[apple/icu.git] / icuSources / tools / gensprep / gensprep.c

/*
*******************************************************************************
*
*   Copyright (C) 2003-2006, International Business Machines
*   Corporation and others.  All Rights Reserved.
*
*******************************************************************************
*   file name:  gensprep.c
*   encoding:   US-ASCII
*   tab size:   8 (not used)
*   indentation:4
*
*   created on: 2003-02-06
*   created by: Ram Viswanadha
*
*   This program reads the Profile.txt files,
*   parses them, and extracts the data for StringPrep profile.
*   It then preprocesses it and writes a binary file for efficient use
*   in various StringPrep conversion processes.
*/

#define USPREP_TYPE_NAMES_ARRAY 1

#include <stdio.h>
#include <stdlib.h>

#include "cmemory.h"
#include "cstring.h"
#include "unewdata.h"
#include "uoptions.h"
#include "uparse.h"
#include "sprpimpl.h"

#include "unicode/udata.h"
#include "unicode/utypes.h"
#include "unicode/putil.h"


U_CDECL_BEGIN
#include "gensprep.h"
U_CDECL_END

UBool beVerbose=FALSE, haveCopyright=TRUE;

#define NORM_CORRECTIONS_FILE_NAME "NormalizationCorrections.txt"

/* prototypes --------------------------------------------------------------- */

static void
parseMappings(const char *filename, UBool reportError, UErrorCode *pErrorCode);

static void
parseNormalizationCorrections(const char *filename, UErrorCode *pErrorCode);


/* -------------------------------------------------------------------------- */

static UOption options[]={
    UOPTION_HELP_H,
    UOPTION_HELP_QUESTION_MARK,
    UOPTION_VERBOSE,
    UOPTION_COPYRIGHT,
    UOPTION_DESTDIR,
    UOPTION_SOURCEDIR,
    UOPTION_ICUDATADIR,
    UOPTION_BUNDLE_NAME,
    { "normalization", NULL, NULL, NULL, 'n', UOPT_REQUIRES_ARG, 0 },
    { "check-bidi", NULL, NULL, NULL,  'k', UOPT_NO_ARG, 0},
    { "unicode", NULL, NULL, NULL, 'u', UOPT_REQUIRES_ARG, 0 },
};

enum{
    HELP,
    HELP_QUESTION_MARK,
    VERBOSE,
    COPYRIGHT,
    DESTDIR,
    SOURCEDIR,
    ICUDATADIR,
    BUNDLE_NAME,
    NORMALIZE,
    CHECK_BIDI,
    UNICODE_VERSION
};

static int printHelp(int argc, char* argv[]){
    /*
     * Broken into chucks because the C89 standard says the minimum
     * required supported string length is 509 bytes.
     */
    fprintf(stderr,
        "Usage: %s [-options] [file_name]\n"
        "\n"
        "Read the files specified and\n"
        "create a binary file [package-name]_[bundle-name]." DATA_TYPE " with the StringPrep profile data\n"
        "\n",
        argv[0]);
    fprintf(stderr,
        "Options:\n"
        "\t-h or -? or --help       print this usage text\n"
        "\t-v or --verbose          verbose output\n"
        "\t-c or --copyright        include a copyright notice\n");
    fprintf(stderr,
        "\t-d or --destdir          destination directory, followed by the path\n"
        "\t-s or --sourcedir        source directory of ICU data, followed by the path\n"
        "\t-b or --bundle-name      generate the ouput data file with the name specified\n"
        "\t-i or --icudatadir       directory for locating any needed intermediate data files,\n"
        "\t                         followed by path, defaults to %s\n",
        u_getDataDirectory());
    fprintf(stderr,
        "\t-n or --normalize        turn on the option for normalization and include mappings\n"
        "\t                         from NormalizationCorrections.txt from the given path,\n"
        "\t                         e.g: /test/icu/source/data/unidata\n"
        "\t-k or --check-bidi       turn on the option for checking for BiDi in the profile\n"
        "\t-u or --unicode          version of Unicode to be used with this profile followed by the version\n"
        );
    return argc<0 ? U_ILLEGAL_ARGUMENT_ERROR : U_ZERO_ERROR;
}


extern int
main(int argc, char* argv[]) {
#if !UCONFIG_NO_IDNA
    char* filename = NULL;
#endif
    const char *srcDir=NULL, *destDir=NULL, *icuUniDataDir=NULL;
    const char *bundleName=NULL, *inputFileName = NULL;
    char *basename=NULL;
    int32_t sprepOptions = 0;

