Next: Introduction, Up: (dir) [Contents][Index]
This manual is for Libidn2 (version 2.3.7, 13 January 2024), an implementation of IDNA2008/TR46 internationalized domain names.
Copyright © 2011–2024 Simon Josefsson
Next: Library Functions, Previous: Libidn2, Up: Libidn2 [Contents][Index]
Libidn2 is a free software implementation of IDNA2008, Punycode and Unicode TR46. Its purpose is to encode and decode internationalized domain names.
The library is a rewrite of the popular but legacy libidn library, and is backwards (API) compatible with it. See Converting from libidn for more information.
For technical reference, see:
Libidn2 uses GNU libunistring (https://www.gnu.org/software/libunistring/) for Unicode processing and optionally GNU libiconv (https://www.gnu.org/software/libiconv/) for character set conversion.
The library is dual-licensed under LGPLv3 or GPLv2, see the file COPYING for detailed information.
Next: Converting from libidn, Previous: Introduction, Up: Libidn2 [Contents][Index]
Below are the interfaces of the Libidn2 library documented.
idn2.hidn2.hTo use the functions documented in this chapter, you need to include the file idn2.h like this:
#include <idn2.h>
When you have the data encoded in UTF-8 form the direct interfaces to the library are as follows.
input: zero terminated input UTF-8 string.
output: pointer to newly allocated output string.
flags: optional idn2_flags to modify behaviour.
Convert UTF-8 domain name to ASCII string using the IDNA2008 rules. The domain name may contain several labels, separated by dots. The output buffer must be deallocated by the caller.
The default behavior of this function (when flags are zero) is to apply
the IDNA2008 rules without the TR46 amendments. As the TR46
non-transitional processing is nowadays ubiquitous, when unsure, it is
recommended to call this function with the IDN2_NONTRANSITIONAL
and the IDN2_NFC_INPUT flags for compatibility with other software.
Return value: Returns IDN2_OK on success, or error code.
Since: 2.0.0
input: Input zero-terminated UTF-8 string.
output: Newly allocated UTF-8 output string.
flags: Currently unused.
Converts a possibly ACE encoded domain name in UTF-8 format into a UTF-8 string (punycode decoding). The output buffer will be zero-terminated and must be deallocated by the caller.
output may be NULL to test lookup of input without allocating memory.
Since: 2.0.0
src: input zero-terminated UTF-8 string in Unicode NFC normalized form.
lookupname: newly allocated output variable with name to lookup in DNS.
flags: optional idn2_flags to modify behaviour.
Perform IDNA2008 lookup string conversion on domain name src , as
described in section 5 of RFC 5891. Note that the input string
must be encoded in UTF-8 and be in Unicode NFC form.
Pass IDN2_NFC_INPUT in flags to convert input to NFC form before
further processing. IDN2_TRANSITIONAL and IDN2_NONTRANSITIONAL
do already imply IDN2_NFC_INPUT .
Pass IDN2_ALABEL_ROUNDTRIP in flags to
convert any input A-labels to U-labels and perform additional
testing. This is default since version 2.2.
To switch this behavior off, pass IDN2_NO_ALABEL_ROUNDTRIP
Pass IDN2_TRANSITIONAL to enable Unicode TR46
transitional processing, and IDN2_NONTRANSITIONAL to enable
Unicode TR46 non-transitional processing.
Multiple flags may be specified by binary or:ing them together.
After version 2.0.3: IDN2_USE_STD3_ASCII_RULES disabled by default.
Previously we were eliminating non-STD3 characters from domain strings
such as _443._tcp.example.com, or IPs 1.2.3.4/24 provided to libidn2
functions. That was an unexpected regression for applications switching
from libidn and thus it is no longer applied by default.
Use IDN2_USE_STD3_ASCII_RULES to enable that behavior again.
After version 0.11: lookupname may be NULL to test lookup of src without allocating memory.
Returns: On successful conversion IDN2_OK is returned, if the
output domain or any label would have been too long
IDN2_TOO_BIG_DOMAIN or IDN2_TOO_BIG_LABEL is returned, or
another error code is returned.
