thodg/libiconv/lib/iconv_open1.h

Branch : - branch - master - tag - 1.11.x-branchpoint clisp-2.25.1-2001-04-06 clisp-2.26-2001-05-23 clisp-2.27-2001-07-17 clisp-2.28-2002-03-03 libcharset-1.1 v0.3 v1.0 v1.1 v1.10 v1.11 v1.11.1 v1.12 v1.13 v1.13.1 v1.14 v1.15 v1.16 v1.17 v1.18 v1.2 v1.3 v1.4 v1.5 v1.5.1 v1.6 v1.6.1 v1.7 v1.8 v1.9 v1.9.1 v1.9.2

• Show log

  Commit

• Author : Bruno Haible
  Date : 2023-04-03 04:12:01
  Hash : 19b6af5e
  Message : Allow overriding the newline conversion for EBCDIC encodings. Reported by Mike Fulton <mikefultonpersonal@gmail.com> in <https://lists.gnu.org/archive/html/bug-gnu-libiconv/2023-04/msg00009.html>. * include/iconv.h.in (ICONV_SURFACE_NONE, ICONV_SURFACE_EBCDIC_ZOS_UNIX): New macros. (ICONV_GET_FROM_SURFACE, ICONV_SET_FROM_SURFACE, ICONV_GET_TO_SURFACE, ICONV_SET_TO_SURFACE): New macros. * lib/converters.h (struct conv_struct): Add the fields isurface, osurface. (swap_x15_x25): New macro. * lib/iconv.c (iconv_open, iconv_open_into): Add local variables from_surface, to_surface. (ALL_SURFACES): New macro. (iconvctl): Adjust ICONV_TRIVIALP implementation. Implement the ICONV_{GET,SET}_{FROM,TO}_SURFACE requests. * lib/iconv_open1.h: Parse a /ZOS_UNIX surface specifier. Set from_surface, to_surface. * lib/iconv_open2.h: Copy the values of from_surface, to_surface into the conversion descriptor. * lib/ebcdic*.h (*_mbtowc): Test the isurface. If requested, call swap_x15_x25 right after fetching an input byte. (*_wctomb): Test the osurface. If requested, call swap_x15_x25 right before storing an output byte. * man/iconvctl.3 (REQUEST VALUES): Document the ICONV_{GET,SET}_{FROM,TO}_SURFACE requests. * src/iconv.c (main): If ICONV_EBCDIC_ZOS_UNIX is set, set the from/to surfaces accordingly. * man/iconv.1 (ENVIRONMENT): New section. * tests/check-ebcdic: New file. * tests/Makefile.in (check): Invoke it. (SOURCE_FILES): Add it. * NEWS: Mention the new functionality.

• lib/iconv_open1.h

• /*
   * Copyright (C) 1999-2008, 2011, 2018, 2020, 2023 Free Software Foundation, Inc.
   * This file is part of the GNU LIBICONV Library.
   *
   * The GNU LIBICONV Library is free software; you can redistribute it
   * and/or modify it under the terms of the GNU Lesser General Public
   * License as published by the Free Software Foundation; either version 2.1
   * of the License, or (at your option) any later version.
   *
   * The GNU LIBICONV Library is distributed in the hope that it will be
   * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of
   * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
   * Lesser General Public License for more details.
   *
   * You should have received a copy of the GNU Lesser General Public
   * License along with the GNU LIBICONV Library; see the file COPYING.LIB.
   * If not, see <https://www.gnu.org/licenses/>.
   */
  
  /* Part 1 of iconv_open.
     Input: const char* tocode, const char* fromcode.
     Output:
       unsigned int from_index;
       int from_wchar;
       unsigned int from_surface;
       unsigned int to_index;
       int to_wchar;
       unsigned int to_surface;
       int transliterate;
       int discard_ilseq;
     Jumps to 'invalid' in case of errror.
   */
  {
    char buf[MAX_WORD_LENGTH+9+9+1];
    const char* cp;
    char* bp;
    const struct alias * ap;
    unsigned int count;
  
    from_surface = ICONV_SURFACE_NONE;
    to_surface = ICONV_SURFACE_NONE;
    transliterate = 0;
    discard_ilseq = 0;
  
