bool
ClParserParser::
isValidAssign()
{
parse_.skipSpace();
if (parse_.eof())
return false;
if (! parse_.isAlpha() && ! parse_.isChar('_'))
return false;
skipIdentifier();
parse_.skipSpace();
while (parse_.isChar('[')) {
parse_.skipChar();
if (! skipArray())
return false;
parse_.skipSpace();
}
while (parse_.isChar('.')) {
parse_.skipChar();
parse_.skipSpace();
if (! parse_.isAlpha() && ! parse_.isChar('_'))
return false;
skipIdentifier();
while (parse_.isChar('[')) {
parse_.skipChar();
if (! skipArray())
return false;
parse_.skipSpace();
}
}
if (! parse_.eof())
return false;
return true;
}
const char *Parser_Perl::findDefinitionOrClass (const char *cp)
{
while (*cp)
{
cp = skipEverything (cp);
if (!strncmp(cp, "sub", 3) || !strncmp(cp, "package", 7))
{
return cp;
}
cp = skipIdentifier (cp);
}
return NULL;
}
bool
ClParserParser::
skipExpression(bool *assign)
{
bool expression = false;
error_code_ = 0;
parse_.skipSpace();
while (! parse_.eof()) {
/* <expression> := <unary_operator> <expression> */
if (! expression && strchr(parser_unary_operator_chars, parse_.getCharAt()) != nullptr) {
skipUnaryOperator();
if (! skipExpression(assign))
return false;
if (*assign) {
error_code_ = int(ClErr::INVALID_LVALUE_FOR_ASSIGNMENT);
return false;
}
expression = true;
}
/* <expression> := '(' <expression> ')' */
else if (! expression && parse_.isChar('(')) {
bool assign1 = false;
parse_.skipChar();
if (! skipExpression(&assign1) && error_code_ != int(ClErr::INVALID_CHARACTER))
return false;
error_code_ = 0;
parse_.skipSpace();
if (! parse_.isChar(')')) {
error_code_ = int(ClErr::MISSING_CLOSE_ROUND_BRACKET);
return false;
}
parse_.skipChar();
expression = true;
}
/* <expression> := <numeric_value> */
else if (! expression &&
(parse_.isDigit() || parse_.isChar('.'))) {
if (! skipNumericValue())
return false;
expression = true;
}
/* <expression> := <string_value> */
else if (! expression && parse_.isOneOf(parser_string_chars)) {
if (! skipStringValue())
return false;
expression = true;
}
/* <expression> := <array_value> */
else if (! expression && parse_.isChar('[')) {
parse_.skipChar();
if (! skipArray())
return false;
expression = true;
}
/* <expression> := <dict_value> */
else if (! expression && parse_.isString("{{")) {
parse_.skipChars(2);
if (! skipDictionary())
return false;
expression = true;
}
/* <expression> := <list_value> */
else if (! expression && parse_.isChar('{')) {
parse_.skipChar();
if (! skipList())
return false;
expression = true;
}
/* <expression> := <expression> <binary_operator> <expression> */
else if (expression && parse_.isOneOf(parser_binary_operator_chars)) {
bool assign1 = false;
skipBinaryOperator(&assign1);
if (error_code_ != 0)
return false;
if (! skipExpression(assign))
return false;
if (! assign1 && *assign) {
error_code_ = int(ClErr::INVALID_LVALUE_FOR_ASSIGNMENT);
return false;
}
if (assign1)
*assign = true;
expression = true;
}
/* <expression> := <expression> '?' <expression> ':' <expression> */
else if (expression && parse_.isChar('?')) {
bool assign1 = false;
bool assign2 = false;
parse_.skipChar();
if (! skipExpression(&assign1) && error_code_ != int(ClErr::INVALID_CHARACTER))
return false;
error_code_ = 0;
parse_.skipSpace();
if (! parse_.isChar(':')) {
error_code_ = int(ClErr::MISSING_COLON_OPERATOR);
return false;
}
parse_.skipChar();
if (! skipExpression(&assign2))
return false;
expression = true;
*assign = assign1 | assign2;
}
/* <expression> := <expression> '[' <expression> ',' .... ']' */
else if (expression && parse_.isChar('[')) {
parse_.skipChar();
if (! skipArray())
return false;
expression = true;
}
/* <expression> := <expression> ',' <expression> */
else if (expression && parse_.isChar(',')) {
bool assign1 = false;
ClParserStackMgrInst->push(CL_PARSER_OP_COMMA);
ClParserStackMgrInst->toNext();
parse_.skipChar();
if (! skipExpression(&assign1))
return false;
expression = true;
}
/* <expression> := <variable> */
/* <expression> := <variable> '.' <variable> '.' <...> */
