2 *******************************************************************************
4 * Copyright (C) 2000-2012, International Business Machines
5 * Corporation and others. All Rights Reserved.
7 *******************************************************************************
10 * tab size: 8 (not used)
13 * created on: 2000apr18
14 * created by: Markus W. Scherer
16 * This file provides a parser for files that are delimited by one single
17 * character like ';' or TAB. Example: the Unicode Character Properties files
18 * like UnicodeData.txt are semicolon-delimited.
21 #include "unicode/utypes.h"
22 #include "unicode/uchar.h"
23 #include "unicode/ustring.h"
24 #include "unicode/utf16.h"
32 U_CAPI const char * U_EXPORT2
33 u_skipWhitespace(const char *s) {
34 while(U_IS_INV_WHITESPACE(*s)) {
40 U_CAPI char * U_EXPORT2
42 char *end=uprv_strchr(s, 0);
43 while(s<end && U_IS_INV_WHITESPACE(*(end-1))) {
50 * If the string starts with # @missing: then return the pointer to the
51 * following non-whitespace character.
52 * Otherwise return the original pointer.
53 * Unicode 5.0 adds such lines in some data files to document
54 * default property values.
55 * Poor man's regex for variable amounts of white space.
58 getMissingLimit(const char *s) {
61 *(s=u_skipWhitespace(s))=='#' &&
62 *(s=u_skipWhitespace(s+1))=='@' &&
63 0==strncmp((s=u_skipWhitespace(s+1)), "missing", 7) &&
64 *(s=u_skipWhitespace(s+7))==':'
66 return u_skipWhitespace(s+1);
73 u_parseDelimitedFile(const char *filename, char delimiter,
74 char *fields[][2], int32_t fieldCount,
75 UParseLineFn *lineFn, void *context,
76 UErrorCode *pErrorCode) {
82 if(U_FAILURE(*pErrorCode)) {
86 if(fields==NULL || lineFn==NULL || fieldCount<=0) {
87 *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
91 if(filename==NULL || *filename==0 || (*filename=='-' && filename[1]==0)) {
93 file=T_FileStream_stdin();
95 file=T_FileStream_open(filename, "r");
98 *pErrorCode=U_FILE_ACCESS_ERROR;
102 while(T_FileStream_readLine(file, line, sizeof(line))!=NULL) {
103 /* remove trailing newline characters */
104 length=(int32_t)(u_rtrim(line)-line);
107 * detect a line with # @missing:
108 * start parsing after that, or else from the beginning of the line
109 * set the default warning for @missing lines
111 start=(char *)getMissingLimit(line);
113 *pErrorCode=U_ZERO_ERROR;
115 *pErrorCode=U_USING_DEFAULT_WARNING;
118 /* skip this line if it is empty or a comment */
119 if(*start==0 || *start=='#') {
123 /* remove in-line comments */
124 limit=uprv_strchr(start, '#');
126 /* get white space before the pound sign */
127 while(limit>start && U_IS_INV_WHITESPACE(*(limit-1))) {
131 /* truncate the line */
135 /* skip lines with only whitespace */
136 if(u_skipWhitespace(start)[0]==0) {
140 /* for each field, call the corresponding field function */
141 for(i=0; i<fieldCount; ++i) {
142 /* set the limit pointer of this field */
144 while(*limit!=delimiter && *limit!=0) {
148 /* set the field start and limit in the fields array */
152 /* set start to the beginning of the next field, if any */
156 } else if(i+1<fieldCount) {
157 *pErrorCode=U_PARSE_ERROR;
164 /* error in a field function? */
165 if(U_FAILURE(*pErrorCode)) {
169 /* call the field function */
170 lineFn(context, fields, fieldCount, pErrorCode);
171 if(U_FAILURE(*pErrorCode)) {
177 T_FileStream_close(file);
182 * parse a list of code points
183 * store them as a UTF-32 string in dest[destCapacity]
184 * return the number of code points
186 U_CAPI int32_t U_EXPORT2
187 u_parseCodePoints(const char *s,
188 uint32_t *dest, int32_t destCapacity,
189 UErrorCode *pErrorCode) {
194 if(U_FAILURE(*pErrorCode)) {
197 if(s==NULL || destCapacity<0 || (destCapacity>0 && dest==NULL)) {
198 *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
204 s=u_skipWhitespace(s);
205 if(*s==';' || *s==0) {
209 /* read one code point */
210 value=(uint32_t)uprv_strtoul(s, &end, 16);
211 if(end<=s || (!U_IS_INV_WHITESPACE(*end) && *end!=';' && *end!=0) || value>=0x110000) {
212 *pErrorCode=U_PARSE_ERROR;
216 /* append it to the destination array */
217 if(count<destCapacity) {
220 *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
223 /* go to the following characters */
229 * parse a list of code points
230 * store them as a string in dest[destCapacity]
231 * set the first code point in *pFirst
232 * @return The length of the string in numbers of UChars.
