Factor out and expose prefixed name predicates

Towards using these in the writer to escape names more precisely.

5 files changed, 63 insertions, 63 deletions

diff --git a/src/n3.c b/src/n3.c
index dc133e69..1374ae7c 100644
--- a/src/n3.c
+++ b/src/n3.c
@@ -4,11 +4,13 @@
 #include "byte_source.h"
 #include "namespaces.h"
 #include "node.h"
+#include "ntriples.h"
 #include "read_ntriples.h"
 #include "reader.h"
 #include "stack.h"
 #include "string_utils.h"
 #include "try.h"
+#include "turtle.h"
 
 #include "serd/node.h"
 #include "serd/reader.h"
@@ -170,33 +172,9 @@ read_PN_LOCAL_ESC(SerdReader* const reader, SerdNode* const dest)
   skip_byte(reader, '\\');
 
   const int c = peek_byte(reader);
-  switch (c) {
-  case '!':
-  case '#':
-  case '$':
-  case '%':
-  case '&':
-  case '\'':
-  case '(':
-  case ')':
-  case '*':
-  case '+':
-  case ',':
-  case '-':
-  case '.':
-  case '/':
-  case ';':
-  case '=':
-  case '?':
-  case '@':
-  case '_':
-  case '~':
-    return push_byte(reader, dest, eat_byte_safe(reader, c));
-  default:
-    break;
-  }
 
-  return r_err(reader, SERD_BAD_SYNTAX, "invalid escape");
+  return is_PN_LOCAL_ESC(c) ? push_byte(reader, dest, eat_byte_safe(reader, c))
+                            : r_err(reader, SERD_BAD_SYNTAX, "invalid escape");
 }
 
 static SerdStatus
@@ -492,7 +470,7 @@ read_verb(SerdReader* const reader, SerdNode** const dest)
   SerdNode* node    = *dest;
   const int next    = peek_byte(reader);
   if (node->length == 1 && serd_node_string(node)[0] == 'a' && next != ':' &&
-      !is_PN_CHARS_BASE((uint32_t)next)) {
+      !is_PN_CHARS_BASE(next)) {
     serd_stack_pop_to(&reader->stack, orig_stack_size);
     return ((*dest = push_node(reader, SERD_URI, NS_RDF "type", 47))
               ? SERD_SUCCESS
diff --git a/src/ntriples.h b/src/ntriples.h
new file mode 100644
index 00000000..404fba71
--- /dev/null
+++ b/src/ntriples.h
@@ -0,0 +1,23 @@
+// Copyright 2011-2021 David Robillard <d@drobilla.net>
+// SPDX-License-Identifier: ISC
+
+#ifndef SERD_SRC_NTRIPLES_H
+#define SERD_SRC_NTRIPLES_H
+
+#include "string_utils.h"
+
+#include <stdbool.h>
+
+static inline bool
+is_PN_CHARS_BASE(const int c)
+{
+  return (is_alpha(c) || in_range(c, 0x000C0U, 0x000D6U) ||
+          in_range(c, 0x000D8U, 0x000F6U) || in_range(c, 0x000F8U, 0x002FFU) ||
+          in_range(c, 0x00370U, 0x0037DU) || in_range(c, 0x0037FU, 0x01FFFU) ||
+          in_range(c, 0x0200CU, 0x0200DU) || in_range(c, 0x02070U, 0x0218FU) ||
+          in_range(c, 0x02C00U, 0x02FEFU) || in_range(c, 0x03001U, 0x0D7FFU) ||
+          in_range(c, 0x0F900U, 0x0FDCFU) || in_range(c, 0x0FDF0U, 0x0FFFDU) ||
+          in_range(c, 0x10000U, 0xEFFFFU));
+}
+
+#endif // SERD_SRC_NTRIPLES_H
diff --git a/src/read_ntriples.c b/src/read_ntriples.c
index 3d3af2ce..96748d64 100644
--- a/src/read_ntriples.c
+++ b/src/read_ntriples.c
@@ -5,6 +5,7 @@
 
 #include "caret.h"
 #include "node.h"
+#include "ntriples.h"
 #include "read_utf8.h"
 #include "reader.h"
 #include "stack.h"
@@ -23,32 +24,6 @@
 #include <stdint.h>
 #include <stdio.h>
 