    UErrorCode errorCode=U_ZERO_ERROR;

    U_MAIN_INIT_ARGS(argc, argv);

    /* preset then read command line options */
    options[DESTDIR].value=u_getDataDirectory();
    options[SOURCEDIR].value="";
    options[UNICODE_VERSION].value="0"; /* don't assume the unicode version */
    options[BUNDLE_NAME].value = DATA_NAME;
    options[NORMALIZE].value = "";

    argc=u_parseArgs(argc, argv, sizeof(options)/sizeof(options[0]), options);

    /* error handling, printing usage message */
    if(argc<0) {
        fprintf(stderr,
            "error in command line argument \"%s\"\n",
            argv[-argc]);
    }
    if(argc<0 || options[HELP].doesOccur || options[HELP_QUESTION_MARK].doesOccur) {
        return printHelp(argc, argv);
        
    }

    /* get the options values */
    beVerbose=options[VERBOSE].doesOccur;
    haveCopyright=options[COPYRIGHT].doesOccur;
    srcDir=options[SOURCEDIR].value;
    destDir=options[DESTDIR].value;
    bundleName = options[BUNDLE_NAME].value;
    icuUniDataDir = options[NORMALIZE].value;

    if(argc<2) {
        /* print the help message */
        return printHelp(argc, argv);
    } else {
        inputFileName = argv[1];
    }
    if(!options[UNICODE_VERSION].doesOccur){
        return printHelp(argc, argv);
    }
    if(options[ICUDATADIR].doesOccur) {
        u_setDataDirectory(options[ICUDATADIR].value);
    }
#if UCONFIG_NO_IDNA

    fprintf(stderr,
        "gensprep writes dummy " U_ICUDATA_NAME "_" DATA_NAME "." DATA_TYPE
        " because UCONFIG_NO_IDNA is set, \n"
        "see icu/source/common/unicode/uconfig.h\n");
    generateData(destDir, bundleName);

#else

    setUnicodeVersion(options[UNICODE_VERSION].value);
    filename = (char* ) uprv_malloc(uprv_strlen(srcDir) + 300); /* hopefully this should be enough */
   
    /* prepare the filename beginning with the source dir */
    if(uprv_strchr(srcDir,U_FILE_SEP_CHAR) == NULL && uprv_strchr(srcDir,U_FILE_ALT_SEP_CHAR) == NULL){
        filename[0] = '.';
        filename[1] = U_FILE_SEP_CHAR;
        uprv_strcpy(filename+2,srcDir);
    }else{
        uprv_strcpy(filename, srcDir);
    }
    
    basename=filename+uprv_strlen(filename);
    if(basename>filename && *(basename-1)!=U_FILE_SEP_CHAR) {
        *basename++=U_FILE_SEP_CHAR;
    }
    
    /* initialize */
    init();

    /* process the file */
    uprv_strcpy(basename,inputFileName);
    parseMappings(filename,FALSE, &errorCode);
    if(U_FAILURE(errorCode)) {
        fprintf(stderr, "Could not open file %s for reading. Error: %s \n", filename, u_errorName(errorCode));
        return errorCode;
    }
    
    if(options[NORMALIZE].doesOccur){
        /* set up directory for NormalizationCorrections.txt */
        uprv_strcpy(filename,icuUniDataDir);
        basename=filename+uprv_strlen(filename);
        if(basename>filename && *(basename-1)!=U_FILE_SEP_CHAR) {
            *basename++=U_FILE_SEP_CHAR;
        }

