X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=src%2Fdata%2Fsys-file-reader.c;h=3369edf738fd2cb10191fbf6e9856f987a25b4ab;hb=refs%2Fbuilds%2F20120730001945%2Fpspp;hp=024b4ae1827994d1d6d60a7820e4535295190f5d;hpb=dff37440177a355bfc0cf9ff56428114e29f5106;p=pspp

diff --git a/src/data/sys-file-reader.c b/src/data/sys-file-reader.c
index 024b4ae182..3369edf738 100644
--- a/src/data/sys-file-reader.c
+++ b/src/data/sys-file-reader.c
@@ -312,12 +312,17 @@ sfm_read_info_destroy (struct sfm_read_info *info)
 /* Opens the system file designated by file handle FH for reading.  Reads the
    system file's dictionary into *DICT.
 
+   Ordinarily the reader attempts to automatically detect the character
+   encoding based on the file's contents.  This isn't always possible,
+   especially for files written by old versions of SPSS or PSPP, so specifying
+   a nonnull ENCODING overrides the choice of character encoding.
+
    If INFO is non-null, then it receives additional info about the system file,
    which the caller must eventually free with sfm_read_info_destroy() when it
    is no longer needed. */
 struct casereader *
-sfm_open_reader (struct file_handle *fh, struct dictionary **dictp,
-                 struct sfm_read_info *infop)
+sfm_open_reader (struct file_handle *fh, const char *volatile encoding,
+                 struct dictionary **dictp, struct sfm_read_info *infop)
 {
   struct sfm_reader *volatile r = NULL;
   struct sfm_read_info info;
@@ -454,8 +459,10 @@ sfm_open_reader (struct file_handle *fh, struct dictionary **dictp,
 
      First, figure out the correct character encoding, because this determines
      how the rest of the header data is to be interpreted. */
-  dict = dict_create (choose_encoding (r, &header, extensions[EXT_INTEGER],
-                                       extensions[EXT_ENCODING]));
+  dict = dict_create (encoding
+                      ? encoding
+                      : choose_encoding (r, &header, extensions[EXT_INTEGER],
+                                         extensions[EXT_ENCODING]));
   r->encoding = dict_get_encoding (dict);
 
   /* These records don't use variables at all. */
@@ -1584,7 +1591,6 @@ parse_long_var_name_map (struct sfm_reader *r,
   while (read_variable_to_value_pair (r, dict, text, &var, &long_name))
     {
       /* Validate long name. */
-      /* XXX need to reencode name to UTF-8 */
       if (!dict_id_is_valid (dict, long_name, false))
         {
           sys_warn (r, record->pos,
@@ -2432,7 +2438,7 @@ text_parse_counted_string (struct sfm_reader *r, struct text_record *text)
 
   start = text->pos;
   n = 0;
-  for (;;)
+  while (text->pos < text->buffer.length)
     {
       int c = text->buffer.string[text->pos];
       if (c < '0' || c > '9')
@@ -2440,7 +2446,7 @@ text_parse_counted_string (struct sfm_reader *r, struct text_record *text)
       n = (n * 10) + (c - '0');
       text->pos++;
     }
-  if (start == text->pos)
+  if (text->pos >= text->buffer.length || start == text->pos)
     {
       sys_warn (r, text->start,
                 _("Expecting digit at offset %zu in MRSETS record."),