Since: 0.1
ulabel: input zero-terminated UTF-8 and Unicode NFC string, or NULL.
alabel: input zero-terminated ACE encoded string (xn–), or NULL.
insertname: newly allocated output variable with name to register in DNS.
flags: optional idn2_flags to modify behaviour.
Perform IDNA2008 register string conversion on domain label ulabel and alabel , as described in section 4 of RFC 5891. Note that the
input ulabel must be encoded in UTF-8 and be in Unicode NFC form.
Pass IDN2_NFC_INPUT in flags to convert input ulabel to NFC form
before further processing.
It is recommended to supply both ulabel and alabel for better
error checking, but supplying just one of them will work. Passing
in only alabel is better than only ulabel . See RFC 5891 section
4 for more information.
After version 0.11: insertname may be NULL to test conversion of src without allocating memory.
Returns: On successful conversion IDN2_OK is returned, when the
given ulabel and alabel does not match each other
IDN2_UALABEL_MISMATCH is returned, when either of the input
labels are too long IDN2_TOO_BIG_LABEL is returned, when alabel does does not appear to be a proper A-label IDN2_INVALID_ALABEL
is returned, or another error code is returned.
As a convenience, the following functions are provided that will convert the input from the locale encoding format to UTF-8 and normalize the string using NFC, and then apply the core functions described earlier.
input: zero terminated input UTF-8 string.
output: pointer to newly allocated output string.
flags: optional idn2_flags to modify behaviour.
Convert a domain name in locale’s encoding to ASCII string using the IDNA2008 rules. The domain name may contain several labels, separated by dots. The output buffer must be deallocated by the caller.
The default behavior of this function (when flags are zero) is to apply
the IDNA2008 rules without the TR46 amendments. As the TR46
non-transitional processing is nowadays ubiquitous, when unsure, it is
recommended to call this function with the IDN2_NONTRANSITIONAL
and the IDN2_NFC_INPUT flags for compatibility with other software.
Returns: IDN2_OK on success, or error code.
Same as described in idn2_lookup_ul() documentation.
Since: 2.0.0
input: Input zero-terminated UTF-8 string.
output: Newly allocated output string in current locale’s character set.
flags: Currently unused.
Converts a possibly ACE encoded domain name in UTF-8 format into a string encoded in the current locale’s character set (punycode decoding). The output buffer will be zero-terminated and must be deallocated by the caller.
output may be NULL to test lookup of input without allocating memory.
Since: 2.0.0
input: Input zero-terminated string encoded in the current locale’s character set.
output: Newly allocated output string in current locale’s character set.
flags: Currently unused.
Converts a possibly ACE encoded domain name in the locale’s character set into a string encoded in the current locale’s character set (punycode decoding). The output buffer will be zero-terminated and must be deallocated by the caller.
output may be NULL to test lookup of input without allocating memory.
Since: 2.0.0
src: input zero-terminated locale encoded string.
lookupname: newly allocated output variable with name to lookup in DNS.
flags: optional idn2_flags to modify behaviour.
Perform IDNA2008 lookup string conversion on domain name src , as
described in section 5 of RFC 5891. Note that the input is assumed
to be encoded in the locale’s default coding system, and will be
transcoded to UTF-8 and NFC normalized by this function.
Pass IDN2_ALABEL_ROUNDTRIP in flags to
convert any input A-labels to U-labels and perform additional
testing. This is default since version 2.2.
To switch this behavior off, pass IDN2_NO_ALABEL_ROUNDTRIP
Pass IDN2_TRANSITIONAL to enable Unicode TR46 transitional processing,
and IDN2_NONTRANSITIONAL to enable Unicode TR46 non-transitional
processing.
Multiple flags may be specified by binary or:ing them together, for
example IDN2_ALABEL_ROUNDTRIP | IDN2_NONTRANSITIONAL .
The IDN2_NFC_INPUT in flags is always enabled in this function.
After version 0.11: lookupname may be NULL to test lookup of src without allocating memory.
Returns: On successful conversion IDN2_OK is returned, if
conversion from locale to UTF-8 fails then IDN2_ICONV_FAIL is
returned, if the output domain or any label would have been too
long IDN2_TOO_BIG_DOMAIN or IDN2_TOO_BIG_LABEL is returned, or
another error code is returned.