        *basename++=U_FILE_SEP_CHAR;
        uprv_strcpy(basename,NORM_CORRECTIONS_FILE_NAME);
    
        parseNormalizationCorrections(filename,&errorCode);
        if(U_FAILURE(errorCode)){
            fprintf(stderr,"Could not open file %s for reading \n", filename);
            return errorCode;
        }
        sprepOptions |= _SPREP_NORMALIZATION_ON;
    }
    
    if(options[CHECK_BIDI].doesOccur){
        sprepOptions |= _SPREP_CHECK_BIDI_ON;
    }

    setOptions(sprepOptions);

    /* process parsed data */
    if(U_SUCCESS(errorCode)) {
        /* write the data file */
        generateData(destDir, bundleName);

        cleanUpData();
    }

    uprv_free(filename);

#endif

    return errorCode;
}

#if !UCONFIG_NO_IDNA

static void U_CALLCONV
normalizationCorrectionsLineFn(void *context,
                    char *fields[][2], int32_t fieldCount,
                    UErrorCode *pErrorCode) {
    uint32_t mapping[40];
    char *end, *s;
    uint32_t code;
    int32_t length;
    UVersionInfo version;
    UVersionInfo thisVersion;

    /* get the character code, field 0 */
    code=(uint32_t)uprv_strtoul(fields[0][0], &end, 16);
    if(U_FAILURE(*pErrorCode)) {
        fprintf(stderr, "gensprep: error parsing NormalizationCorrections.txt mapping at %s\n", fields[0][0]);
        exit(*pErrorCode);
    }
    /* Original (erroneous) decomposition */
    s = fields[1][0];

    /* parse the mapping string */
    length=u_parseCodePoints(s, mapping, sizeof(mapping)/4, pErrorCode);

    /* ignore corrected decomposition */

    u_versionFromString(version,fields[3][0] );
    u_versionFromString(thisVersion, "3.2.0");


    if(U_FAILURE(*pErrorCode)) {
        fprintf(stderr, "gensprep error parsing NormalizationCorrections.txt of U+%04lx - %s\n",
                (long)code, u_errorName(*pErrorCode));
        exit(*pErrorCode);
    }

    /* store the mapping */
    if( version[0] > thisVersion[0] || 
        ((version[0]==thisVersion[0]) && (version[1] > thisVersion[1]))
        ){
        storeMapping(code,mapping, length, USPREP_MAP, pErrorCode);
    }
    setUnicodeVersionNC(version);
}

static void
parseNormalizationCorrections(const char *filename, UErrorCode *pErrorCode) {
    char *fields[4][2];

    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
        return;
    }

    u_parseDelimitedFile(filename, ';', fields, 4, normalizationCorrectionsLineFn, NULL, pErrorCode);

    /* fprintf(stdout,"Number of code points that have NormalizationCorrections mapping with length >1 : %i\n",len); */

    if(U_FAILURE(*pErrorCode) && ( *pErrorCode!=U_FILE_ACCESS_ERROR)) {
        fprintf(stderr, "gensprep error: u_parseDelimitedFile(\"%s\") failed - %s\n", filename, u_errorName(*pErrorCode));
        exit(*pErrorCode);
    }
}

static void U_CALLCONV
strprepProfileLineFn(void *context,
              char *fields[][2], int32_t fieldCount,
              UErrorCode *pErrorCode) {
    uint32_t mapping[40];
    char *end, *map;
    uint32_t code;
    int32_t length;
   /*UBool* mapWithNorm = (UBool*) context;*/
    const char* typeName;
    uint32_t rangeStart=0,rangeEnd =0;
    const char* filename = (const char*) context;
 
    typeName = fields[2][0];
    map = fields[1][0];

    if(uprv_strstr(typeName, usprepTypeNames[USPREP_UNASSIGNED])!=NULL){

        u_parseCodePointRange(fields[0][0], &rangeStart,&rangeEnd, pErrorCode);
        if(U_FAILURE(*pErrorCode)){
            fprintf(stderr, "Could not parse code point range. Error: %s\n",u_errorName(*pErrorCode));
            return;
        }