234 U_CAPI int32_t U_EXPORT2
235 u_parseString(const char *s,
236 UChar *dest, int32_t destCapacity,
238 UErrorCode *pErrorCode) {
243 if(U_FAILURE(*pErrorCode)) {
246 if(s==NULL || destCapacity<0 || (destCapacity>0 && dest==NULL)) {
247 *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
257 s=u_skipWhitespace(s);
258 if(*s==';' || *s==0) {
259 if(destLength<destCapacity) {
261 } else if(destLength==destCapacity) {
262 *pErrorCode=U_STRING_NOT_TERMINATED_WARNING;
264 *pErrorCode=U_BUFFER_OVERFLOW_ERROR;
269 /* read one code point */
270 value=(uint32_t)uprv_strtoul(s, &end, 16);
271 if(end<=s || (!U_IS_INV_WHITESPACE(*end) && *end!=';' && *end!=0) || value>=0x110000) {
272 *pErrorCode=U_PARSE_ERROR;
276 /* store the first code point */
282 /* append it to the destination array */
283 if((destLength+U16_LENGTH(value))<=destCapacity) {
284 U16_APPEND_UNSAFE(dest, destLength, value);
286 destLength+=U16_LENGTH(value);
289 /* go to the following characters */
294 /* read a range like start or start..end */
295 U_CAPI int32_t U_EXPORT2
296 u_parseCodePointRangeAnyTerminator(const char *s,
297 uint32_t *pStart, uint32_t *pEnd,
298 const char **terminator,
299 UErrorCode *pErrorCode) {
303 if(U_FAILURE(*pErrorCode)) {
306 if(s==NULL || pStart==NULL || pEnd==NULL) {
307 *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
311 /* read the start code point */
312 s=u_skipWhitespace(s);
313 value=(uint32_t)uprv_strtoul(s, &end, 16);
314 if(end<=s || value>=0x110000) {
315 *pErrorCode=U_PARSE_ERROR;
320 /* is there a "..end"? */
321 s=u_skipWhitespace(end);
322 if(*s!='.' || s[1]!='.') {
326 s=u_skipWhitespace(s+2);
328 /* read the end code point */
329 value=(uint32_t)uprv_strtoul(s, &end, 16);
330 if(end<=s || value>=0x110000) {
331 *pErrorCode=U_PARSE_ERROR;
336 /* is this a valid range? */
338 *pErrorCode=U_PARSE_ERROR;
343 return value-*pStart+1;
346 U_CAPI int32_t U_EXPORT2
347 u_parseCodePointRange(const char *s,
348 uint32_t *pStart, uint32_t *pEnd,
349 UErrorCode *pErrorCode) {
350 const char *terminator;
352 u_parseCodePointRangeAnyTerminator(s, pStart, pEnd, &terminator, pErrorCode);
353 if(U_SUCCESS(*pErrorCode)) {
354 terminator=u_skipWhitespace(terminator);
355 if(*terminator!=';' && *terminator!=0) {
356 *pErrorCode=U_PARSE_ERROR;
363 U_CAPI int32_t U_EXPORT2
364 u_parseUTF8(const char *source, int32_t sLen, char *dest, int32_t destCapacity, UErrorCode *status) {
365 const char *read = source;
367 unsigned int value = 0;
369 sLen = (int32_t)strlen(source);
372 while(read < source+sLen) {
373 sscanf(read, "%2x", &value);
374 if(i < destCapacity) {
375 dest[i] = (char)value;
380 return u_terminateChars(dest, destCapacity, i, status);