    /* Before calling aliases_lookup, convert the input string to upper case,
     * and check whether it's entirely ASCII (we call gperf with option "-7"
     * to achieve a smaller table) and non-empty. If it's not entirely ASCII,
     * or if it's too long, it is not a valid encoding name.
     */
    for (to_wchar = 0;;) {
      /* Search tocode in the table. */
      for (cp = tocode, bp = buf, count = MAX_WORD_LENGTH+9+9+1; ; cp++, bp++) {
        unsigned char c = (unsigned char) *cp;
        if (c >= 0x80)
          goto invalid;
        if (c >= 'a' && c <= 'z')
          c -= 'a'-'A';
        *bp = c;
        if (c == '\0')
          break;
        if (--count == 0)
          goto invalid;
      }
      for (;;) {
        char *sp = bp;
        int parsed_translit = 0;
        int parsed_ignore = 0;
        if (sp-buf > 9 && memcmp(sp-9,"/TRANSLIT",9)==0) {
          sp = sp - 9;
          parsed_translit = 1;
        } else if (sp-buf > 7 && memcmp(sp-7,"/IGNORE",7)==0) {
          sp = sp - 7;
          parsed_ignore = 1;
        }
        if (sp > buf && memcmp(sp-1,"/",1) == 0) {
          bp = sp - 1;
        } else if (sp-buf >= 9 && memcmp(sp-9,"/ZOS_UNIX",9)==0) {
          bp = sp - 9;
          to_surface = ICONV_SURFACE_EBCDIC_ZOS_UNIX;
        } else
          break;
        *bp = '\0';
        if (parsed_translit)
          transliterate = 1;
        if (parsed_ignore)
          discard_ilseq = 1;
        break;
      }
      if (buf[0] == '\0') {
        tocode = locale_charset();
        /* Avoid an endless loop that could occur when using an older version
           of localcharset.c. */
        if (tocode[0] == '\0')
          goto invalid;
        continue;
      }
      ap = aliases_lookup(buf,bp-buf);
      if (ap == NULL) {
        ap = aliases2_lookup(buf);
        if (ap == NULL)
          goto invalid;
      }
      if (ap->encoding_index == ei_local_char) {
        tocode = locale_charset();
        /* Avoid an endless loop that could occur when using an older version
           of localcharset.c. */
        if (tocode[0] == '\0')
          goto invalid;
        continue;
      }
      if (ap->encoding_index == ei_local_wchar_t) {
        /* On systems which define __STDC_ISO_10646__, wchar_t is Unicode.
           This is also the case on native Woe32 systems and Cygwin >= 1.7, where
           we know that it is UTF-16.  */
  #if (defined _WIN32 && !defined __CYGWIN__) || (defined __CYGWIN__ && CYGWIN_VERSION_DLL_MAJOR >= 1007)
        if (sizeof(wchar_t) == 4) {
          to_index = ei_ucs4internal;
          break;
        }
        if (sizeof(wchar_t) == 2) {
  # if WORDS_LITTLEENDIAN
          to_index = ei_utf16le;
  # else
          to_index = ei_utf16be;
  # endif
          break;
        }
  #elif __STDC_ISO_10646__
        if (sizeof(wchar_t) == 4) {
          to_index = ei_ucs4internal;
          break;
        }
        if (sizeof(wchar_t) == 2) {
          to_index = ei_ucs2internal;
          break;
        }
        if (sizeof(wchar_t) == 1) {
          to_index = ei_iso8859_1;
          break;
        }
  #endif
  #if HAVE_MBRTOWC
        to_wchar = 1;
        tocode = locale_charset();
        continue;
  #endif
        goto invalid;
      }
      to_index = ap->encoding_index;
      break;
    }
    for (from_wchar = 0;;) {
      /* Search fromcode in the table. */
      for (cp = fromcode, bp = buf, count = MAX_WORD_LENGTH+9+9+1; ; cp++, bp++) {
        unsigned char c = (unsigned char) *cp;
        if (c >= 0x80)
          goto invalid;
        if (c >= 'a' && c <= 'z')
          c -= 'a'-'A';
        *bp = c;
        if (c == '\0')
          break;
        if (--count == 0)
          goto invalid;
      }
      for (;;) {
        char *sp = bp;
        int parsed_translit = 0;
        int parsed_ignore = 0;
        if (sp-buf > 9 && memcmp(sp-9,"/TRANSLIT",9)==0) {
          sp = sp - 9;
          parsed_translit = 1;
        } else if (sp-buf > 7 && memcmp(sp-7,"/IGNORE",7)==0) {
          sp = sp - 7;
          parsed_ignore = 1;
        }
        if (sp > buf && memcmp(sp-1,"/",1) == 0) {
          bp = sp - 1;
        } else if (sp-buf >= 9 && memcmp(sp-9,"/ZOS_UNIX",9)==0) {
          bp = sp - 9;
          from_surface = ICONV_SURFACE_EBCDIC_ZOS_UNIX;
        } else
          break;
        *bp = '\0';
        if (parsed_translit)
          transliterate = 1;
        if (parsed_ignore)
          discard_ilseq = 1;
        break;
      }
      if (buf[0] == '\0') {
        fromcode = locale_charset();
        /* Avoid an endless loop that could occur when using an older version
           of localcharset.c. */
        if (fromcode[0] == '\0')
          goto invalid;
        continue;
      }
      ap = aliases_lookup(buf,bp-buf);
      if (ap == NULL) {
        ap = aliases2_lookup(buf);
        if (ap == NULL)
          goto invalid;
      }
      if (ap->encoding_index == ei_local_char) {
        fromcode = locale_charset();
        /* Avoid an endless loop that could occur when using an older version
           of localcharset.c. */
        if (fromcode[0] == '\0')
          goto invalid;
        continue;
      }
      if (ap->encoding_index == ei_local_wchar_t) {
        /* On systems which define __STDC_ISO_10646__, wchar_t is Unicode.
           This is also the case on native Woe32 systems and Cygwin >= 1.7, where
           we know that it is UTF-16.  */
  #if (defined _WIN32 && !defined __CYGWIN__) || (defined __CYGWIN__ && CYGWIN_VERSION_DLL_MAJOR >= 1007)
        if (sizeof(wchar_t) == 4) {
          from_index = ei_ucs4internal;
          break;
        }
        if (sizeof(wchar_t) == 2) {
  # if WORDS_LITTLEENDIAN
          from_index = ei_utf16le;
  # else
          from_index = ei_utf16be;
  # endif
          break;
        }
  #elif __STDC_ISO_10646__
        if (sizeof(wchar_t) == 4) {
          from_index = ei_ucs4internal;
          break;
        }
        if (sizeof(wchar_t) == 2) {
          from_index = ei_ucs2internal;
          break;
        }
        if (sizeof(wchar_t) == 1) {
          from_index = ei_iso8859_1;
          break;
        }
  #endif
  #if HAVE_WCRTOMB
        from_wchar = 1;
        fromcode = locale_charset();
        continue;
  #endif
        goto invalid;
      }
      from_index = ap->encoding_index;
      break;
    }
  }