Fix SPARQL syntax to support internal dot characters

fastobo · Feb 25, 2024 · 4bed022 · 4bed022
1 parent e66b6df
commit 4bed022
Showing 1 changed file with 17 additions and 6 deletions.
diff --git a/src/sparql.pest b/src/sparql.pest
@@ -4,16 +4,27 @@ SPARQL_BlankNodeLabel = ${ "_:" ~ SPARQL_PnLocal }
 SPARQL_PnameNs = ${ SPARQL_PnPrefix? ~ ":" }
 SPARQL_PnameLn = ${ SPARQL_PnameNs ~ SPARQL_PnLocal }
 
-SPARQL_PnPrefixDotEnding = @{ SPARQL_PnCharsBase ~ SPARQL_PnChars* ~ "." }
-SPARQL_PnPrefix = @{ !SPARQL_PnPrefixDotEnding ~ SPARQL_PnCharsBase ~ SPARQL_PnChars* }
-SPARQL_PnLocalDotEnding = @{ (SPARQL_PnCharsU | '0'..'9') ~ SPARQL_PnChars* ~ "." }
-SPARQL_PnLocal = @{ !SPARQL_PnLocalDotEnding ~ (SPARQL_PnCharsU | '0'..'9') ~ SPARQL_PnChars* } // FIXME
+// NOTE: In the SPARQL grammar a local identifier can contain or end with a
+//       semicolon, however this conflicts with most of the rules including
+//       several keyword terminals, so I made it so that the rule only accepts
+//       escaped semi-colons.
+SPARQL_PnLocal        = @{ (SPARQL_PnCharsU | ":" | ASCII_DIGIT | SPARQL_Plx) ~ (("." ~ (SPARQL_PnChars | SPARQL_Plx)) | SPARQL_PnChars | SPARQL_Plx)* }
+SPARQL_PnPrefix       = @{ SPARQL_PnCharsBase ~ (("." ~ SPARQL_PnChars) | SPARQL_PnChars)* }
 
-SPARQL_PnChars = @{ SPARQL_PnCharsU | "-" | '0'..'9' | "\u{00B7}" | '\u{0300}'..'\u{036F}' |  '\u{203F}'..'\u{2040}' }
+SPARQL_PnChars = @{ SPARQL_PnCharsU | "-" | ASCII_DIGIT | "\u{00B7}" | '\u{0300}'..'\u{036F}' |  '\u{203F}'..'\u{2040}' }
 SPARQL_PnCharsU = @{ SPARQL_PnCharsBase | "_" }
 SPARQL_PnCharsBase = @{
-  'A'..'Z' | 'a'..'z' | '\u{00C0}'..'\u{00D6}' | '\u{00D8}'..'\u{00F6}' | '\u{00F8}'..'\u{02FF}'
+  ASCII_ALPHA | '\u{00C0}'..'\u{00D6}' | '\u{00D8}'..'\u{00F6}' | '\u{00F8}'..'\u{02FF}'
   | '\u{0370}'..'\u{037D}' | '\u{037F}'..'\u{1FFF}' | '\u{200C}'..'\u{200D}' | '\u{2070}'..'\u{218F}'
   | '\u{2C00}'..'\u{2FEF}' | '\u{3001}'..'\u{D7FF}' | '\u{F900}'..'\u{FDCF}' | '\u{FDF0}'..'\u{FFFD}'
   | '\u{10000}'..'\u{EFFFF}'
+}
+
+SPARQL_Plx            = @{ SPARQL_Percent | SPARQL_PnLocalEsc }
+SPARQL_Percent        = @{ "%" ~ SPARQL_Hex ~ SPARQL_Hex }
+SPARQL_Hex            = @{ ASCII_HEX_DIGIT }
+
+SPARQL_PnLocalEsc     = @{ "\\" ~ SPARQL_PnLocalEscChar }
+SPARQL_PnLocalEscChar = @{
+  "_" | "~" | "." | "-" | "!" | "$" | "&" | "\"" | "(" | ")" | "*" | "+" | "," | ";" | "=" | "/" | "?" | "#" | "@" | "%" | ":"
 }