IsisReader: contrib/openisis/lcs.h Source File

00001 /*
00002         openisis - an open implementation of the CDS/ISIS database
00003         Version 0.8.x (patchlevel see file Version)
00004         Copyright (C) 2001-2003 by Erik Grziwotz, erik@openisis.org
00005 
00006         This library is free software; you can redistribute it and/or
00007         modify it under the terms of the GNU Lesser General Public
00008         License as published by the Free Software Foundation; either
00009         version 2.1 of the License, or (at your option) any later version.
00010 
00011         This library is distributed in the hope that it will be useful,
00012         but WITHOUT ANY WARRANTY; without even the implied warranty of
00013         MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
00014         Lesser General Public License for more details.
00015 
00016         You should have received a copy of the GNU Lesser General Public
00017         License along with this library; if not, write to the Free Software
00018         Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
00019 
00020         see README for more information
00021 EOH */
00022 #ifndef LCS_H
00023 
00024 /*
00025         $Id: lcs.h,v 1.3 2003/04/08 00:20:52 kripke Exp $
00026         charset support
00027 */
00028 
00029 typedef enum {
00030         LCS_SINGLE, /* some ASCII-compliant single byte charset */
00031         LCS_UTF8, /* unicode in 8bit transfer encoding */
00032         LCS_UCS2 /* the unicode base multilingual plane (BMP, first 64K chars)
00033                 in 2byte encoding, native (platform) byte order */
00034 }       lcs_type;
00035 
00036 typedef enum {
00037         LCS_LATIN1, /* the ISO-8859-1 (Latin 1) charset */
00038         LCS_CP850, /* MS-DOS Latin 1, known as the "OEM" charset */
00039         LCS__SBCS  /* # of single byte charsets */
00040 }       lcs_single;
00041 
00042 
00043 typedef enum {
00044         LCS_CTYPE, /* flags see below */
00045         LCS_UNICO, /* 128 unicodes */
00046         LCS_UCASE, /* each characters uppercase code */
00047         LCS_INDEX, /* single byte recoding for index ("strxfrm") */
00048         LCS_XEDNI, /* inverse */
00049         LCS_RECOD, /* single byte recoding to extern (typically Latin1) */
00050         LCS_DOCER, /* inverse */
00051         LCS__TABS
00052 } lcs_tabid;
00053 
00054 
00055 /* some conversion table on a single byte charset
00056         c may hold an actab (!0 for any character deemed "alpha"),
00057         uctab ("uppercase" replacement code, typically removing diacriticals),
00058         code-to-code conversion or other.
00059         u may hold unicode values for the upper half of an
00060         ASCII compatible single byte charset.
00061 */
00062 
00063 typedef union {
00064         unsigned char  c[256];
00065         unsigned short u[128];
00066 }       LcsTab;
00067 
00068 enum {
00069         /* controls (0..31,127) */
00070         LCS_R = 0x00, /* record separator (FF,GS), */
00071         LCS_F = 0x10, /* field  separator (NUL,CR,LF,RS) */
00072         LCS_C = 0x28, /* other control */
00073         /* other separators */
00074         LCS_S = 0x48, /* space(blank,nbsp), */
00075         LCS_T = 0x68, /* other token separator (,:;=), */
00076         /* other punctuation assumed "word" characters */
00077         LCS_P = 0x80, 
00078         LCS_Y = 0xb0, /* symbol */
00079         /* identifiers */
00080         LCS_I = 0xc0, /* C-ident (_). */
00081         LCS_D = 0xd0, /* digits ('0'..'9') */
00082         LCS_A = 0xe0, /* ASCII alpha */
00083         LCS_L = 0xe4, /* ... among those formatting literals a,c,i,x */
00084         LCS_N = 0xf0  /* other alpha ("national"/non-ASCII) */
00085 };
00086 
00087 #define LCS_ISCONTR( t ) (!(0xc0 & (t)))
00088 #define LCS_ISSPACE( t ) (LCS_S >= (t))
00089 #define LCS_ISPRINT( t ) (0xc0 & (t)) /* (LCS_S <= (t)) */
00090 #define LCS_ISIDENT( t ) (0xc0 == (0xc0 & (t)))
00091 #define LCS_ISALPHA( t ) (0xe0 == (0xe0 & (t)))
00092 
00093 /* for record parsing */
00094 #define LCS_ISSEP( t )   (!(0x80 & (t)))
00095 #define LCS_ISWORD( t )  (0x80 & (t))
00096 #define LCS_ISFR( t )    (!(0xe0 & (t))) /* field or record separator */
00097 #define LCS_ISCST( t )   (0x08 & (t)) /* other separator */
00098 
00099 
00100 enum {
00101         LCS_MKUNI = -1 /* as "bits" value, have mktab create unicode table */
00102 };
00103 
00111 extern int lcs_mktab ( LcsTab *dst, char *numbers, int len, int bits );
00112 
00116 extern int lcs_mkrecod ( unsigned char *dst, unsigned short *from,
00117         unsigned short *to );
00118 
00119 extern unsigned char lcs_latin1_uc[256];
00120 extern unsigned char lcs_latin1_ct[256];
00121 
00122 #define LCS_H
00123 #endif /* LCS_H */