Since: 0.1
ulabel: input zero-terminated locale encoded string, or NULL.
alabel: input zero-terminated ACE encoded string (xn–), or NULL.
insertname: newly allocated output variable with name to register in DNS.
flags: optional idn2_flags to modify behaviour.
Perform IDNA2008 register string conversion on domain label ulabel and alabel , as described in section 4 of RFC 5891. Note that the
input ulabel is assumed to be encoded in the locale’s default
coding system, and will be transcoded to UTF-8 and NFC normalized
by this function.
It is recommended to supply both ulabel and alabel for better
error checking, but supplying just one of them will work. Passing
in only alabel is better than only ulabel . See RFC 5891 section
4 for more information.
After version 0.11: insertname may be NULL to test conversion of src without allocating memory.
Returns: On successful conversion IDN2_OK is returned, when the
given ulabel and alabel does not match each other
IDN2_UALABEL_MISMATCH is returned, when either of the input
labels are too long IDN2_TOO_BIG_LABEL is returned, when alabel does does not appear to be a proper A-label IDN2_INVALID_ALABEL
is returned, when ulabel locale to UTF-8 conversion failed
IDN2_ICONV_FAIL is returned, or another error code is returned.
The flags parameter can take on the following values, or a
bit-wise inclusive or of any subset of the parameters:
Apply NFC normalization on input.
Apply additional round-trip conversion of A-label inputs.
Perform Unicode TR46 transitional processing.
Perform Unicode TR46 non-transitional processing (default).
Disable any TR#46 transitional or non-transitional processing.
Use STD3 ASCII rules. This is a TR#46 flag and is a no-op when IDN2_NO_TR46 is specified.
rc: return code from another libidn2 function.
Convert internal libidn2 error code to a humanly readable string. The returned pointer must not be de-allocated by the caller.
Return value: A humanly readable string describing error.
rc: return code from another libidn2 function.
Convert internal libidn2 error code to a string corresponding to internal header file symbols. For example, idn2_strerror_name(IDN2_MALLOC) will return the string "IDN2_MALLOC".
The caller must not attempt to de-allocate the returned string.
Return value: A string corresponding to error code symbol.
The functions normally return 0 on success or a negative error code.
Successful return.
Memory allocation error.
Could not determine locale string encoding format.
Could not transcode locale string to UTF-8.
Unicode data encoding error.
Error normalizing string.
Punycode invalid input.
Punycode output buffer too small.
Punycode conversion would overflow.
Domain name longer than 255 characters.
Domain label longer than 63 characters.
Input A-label is not valid.
Input A-label and U-label does not match.
Invalid combination of flags.
String is not NFC.
String has forbidden two hyphens.
String has forbidden starting/ending hyphen.
String has forbidden leading combining character.
String has disallowed character.
String has forbidden context-j character.
String has context-j character with no rull.
String has forbidden context-o character.
String has context-o character with no rull.
String has forbidden unassigned character.
String has forbidden bi-directional properties.
Label has forbidden dot (TR46).
Label has character forbidden in transitional mode (TR46).
Label has character forbidden in non-transitional mode (TR46).
ptr: pointer to deallocate
Call free(3) on the given pointer.
This function is typically only useful on systems where the library malloc heap is different from the library caller malloc heap, which happens on Windows when the library is a separate DLL.
It is often desirable to check that the version of Libidn2 used is indeed one which fits all requirements. Even with binary compatibility new features may have been introduced but due to problem with the dynamic linker an old version is actually used. So you may want to check that the version is okay right after program startup.
req_version: version string to compare with, or NULL.
Check IDN2 library version. This function can also be used to read
out the version of the library code used. See IDN2_VERSION for a
suitable req_version string, it corresponds to the idn2.h header
file version. Normally these two version numbers match, but if you
are using an application built against an older libidn2 with a
newer libidn2 shared library they will be different.
Return value: Check that the version of the library is at
minimum the one given as a string in req_version and return the
actual version string of the library; return NULL if the
condition is not met. If NULL is passed to this function no
check is done and only the version string is returned.
