sys-file-reader: Break reading a system file into two stages.

[pspp] / src / data / any-reader.c

diff --git a/src/data/any-reader.c b/src/data/any-reader.c
index 731f552a9f9021bf05ffd524c5fbebad11946405..ad3fcb5b7a91a99c6acd22798c459037bc0b22c5 100644 (file)
--- a/src/data/any-reader.c
+++ b/src/data/any-reader.c
@@ -1,5 +1,5 @@
 /* PSPP - a program for statistical analysis.
-   Copyright (C) 2006, 2010 Free Software Foundation, Inc.
+   Copyright (C) 2006, 2010, 2011, 2012, 2014 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
@@ -15,33 +15,29 @@
    along with this program.  If not, see <http://www.gnu.org/licenses/>. */
 
 #include <config.h>
-#include "any-reader.h"
+
+#include "data/any-reader.h"
+
 #include <assert.h>
 #include <errno.h>
 #include <stdbool.h>
 #include <stdio.h>
 #include <stdlib.h>
-#include <libpspp/assertion.h>
-#include <libpspp/message.h>
-#include "file-handle-def.h"
-#include "file-name.h"
-#include "por-file-reader.h"
-#include "sys-file-reader.h"
-#include <libpspp/str.h>
-#include "scratch-reader.h"
-#include "xalloc.h"
+
+#include "data/dataset-reader.h"
+#include "data/file-handle-def.h"
+#include "data/file-name.h"
+#include "data/por-file-reader.h"
+#include "data/sys-file-reader.h"
+#include "libpspp/assertion.h"
+#include "libpspp/message.h"
+#include "libpspp/str.h"
+
+#include "gl/xalloc.h"
 
 #include "gettext.h"
 #define _(msgid) gettext (msgid)
 
-/* Result of type detection. */
-enum detect_result
-  {
-    YES,                        /* It is this type. */
-    NO,                         /* It is not this type. */
-    IO_ERROR                    /* File couldn't be opened. */
-  };
-
 /* Tries to detect whether FILE is a given type of file, by opening the file
    and passing it to DETECT, and returns a detect_result. */
 static enum detect_result
@@ -53,32 +49,42 @@ try_detect (const char *file_name, bool (*detect) (FILE *))
   file = fn_open (file_name, "rb");
   if (file == NULL)
     {
-      msg (ME, _("An error occurred while opening \"%s\": %s."),
+      msg (ME, _("An error occurred while opening `%s': %s."),
            file_name, strerror (errno));
-      return IO_ERROR;
+      return ANY_ERROR;
     }
 
   is_type = detect (file);
 
   fn_close (file_name, file);
 
-  return is_type ? YES : NO;
+  return is_type ? ANY_YES : ANY_NO;
 }
 
 /* Returns true if any_reader_open() would be able to open FILE as a data
    file, false otherwise. */
-bool
+enum detect_result
 any_reader_may_open (const char *file)
 {
-  return (try_detect (file, sfm_detect) == YES
-          || try_detect (file, pfm_detect) == YES);
+  enum detect_result res = try_detect (file, sfm_detect);
+  
+  if (res == ANY_NO)
+    res = try_detect (file, pfm_detect);
+
+  return res;
 }
 
 /* Returns a casereader for HANDLE.  On success, returns the new
    casereader and stores the file's dictionary into *DICT.  On
-   failure, returns a null pointer. */
+   failure, returns a null pointer.
+
+   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.  */
 struct casereader *
-any_reader_open (struct file_handle *handle, struct dictionary **dict)
+any_reader_open (struct file_handle *handle, const char *encoding,
+                 struct dictionary **dict)
 {
   switch (fh_get_referent (handle))
     {
@@ -87,18 +93,26 @@ any_reader_open (struct file_handle *handle, struct dictionary **dict)
         enum detect_result result;
 
         result = try_detect (fh_get_file_name (handle), sfm_detect);
-        if (result == IO_ERROR)
+        if (result == ANY_ERROR)
           return NULL;
-        else if (result == YES)
-          return sfm_open_reader (handle, dict, NULL);
+        else if (result == ANY_YES)
+          {
+            struct sfm_reader *r;
+
+            r = sfm_open (handle);
+            if (r == NULL)
+              return NULL;
+
+            return sfm_decode (r, encoding, dict, NULL);
+          }
 
         result = try_detect (fh_get_file_name (handle), pfm_detect);
-        if (result == IO_ERROR)
+        if (result == ANY_ERROR)
           return NULL;
-        else if (result == YES)
+        else if (result == ANY_YES)
           return pfm_open_reader (handle, dict, NULL);
 
-        msg (SE, _("\"%s\" is not a system or portable file."),
+        msg (SE, _("`%s' is not a system or portable file."),
              fh_get_file_name (handle));
         return NULL;
       }
@@ -107,8 +121,8 @@ any_reader_open (struct file_handle *handle, struct dictionary **dict)
       msg (SE, _("The inline file is not allowed here."));
       return NULL;
 
-    case FH_REF_SCRATCH:
-      return scratch_reader_open (handle, dict);
+    case FH_REF_DATASET:
+      return dataset_reader_open (handle, dict);
     }
   NOT_REACHED ();
 }