        /* store the range */
        storeRange(rangeStart,rangeEnd,USPREP_UNASSIGNED, pErrorCode);

    }else if(uprv_strstr(typeName, usprepTypeNames[USPREP_PROHIBITED])!=NULL){

        u_parseCodePointRange(fields[0][0], &rangeStart,&rangeEnd, pErrorCode);
        if(U_FAILURE(*pErrorCode)){
            fprintf(stderr, "Could not parse code point range. Error: %s\n",u_errorName(*pErrorCode));
            return;
        }

        /* store the range */
        storeRange(rangeStart,rangeEnd,USPREP_PROHIBITED, pErrorCode);

    }else if(uprv_strstr(typeName, usprepTypeNames[USPREP_MAP])!=NULL){

        /* get the character code, field 0 */
        code=(uint32_t)uprv_strtoul(fields[0][0], &end, 16);
        if(end<=fields[0][0] || end!=fields[0][1]) {
            fprintf(stderr, "gensprep: syntax error in field 0 at %s\n", fields[0][0]);
            *pErrorCode=U_PARSE_ERROR;
            exit(U_PARSE_ERROR);
        }

        /* parse the mapping string */
        length=u_parseCodePoints(map, mapping, sizeof(mapping)/4, pErrorCode);
        
        /* store the mapping */
        storeMapping(code,mapping, length,USPREP_MAP, pErrorCode);

    }else{
        *pErrorCode = U_INVALID_FORMAT_ERROR;
    }
    
    if(U_FAILURE(*pErrorCode)) {
        fprintf(stderr, "gensprep error parsing  %s line %s at %s. Error: %s\n",filename,
               fields[0][0],fields[2][0],u_errorName(*pErrorCode));
        exit(*pErrorCode);
    }

}

static void
parseMappings(const char *filename, UBool reportError, UErrorCode *pErrorCode) {
    char *fields[3][2];

    if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
        return;
    }

    u_parseDelimitedFile(filename, ';', fields, 3, strprepProfileLineFn, (void*)filename, pErrorCode);

    /*fprintf(stdout,"Number of code points that have mappings with length >1 : %i\n",len);*/

    if(U_FAILURE(*pErrorCode) && (reportError || *pErrorCode!=U_FILE_ACCESS_ERROR)) {
        fprintf(stderr, "gensprep error: u_parseDelimitedFile(\"%s\") failed - %s\n", filename, u_errorName(*pErrorCode));
        exit(*pErrorCode);
    }
}