The normal way to use the function is to put something similar to the
following first in your main:
if (!idn2_check_version (IDN2_VERSION))
{
printf ("idn2_check_version() failed:\n"
"Header file incompatible with shared library.\n");
exit(EXIT_FAILURE);
}
Next: Examples, Previous: Library Functions, Up: Libidn2 [Contents][Index]
This library is backwards (API) compatible with the libidn library (https://www.gnu.org/software/libidn/).
Although it is recommended for new software to use the native libidn2
functions (i.e., the ones prefixed with idn2), old software
isn’t always feasible to modify.
As such, libidn2, provides compatibility macros which switch all libidn
functions, to libidn2 functions in a backwards compatible way. To take
advantage of these compatibility functions, it is sufficient to replace
the idna.h header in legacy code, with idn2.h. That
would transform the software from using libidn, i.e., IDNA2003,
to using libidn2 with IDNA2008 non-transitional encoding.
However, it is recommended to switch applications to the IDN2 native APIs. The following table provides a mapping of libidn code snippets to libidn2, for switching to IDNA2008.
| libidn | libidn2 |
|---|---|
rc = idna_to_ascii_8z (buf, &p, IDNA_USE_STD3_ASCII_RULES); if (rc != IDNA_SUCCESS) | rc = idn2_to_ascii_8z (buf, &p, IDN2_USE_STD3_ASCII_RULES); if (rc != IDN2_OK) |
rc = idna_to_ascii_8z (buf, &p, 0 /* any other flags */); if (rc != IDNA_SUCCESS) | /* we recommend to use the default flags (0), so that * the default behavior of libidn2 applies. */ rc = idn2_to_ascii_8z (buf, &p, 0); if (rc != IDN2_OK) |
rc = idna_to_unicode_8z8z (buf, &p, 0 /* any flags */); if (rc != IDNA_SUCCESS) | rc = idn2_to_unicode_8z8z (buf, &p, 0); if (rc != IDN2_OK) |
Note that, although the table only lists the UTF-8 functions, the mapping
is identical for every other one on the family of toUnicode and toAscii.
As the IDNA2003 details differ signicantly to IDNA2008, not all flags used in
the libidn functions map to any specific flags; it is typically safe to use
the suggested libidn2 flags. Exceptionally the libidn flag IDNA_USE_STD3_ASCII_RULES
is mapped to IDN2_USE_STD3_ASCII_RULES.
In several cases where IDNA2008 mappings do not exist whereas IDNA2003 mappings do, software like browsers take a backwards compatible approach. That is convert the domain to IDNA2008 form, and if that fails try the IDNA2003 conversion. The following example demonstrates that approach.
rc = idn2_to_ascii_8z (buf, &p, IDN2_NONTRANSITIONAL); /* IDNA2008 */ if (rc == IDN2_DISALLOWED) rc = idn2_to_ascii_8z (buf, &p, IDN2_TRANSITIONAL); /* IDNA2003 - compatible */
In the special case of software that needs to support both
libraries (e.g., both IDNA2003 and IDNA2008), you must define
IDN2_SKIP_LIBIDN_COMPAT prior to including idn2.h
in order to disable compatibility code which overlaps with libidn
functionality. That would allow software to use both libraries’ functions.
The original libidn library includes functionality for the stringprep
processing in stringprep.h. That functionality was an integral part
of an IDNA2003 implementation, but it does not apply to IDNA2008. Furthermore,
stringprep processing has been replaced by the PRECIS framework (RFC8264).
For the reasons above, libidn2 does not implement stringprep or any other string processing protocols unrelated to IDNA2008. Applications requiring the stringprep processing should continue using the original libidn, and new applications should consider using the PRECIS framework.
Next: Invoking idn2, Previous: Converting from libidn, Up: Libidn2 [Contents][Index]
This chapter contains example code which illustrate how Libidn2 is used when you write your own application.
Next: ToUnicode example, Up: Examples [Contents][Index]
This example demonstrates how the library is used to convert internationalized domain names into ASCII compatible names (ACE). It expects input to be in UTF-8 form.
/* example-toascii.c --- Example ToASCII() code showing how to use Libidn2.
*
* This code is placed under public domain.
