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

/*
*******************************************************************************
*
*   Copyright (C) 2003-2004, 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

#ifdef WIN32
#   pragma warning(disable: 4100)
#endif

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_PACKAGE_NAME,
    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,
    PACKAGE_NAME,
    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-p or --package-name     prepend the output data file name with the package name specified\n"
        "\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 *packageName=NULL, *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[PACKAGE_NAME].value = NULL;
    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;
    packageName = options[PACKAGE_NAME].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, packageName, 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, packageName, 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	*
	4	* Copyright (C) 2003-2004, International Business Machines
	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
	43	#ifdef WIN32
	44	# pragma warning(disable: 4100)
	45	#endif
	46
	47	UBool beVerbose=FALSE, haveCopyright=TRUE;
	48
	49	#define NORM_CORRECTIONS_FILE_NAME "NormalizationCorrections.txt"
	50
	51	/* prototypes --------------------------------------------------------------- */
	52
	53	static void
	54	parseMappings(const char filename, UBool reportError, UErrorCode pErrorCode);
	55
	56	static void
	57	parseNormalizationCorrections(const char filename, UErrorCode pErrorCode);
	58
	59
	60	/* -------------------------------------------------------------------------- */
	61
	62	static UOption options[]={
	63	UOPTION_HELP_H,
	64	UOPTION_HELP_QUESTION_MARK,
65	UOPTION_VERBOSE,
66	UOPTION_COPYRIGHT,
67	UOPTION_DESTDIR,
68	UOPTION_SOURCEDIR,
69	UOPTION_ICUDATADIR,
70	UOPTION_PACKAGE_NAME,
71	UOPTION_BUNDLE_NAME,
72	{ "normalization", NULL, NULL, NULL, 'n', UOPT_REQUIRES_ARG, 0 },
73	{ "check-bidi", NULL, NULL, NULL, 'k', UOPT_NO_ARG, 0},
74	{ "unicode", NULL, NULL, NULL, 'u', UOPT_REQUIRES_ARG, 0 },
75	};
76
77	enum{
78	HELP,
79	HELP_QUESTION_MARK,
80	VERBOSE,
81	COPYRIGHT,
82	DESTDIR,
83	SOURCEDIR,
84	ICUDATADIR,
85	PACKAGE_NAME,
86	BUNDLE_NAME,
87	NORMALIZE,
88	CHECK_BIDI,
89	UNICODE_VERSION
90	};
91
92	static int printHelp(int argc, char* argv[]){
93	/*
94	* Broken into chucks because the C89 standard says the minimum
95	* required supported string length is 509 bytes.
96	*/
97	fprintf(stderr,
98	"Usage: %s [-options] [file_name]\n"
99	"\n"
100	"Read the files specified and\n"
101	"create a binary file [package-name]_[bundle-name]." DATA_TYPE " with the StringPrep profile data\n"
102	"\n",
103	argv[0]);
104	fprintf(stderr,
105	"Options:\n"
106	"\t-h or -? or --help print this usage text\n"
107	"\t-v or --verbose verbose output\n"
108	"\t-c or --copyright include a copyright notice\n");
109	fprintf(stderr,
110	"\t-d or --destdir destination directory, followed by the path\n"
111	"\t-s or --sourcedir source directory of ICU data, followed by the path\n"
112	"\t-b or --bundle-name generate the ouput data file with the name specified\n"
