doc: Improve description of extension subtype 5.

[pspp] / tests / language / lexer / scan-test.c

diff --git a/tests/language/lexer/scan-test.c b/tests/language/lexer/scan-test.c
index 84eff8171f9520d7a80657ac0155470812fcefb2..95c52a42be48cc204f9bd79ef0ef0a692213d1ce 100644 (file)
--- a/tests/language/lexer/scan-test.c
+++ b/tests/language/lexer/scan-test.c
@@ -1,5 +1,5 @@
 /* PSPP - a program for statistical analysis.
-   Copyright (C) 2010, 2011 Free Software Foundation, Inc.
+   Copyright (C) 2010, 2011, 2013 Free Software Foundation, Inc.
 
    This program is free software: you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
@@ -39,6 +39,10 @@
 /* -a/--auto, -b/--batch, -i/--interactive: syntax mode. */
 static enum segmenter_mode mode = SEG_MODE_AUTO;
 
+/* -s, --strip-trailing-newline: Strip trailing newline from last line of
+    input. */
+static bool strip_trailing_newline;
+
 static const char *parse_options (int argc, char **argv);
 static void usage (void) NO_RETURN;
 
@@ -50,46 +54,84 @@ main (int argc, char *argv[])
   char *input;
 
   struct string_lexer slex;
-  bool more;
 
   set_program_name (argv[0]);
   file_name = parse_options (argc, argv);
 
-  /* Read from stdin into 'input'.  Ensure that 'input' ends in a new-line
-     followed by a null byte. */
+  /* Read from stdin into 'input'. */
   input = (!strcmp (file_name, "-")
-           ? fread_file (stdin, &length)
-           : read_file (file_name, &length));
+           ? fread_file (stdin, 0, &length)
+           : read_file (file_name, 0, &length));
   if (input == NULL)
     error (EXIT_FAILURE, errno, "reading %s failed", file_name);
-  input = xrealloc (input, length + 3);
-  if (length == 0 || input[length - 1] != '\n')
-    input[length++] = '\n';
-  input[length++] = '\0';
 
-  string_lexer_init (&slex, input, mode);
-  do
+  if (strip_trailing_newline && length && input[length - 1] == '\n')
+    {
+      length--;
+      if (length && input[length - 1] == '\r')
+        length--;
+    }
+
+  struct token *tokens = NULL;
+  size_t n_tokens = 0;
+  size_t allocated_tokens = 0;
+  string_lexer_init (&slex, input, length, mode, false);
+  for (;;)
     {
-      struct token token;
+      if (n_tokens >= allocated_tokens)
+        tokens = x2nrealloc (tokens, &allocated_tokens, sizeof *tokens);
+      enum string_lexer_result result
+        = string_lexer_next (&slex, &tokens[n_tokens]);
 
-      more = string_lexer_next (&slex, &token);
+      if (result == SLR_ERROR)
+        tokens[n_tokens].type = T_STOP;
+      n_tokens++;
 
-      printf ("%s", scan_type_to_string (token.type));
-      if (token.number != 0.0)
+      if (result == SLR_END)
+        break;
+    }
+
+  for (size_t i = 0; i < n_tokens; )
+    {
+      struct merger m = MERGER_INIT;
+      int retval;
+      struct token out;
+      for (size_t j = i; ; j++)
         {
-          char s[DBL_BUFSIZE_BOUND];
+          assert (j < n_tokens);
+          retval = merger_add (&m, &tokens[j], &out);
+          if (retval != -1)
+            break;
+        }
+
+      const struct token *t = retval ? &out : &tokens[i];
+
+      fputs (token_type_to_name (t->type), stdout);
+      if (t->number != 0.0)
+        {
+          double x = t->number;
 
-          dtoastr (s, sizeof s, 0, 0, token.number);
-          printf (" %s", s);
+          if (x > LONG_MIN && x <= LONG_MAX && floor (x) == x)
+            printf (" %ld", (long int) x);
+          else
+            printf (" %.3g", x);
         }
-      if (token.string.string != NULL || token.string.length > 0)
-        printf (" \"%.*s\"", (int) token.string.length, token.string.string);
+      if (t->string.string != NULL || t->string.length > 0)
+        printf (" \"%.*s\"", (int) t->string.length, t->string.string);
       printf ("\n");
 
-      token_destroy (&token);
+      if (retval)
+        {
+          i += retval;
+          token_uninit (&out);
+        }
+      else
+        i++;
     }
-  while (more);
 
+  for (size_t i = 0; i < n_tokens; i++)
+    token_uninit (&tokens[i]);
+  free (tokens);
   free (input);
 
   return 0;
@@ -105,11 +147,12 @@ parse_options (int argc, char **argv)
           {"auto", no_argument, NULL, 'a'},
           {"batch", no_argument, NULL, 'b'},
           {"interactive", no_argument, NULL, 'i'},
+          {"strip-trailing-newline", no_argument, NULL, 's'},
           {"help", no_argument, NULL, 'h'},
           {NULL, 0, NULL, 0},
         };
 
-      int c = getopt_long (argc, argv, "abih", options, NULL);
+      int c = getopt_long (argc, argv, "sabih", options, NULL);
       if (c == -1)
         break;
 
@@ -127,6 +170,10 @@ parse_options (int argc, char **argv)
           mode = SEG_MODE_INTERACTIVE;
           break;
 
+        case 's':
+          strip_trailing_newline = true;
+          break;
+
         case 'h':
           usage ();
 
@@ -157,10 +204,10 @@ usage (void)
 usage: %s [OPTIONS] INPUT\n\
 \n\
 Options:\n\
-  -1, --one-segment   feed one segment at a time\n\
   -a, --auto          use \"auto\" syntax mode\n\
   -b, --batch         use \"batch\" syntax mode\n\
   -i, --interactive   use \"interactive\" syntax mode (default)\n\
+  -s, --strip-trailing-newline  remove newline from end of input\n\
   -v, --verbose       include rows and column numbers in output\n\
   -h, --help          print this help message\n",
           program_name, program_name);