segment: Distinguish snippets from full files.

[pspp] / src / language / lexer / scan.c

diff --git a/src/language/lexer/scan.c b/src/language/lexer/scan.c
index 2ad467d978bde55ae2a7dc2330b75b831cb3851c..0acff46cdd119925f98e5845f67e8bfb975eae94 100644 (file)
--- a/src/language/lexer/scan.c
+++ b/src/language/lexer/scan.c
@@ -25,11 +25,15 @@
 #include "language/lexer/token.h"
 #include "libpspp/assertion.h"
 #include "libpspp/cast.h"
+#include "libpspp/i18n.h"
 
 #include "gl/c-ctype.h"
 #include "gl/c-strtod.h"
 #include "gl/xmemdup0.h"
 
+#include "gettext.h"
+#define _(msgid) gettext (msgid)
+
 enum
   {
     S_START,
@@ -421,6 +425,48 @@ is_scan_type (enum scan_type type)
   return type > SCAN_FIRST && type < SCAN_LAST;
 }
 
+/* If TOKEN has the type of a scan error (a subset of those identified by
+   is_scan_type()), returns an appropriate error message.  Otherwise, returns
+   NULL. */
+char *
+scan_token_to_error (const struct token *token)
+{
+  switch (token->type)
+    {
+    case SCAN_BAD_HEX_LENGTH:
+      return xasprintf (_("String of hex digits has %d characters, which "
+                          "is not a multiple of 2."), (int) token->number);
+
+    case SCAN_BAD_HEX_DIGIT:
+    case SCAN_BAD_UNICODE_DIGIT:
+      return xasprintf (_("`%c' is not a valid hex digit."),
+                        (int) token->number);
+
+    case SCAN_BAD_UNICODE_LENGTH:
+      return xasprintf (_("Unicode string contains %d bytes, which is "
+                          "not in the valid range of 1 to 8 bytes."),
+                        (int) token->number);
+
+    case SCAN_BAD_UNICODE_CODE_POINT:
+      return xasprintf (_("U+%04X is not a valid Unicode code point."),
+                        (int) token->number);
+
+    case SCAN_EXPECTED_QUOTE:
+      return xasprintf (_("Unterminated string constant."));
+
+    case SCAN_EXPECTED_EXPONENT:
+      return xasprintf (_("Missing exponent following `%s'."),
+                        token->string.string);
+
+    case SCAN_UNEXPECTED_CHAR:
+      char c_name[16];
+      return xasprintf (_("Bad character %s in input."),
+                        uc_name (token->number, c_name));
+    }
+
+  return NULL;
+}
+
 static enum scan_result
 scan_start__ (struct scanner *scanner, enum segment_type type,
               struct substring s, struct token *token)
@@ -506,10 +552,6 @@ scan_start__ (struct scanner *scanner, enum segment_type type,
       ss_alloc_substring (&token->string, s);
       return SCAN_DONE;
 
-    case SEG_UNEXPECTED_DOT:
-      token->type = SCAN_UNEXPECTED_DOT;
-      return SCAN_DONE;
-
     case SEG_UNEXPECTED_CHAR:
       return scan_unexpected_char (&s, token);
     }
@@ -609,12 +651,14 @@ scanner_push (struct scanner *scanner, enum segment_type type,
    INPUT must not be modified or freed while SLEX is still in use. */
 void
 string_lexer_init (struct string_lexer *slex, const char *input, size_t length,
-                   enum segmenter_mode mode)
+                   enum segmenter_mode mode, bool is_snippet)
 {
-  slex->input = input;
-  slex->length = length;
-  slex->offset = 0;
-  segmenter_init (&slex->segmenter, mode);
+  *slex = (struct string_lexer) {
+    .input = input,
+    .length = length,
+    .offset = 0,
+    .segmenter = segmenter_init (mode, is_snippet),
+  };
 }
 
 /*  */