113	"\t-i or --icudatadir directory for locating any needed intermediate data files,\n"
114	"\t followed by path, defaults to %s\n",
115	u_getDataDirectory());
116	fprintf(stderr,
117	"\t-p or --package-name prepend the output data file name with the package name specified\n"
118	"\t-n or --normalize turn on the option for normalization and include mappings\n"
119	"\t from NormalizationCorrections.txt from the given path,\n"
120	"\t e.g: /test/icu/source/data/unidata\n"
121	"\t-k or --check-bidi turn on the option for checking for BiDi in the profile\n"
122	"\t-u or --unicode version of Unicode to be used with this profile followed by the version\n"
123	);
124	return argc<0 ? U_ILLEGAL_ARGUMENT_ERROR : U_ZERO_ERROR;
125	}
126
127
128	extern int
129	main(int argc, char* argv[]) {
130	#if !UCONFIG_NO_IDNA
131	char* filename = NULL;
132	#endif
133	const char srcDir=NULL, destDir=NULL, *icuUniDataDir=NULL;
134	const char packageName=NULL, bundleName=NULL, *inputFileName = NULL;
135	char *basename=NULL;
136	int32_t sprepOptions = 0;
137
138	UErrorCode errorCode=U_ZERO_ERROR;
139
140	U_MAIN_INIT_ARGS(argc, argv);
141
142	/* preset then read command line options */
143	options[DESTDIR].value=u_getDataDirectory();
144	options[SOURCEDIR].value="";
145	options[UNICODE_VERSION].value="0"; /* don't assume the unicode version */
146	options[BUNDLE_NAME].value = DATA_NAME;
147	options[PACKAGE_NAME].value = NULL;
148	options[NORMALIZE].value = "";
149
150	argc=u_parseArgs(argc, argv, sizeof(options)/sizeof(options[0]), options);
151
152	/* error handling, printing usage message */
153	if(argc<0) {
154	fprintf(stderr,
155	"error in command line argument \"%s\"\n",
156	argv[-argc]);
157	}
158	if(argc<0 \|\| options[HELP].doesOccur \|\| options[HELP_QUESTION_MARK].doesOccur) {
159	return printHelp(argc, argv);
160
161	}
162
163	/* get the options values */
164	beVerbose=options[VERBOSE].doesOccur;
165	haveCopyright=options[COPYRIGHT].doesOccur;
166	srcDir=options[SOURCEDIR].value;
167	destDir=options[DESTDIR].value;
168	packageName = options[PACKAGE_NAME].value;
169	bundleName = options[BUNDLE_NAME].value;
170	icuUniDataDir = options[NORMALIZE].value;
171
172	if(argc<2) {
173	/* print the help message */
174	return printHelp(argc, argv);
175	} else {
176	inputFileName = argv[1];
177	}
178	if(!options[UNICODE_VERSION].doesOccur){
179	return printHelp(argc, argv);
180	}
181	if(options[ICUDATADIR].doesOccur) {
182	u_setDataDirectory(options[ICUDATADIR].value);
183	}
184	#if UCONFIG_NO_IDNA
185
186	fprintf(stderr,
187	"gensprep writes dummy " U_ICUDATA_NAME "_" DATA_NAME "." DATA_TYPE
188	" because UCONFIG_NO_IDNA is set, \n"
189	"see icu/source/common/unicode/uconfig.h\n");
190	generateData(destDir, packageName, bundleName);
191
192	#else
193
194	setUnicodeVersion(options[UNICODE_VERSION].value);
195	filename = (char* ) uprv_malloc(uprv_strlen(srcDir) + 300); /* hopefully this should be enough */
196
197	/* prepare the filename beginning with the source dir */
198	if(uprv_strchr(srcDir,U_FILE_SEP_CHAR) == NULL && uprv_strchr(srcDir,U_FILE_ALT_SEP_CHAR) == NULL){
199	filename[0] = '.';
200	filename[1] = U_FILE_SEP_CHAR;
201	uprv_strcpy(filename+2,srcDir);
202	}else{
203	uprv_strcpy(filename, srcDir);
204	}