*/
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <idn2.h> /* idn2_to_ascii_8z() */
/*
* Compiling using pkg-config is recommended:
*
* $ cc -o example-toascii example-toascii.c $(pkg-config --cflags --libs libidn2)
* $ ./example-toascii
* Input domain encoded as `UTF-8': βόλος.com
* Read string (length 15): ce b2 cf 8c ce bb ce bf cf 82 2e 63 6f 6d 0a
* ACE label (length 17): 'xn--nxasmm1c.com'
*
*/
int
main (void)
{
char buf[BUFSIZ];
char *p;
int rc;
size_t i;
if (!fgets (buf, BUFSIZ, stdin))
perror ("fgets");
buf[strlen (buf) - 1] = '\0';
printf ("Read string (length %ld): ", (long int) strlen (buf));
for (i = 0; i < strlen (buf); i++)
printf ("%02x ", (unsigned) buf[i] & 0xFF);
printf ("\n");
/* Use non-transitional IDNA2008 */
rc = idn2_to_ascii_8z (buf, &p, IDN2_NONTRANSITIONAL);
if (rc != IDNA_SUCCESS)
{
printf ("ToASCII() failed (%d): %s\n", rc, idn2_strerror (rc));
return EXIT_FAILURE;
}
printf ("ACE label (length %ld): '%s'\n", (long int) strlen (p), p);
free (p); /* or idn2_free() */
return 0;
}
Next: Lookup, Previous: ToASCII example, Up: Examples [Contents][Index]
This example demonstrates how the library is used to convert ASCII compatible names (ACE) to internationalized domain names. Both input and output are in UTF-8 form.
/* example-tounicode.c --- Example ToUnicode() code showing how to use Libidn2.
*
* This code is placed under public domain.
*/
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <idn2.h> /* idn2_to_unicode_8z8z() */
/*
* Compiling using pkg-config is recommended:
*
* $ cc -o example-to-unicode example-to-unicode.c $(pkg-config --cflags --libs libidn2)
* $ ./example-tounicode
* Input domain (ACE) encoded as `UTF-8': xn--nxasmm1c.com
*
* Read string (length 16): 78 6e 2d 2d 6e 78 61 73 6d 6d 31 63 2e 63 6f 6d
* ACE label (length 14): 'βόλος.com'
*
*/
int
main (void)
{
char buf[BUFSIZ];
char *p;
int rc;
size_t i;
if (!fgets (buf, BUFSIZ, stdin))
perror ("fgets");
buf[strlen (buf) - 1] = '\0';
printf ("Read string (length %ld): ", (long int) strlen (buf));
for (i = 0; i < strlen (buf); i++)
printf ("%02x ", (unsigned) buf[i] & 0xFF);
printf ("\n");
rc = idn2_to_unicode_8z8z (buf, &p, 0);
if (rc != IDNA_SUCCESS)
{
printf ("ToUnicode() failed (%d): %s\n", rc, idn2_strerror (rc));
return EXIT_FAILURE;
}
printf ("ACE label (length %ld): '%s'\n", (long int) strlen (p), p);
free (p); /* or idn2_free() */
return 0;
}
Next: Register, Previous: ToUnicode example, Up: Examples [Contents][Index]
This example demonstrates how a domain name is processed before it is lookup in the DNS. The input expected is in the locale encoding.
#include <stdio.h> /* printf, fflush, fgets, stdin, perror, fprintf */
#include <string.h> /* strlen */
#include <locale.h> /* setlocale */
#include <stdlib.h> /* free */
#include <idn2.h> /* idn2_lookup_ul, IDN2_OK, idn2_strerror, idn2_strerror_name */
int
main (int argc, char *argv[])
{
int rc;
char src[BUFSIZ];
char *lookupname;
setlocale (LC_ALL, "");
printf ("Enter (possibly non-ASCII) domain name to lookup: ");
fflush (stdout);
if (!fgets (src, sizeof (src), stdin))
{
perror ("fgets");
return 1;
}
src[strlen (src) - 1] = '\0';
rc = idn2_lookup_ul (src, &lookupname, 0);
if (rc != IDN2_OK)
{
fprintf (stderr, "error: %s (%s, %d)\n",
idn2_strerror (rc), idn2_strerror_name (rc), rc);
return 1;
}
printf ("IDNA2008 domain name to lookup in DNS: %s\n", lookupname);
free (lookupname);
return 0;
}
This example demonstrates how a domain label is processed before it is registered in the DNS. The input expected is in the locale encoding.