#endif /* #if !UCONFIG_NO_IDNA */

/*
 * Hey, Emacs, please set the following:
 *
 * Local Variables:
 * indent-tabs-mode: nil
 * End:
 *
 */
Commit	Line	Data
374ca955 A	1	/*
	2	*******************************************************************************
	3	*
73c04bcf	4	* Copyright (C) 2003-2006, International Business Machines
374ca955 A	5	* Corporation and others. All Rights Reserved.
	6	*
	7	*******************************************************************************
	8	* file name: gensprep.c
	9	* encoding: US-ASCII
	10	* tab size: 8 (not used)
	11	* indentation:4
	12	*
	13	* created on: 2003-02-06
	14	* created by: Ram Viswanadha
	15	*
	16	* This program reads the Profile.txt files,
	17	* parses them, and extracts the data for StringPrep profile.
	18	* It then preprocesses it and writes a binary file for efficient use
	19	* in various StringPrep conversion processes.
	20	*/
	21
	22	#define USPREP_TYPE_NAMES_ARRAY 1
	23
	24	#include <stdio.h>
	25	#include <stdlib.h>
	26
	27	#include "cmemory.h"
	28	#include "cstring.h"
	29	#include "unewdata.h"
	30	#include "uoptions.h"
	31	#include "uparse.h"
	32	#include "sprpimpl.h"
	33
	34	#include "unicode/udata.h"
	35	#include "unicode/utypes.h"
	36	#include "unicode/putil.h"
	37
	38
	39	U_CDECL_BEGIN
	40	#include "gensprep.h"
	41	U_CDECL_END
	42
374ca955 A	43	UBool beVerbose=FALSE, haveCopyright=TRUE;
	44
	45	#define NORM_CORRECTIONS_FILE_NAME "NormalizationCorrections.txt"
	46
	47	/* prototypes --------------------------------------------------------------- */
	48
	49	static void
	50	parseMappings(const char filename, UBool reportError, UErrorCode pErrorCode);
	51
	52	static void
	53	parseNormalizationCorrections(const char filename, UErrorCode pErrorCode);
	54
	55
	56	/* -------------------------------------------------------------------------- */
	57
	58	static UOption options[]={
	59	UOPTION_HELP_H,
	60	UOPTION_HELP_QUESTION_MARK,
	61	UOPTION_VERBOSE,
	62	UOPTION_COPYRIGHT,
	63	UOPTION_DESTDIR,
	64	UOPTION_SOURCEDIR,
	65	UOPTION_ICUDATADIR,
374ca955 A	66	UOPTION_BUNDLE_NAME,
	67	{ "normalization", NULL, NULL, NULL, 'n', UOPT_REQUIRES_ARG, 0 },
	68	{ "check-bidi", NULL, NULL, NULL, 'k', UOPT_NO_ARG, 0},
	69	{ "unicode", NULL, NULL, NULL, 'u', UOPT_REQUIRES_ARG, 0 },
	70	};
	71
	72	enum{
	73	HELP,
	74	HELP_QUESTION_MARK,
	75	VERBOSE,
	76	COPYRIGHT,
	77	DESTDIR,
	78	SOURCEDIR,
	79	ICUDATADIR,
374ca955 A	80	BUNDLE_NAME,
	81	NORMALIZE,
	82	CHECK_BIDI,
	83	UNICODE_VERSION
	84	};
	85
	86	static int printHelp(int argc, char* argv[]){
	87	/*
	88	* Broken into chucks because the C89 standard says the minimum
	89	* required supported string length is 509 bytes.
	90	*/
	91	fprintf(stderr,
	92	"Usage: %s [-options] [file_name]\n"
	93	"\n"
	94	"Read the files specified and\n"
	95	"create a binary file [package-name]_[bundle-name]." DATA_TYPE " with the StringPrep profile data\n"
	96	"\n",
	97	argv[0]);
	98	fprintf(stderr,
	99	"Options:\n"
	100	"\t-h or -? or --help print this usage text\n"
	101	"\t-v or --verbose verbose output\n"
	102	"\t-c or --copyright include a copyright notice\n");
	103	fprintf(stderr,
	104	"\t-d or --destdir destination directory, followed by the path\n"
	105	"\t-s or --sourcedir source directory of ICU data, followed by the path\n"
