2017-07-18 14:08:31 +02:00
|
|
|
This hack makes Guile default to UTF-8. This avoids calls to
|
|
|
|
`iconv_open'; `iconv_open' tries to open shared objects that aren't
|
|
|
|
available during bootstrap, so using UTF-8 avoids that (and UTF-8 has
|
|
|
|
built-in conversions in glibc, too.)
|
|
|
|
|
|
|
|
diff --git a/libguile/bytevectors.c b/libguile/bytevectors.c
|
2019-07-01 14:25:56 +02:00
|
|
|
index 0ac5ea6a6..f73301e2f 100644
|
2017-07-18 14:08:31 +02:00
|
|
|
--- a/libguile/bytevectors.c
|
|
|
|
+++ b/libguile/bytevectors.c
|
2019-07-01 14:25:56 +02:00
|
|
|
@@ -1931,7 +1931,7 @@ utf_encoding_name (char *name, size_t utf_width, SCM endianness)
|
2017-07-18 14:08:31 +02:00
|
|
|
if (scm_i_is_narrow_string (str)) \
|
|
|
|
{ \
|
|
|
|
err = mem_iconveh (scm_i_string_chars (str), c_strlen, \
|
|
|
|
- "ISO-8859-1", c_utf_name, \
|
|
|
|
+ "UTF-8", c_utf_name, \
|
|
|
|
iconveh_question_mark, NULL, \
|
|
|
|
&c_utf, &c_utf_len); \
|
|
|
|
if (SCM_UNLIKELY (err)) \
|
|
|
|
diff --git a/libguile/ports.c b/libguile/ports.c
|
2019-07-01 14:25:56 +02:00
|
|
|
index 45e62f4e4..42012f3aa 100644
|
2017-07-18 14:08:31 +02:00
|
|
|
--- a/libguile/ports.c
|
|
|
|
+++ b/libguile/ports.c
|
2019-07-01 14:25:56 +02:00
|
|
|
@@ -974,7 +974,9 @@ canonicalize_encoding (const char *enc)
|
2017-07-18 14:08:31 +02:00
|
|
|
char *ret;
|
|
|
|
int i;
|
|
|
|
|
|
|
|
- if (!enc || encoding_matches (enc, sym_ISO_8859_1))
|
|
|
|
+ if (enc == NULL)
|
|
|
|
+ return sym_UTF_8;
|
|
|
|
+ if (encoding_matches (enc, sym_ISO_8859_1))
|
|
|
|
return sym_ISO_8859_1;
|
|
|
|
if (encoding_matches (enc, sym_UTF_8))
|
|
|
|
return sym_UTF_8;
|
2019-07-01 14:25:56 +02:00
|
|
|
@@ -4198,7 +4200,7 @@ scm_init_ports (void)
|
2017-07-18 14:08:31 +02:00
|
|
|
scm_c_define ("%default-port-conversion-strategy",
|
|
|
|
scm_make_fluid_with_default (sym_substitute));
|
|
|
|
/* Use the locale as the default port encoding. */
|
|
|
|
- scm_i_set_default_port_encoding (locale_charset ());
|
|
|
|
+ scm_i_set_default_port_encoding ("UTF-8");
|
|
|
|
|
|
|
|
scm_c_register_extension ("libguile-" SCM_EFFECTIVE_VERSION,
|
|
|
|
"scm_init_ice_9_ports",
|
|
|
|
diff --git a/libguile/strings.c b/libguile/strings.c
|
2019-07-01 14:25:56 +02:00
|
|
|
index 056b4c99f..63a6c050d 100644
|
2017-07-18 14:08:31 +02:00
|
|
|
--- a/libguile/strings.c
|
|
|
|
+++ b/libguile/strings.c
|
2019-07-01 14:25:56 +02:00
|
|
|
@@ -1579,7 +1579,7 @@ scm_i_default_string_failed_conversion_handler (void)
|
2017-07-18 14:08:31 +02:00
|
|
|
SCM
|
|
|
|
scm_from_locale_stringn (const char *str, size_t len)
|
|
|
|
{
|
|
|
|
- return scm_from_stringn (str, len, locale_charset (),
|
|
|
|
+ return scm_from_stringn (str, len, "UTF-8",
|
|
|
|
scm_i_default_string_failed_conversion_handler ());
|
|
|
|
}
|
|
|
|
|
2019-07-01 14:25:56 +02:00
|
|
|
@@ -1907,7 +1907,7 @@ char *
|
2017-07-18 14:08:31 +02:00
|
|
|
scm_to_locale_stringn (SCM str, size_t *lenp)
|
|
|
|
{
|
|
|
|
return scm_to_stringn (str, lenp,
|
|
|
|
- locale_charset (),
|
|
|
|
+ "UTF-8",
|
|
|
|
scm_i_default_string_failed_conversion_handler ());
|
|
|
|
}
|
|
|
|
|
2019-07-01 14:25:56 +02:00
|
|
|
@@ -2195,7 +2195,7 @@ scm_to_stringn (SCM str, size_t *lenp, const char *encoding,
|
|
|
|
scm_wrong_type_arg_msg (NULL, 0, str, "string");
|
|
|
|
|
|
|
|
if (encoding == NULL)
|
|
|
|
- encoding = "ISO-8859-1";
|
|
|
|
+ encoding = "UTF-8";
|
|
|
|
|
|
|
|
if (c_strcasecmp (encoding, "UTF-8") == 0)
|
|
|
|
/* This is the most common case--e.g., when calling libc bindings
|
|
|
|
@@ -2247,7 +2247,7 @@ scm_to_stringn (SCM str, size_t *lenp, const char *encoding,
|
2017-07-18 14:08:31 +02:00
|
|
|
if (scm_i_is_narrow_string (str))
|
|
|
|
{
|
|
|
|
ret = mem_iconveh (scm_i_string_chars (str), ilen,
|
2019-07-01 14:25:56 +02:00
|
|
|
- "ISO-8859-1", encoding,
|
|
|
|
+ "UTF-8", encoding,
|
2017-07-18 14:08:31 +02:00
|
|
|
(enum iconv_ilseq_handler) handler, NULL,
|
|
|
|
&buf, &len);
|
|
|
|
|