#include <stdio.h> /* printf, fflush, fgets, stdin, perror, fprintf */
#include <string.h> /* strlen */
#include <locale.h> /* setlocale */
#include <stdlib.h> /* free */
#include <idn2.h> /* idn2_register_ul, IDN2_OK, idn2_strerror, idn2_strerror_name */
int
main (int argc, char *argv[])
{
int rc;
char src[BUFSIZ];
char *insertname;
setlocale (LC_ALL, "");
printf ("Enter (possibly non-ASCII) label to register: ");
fflush (stdout);
if (!fgets (src, sizeof (src), stdin))
{
perror ("fgets");
return 1;
}
src[strlen (src) - 1] = '\0';
rc = idn2_register_ul (src, NULL, &insertname, 0);
if (rc != IDN2_OK)
{
fprintf (stderr, "error: %s (%s, %d)\n",
idn2_strerror (rc), idn2_strerror_name (rc), rc);
return 1;
}
printf ("IDNA2008 label to register in DNS: %s\n", insertname);
free (insertname);
return 0;
}
Next: Interface Index, Previous: Examples, Up: Libidn2 [Contents][Index]
idn2 translates internationalized domain names to the
IDNA2008 encoded format, either for lookup or registration.
If strings are specified on the command line, they are used as input
and the computed output is printed to standard output stdout.
If no strings are specified on the command line, the program read
data, line by line, from the standard input stdin, and print
the computed output to standard output. What processing is performed
(e.g., lookup or register) is indicated by options. If any errors are
encountered, the execution of the applications is aborted.
All strings are expected to be encoded in the preferred charset used
by your locale. Use --debug to find out what this charset is.
On POSIX systems you may use the LANG environment variable to
specify a different locale.
To process a string that starts with -, for example
-foo, use -- to signal the end of parameters, as in
idn2 -r -- -foo.
idn2 recognizes these commands:
-h, --help Print help and exit
-V, --version Print version and exit
-d, --decode Decode (punycode) domain name
-l, --lookup Lookup domain name (default)
-r, --register Register label
-T, --tr46t Enable TR46 transitional processing
-N, --tr46nt Enable TR46 non-transitional processing
--no-tr46 Disable TR46 processing
--usestd3asciirules Enable STD3 ASCII rules
--no-alabelroundtrip Disable A-label roundtrip for lookups
--debug Print debugging information
--quiet Silent operation
On POSIX systems the LANG environment variable can be used to override the system locale for the command being invoked. The system locale may influence what character set is used to decode data (i.e., strings on the command line or data read from the standard input stream), and to encode data to the standard output. If your system is set up correctly, however, the application will use the correct locale and character set automatically. Example usage:
$ LANG=en_US.UTF-8 idn2 ...
Standard usage, reading input from standard input and disabling license and usage instructions:
jas@latte:~$ idn2 --quiet räksmörgås.se xn--rksmrgs-5wao1o.se ...
Reading input from the command line:
jas@latte:~$ idn2 räksmörgås.se blåbærgrød.no xn--rksmrgs-5wao1o.se xn--blbrgrd-fxak7p.no jas@latte:~$
Testing the IDNA2008 Register function:
jas@latte:~$ idn2 --register fußball xn--fuball-cta jas@latte:~$
Getting character data encoded right, and making sure Libidn2 use the
same encoding, can be difficult. The reason for this is that most
systems may encode character data in more than one character encoding,
i.e., using UTF-8 together with ISO-8859-1 or
ISO-2022-JP. This problem is likely to continue to exist until
only one character encoding come out as the evolutionary winner, or
(more likely, at least to some extents) forever.
The first step to troubleshooting character encoding problems with Libidn2 is to use the ‘--debug’ parameter to find out which character set encoding ‘idn2’ believe your locale uses.
jas@latte:~$ idn2 --debug --quiet "" Charset: UTF-8 jas@latte:~$
If it prints ANSI_X3.4-1968 (i.e., US-ASCII), this
indicate you have not configured your locale properly. To configure
the locale, you can, for example, use ‘LANG=sv_SE.UTF-8; export
LANG’ at a /bin/sh prompt, to set up your locale for a Swedish
environment using UTF-8 as the encoding.
