @constant)
+(references_field                 name: (field_name) @constant)
+(received_field                   name: (field_name) @constant)
+(return_path_field                name: (field_name) @constant)
+(subject_field                    name: (field_name) @constant)
+(comments_field                   name: (field_name) @constant)
+(content_type_field               name: (field_name) @constant)
+(content_transfer_encoding_field  name: (field_name) @constant)
+(content_disposition_field        name: (field_name) @constant)
+(mime_version_field               name: (field_name) @constant)
+(content_description_field        name: (field_name) @constant)
+
+; Covers non-ASCII field names as well, which the grammar degrades to
+; optional_field
+(optional_field name: (field_name) @property)
+
+(subject_field value: (unstructured) @markup.heading)
+(subject_field prefix: (reply_forward_marker) @comment)
+
+(comments_field            value: (unstructured) @string)
+(content_description_field value: (unstructured) @string)
+(optional_field            value: (unstructured) @string)
+(content_type_field        value: (unstructured) @string)
+(mime_version_field        value: (unstructured) @string)
+
+; Obsolete forms (2-digit year, named/military timezones like GMT) share
+; the same `year`/`zone` node types as their modern equivalents.
+(date_time)   @string.special
+(day_of_week) @string.special
+(day)         @string.special
+(month)       @string.special
+(year)        @string.special
+(hour)        @string.special
+(minute)      @string.special
+(second)      @string.special
+(zone)        @string.special
+
+(msg_id) @string.special
+(msg_id "<" @punctuation.bracket)
+(msg_id ">" @punctuation.bracket)
+(msg_id "@" @punctuation.delimiter)
+
+; addr_spec (local@domain) captured without surrounding <> brackets.
+(addr_spec) @markup.link.url
+(addr_spec "@" @punctuation.delimiter)
+
+(domain_literal) @string.special
+
+(angle_addr "<" @punctuation.bracket)
+(angle_addr ">" @punctuation.bracket)
+
+(mailbox display_name: (display_name) @string)
+(group   display_name: (display_name) @string)
+
+(group ":" @punctuation.delimiter)
+(group ";" @punctuation.delimiter)
+
+(encoded_word) @constant
+
+(mailbox_sep "," @punctuation.delimiter)
+
+(route) @string.special
+
+(return_path_value) @string.special
+
+(mime_type)    @type
+(mime_subtype) @type
+(content_type "/" @punctuation.delimiter)
+
+(parameter_name)  @property
+(parameter_value) @string
+
+; The quoted form (e.g. filename="report.pdf") needs its own capture.
+(mime_parameter value: (quoted_string) @string)
+(mime_parameter          "=" @punctuation.delimiter)
+(mime_parameter          ";" @punctuation.special)
+(mime_boundary_parameter "=" @punctuation.delimiter)
+(mime_boundary_parameter ";" @punctuation.special)
+
+(encoding_mechanism) @string.special
+(disposition_type)   @string.special
+(mime_version)       @string.special
+
+; Highlight attachment filenames distinctly.
+(content_disposition
+  type: (disposition_type) @_disposition_type
+  (mime_parameter
+    name: (parameter_name) @_param_name
+    value: (_) @label)
+  (#match? @_disposition_type "^[Aa][Tt][Tt][Aa][Cc][Hh][Mm][Ee][Nn][Tt]$")
+  (#match? @_param_name "^[Ff][Ii][Ll][Ee][Nn][Aa][Mm][Ee]"))
+
+(dash_boundary)   @punctuation.special
+(close_delimiter) @punctuation.special
+
+; Base case: every quote marker/line gets markup.quote regardless of depth.
+(quote_marker)   @markup.quote
 (quote_contents) @markup.quote
 
+; Apply progressively deeper quote captures through six nesting levels.
+(quote_group
+  (quoted_block
+    [(quote_marker) (quote_contents)] @markup.quote.1))
+
+(quote_group
+  (quoted_block
+    (quoted_block
+      [(quote_marker) (quote_contents)] @markup.quote.2)))
+
+(quote_group
+  (quoted_block
+    (quoted_block
+      (quoted_block
+        [(quote_marker) (quote_contents)] @markup.quote.3))))
+
+(quote_group
+  (quoted_block
+    (quoted_block
+      (quoted_block
+        (quoted_block
+          [(quote_marker) (quote_contents)] @markup.quote.4)))))
+
+(quote_group
+  (quoted_block
+    (quoted_block
+      (quoted_block
+        (quoted_block
+          (quoted_block
+            [(quote_marker) (quote_contents)] @markup.quote.5))))))
+
+(quote_group
+  (quoted_block
+    (quoted_block
+      (quoted_block
+        (quoted_block
+          (quoted_block
+            (quoted_block
+              [(quote_marker) (quote_contents)] @markup.quote.6)))))))
+
+(signature_separator) @punctuation.special
+(signature)           @comment
+
+(comment) @comment
+
+(received_tokens) @string

runtime/queries/mail/indents.scm

@@ -0,0 +1 @@
+; Email is not an indented language.

