/* PSPP - a program for statistical analysis.
- Copyright (C) 2006, 2009, 2010, 2011 Free Software Foundation, Inc.
+ Copyright (C) 2006, 2009, 2010, 2011, 2012 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
#include <unigbrk.h>
#include "libpspp/assertion.h"
+#include "libpspp/compiler.h"
#include "libpspp/hmapx.h"
#include "libpspp/hash-functions.h"
#include "libpspp/pool.h"
#include "gl/relocatable.h"
#include "gl/xstrndup.h"
+#include "gettext.h"
+#define _(msgid) gettext (msgid)
+
struct converter
{
char *tocode;
char *fromcode;
iconv_t conv;
+ int error;
};
static char *default_encoding;
static struct hmapx map;
/* A wrapper around iconv_open */
-static iconv_t
-create_iconv (const char* tocode, const char* fromcode)
+static struct converter *
+create_iconv__ (const char* tocode, const char* fromcode)
{
size_t hash;
struct hmapx_node *node;
HMAPX_FOR_EACH_WITH_HASH (converter, node, hash, &map)
if (!strcmp (tocode, converter->tocode)
&& !strcmp (fromcode, converter->fromcode))
- return converter->conv;
+ return converter;
converter = xmalloc (sizeof *converter);
converter->tocode = xstrdup (tocode);
converter->fromcode = xstrdup (fromcode);
converter->conv = iconv_open (tocode, fromcode);
+ converter->error = converter->conv == (iconv_t) -1 ? errno : 0;
hmapx_insert (&map, converter, hash);
+ return converter;
+}
+
+static iconv_t
+create_iconv (const char* tocode, const char* fromcode)
+{
+ struct converter *converter;
+
+ converter = create_iconv__ (tocode, fromcode);
+
/* I don't think it's safe to translate this string or to use messaging
as the converters have not yet been set up */
- if ( (iconv_t) -1 == converter->conv && 0 != strcmp (tocode, fromcode))
+ if (converter->error && strcmp (tocode, fromcode))
{
- const int err = errno;
fprintf (stderr,
"Warning: "
"cannot create a converter for `%s' to `%s': %s\n",
- fromcode, tocode, strerror (err));
+ fromcode, tocode, strerror (converter->error));
+ converter->error = 0;
}
return converter->conv;
i18n_init (void)
{
setlocale (LC_CTYPE, "");
+ setlocale (LC_COLLATE, "");
setlocale (LC_MESSAGES, "");
#if HAVE_LC_PAPER
setlocale (LC_PAPER, "");
"ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`"
"abcdefghijklmnopqrstuvwxyz{|}~");
- struct substring out, cr, lf;
+ struct substring out, cr, lf, space;
bool ok;
memset (e, 0, sizeof *e);
cr = recode_substring_pool (name, "UTF-8", ss_cstr ("\r"), NULL);
lf = recode_substring_pool (name, "UTF-8", ss_cstr ("\n"), NULL);
- ok = cr.length >= 1 && cr.length <= MAX_UNIT && cr.length == lf.length;
+ space = recode_substring_pool (name, "UTF-8", ss_cstr (" "), NULL);
+ ok = (cr.length >= 1
+ && cr.length <= MAX_UNIT
+ && cr.length == lf.length
+ && cr.length == space.length);
if (!ok)
{
fprintf (stderr, "warning: encoding `%s' is not supported.\n", name);
ss_dealloc (&cr);
ss_dealloc (&lf);
+ ss_dealloc (&space);
ss_alloc_substring (&cr, ss_cstr ("\r"));
ss_alloc_substring (&lf, ss_cstr ("\n"));
+ ss_alloc_substring (&space, ss_cstr (" "));
}
e->unit = cr.length;
memcpy (e->cr, cr.string, e->unit);
memcpy (e->lf, lf.string, e->unit);
+ memcpy (e->space, space.string, e->unit);
ss_dealloc (&cr);
ss_dealloc (&lf);
+ ss_dealloc (&space);
out = recode_substring_pool ("UTF-8", name, in, NULL);
e->is_ascii_compatible = ss_equals (in, out);
ss_dealloc (&out);
+ if (!e->is_ascii_compatible && e->unit == 1)
+ {
+ out = recode_substring_pool ("UTF-8", name, ss_cstr ("A"), NULL);
+ e->is_ebcdic_compatible = (out.length == 1
+ && (uint8_t) out.string[0] == 0xc1);
+ ss_dealloc (&out);
+ }
+ else
+ e->is_ebcdic_compatible = false;
+
return ok;
}
get_encoding_info (&e, encoding);
return e.is_ascii_compatible;
}
+
+bool
+is_encoding_ebcdic_compatible (const char *encoding)
+{
+ struct encoding_info e;
+
+ get_encoding_info (&e, encoding);
+ return e.is_ebcdic_compatible;
+}
+
+/* Returns true if iconv can convert ENCODING to and from UTF-8,
+ otherwise false. */
+bool
+is_encoding_supported (const char *encoding)
+{
+ return (create_iconv__ ("UTF-8", encoding)->conv != (iconv_t) -1
+ && create_iconv__ (encoding, "UTF-8")->conv != (iconv_t) -1);
+}
+
+/* Returns true if E is the name of a UTF-8 encoding.
+
+ XXX Possibly we should test not E as a string but its properties via
+ iconv. */
+bool
+is_encoding_utf8 (const char *e)
+{
+ return ((e[0] == 'u' || e[0] == 'U')
+ && (e[1] == 't' || e[1] == 'T')
+ && (e[2] == 'f' || e[2] == 'F')
+ && ((e[3] == '8' && e[4] == '\0')
+ || (e[3] == '-' && e[4] == '8' && e[5] == '\0')));
+}
+\f
+static struct encoding_category *categories;
+static int n_categories;
+
+static void SENTINEL (0)
+add_category (size_t *allocated_categories, const char *category, ...)