205
206	basename=filename+uprv_strlen(filename);
207	if(basename>filename && *(basename-1)!=U_FILE_SEP_CHAR) {
208	*basename++=U_FILE_SEP_CHAR;
209	}
210
211	/* initialize */
212	init();
213
214	/* process the file */
215	uprv_strcpy(basename,inputFileName);
216	parseMappings(filename,FALSE, &errorCode);
217	if(U_FAILURE(errorCode)) {
218	fprintf(stderr, "Could not open file %s for reading. Error: %s \n", filename, u_errorName(errorCode));
219	return errorCode;
220	}
221
222	if(options[NORMALIZE].doesOccur){
223	/* set up directory for NormalizationCorrections.txt */
224	uprv_strcpy(filename,icuUniDataDir);
225	basename=filename+uprv_strlen(filename);
226	if(basename>filename && *(basename-1)!=U_FILE_SEP_CHAR) {
227	*basename++=U_FILE_SEP_CHAR;
228	}
229
230	*basename++=U_FILE_SEP_CHAR;
231	uprv_strcpy(basename,NORM_CORRECTIONS_FILE_NAME);
232
233	parseNormalizationCorrections(filename,&errorCode);
234	if(U_FAILURE(errorCode)){
235	fprintf(stderr,"Could not open file %s for reading \n", filename);
236	return errorCode;
237	}
238	sprepOptions \|= _SPREP_NORMALIZATION_ON;
239	}
240
241	if(options[CHECK_BIDI].doesOccur){
242	sprepOptions \|= _SPREP_CHECK_BIDI_ON;
243	}
244
245	setOptions(sprepOptions);
246
247	/* process parsed data */
248	if(U_SUCCESS(errorCode)) {
249	/* write the data file */
250	generateData(destDir, packageName, bundleName);
251
252	cleanUpData();
253	}
254
255	uprv_free(filename);
256
257	#endif
258
259	return errorCode;
260	}
261
262	#if !UCONFIG_NO_IDNA
263
264	static void U_CALLCONV
265	normalizationCorrectionsLineFn(void *context,
266	char *fields[][2], int32_t fieldCount,
267	UErrorCode *pErrorCode) {
268	uint32_t mapping[40];
269	char end, s;
270	uint32_t code;
271	int32_t length;
272	UVersionInfo version;
273	UVersionInfo thisVersion;
274
275	/* get the character code, field 0 */
276	code=(uint32_t)uprv_strtoul(fields[0][0], &end, 16);
277	if(U_FAILURE(*pErrorCode)) {
278	fprintf(stderr, "gensprep: error parsing NormalizationCorrections.txt mapping at %s\n", fields[0][0]);
279	exit(*pErrorCode);
280	}
281	/* Original (erroneous) decomposition */
282	s = fields[1][0];
283
284	/* parse the mapping string */
285	length=u_parseCodePoints(s, mapping, sizeof(mapping)/4, pErrorCode);
286
287	/* ignore corrected decomposition */
288
289	u_versionFromString(version,fields[3][0] );
290	u_versionFromString(thisVersion, "3.2.0");
291
292
293
294	if(U_FAILURE(*pErrorCode)) {
295	fprintf(stderr, "gensprep error parsing NormalizationCorrections.txt of U+%04lx - %s\n",
296	(long)code, u_errorName(*pErrorCode));
297	exit(*pErrorCode);
298	}
299
300	/* store the mapping */
301	if( version[0] > thisVersion[0] \|\|
302	((version[0]==thisVersion[0]) && (version[1] > thisVersion[1]))
303	){
304	storeMapping(code,mapping, length, USPREP_MAP, pErrorCode);
305	}
306	setUnicodeVersionNC(version);
307	}
308
309	static void
310	parseNormalizationCorrections(const char filename, UErrorCode pErrorCode) {
311	char *fields[4][2];
312
313	if(pErrorCode==NULL \|\| U_FAILURE(*pErrorCode)) {
314	return;
315	}
316
317	u_parseDelimitedFile(filename, ';', fields, 4, normalizationCorrectionsLineFn, NULL, pErrorCode);
318
319	/* fprintf(stdout,"Number of code points that have NormalizationCorrections mapping with length >1 : %i\n",len); */
