static int test_eof(lua_State *L, UFILE *ufile) {
UChar32 c = u_fgetcx(ufile);
u_fungetc(c, ufile);
lua_pushlstring(L, NULL, 0);
icu4lua_internrawustring(L, UFILE_UV_USTRING_META, UFILE_UV_USTRING_POOL);
return (c != EOF);
}
/* TODO: Is always skipping the prefix symbol as a positive sign a good idea in all locales? */
static int32_t
u_scanf_skip_leading_positive_sign(UFILE *input,
UNumberFormat *format,
UErrorCode *status)
{
UChar c;
int32_t count = 0;
UBool isNotEOF;
UChar plusSymbol[USCANF_SYMBOL_BUFFER_SIZE];
int32_t symbolLen;
UErrorCode localStatus = U_ZERO_ERROR;
if (U_SUCCESS(*status)) {
symbolLen = unum_getSymbol(format,
UNUM_PLUS_SIGN_SYMBOL,
plusSymbol,
UPRV_LENGTHOF(plusSymbol),
&localStatus);
if (U_SUCCESS(localStatus)) {
/* skip all leading ws in the input */
while( (isNotEOF = ufile_getch(input, &c)) && (count < symbolLen && c == plusSymbol[count]) )
{
count++;
}
/* put the final character back on the input */
if(isNotEOF) {
u_fungetc(c, input);
}
}
}
return count;
}
static int32_t
u_scanf_ustring_handler(UFILE *input,
u_scanf_spec_info *info,
ufmt_args *args,
const UChar *fmt,
int32_t *fmtConsumed,
int32_t *argConverted)
{
UChar *arg = (UChar*)(args[0].ptrValue);
UChar *alias = arg;
int32_t count;
int32_t skipped = 0;
UChar c;
UBool isNotEOF = FALSE;
/* skip all ws in the input */
if (info->fIsString) {
skipped = u_scanf_skip_leading_ws(input, info->fPadChar);
}
/* get the string one character at a time, truncating to the width */
count = 0;
while( (info->fWidth == -1 || count < info->fWidth)
&& (isNotEOF = ufile_getch(input, &c))
&& (!info->fIsString || (c != info->fPadChar && !u_isWhitespace(c))))
{
/* put the character from the input onto the target */
if (!info->fSkipArg) {
*alias++ = c;
}
/* increment the count */
++count;
}
/* put the final character we read back on the input */
if (!info->fSkipArg) {
if((info->fWidth == -1 || count < info->fWidth) && isNotEOF) {
u_fungetc(c, input);
}
/* add the terminator */
if (info->fIsString) {
*alias = 0x0000;
}
}
/* we converted 1 arg */
*argConverted = !info->fSkipArg;
return count + skipped;
}
static int32_t
u_scanf_skip_leading_ws(UFILE *input,
UChar pad)
{
UChar c;
int32_t count = 0;
UBool isNotEOF;
/* skip all leading ws in the input */
while( (isNotEOF = ufile_getch(input, &c)) && (c == pad || u_isWhitespace(c)) )
{
count++;
}
/* put the final character back on the input */
if(isNotEOF)
u_fungetc(c, input);
return count;
}
static int32_t
u_scanf_string_handler(UFILE *input,
u_scanf_spec_info *info,
ufmt_args *args,
const UChar *fmt,
int32_t *fmtConsumed,
int32_t *argConverted)
{
const UChar *source;
UConverter *conv;
char *arg = (char*)(args[0].ptrValue);
char *alias = arg;
char *limit;
UErrorCode status = U_ZERO_ERROR;
int32_t count;
int32_t skipped = 0;
UChar c;
UBool isNotEOF = FALSE;
/* skip all ws in the input */
if (info->fIsString) {
skipped = u_scanf_skip_leading_ws(input, info->fPadChar);
}
/* get the string one character at a time, truncating to the width */
count = 0;
/* open the default converter */
conv = u_getDefaultConverter(&status);
if(U_FAILURE(status))
return -1;
while( (info->fWidth == -1 || count < info->fWidth)
&& (isNotEOF = ufile_getch(input, &c))
&& (!info->fIsString || (c != info->fPadChar && !u_isWhitespace(c))))
{
if (!info->fSkipArg) {
/* put the character from the input onto the target */
source = &c;
/* Since we do this one character at a time, do it this way. */
if (info->fWidth > 0) {
limit = alias + info->fWidth - count;
}
else {
limit = alias + ucnv_getMaxCharSize(conv);
}
/* convert the character to the default codepage */
ucnv_fromUnicode(conv, &alias, limit, &source, source + 1,
NULL, TRUE, &status);
if(U_FAILURE(status)) {
/* clean up */
u_releaseDefaultConverter(conv);
return -1;
}
}
/* increment the count */
++count;
}
/* put the final character we read back on the input */
if (!info->fSkipArg) {
if ((info->fWidth == -1 || count < info->fWidth) && isNotEOF)
u_fungetc(c, input);
/* add the terminator */
if (info->fIsString) {
*alias = 0x00;
}
}
/* clean up */
u_releaseDefaultConverter(conv);
/* we converted 1 arg */
*argConverted = !info->fSkipArg;
return count + skipped;
}
static int32_t
u_scanf_scanset_handler(UFILE *input,
u_scanf_spec_info *info,
ufmt_args *args,
const UChar *fmt,
int32_t *fmtConsumed,
int32_t *argConverted)
{
USet *scanset;
UErrorCode status = U_ZERO_ERROR;
int32_t chLeft = INT32_MAX;
UChar32 c;
UChar *alias = (UChar*) (args[0].ptrValue);
UBool isNotEOF = FALSE;
UBool readCharacter = FALSE;
/* Create an empty set */
scanset = uset_open(0, -1);
/* Back up one to get the [ */
fmt--;
/* truncate to the width, if specified and alias the target */
if(info->fWidth >= 0) {
chLeft = info->fWidth;
}
/* parse the scanset from the fmt string */
*fmtConsumed = uset_applyPattern(scanset, fmt, -1, 0, &status);
/* verify that the parse was successful */
if (U_SUCCESS(status)) {
c=0;
/* grab characters one at a time and make sure they are in the scanset */
while(chLeft > 0) {
if ((isNotEOF = ufile_getch32(input, &c)) && uset_contains(scanset, c)) {
readCharacter = TRUE;
if (!info->fSkipArg) {
int32_t idx = 0;
UBool isError = FALSE;
U16_APPEND(alias, idx, chLeft, c, isError);
if (isError) {
break;
}
alias += idx;
}
chLeft -= (1 + U_IS_SUPPLEMENTARY(c));
}
else {
/* if the character's not in the scanset, break out */
break;
}
}
/* put the final character we read back on the input */
if(isNotEOF && chLeft > 0) {
u_fungetc(c, input);
}
}
uset_close(scanset);
/* if we didn't match at least 1 character, fail */
if(!readCharacter)
return -1;
/* otherwise, add the terminator */
else if (!info->fSkipArg) {
*alias = 0x00;
}
/* we converted 1 arg */
*argConverted = !info->fSkipArg;
return (info->fWidth >= 0 ? info->fWidth : INT32_MAX) - chLeft;
}
