Reduce size of character classification code

3 files changed, 24 insertions, 121 deletions

diff --git a/src/n3.c b/src/n3.c
index 64aa8d7b..073f84d4 100644
--- a/src/n3.c
+++ b/src/n3.c
@@ -56,14 +56,11 @@ read_UCHAR(SerdReader* const reader, const Ref dest, uint32_t* const char_code)
 {
   const int b      = peek_byte(reader);
   unsigned  length = 0;
-  switch (b) {
-  case 'U':
+  if (b == 'U') {
     length = 8;
-    break;
-  case 'u':
+  } else if (b == 'u') {
     length = 4;
-    break;
-  default:
+  } else {
     return SERD_ERR_BAD_SYNTAX;
   }
 
@@ -239,18 +236,12 @@ read_character(SerdReader* const    reader,
                const uint8_t        c)
 {
   if (!(c & 0x80)) {
-    switch (c) {
-    case 0xA:
-    case 0xD:
+    if (c == 0xA || c == 0xD) {
       *flags |= SERD_HAS_NEWLINE;
-      break;
-    case '"':
-    case '\'':
+    } else if (c == '"' || c == '\'') {
       *flags |= SERD_HAS_QUOTE;
-      break;
-    default:
-      break;
     }
+
     return push_byte(reader, dest, c);
   }
 
@@ -537,34 +528,10 @@ read_PN_LOCAL_ESC(SerdReader* const reader, const Ref 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 '~':
-    push_byte(reader, dest, eat_byte_safe(reader, c));
-    break;
-  default:
-    return r_err(reader, SERD_ERR_BAD_SYNTAX, "invalid escape\n");
-  }
-
-  return SERD_SUCCESS;
+  return ((c == '!') || in_range(c, '#', '/') || (c == ';') || (c == '=') ||
+          (c == '?') || (c == '@') || (c == '_') || (c == '~'))
+           ? push_byte(reader, dest, eat_byte_safe(reader, c))
+           : r_err(reader, SERD_ERR_BAD_SYNTAX, "invalid escape\n");
 }
 
 static SerdStatus
diff --git a/src/node.c b/src/node.c
index 6e0ba486..e435093e 100644
--- a/src/node.c
+++ b/src/node.c
@@ -142,37 +142,7 @@ serd_node_new_uri_from_string(const uint8_t* const str,
 static bool
 is_uri_path_char(const uint8_t c)
 {
-  if (is_alpha(c) || is_digit(c)) {
-    return true;
-  }
-
-  switch (c) {
-  // unreserved:
-  case '-':
-  case '.':
-  case '_':
-  case '~':
-  case ':':
-
-  case '@': // pchar
-  case '/': // separator
-
-  // sub-delimiters:
-  case '!':
-  case '$':
-  case '&':
-  case '\'':
-  case '(':
-  case ')':
-  case '*':
-  case '+':
-  case ',':
-  case ';':
-  case '=':
-    return true;
-  default:
-    return false;
-  }
+  return is_alpha(c) || is_digit(c) || strchr("!$&\'()*+,-./:;=@_~", c);
 }
 
 static bool
diff --git a/src/writer.c b/src/writer.c
index 08192bb1..e4ef5651 100644
--- a/src/writer.c
+++ b/src/writer.c
@@ -283,21 +283,8 @@ write_character(SerdWriter*    writer,
 SERD_NODISCARD static bool
 uri_must_escape(const uint8_t c)
 {
-  switch (c) {
-  case ' ':
-  case '"':
-  case '<':
-  case '>':
-  case '\\':
-  case '^':
-  case '`':
-  case '{':
-  case '|':
-  case '}':
-    return true;
-  default:
-    return !in_range(c, 0x20, 0x7E);
-  }
+  return (c == '"') || (c == '<') || (c == '>') || (c == '\\') || (c == '^') ||
+         (c == '`') || in_range(c, '{', '}') || !in_range(c, 0x21, 0x7E);
 }
 
 static size_t
@@ -368,38 +355,17 @@ write_uri_from_node(SerdWriter* writer, const SerdNode* node)
 static bool
 lname_must_escape(const uint8_t c)
 {
-  /* This arbitrary list of characters, most of which have nothing to do with
-     Turtle, must be handled as special cases here because the RDF and SPARQL
-     WGs are apparently intent on making the once elegant Turtle a baroque
-     and inconsistent mess, throwing elegance and extensibility completely
-     out the window for no good reason.
-
-     Note '-', '.', and '_' are also in PN_LOCAL_ESC, but are valid unescaped
-     in local names, so they are not escaped here. */
-
-  switch (c) {
-  case '\'':
-  case '!':
-  case '#':
-  case '$':
-  case '%':
-  case '&':
-  case '(':
-  case ')':
-  case '*':
-  case '+':
-  case ',':
-  case '/':
-  case ';':
-  case '=':
-  case '?':
-  case '@':
-  case '~':
-    return true;
-  default:
-    break;
-  }
-  return false;
+  /* Most of these characters have nothing to do with Turtle, but were taken
+     from SPARQL and mashed into the Turtle grammar (despite not being used)
+     with RDF 1.1.  So now Turtle is a mess because the SPARQL grammar is
+     poorly designed and didn't use a leading character to distinguish things
+     like path patterns like it should have.
+
+     Note that '-', '.', and '_' are also in PN_LOCAL_ESC, but are valid
+     unescaped in local names, so they are not escaped here. */
+
+  return (c == '!') || (c == '/') || (c == ';') || (c == '=') || (c == '?') ||
+         (c == '@') || (c == '~') || in_range(c, '#', ',');
 }
 
 SERD_NODISCARD static SerdStatus