	106	"\t-b or --bundle-name generate the ouput data file with the name specified\n"
	107	"\t-i or --icudatadir directory for locating any needed intermediate data files,\n"
	108	"\t followed by path, defaults to %s\n",
	109	u_getDataDirectory());
	110	fprintf(stderr,
374ca955 A	111	"\t-n or --normalize turn on the option for normalization and include mappings\n"
	112	"\t from NormalizationCorrections.txt from the given path,\n"
	113	"\t e.g: /test/icu/source/data/unidata\n"
	114	"\t-k or --check-bidi turn on the option for checking for BiDi in the profile\n"
	115	"\t-u or --unicode version of Unicode to be used with this profile followed by the version\n"
	116	);
	117	return argc<0 ? U_ILLEGAL_ARGUMENT_ERROR : U_ZERO_ERROR;
	118	}
	119
	120
	121	extern int
	122	main(int argc, char* argv[]) {
	123	#if !UCONFIG_NO_IDNA
	124	char* filename = NULL;
	125	#endif
	126	const char srcDir=NULL, destDir=NULL, *icuUniDataDir=NULL;
73c04bcf	127	const char bundleName=NULL, inputFileName = NULL;
374ca955 A	128	char *basename=NULL;
	129	int32_t sprepOptions = 0;
	130
	131	UErrorCode errorCode=U_ZERO_ERROR;
	132
	133	U_MAIN_INIT_ARGS(argc, argv);
	134
	135	/* preset then read command line options */
	136	options[DESTDIR].value=u_getDataDirectory();
	137	options[SOURCEDIR].value="";
	138	options[UNICODE_VERSION].value="0"; /* don't assume the unicode version */
	139	options[BUNDLE_NAME].value = DATA_NAME;
374ca955 A	140	options[NORMALIZE].value = "";
	141
	142	argc=u_parseArgs(argc, argv, sizeof(options)/sizeof(options[0]), options);
	143
	144	/* error handling, printing usage message */
	145	if(argc<0) {
	146	fprintf(stderr,
	147	"error in command line argument \"%s\"\n",
	148	argv[-argc]);
	149	}
	150	if(argc<0 \|\| options[HELP].doesOccur \|\| options[HELP_QUESTION_MARK].doesOccur) {
	151	return printHelp(argc, argv);
	152
	153	}
	154
	155	/* get the options values */
	156	beVerbose=options[VERBOSE].doesOccur;
	157	haveCopyright=options[COPYRIGHT].doesOccur;
	158	srcDir=options[SOURCEDIR].value;
	159	destDir=options[DESTDIR].value;
374ca955 A	160	bundleName = options[BUNDLE_NAME].value;
	161	icuUniDataDir = options[NORMALIZE].value;
	162
	163	if(argc<2) {
	164	/* print the help message */
	165	return printHelp(argc, argv);
	166	} else {
	167	inputFileName = argv[1];
	168	}
	169	if(!options[UNICODE_VERSION].doesOccur){
	170	return printHelp(argc, argv);
	171	}
	172	if(options[ICUDATADIR].doesOccur) {
	173	u_setDataDirectory(options[ICUDATADIR].value);
	174	}
	175	#if UCONFIG_NO_IDNA
	176
	177	fprintf(stderr,
	178	"gensprep writes dummy " U_ICUDATA_NAME "_" DATA_NAME "." DATA_TYPE
	179	" because UCONFIG_NO_IDNA is set, \n"
	180	"see icu/source/common/unicode/uconfig.h\n");
73c04bcf	181	generateData(destDir, bundleName);
374ca955 A	182
	183	#else
	184
	185	setUnicodeVersion(options[UNICODE_VERSION].value);
	186	filename = (char* ) uprv_malloc(uprv_strlen(srcDir) + 300); /* hopefully this should be enough */
	187
	188	/* prepare the filename beginning with the source dir */
	189	if(uprv_strchr(srcDir,U_FILE_SEP_CHAR) == NULL && uprv_strchr(srcDir,U_FILE_ALT_SEP_CHAR) == NULL){
	190	filename[0] = '.';
	191	filename[1] = U_FILE_SEP_CHAR;
	192	uprv_strcpy(filename+2,srcDir);
	193	}else{
	194	uprv_strcpy(filename, srcDir);
	195	}
	196
	197	basename=filename+uprv_strlen(filename);
