From 2d72a1aa64a66c486a2329999123c14afcddeb32 Mon Sep 17 00:00:00 2001
From: Taylan Kammer <taylan.kammer@gmail.com>
Date: Fri, 9 Jan 2026 18:09:59 +0100
Subject: More grammar fuckery.  BNF is horrible!

---
 spec/syntax.abnf | 65 +++++++++++++++++++++++++++++++++++---------------------
 1 file changed, 41 insertions(+), 24 deletions(-)

(limited to 'spec/syntax.abnf')

diff --git a/spec/syntax.abnf b/spec/syntax.abnf
index a083eda..132deeb 100644
--- a/spec/syntax.abnf
+++ b/spec/syntax.abnf
@@ -6,42 +6,52 @@ File          = [Unit] *( Blank Unit ) *Blank [Trail]
 
 Unit          = *Blank Datum
 
-Blank         = HTAB / LF / %x0b / %x0c / CR / Comment
+Blank         = HTAB / LF / %x0b / %x0c / CR / SP / Comment
 
 Trail         = SkipLine / SkipUnit
 
 
+Datum         = BareString
+              / DottedString
+              / CladDatum
+              / HashExpr
+              / HashDotExpr
+              / QuoteExpr
+              / JoinExpr
+
 Comment       = SkipLine LF / SkipUnit Blank
 
 SkipLine      = ';' [ SkipLStart *AnyButLF ]
 
-SkipUnit      = ';' '~' Unit
-
-
 SkipLStart    = %x00-09 / %x0b-7d / %x7f-ff
               ; any but LF or '~'
 
 AnyButLF      = %x00-09 / %x0b-ff
 
-
-Datum         = SingleDatum
-              / JoinedDatum *( [ '.' / ':' ] JoinedDatum )
-
-
-SingleDatum   = BareString / CladDatum / DottedString
-
-JoinedDatum   = BareString / CladDatum
+SkipUnit      = ';' '~' Unit
 
 
 BareString    = BareChar *( BareChar / Numeric )
 
+DottedString  = ( '.' / Numeric ) *( '.' / Numeric / BareChar )
+
 CladDatum     = '|' *( PipeStrChar / '\' StringEsc ) '|'
               / '"' *( QuotStrChar / '\' StringEsc ) '"'
-              / '#' HashExpr
               / '(' List ')' / '[' List ']' / '{' List '}'
-              / "'" Datum / '`' Datum / ',' Datum
 
-DottedString  = ( '.' / Numeric ) *( '.' / Numeric / BareChar )
+HashExpr      = LabelExpr / RuneExpr / HashDatum
+
+HashDotExpr   = RuneDotExpr / HashDotDatum
+
+QuoteExpr     = "'" Datum / '`' Datum / ',' Datum
+
+JoinExpr      = Datum LeftCladDatum
+              / Datum ':' Datum
+              / DotlessDatum '.' Datum
+
+LeftCladDatum = CladDatum / HashExpr / QuoteExpr
+
+DotlessDatum  = BareString / CladDatum / RuneExpr / HashDatum
 
 
 BareChar      = '!' / '$' / '%' / '*' / '/' / '<' / '=' / '>'
@@ -49,29 +59,36 @@ BareChar      = '!' / '$' / '%' / '*' / '/' / '<' / '=' / '>'
 
 Numeric       = '+' / '-' / DIGIT
 
-
 PipeStrChar   = %x00-5b / %x5d-7b / %x7d-ff
               ; any but '|' or '\'
 
 QuotStrChar   = %x00-21 / %x23-5b / %x5d-ff
               ; any but '"' or '\'
 
-HashExpr      = Rune [ '\' BareString / CladDatum ]
-              / '\' BareString
-              / '%' Label ( '%' / '=' Datum )
-              / CladDatum
-
 List          = [Unit] *( Blank Unit ) *Blank [Tail] [SkipUnit]
 
 Tail          = '&' Unit *Blank
 
+LabelExpr     = '#' '%' Label ( '%' / '=' Datum )
+
+RuneExpr      = '#' Rune [ '\' BareString / CladDatum ]
+
+RuneDotExpr   = '#' Rune '\' DottedString
+
+HashDatum     = '#' '\' BareString / CladDatum
+
+HashDotDatum  = '#' '\' DottedString
+
 
+; Unicode escapes must not represent surrogate code points.
+; This is difficult to express in ABNF.  But we do at least
+; disallow code points greater than \u10FFFF which are also
+; invalid, since U+10FFFF is the highest allowed.
 StringEsc     = '\' / '|' / '"' / *( HTAB / SP ) LF *( HTAB / SP )
               / 'a' / 'b' / 't' / 'n' / 'v' / 'f' / 'r' / 'e'
               / 'x' 1*( 2HEXDIG ) ';'
-              / 'u' 1*5HEXDIG ';'
-              / 'u' '0' 1*5HEXDIG ';'
-              / 'u' '1' '0' 1*4HEXDIG ';'
+              / 'u' ['0'] 1*5HEXDIG ';'
+              / 'u' '1' '0' 4HEXDIG ';'
 
 
 Rune          = ALPHA *5( ALPHA / DIGIT )
-- 
cgit v1.2.3