+{
+ struct encoding_category *c;
+ const char *encodings[16];
+ va_list args;
+ int i, n;
+
+ /* Count encoding arguments. */
+ va_start (args, category);
+ n = 0;
+ while ((encodings[n] = va_arg (args, const char *)) != NULL)
+ {
+ const char *encoding = encodings[n];
+ if (!strcmp (encoding, "Auto") || is_encoding_supported (encoding))
+ n++;
+ }
+ assert (n < sizeof encodings / sizeof *encodings);
+ va_end (args);
+
+ if (n == 0)
+ return;
+
+ if (n_categories >= *allocated_categories)
+ categories = x2nrealloc (categories,
+ allocated_categories, sizeof *categories);
+
+ c = &categories[n_categories++];
+ c->category = category;
+ c->encodings = xmalloc (n * sizeof *c->encodings);
+ for (i = 0; i < n; i++)
+ c->encodings[i] = encodings[i];
+ c->n_encodings = n;
+}
+
+static void
+init_encoding_categories (void)
+{
+ static bool inited;
+ size_t alloc;
+
+ if (inited)
+ return;
+ inited = true;
+
+ alloc = 0;
+ add_category (&alloc, "Unicode", "UTF-8", "UTF-16", "UTF-16BE", "UTF-16LE",
+ "UTF-32", "UTF-32BE", "UTF-32LE", NULL_SENTINEL);
+ add_category (&alloc, _("Arabic"), "IBM864", "ISO-8859-6", "Windows-1256",
+ NULL_SENTINEL);
+ add_category (&alloc, _("Armenian"), "ARMSCII-8", NULL_SENTINEL);
+ add_category (&alloc, _("Baltic"), "ISO-8859-13", "ISO-8859-4",
+ "Windows-1257", NULL_SENTINEL);
+ add_category (&alloc, _("Celtic"), "ISO-8859-14", NULL_SENTINEL);
+ add_category (&alloc, _("Central European"), "IBM852", "ISO-8859-2",
+ "Mac-CentralEurope", "Windows-1250", NULL_SENTINEL);
+ add_category (&alloc, _("Chinese Simplified"), "GB18030", "GB2312", "GBK",
+ "HZ-GB-2312", "ISO-2022-CN", NULL_SENTINEL);
+ add_category (&alloc, _("Chinese Traditional"), "Big5", "Big5-HKSCS",
+ "EUC-TW", NULL_SENTINEL);
+ add_category (&alloc, _("Croatian"), "MacCroatian", NULL_SENTINEL);
+ add_category (&alloc, _("Cyrillic"), "IBM855", "ISO-8859-5", "ISO-IR-111",
+ "KOI8-R", "MacCyrillic", NULL_SENTINEL);
+ add_category (&alloc, _("Cyrillic/Russian"), "IBM866", NULL_SENTINEL);
+ add_category (&alloc, _("Cyrillic/Ukrainian"), "KOI8-U", "MacUkrainian",
+ NULL_SENTINEL);
+ add_category (&alloc, _("Georgian"), "GEOSTD8", NULL_SENTINEL);
+ add_category (&alloc, _("Greek"), "ISO-8859-7", "MacGreek", NULL_SENTINEL);
+ add_category (&alloc, _("Gujarati"), "MacGujarati", NULL_SENTINEL);
+ add_category (&alloc, _("Gurmukhi"), "MacGurmukhi", NULL_SENTINEL);
+ add_category (&alloc, _("Hebrew"), "IBM862", "ISO-8859-8-I", "Windows-1255",
+ NULL_SENTINEL);
+ add_category (&alloc, _("Hebrew Visual"), "ISO-8859-8", NULL_SENTINEL);
+ add_category (&alloc, _("Hindi"), "MacDevangari", NULL_SENTINEL);
+ add_category (&alloc, _("Icelandic"), "MacIcelandic", NULL_SENTINEL);
+ add_category (&alloc, _("Japanese"), "EUC-JP", "ISO-2022-JP", "Shift_JIS",
+ NULL_SENTINEL);
+ add_category (&alloc, _("Korean"), "EUC-KR", "ISO-2022-KR", "JOHAB", "UHC",
+ NULL_SENTINEL);
+ add_category (&alloc, _("Nordic"), "ISO-8859-10", NULL_SENTINEL);
+ add_category (&alloc, _("Romanian"), "ISO-8859-16", "MacRomanian",
+ NULL_SENTINEL);
+ add_category (&alloc, _("South European"), "ISO-8859-3", NULL_SENTINEL);
+ add_category (&alloc, _("Thai"), "ISO-8859-11", "TIS-620", "Windows-874",
+ NULL_SENTINEL);
+ add_category (&alloc, _("Turkish"), "IBM857", "ISO-8859-9", "Windows-1254",
+ NULL_SENTINEL);
+ add_category (&alloc, _("Vietnamese"), "TVCN", "VISCII", "VPS",
+ "Windows-1258", NULL_SENTINEL);
+ add_category (&alloc, _("Western European"), "ISO-8859-1", "ISO-8859-15",
+ "Windows-1252", "IBM850", "MacRoman", NULL_SENTINEL);
+}
+
+/* Returns an array of "struct encoding_category" that contains only the
+ categories and encodings that the system supports. */
+struct encoding_category *
+get_encoding_categories (void)
+{
+ init_encoding_categories ();
+ return categories;
+}
+
+/* Returns the number of elements in the array returned by
+ get_encoding_categories(). */
+size_t
+get_n_encoding_categories (void)
+{
+ init_encoding_categories ();
+ return n_categories;
+}
projects / pspp / blobdiff