-// Utilities
-
-static inline bool
-codepoint_in_range(const uint32_t c, const uint32_t min, const uint32_t max)
-{
-  return c >= min && c <= max;
-}
-
-bool
-is_PN_CHARS_BASE(const uint32_t c)
-{
-  return (codepoint_in_range(c, 'A', 'Z') || codepoint_in_range(c, 'a', 'z') ||
-          codepoint_in_range(c, 0x000C0U, 0x000D6U) ||
-          codepoint_in_range(c, 0x000D8U, 0x000F6U) ||
-          codepoint_in_range(c, 0x000F8U, 0x002FFU) ||
-          codepoint_in_range(c, 0x00370U, 0x0037DU) ||
-          codepoint_in_range(c, 0x0037FU, 0x01FFFU) ||
-          codepoint_in_range(c, 0x0200CU, 0x0200DU) ||
-          codepoint_in_range(c, 0x02070U, 0x0218FU) ||
-          codepoint_in_range(c, 0x02C00U, 0x02FEFU) ||
-          codepoint_in_range(c, 0x03001U, 0x0D7FFU) ||
-          codepoint_in_range(c, 0x0F900U, 0x0FDCFU) ||
-          codepoint_in_range(c, 0x0FDF0U, 0x0FFFDU) ||
-          codepoint_in_range(c, 0x10000U, 0xEFFFFU));
-}
-
 /**
    Read an initial prefixed name character.
 
@@ -484,7 +459,7 @@ read_PN_CHARS_BASE(SerdReader* const reader, SerdNode* const dest)
 
   TRY(st, read_utf8_code_point(reader, dest, &code, (uint8_t)c));
 
-  if (!is_PN_CHARS_BASE(code)) {
+  if (!is_PN_CHARS_BASE((int)code)) {
     r_err(
       reader, SERD_BAD_SYNTAX, "U+%04X is not a valid name character", code);
     if (reader->strict) {
@@ -526,7 +501,7 @@ read_PN_CHARS(SerdReader* const reader, SerdNode* const dest)
   uint32_t code = 0U;
   TRY(st, read_utf8_code_point(reader, dest, &code, (uint8_t)c));
 
-  if (!is_PN_CHARS_BASE(code) && code != 0xB7 &&
+  if (!is_PN_CHARS_BASE((int)code) && code != 0xB7 &&
       !(code >= 0x0300 && code <= 0x036F) &&
       !(code >= 0x203F && code <= 0x2040)) {
     return r_err(
diff --git a/src/read_ntriples.h b/src/read_ntriples.h
index e6051fde..58daae0c 100644
--- a/src/read_ntriples.h
+++ b/src/read_ntriples.h
@@ -14,14 +14,6 @@
 // Utilities
 
 /**
-   Return true if the codepoint `c` is a valid PN_CHARS_BASE character.
-
-   RDF 1.1 NTriples: [157s] PN_CHARS_BASE
-*/
-bool
-is_PN_CHARS_BASE(uint32_t c);
-
-/**
    Read one (possibly multi-byte) character.
 
    The caller must have already eaten the first byte, `c`.
diff --git a/src/turtle.h b/src/turtle.h
new file mode 100644
index 00000000..6e7e3a8d
--- /dev/null
+++ b/src/turtle.h
@@ -0,0 +1,32 @@
+// Copyright 2011-2020 David Robillard <d@drobilla.net>
+// SPDX-License-Identifier: ISC
+
+#ifndef SERD_SRC_TURTLE_H
+#define SERD_SRC_TURTLE_H
+
+#include "ntriples.h"
+#include "string_utils.h"
+
+#include <stdbool.h>
+#include <string.h>
+
+static inline bool
+is_PN_CHARS_U(const int c)
+{
+  return c == '_' || is_PN_CHARS_BASE(c);
+}
+
+static inline bool
+is_PN_CHARS(const int c)
+{
+  return (is_PN_CHARS_U(c) || c == '-' || in_range(c, '0', '9') || c == 0xB7 ||
+          (c >= 0x0300 && c <= 0x036F) || (c >= 0x203F && c <= 0x2040));
+}
+
+static inline bool
+is_PN_LOCAL_ESC(const int c)
+{
+  return strchr("!#$%&\'()*+,-./;=?@_~", c) != NULL;
+}
+
+#endif // SERD_SRC_TURTLE_H