	198	if(basename>filename && *(basename-1)!=U_FILE_SEP_CHAR) {
	199	*basename++=U_FILE_SEP_CHAR;
	200	}
	201
	202	/* initialize */
	203	init();
	204
	205	/* process the file */
	206	uprv_strcpy(basename,inputFileName);
	207	parseMappings(filename,FALSE, &errorCode);
	208	if(U_FAILURE(errorCode)) {
	209	fprintf(stderr, "Could not open file %s for reading. Error: %s \n", filename, u_errorName(errorCode));
	210	return errorCode;
	211	}
	212
	213	if(options[NORMALIZE].doesOccur){
	214	/* set up directory for NormalizationCorrections.txt */
	215	uprv_strcpy(filename,icuUniDataDir);
	216	basename=filename+uprv_strlen(filename);
	217	if(basename>filename && *(basename-1)!=U_FILE_SEP_CHAR) {
	218	*basename++=U_FILE_SEP_CHAR;
	219	}
	220
	221	*basename++=U_FILE_SEP_CHAR;
	222	uprv_strcpy(basename,NORM_CORRECTIONS_FILE_NAME);
	223
	224	parseNormalizationCorrections(filename,&errorCode);
	225	if(U_FAILURE(errorCode)){
	226	fprintf(stderr,"Could not open file %s for reading \n", filename);
	227	return errorCode;
	228	}
	229	sprepOptions \|= _SPREP_NORMALIZATION_ON;
	230	}
	231
	232	if(options[CHECK_BIDI].doesOccur){
	233	sprepOptions \|= _SPREP_CHECK_BIDI_ON;
	234	}
	235
	236	setOptions(sprepOptions);
	237
	238	/* process parsed data */
	239	if(U_SUCCESS(errorCode)) {
	240	/* write the data file */
73c04bcf	241	generateData(destDir, bundleName);
374ca955	242
73c04bcf	243	cleanUpData();
374ca955 A	244	}
	245
	246	uprv_free(filename);
	247
	248	#endif
	249
	250	return errorCode;
	251	}
	252
	253	#if !UCONFIG_NO_IDNA
	254
	255	static void U_CALLCONV
	256	normalizationCorrectionsLineFn(void *context,
	257	char *fields[][2], int32_t fieldCount,
	258	UErrorCode *pErrorCode) {
	259	uint32_t mapping[40];
	260	char end, s;
	261	uint32_t code;
	262	int32_t length;
	263	UVersionInfo version;
	264	UVersionInfo thisVersion;
	265
	266	/* get the character code, field 0 */
	267	code=(uint32_t)uprv_strtoul(fields[0][0], &end, 16);
	268	if(U_FAILURE(*pErrorCode)) {
	269	fprintf(stderr, "gensprep: error parsing NormalizationCorrections.txt mapping at %s\n", fields[0][0]);
	270	exit(*pErrorCode);
	271	}
	272	/* Original (erroneous) decomposition */
	273	s = fields[1][0];
	274
	275	/* parse the mapping string */
	276	length=u_parseCodePoints(s, mapping, sizeof(mapping)/4, pErrorCode);
	277
	278	/* ignore corrected decomposition */
	279
	280	u_versionFromString(version,fields[3][0] );
	281	u_versionFromString(thisVersion, "3.2.0");
	282
	283
	284
	285	if(U_FAILURE(*pErrorCode)) {
	286	fprintf(stderr, "gensprep error parsing NormalizationCorrections.txt of U+%04lx - %s\n",
	287	(long)code, u_errorName(*pErrorCode));
	288	exit(*pErrorCode);
	289	}
	290
	291	/* store the mapping */
	292	if( version[0] > thisVersion[0] \|\|
	293	((version[0]==thisVersion[0]) && (version[1] > thisVersion[1]))
	294	){
	295	storeMapping(code,mapping, length, USPREP_MAP, pErrorCode);
	296	}
	297	setUnicodeVersionNC(version);
	298	}
	299
	300	static void
	301	parseNormalizationCorrections(const char filename, UErrorCode pErrorCode) {
	302	char *fields[4][2];
	303
	304	if(pErrorCode==NULL \|\| U_FAILURE(*pErrorCode)) {
	305	return;
	306	}
	307
308	u_parseDelimitedFile(filename, ';', fields, 4, normalizationCorrectionsLineFn, NULL, pErrorCode);
309