320
321	if(U_FAILURE(pErrorCode) && ( pErrorCode!=U_FILE_ACCESS_ERROR)) {
322	fprintf(stderr, "gensprep error: u_parseDelimitedFile(\"%s\") failed - %s\n", filename, u_errorName(*pErrorCode));
323	exit(*pErrorCode);
324	}
325	}
326
327	static void U_CALLCONV
328	strprepProfileLineFn(void *context,
329	char *fields[][2], int32_t fieldCount,
330	UErrorCode *pErrorCode) {
331	uint32_t mapping[40];
332	char end, map;
333	uint32_t code;
334	int32_t length;
335	/UBool mapWithNorm = (UBool) context;/
336	const char* typeName;
337	uint32_t rangeStart=0,rangeEnd =0;
338	const char* filename = (const char*) context;
339
340	typeName = fields[2][0];
341	map = fields[1][0];
342
343	if(uprv_strstr(typeName, usprepTypeNames[USPREP_UNASSIGNED])!=NULL){
344
345	u_parseCodePointRange(fields[0][0], &rangeStart,&rangeEnd, pErrorCode);
346	if(U_FAILURE(*pErrorCode)){
347	fprintf(stderr, "Could not parse code point range. Error: %s\n",u_errorName(*pErrorCode));
348	return;
349	}
350
351	/* store the range */
352	storeRange(rangeStart,rangeEnd,USPREP_UNASSIGNED, pErrorCode);
353
354	}else if(uprv_strstr(typeName, usprepTypeNames[USPREP_PROHIBITED])!=NULL){
355
356	u_parseCodePointRange(fields[0][0], &rangeStart,&rangeEnd, pErrorCode);
357	if(U_FAILURE(*pErrorCode)){
358	fprintf(stderr, "Could not parse code point range. Error: %s\n",u_errorName(*pErrorCode));
359	return;
360	}
361
362	/* store the range */
363	storeRange(rangeStart,rangeEnd,USPREP_PROHIBITED, pErrorCode);
364
365	}else if(uprv_strstr(typeName, usprepTypeNames[USPREP_MAP])!=NULL){
366
367	/* get the character code, field 0 */
368	code=(uint32_t)uprv_strtoul(fields[0][0], &end, 16);
369	if(end<=fields[0][0] \|\| end!=fields[0][1]) {
370	fprintf(stderr, "gensprep: syntax error in field 0 at %s\n", fields[0][0]);
371	*pErrorCode=U_PARSE_ERROR;
372	exit(U_PARSE_ERROR);
373	}
374
375	/* parse the mapping string */
376	length=u_parseCodePoints(map, mapping, sizeof(mapping)/4, pErrorCode);
377
378	/* store the mapping */
379	storeMapping(code,mapping, length,USPREP_MAP, pErrorCode);
380
381	}else{
382	*pErrorCode = U_INVALID_FORMAT_ERROR;
383	}
384
385	if(U_FAILURE(*pErrorCode)) {
386	fprintf(stderr, "gensprep error parsing %s line %s at %s. Error: %s\n",filename,
387	fields[0][0],fields[2][0],u_errorName(*pErrorCode));
388	exit(*pErrorCode);
389	}
390
391	}
392
393	static void
394	parseMappings(const char filename, UBool reportError, UErrorCode pErrorCode) {
395	char *fields[3][2];
396
397	if(pErrorCode==NULL \|\| U_FAILURE(*pErrorCode)) {
398	return;
399	}
400
401	u_parseDelimitedFile(filename, ';', fields, 3, strprepProfileLineFn, (void*)filename, pErrorCode);
402
403	/fprintf(stdout,"Number of code points that have mappings with length >1 : %i\n",len);/
404
405	if(U_FAILURE(pErrorCode) && (reportError \|\| pErrorCode!=U_FILE_ACCESS_ERROR)) {
406	fprintf(stderr, "gensprep error: u_parseDelimitedFile(\"%s\") failed - %s\n", filename, u_errorName(*pErrorCode));
407	exit(*pErrorCode);
408	}
409	}
410
411
412	#endif /* #if !UCONFIG_NO_IDNA */
413
414	/*
415	* Hey, Emacs, please set the following:
416	*
417	* Local Variables:
418	* indent-tabs-mode: nil
419	* End:
420	*
421	*/