From 9b1139fe7045a0630e87501235af21803860b80c Mon Sep 17 00:00:00 2001
From: David Robillard <d@drobilla.net>
Date: Fri, 1 Dec 2023 21:59:18 -0500
Subject: [WIP] Add support for reading and writing variables

[WIP] Command line option, move later?

This adds a reader flag and serdi option for extending a syntax with support
for SPARQL-like variables, for storing things like patterns or simple queries.
---
 src/read_ntriples.c | 52 +++++++++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 51 insertions(+), 1 deletion(-)

(limited to 'src/read_ntriples.c')

diff --git a/src/read_ntriples.c b/src/read_ntriples.c
index da726658..bec59c13 100644
--- a/src/read_ntriples.c
+++ b/src/read_ntriples.c
@@ -520,6 +520,51 @@ read_HEX(SerdReader* const reader)
   return 0;
 }
 
+/**
+   Read a variable name, starting after the '?' or '$'.
+
+   This is an extension that serd uses in certain contexts to support patterns.
+
+   Restricted version of SPARQL 1.1: [166] VARNAME
+*/
+static SerdStatus
+read_VARNAME(SerdReader* const reader, SerdNode** const dest)
+{
+  // Simplified from SPARQL: VARNAME ::= (PN_CHARS_U | [0-9])+
+  SerdNode*  n  = *dest;
+  SerdStatus st = SERD_SUCCESS;
+  int        c  = 0;
+
+  while ((c = peek_byte(reader))) {
+    if (is_digit(c) || c == '_') {
+      st = push_byte(reader, n, eat_byte_safe(reader, c));
+    } else if ((st = read_PN_CHARS(reader, n))) {
+      st = st > SERD_FAILURE ? st : SERD_SUCCESS;
+      break;
+    }
+  }
+
+  return st;
+}
+
+SerdStatus
+read_Var(SerdReader* const reader, SerdNode** const dest)
+{
+  if (!(reader->flags & SERD_READ_VARIABLES)) {
+    return r_err(reader, SERD_BAD_SYNTAX, "syntax does not support variables");
+  }
+
+  const int c = peek_byte(reader);
+  assert(c == '$' || c == '?');
+  skip_byte(reader, c);
+
+  if (!(*dest = push_node(reader, SERD_VARIABLE, "", 0))) {
+    return SERD_BAD_STACK;
+  }
+
+  return read_VARNAME(reader, dest);
+}
+
 // Nonterminals
 
 // comment ::= '#' ( [^#xA #xD] )*
@@ -576,6 +621,8 @@ read_nt_subject(SerdReader* const reader, SerdNode** const dest)
   switch (peek_byte(reader)) {
   case '<':
     return read_IRI(reader, dest);
+  case '?':
+    return read_Var(reader, dest);
   case '_':
     return read_BLANK_NODE_LABEL(reader, dest, &ate_dot);
   default:
@@ -589,7 +636,8 @@ read_nt_subject(SerdReader* const reader, SerdNode** const dest)
 SerdStatus
 read_nt_predicate(SerdReader* const reader, SerdNode** const dest)
 {
-  return read_IRI(reader, dest);
+  return (peek_byte(reader) == '?') ? read_Var(reader, dest)
+                                    : read_IRI(reader, dest);
 }
 
 /// [4] object
@@ -605,6 +653,8 @@ read_nt_object(SerdReader* const reader,
     return read_literal(reader, dest);
   case '<':
     return read_IRI(reader, dest);
+  case '?':
+    return read_Var(reader, dest);
   case '_':
     return read_BLANK_NODE_LABEL(reader, dest, ate_dot);
   default:
-- 
cgit v1.2.1