Sometimes ‘idn2’ appear to be unable to translate from your
system locale into UTF-8 (which is used internally), and you
will get an error message like this:
idn2: lookup: could not convert string to UTF-8
One explanation is that you didn’t install the ‘iconv’ conversion tools. You can find it as a standalone library in GNU Libiconv (https://www.gnu.org/software/libiconv/). On many GNU/Linux systems, this library is part of the system, but you may have to install additional packages to be able to use it.
Another explanation is that the error is correct and you are feeding
‘idn2’ invalid data. This can happen inadvertently if you are
not careful with the character set encoding you use. For example, if
your shell run in a ISO-8859-1 environment, and you invoke
‘idn2’ with the ‘LANG’ environment variable as follows, you
will feed it ISO-8859-1 characters but force it to believe they
are UTF-8. Naturally this will lead to an error, unless the
byte sequences happen to be valid UTF-8. Note that even if you
don’t get an error, the output may be incorrect in this situation,
because ISO-8859-1 and UTF-8 does not in general encode
the same characters as the same byte sequences.
jas@latte:~$ idn2 --quiet --debug "" Charset: ISO-8859-1 jas@latte:~$ LANG=sv_SE.UTF-8 idn2 --debug räksmörgås Charset: UTF-8 input[0] = 0x72 input[1] = 0xc3 input[2] = 0xa4 input[3] = 0xc3 input[4] = 0xa4 input[5] = 0x6b input[6] = 0x73 input[7] = 0x6d input[8] = 0xc3 input[9] = 0xb6 input[10] = 0x72 input[11] = 0x67 input[12] = 0xc3 input[13] = 0xa5 input[14] = 0x73 UCS-4 input[0] = U+0072 UCS-4 input[1] = U+00e4 UCS-4 input[2] = U+00e4 UCS-4 input[3] = U+006b UCS-4 input[4] = U+0073 UCS-4 input[5] = U+006d UCS-4 input[6] = U+00f6 UCS-4 input[7] = U+0072 UCS-4 input[8] = U+0067 UCS-4 input[9] = U+00e5 UCS-4 input[10] = U+0073 output[0] = 0x72 output[1] = 0xc3 output[2] = 0xa4 output[3] = 0xc3 output[4] = 0xa4 output[5] = 0x6b output[6] = 0x73 output[7] = 0x6d output[8] = 0xc3 output[9] = 0xb6 output[10] = 0x72 output[11] = 0x67 output[12] = 0xc3 output[13] = 0xa5 output[14] = 0x73 UCS-4 output[0] = U+0072 UCS-4 output[1] = U+00e4 UCS-4 output[2] = U+00e4 UCS-4 output[3] = U+006b UCS-4 output[4] = U+0073 UCS-4 output[5] = U+006d UCS-4 output[6] = U+00f6 UCS-4 output[7] = U+0072 UCS-4 output[8] = U+0067 UCS-4 output[9] = U+00e5 UCS-4 output[10] = U+0073 xn--rksmrgs-5waap8p jas@latte:~$
The sense moral here is to forget about ‘LANG’ (instead, configure your system locale properly) unless you know what you are doing, and if you want to use ‘LANG’, do it carefully and after verifying with ‘--debug’ that you get the desired results.
Next: Concept Index, Previous: Invoking idn2, Up: Libidn2 [Contents][Index]
| Jump to: | I |
|---|
| Jump to: | I |
|---|
Previous: Interface Index, Up: Libidn2 [Contents][Index]
| Jump to: | C E I L |
|---|
| Index Entry | Section | ||
|---|---|---|---|
| | |||
| C | |||
| command line: | Invoking idn2 | ||
| | |||
| E | |||
| Examples: | Examples | ||
| | |||
| I | |||
idn2: | Invoking idn2 | ||
invoking idn2: | Invoking idn2 | ||
| | |||
| L | |||
| libidn: | Converting from libidn | ||
| Library Functions: | Library Functions | ||
| | |||
| Jump to: | C E I L |
|---|