310	/* fprintf(stdout,"Number of code points that have NormalizationCorrections mapping with length >1 : %i\n",len); */
311
312	if(U_FAILURE(pErrorCode) && ( pErrorCode!=U_FILE_ACCESS_ERROR)) {
313	fprintf(stderr, "gensprep error: u_parseDelimitedFile(\"%s\") failed - %s\n", filename, u_errorName(*pErrorCode));
314	exit(*pErrorCode);
315	}
316	}
317
318	static void U_CALLCONV
319	strprepProfileLineFn(void *context,
320	char *fields[][2], int32_t fieldCount,
321	UErrorCode *pErrorCode) {
322	uint32_t mapping[40];
323	char end, map;
324	uint32_t code;
325	int32_t length;
326	/UBool mapWithNorm = (UBool) context;/
327	const char* typeName;
328	uint32_t rangeStart=0,rangeEnd =0;
329	const char* filename = (const char*) context;
330
331	typeName = fields[2][0];
332	map = fields[1][0];
333
334	if(uprv_strstr(typeName, usprepTypeNames[USPREP_UNASSIGNED])!=NULL){
335
336	u_parseCodePointRange(fields[0][0], &rangeStart,&rangeEnd, pErrorCode);
337	if(U_FAILURE(*pErrorCode)){
338	fprintf(stderr, "Could not parse code point range. Error: %s\n",u_errorName(*pErrorCode));
339	return;
340	}
341
342	/* store the range */
343	storeRange(rangeStart,rangeEnd,USPREP_UNASSIGNED, pErrorCode);
344
345	}else if(uprv_strstr(typeName, usprepTypeNames[USPREP_PROHIBITED])!=NULL){
346
347	u_parseCodePointRange(fields[0][0], &rangeStart,&rangeEnd, pErrorCode);
348	if(U_FAILURE(*pErrorCode)){
349	fprintf(stderr, "Could not parse code point range. Error: %s\n",u_errorName(*pErrorCode));
350	return;
351	}
352
353	/* store the range */
354	storeRange(rangeStart,rangeEnd,USPREP_PROHIBITED, pErrorCode);
355
356	}else if(uprv_strstr(typeName, usprepTypeNames[USPREP_MAP])!=NULL){
357
358	/* get the character code, field 0 */
359	code=(uint32_t)uprv_strtoul(fields[0][0], &end, 16);
360	if(end<=fields[0][0] \|\| end!=fields[0][1]) {
361	fprintf(stderr, "gensprep: syntax error in field 0 at %s\n", fields[0][0]);
362	*pErrorCode=U_PARSE_ERROR;
363	exit(U_PARSE_ERROR);
364	}
365
366	/* parse the mapping string */
367	length=u_parseCodePoints(map, mapping, sizeof(mapping)/4, pErrorCode);
368
369	/* store the mapping */
370	storeMapping(code,mapping, length,USPREP_MAP, pErrorCode);
371
372	}else{
373	*pErrorCode = U_INVALID_FORMAT_ERROR;
374	}
375
376	if(U_FAILURE(*pErrorCode)) {
377	fprintf(stderr, "gensprep error parsing %s line %s at %s. Error: %s\n",filename,
378	fields[0][0],fields[2][0],u_errorName(*pErrorCode));
379	exit(*pErrorCode);
380	}
381
382	}
383
384	static void
385	parseMappings(const char filename, UBool reportError, UErrorCode pErrorCode) {
386	char *fields[3][2];
387
388	if(pErrorCode==NULL \|\| U_FAILURE(*pErrorCode)) {
389	return;
390	}
391
392	u_parseDelimitedFile(filename, ';', fields, 3, strprepProfileLineFn, (void*)filename, pErrorCode);
393
394	/fprintf(stdout,"Number of code points that have mappings with length >1 : %i\n",len);/
395
396	if(U_FAILURE(pErrorCode) && (reportError \|\| pErrorCode!=U_FILE_ACCESS_ERROR)) {
397	fprintf(stderr, "gensprep error: u_parseDelimitedFile(\"%s\") failed - %s\n", filename, u_errorName(*pErrorCode));
398	exit(*pErrorCode);
399	}
400	}
401
402
403	#endif /* #if !UCONFIG_NO_IDNA */
404
405	/*
406	* Hey, Emacs, please set the following:
407	*
408	* Local Variables:
409	* indent-tabs-mode: nil
410	* End:
411	*
412	*/