runtime/queries/mail/injections.scm

@@ -0,0 +1,64 @@
+; Inject MIME parts based on their Content-Type subtype. A subtype that does
+; not correspond to a configured Helix language is left uninterpreted.
+(mime_part
+  (part_headers
+    (content_type_field
+      value: (content_type
+        subtype: (mime_subtype) @injection.language)))
+  (part_body) @injection.content
+  (#not-any-of? @injection.language "plain" "md" "x-markdown" "rtf" "x-rtf"))
+
+; MIME aliases whose subtype does not match Helix's language name.
+(mime_part
+  (part_headers
+    (content_type_field
+      value: (content_type
+        subtype: (mime_subtype) @_subtype)))
+  (part_body) @injection.content
+  (#any-of? @_subtype "md" "x-markdown")
+  (#set! injection.language "markdown"))
+
+(mime_part
+  (part_headers
+    (content_type_field
+      value: (content_type
+        subtype: (mime_subtype) @_subtype)))
+  (part_body) @injection.content
+  (#any-of? @_subtype "rtf" "x-rtf")
+  (#set! injection.language "rtf"))
+
+; Diff-bearing plain-text bodies receive diff highlighting; other text bodies
+; are rendered as Markdown.
+([
+  (body_block)
+  (epilogue)
+] @injection.content
+ (#match? @injection.content "diff --git ")
+ (#set! injection.language "diff"))
+
+([
+  (body_block)
+  (epilogue)
+] @injection.content
+ (#not-match? @injection.content "diff --git ")
+ (#set! injection.language "markdown"))
+
+(mime_part
+  (part_headers
+    (content_type_field
+      value: (content_type
+        subtype: (mime_subtype) @_subtype)))
+  (part_body) @injection.content
+  (#match? @_subtype "^[Pp][Ll][Aa][Ii][Nn]$")
+  (#match? @injection.content "diff --git ")
+  (#set! injection.language "diff"))
+
+(mime_part
+  (part_headers
+    (content_type_field
+      value: (content_type
+        subtype: (mime_subtype) @_subtype)))
+  (part_body) @injection.content
+  (#match? @_subtype "^[Pp][Ll][Aa][Ii][Nn]$")
+  (#not-match? @injection.content "diff --git ")
+  (#set! injection.language "markdown"))

runtime/queries/mail/locals.scm

@@ -0,0 +1 @@
+; Mail does not have variable scoping.

runtime/queries/mail/tags.scm

@@ -0,0 +1,9 @@
+; Threading headers provide useful definitions and references.
+(message_id_field
+  value: (msg_id) @name) @definition.message
+
+(in_reply_to_field
+  (msg_id) @name) @reference.reply
+
+(references_field
+  (msg_id) @name) @reference.thread

runtime/queries/mail/textobjects.scm

@@ -1,13 +1,59 @@
-(atom_block
-  (atom) @entry.inside) @entry.around
+(address_field
+  value: (address_list) @field.inner @entry.inside) @field.outer @entry.around
 
-(email_address) @entry.around
-(header_other
-  (header_unstructured) @entry.around)
+(subject_field
+  value: (unstructured) @field.inner @entry.inside) @field.outer @entry.around
 
-(quoted_block)+ @comment.around
+(comments_field
+  value: (unstructured) @field.inner @entry.inside) @field.outer @entry.around
 
-(body_block)+ @function.around
+(date_field
+  value: (_) @field.inner @entry.inside) @field.outer @entry.around
 
-(header_subject
-  (subject) @function.around)
+(message_id_field
+  value: (msg_id) @field.inner @entry.inside) @field.outer @entry.around
+
+(in_reply_to_field) @field.outer @entry.around
+(references_field)  @field.outer @entry.around
+
+(received_field) @field.outer @entry.around
+
+(return_path_field
+  value: (return_path_value) @field.inner @entry.inside) @field.outer @entry.around
+
+(content_type_field
+  value: (_) @field.inner @entry.inside) @field.outer @entry.around
+
+(content_transfer_encoding_field
+  value: (encoding_mechanism) @field.inner @entry.inside) @field.outer @entry.around
+
+(content_disposition_field
+  value: (content_disposition) @field.inner @entry.inside) @field.outer @entry.around
+
+(mime_version_field
+  value: (_) @field.inner @entry.inside) @field.outer @entry.around
+
+(content_description_field
+  value: (unstructured) @field.inner @entry.inside) @field.outer @entry.around
+
+(optional_field
+  value: (unstructured) @field.inner @entry.inside) @field.outer @entry.around
+
+(comment) @comment.outer @comment.around
+(comment) @comment.inner @comment.inside
+
+(quote_group) @quote.outer @comment.outer @comment.around
+
+(quoted_block
+  (quote_contents) @quote.inner @comment.inner @comment.inside)
+
+(body
+  (signature_separator)
+  (signature) @signature.inner) @signature.outer
+
+(mime_part
+  (part_body) @mime_part.inner) @mime_part.outer
+
+(body) @body.outer
+(body_block) @body.inner
+(epilogue)   @body.inner