/* <expression> := <variable> '[' <expression> ',' .... ']' */
/* <expression> := <function> '(' <expression> ',' .... ')' */
/* <expression> := <type> '(' <expression> ',' .... ')' */
else if (! expression && (parse_.isAlpha() || parse_.isChar('_'))) {
skipIdentifier();
if (error_code_ != 0)
return false;
parse_.skipSpace();
if (parse_.isChar('(')) {
parse_.skipChar();
if (! skipArgList())
return false;
expression = true;
}
else {
if (parse_.isChar('[')) {
parse_.skipChar();
if (! skipArray())
return false;
parse_.skipSpace();
}
while (parse_.isChar('.')) {
parse_.skipChar();
parse_.skipSpace();
skipIdentifier();
if (error_code_ != 0)
return false;
if (parse_.isChar('[')) {
parse_.skipChar();
if (! skipArray())
return false;
parse_.skipSpace();
}
}
expression = true;
}
}
else {
error_code_ = int(ClErr::INVALID_CHARACTER);
return false;
}
parse_.skipSpace();
}
if (! expression) {
error_code_ = int(ClErr::NULL_EXPRESSION);
return false;
}
return true;
}
int32_t
MessagePattern::parsePluralOrSelectStyle(UMessagePatternArgType argType,
int32_t index, int32_t nestingLevel,
UParseError *parseError, UErrorCode &errorCode) {
if(U_FAILURE(errorCode)) {
return 0;
}
int32_t start=index;
UBool isEmpty=TRUE;
UBool hasOther=FALSE;
for(;;) {
// First, collect the selector looking for a small set of terminators.
// It would be a little faster to consider the syntax of each possible
// token right here, but that makes the code too complicated.
index=skipWhiteSpace(index);
UBool eos=index==msg.length();
if(eos || msg.charAt(index)==u_rightCurlyBrace) {
if(eos==inMessageFormatPattern(nestingLevel)) {
setParseError(parseError, start); // Bad plural/select pattern syntax.
errorCode=U_PATTERN_SYNTAX_ERROR;
return 0;
}
if(!hasOther) {
setParseError(parseError, 0); // Missing 'other' keyword in plural/select pattern.
errorCode=U_DEFAULT_KEYWORD_MISSING;
return 0;
}
return index;
}
int32_t selectorIndex=index;
if(UMSGPAT_ARG_TYPE_HAS_PLURAL_STYLE(argType) && msg.charAt(selectorIndex)==u_equal) {
// explicit-value plural selector: =double
index=skipDouble(index+1);
int32_t length=index-selectorIndex;
if(length==1) {
setParseError(parseError, start); // Bad plural/select pattern syntax.
errorCode=U_PATTERN_SYNTAX_ERROR;
return 0;
}
if(length>Part::MAX_LENGTH) {
setParseError(parseError, selectorIndex); // Argument selector too long.
errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
return 0;
}
addPart(UMSGPAT_PART_TYPE_ARG_SELECTOR, selectorIndex, length, 0, errorCode);
parseDouble(selectorIndex+1, index, FALSE,
parseError, errorCode); // adds ARG_INT or ARG_DOUBLE
} else {
index=skipIdentifier(index);
int32_t length=index-selectorIndex;
if(length==0) {
setParseError(parseError, start); // Bad plural/select pattern syntax.
errorCode=U_PATTERN_SYNTAX_ERROR;
return 0;
}
// Note: The ':' in "offset:" is just beyond the skipIdentifier() range.
if( UMSGPAT_ARG_TYPE_HAS_PLURAL_STYLE(argType) && length==6 && index<msg.length() &&
0==msg.compare(selectorIndex, 7, kOffsetColon, 0, 7)
) {
// plural offset, not a selector
if(!isEmpty) {
// Plural argument 'offset:' (if present) must precede key-message pairs.
setParseError(parseError, start);
errorCode=U_PATTERN_SYNTAX_ERROR;
return 0;
}
// allow whitespace between offset: and its value
int32_t valueIndex=skipWhiteSpace(index+1); // The ':' is at index.
index=skipDouble(valueIndex);
if(index==valueIndex) {
setParseError(parseError, start); // Missing value for plural 'offset:'.
errorCode=U_PATTERN_SYNTAX_ERROR;
return 0;
}
if((index-valueIndex)>Part::MAX_LENGTH) {
setParseError(parseError, valueIndex); // Plural offset value too long.
errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
return 0;
}
parseDouble(valueIndex, index, FALSE,
parseError, errorCode); // adds ARG_INT or ARG_DOUBLE
if(U_FAILURE(errorCode)) {
return 0;
}
isEmpty=FALSE;
continue; // no message fragment after the offset
} else {
// normal selector word
if(length>Part::MAX_LENGTH) {
setParseError(parseError, selectorIndex); // Argument selector too long.
errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
return 0;
}
addPart(UMSGPAT_PART_TYPE_ARG_SELECTOR, selectorIndex, length, 0, errorCode);
if(0==msg.compare(selectorIndex, length, kOther, 0, 5)) {
hasOther=TRUE;
}
}
}
if(U_FAILURE(errorCode)) {
return 0;
}
// parse the message fragment following the selector
index=skipWhiteSpace(index);
if(index==msg.length() || msg.charAt(index)!=u_leftCurlyBrace) {
setParseError(parseError, selectorIndex); // No message fragment after plural/select selector.
errorCode=U_PATTERN_SYNTAX_ERROR;
return 0;
}
index=parseMessage(index, 1, nestingLevel+1, argType, parseError, errorCode);
if(U_FAILURE(errorCode)) {
return 0;
}
isEmpty=FALSE;
}
}
int32_t
MessagePattern::parseArg(int32_t index, int32_t argStartLength, int32_t nestingLevel,
UParseError *parseError, UErrorCode &errorCode) {
int32_t argStart=partsLength;
UMessagePatternArgType argType=UMSGPAT_ARG_TYPE_NONE;
addPart(UMSGPAT_PART_TYPE_ARG_START, index, argStartLength, argType, errorCode);
if(U_FAILURE(errorCode)) {
return 0;
}
int32_t nameIndex=index=skipWhiteSpace(index+argStartLength);
if(index==msg.length()) {
setParseError(parseError, 0); // Unmatched '{' braces in message.
errorCode=U_UNMATCHED_BRACES;
return 0;
}
// parse argument name or number
index=skipIdentifier(index);
int32_t number=parseArgNumber(nameIndex, index);
if(number>=0) {
int32_t length=index-nameIndex;
if(length>Part::MAX_LENGTH || number>Part::MAX_VALUE) {
setParseError(parseError, nameIndex); // Argument number too large.
errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
return 0;
}
hasArgNumbers=TRUE;
addPart(UMSGPAT_PART_TYPE_ARG_NUMBER, nameIndex, length, number, errorCode);
} else if(number==UMSGPAT_ARG_NAME_NOT_NUMBER) {
int32_t length=index-nameIndex;
if(length>Part::MAX_LENGTH) {
setParseError(parseError, nameIndex); // Argument name too long.
errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
return 0;
}
hasArgNames=TRUE;
addPart(UMSGPAT_PART_TYPE_ARG_NAME, nameIndex, length, 0, errorCode);
} else { // number<-1 (ARG_NAME_NOT_VALID)
setParseError(parseError, nameIndex); // Bad argument syntax.
errorCode=U_PATTERN_SYNTAX_ERROR;
return 0;
}
index=skipWhiteSpace(index);
if(index==msg.length()) {
setParseError(parseError, 0); // Unmatched '{' braces in message.
errorCode=U_UNMATCHED_BRACES;
return 0;
}
UChar c=msg.charAt(index);
if(c==u_rightCurlyBrace) {
// all done
} else if(c!=u_comma) {
setParseError(parseError, nameIndex); // Bad argument syntax.
errorCode=U_PATTERN_SYNTAX_ERROR;
return 0;
} else /* ',' */ {
// parse argument type: case-sensitive a-zA-Z
int32_t typeIndex=index=skipWhiteSpace(index+1);
while(index<msg.length() && isArgTypeChar(msg.charAt(index))) {
++index;
}
int32_t length=index-typeIndex;
index=skipWhiteSpace(index);
if(index==msg.length()) {
setParseError(parseError, 0); // Unmatched '{' braces in message.
errorCode=U_UNMATCHED_BRACES;
return 0;
}
if(length==0 || ((c=msg.charAt(index))!=u_comma && c!=u_rightCurlyBrace)) {
setParseError(parseError, nameIndex); // Bad argument syntax.
errorCode=U_PATTERN_SYNTAX_ERROR;
return 0;
}
if(length>Part::MAX_LENGTH) {
setParseError(parseError, nameIndex); // Argument type name too long.
errorCode=U_INDEX_OUTOFBOUNDS_ERROR;
return 0;
}
argType=UMSGPAT_ARG_TYPE_SIMPLE;
if(length==6) {
// case-insensitive comparisons for complex-type names
if(isChoice(typeIndex)) {
argType=UMSGPAT_ARG_TYPE_CHOICE;
} else if(isPlural(typeIndex)) {
argType=UMSGPAT_ARG_TYPE_PLURAL;
} else if(isSelect(typeIndex)) {
argType=UMSGPAT_ARG_TYPE_SELECT;
}
} else if(length==13) {
if(isSelect(typeIndex) && isOrdinal(typeIndex+6)) {
argType=UMSGPAT_ARG_TYPE_SELECTORDINAL;
}
}
// change the ARG_START type from NONE to argType
partsList->a[argStart].value=(int16_t)argType;
if(argType==UMSGPAT_ARG_TYPE_SIMPLE) {
addPart(UMSGPAT_PART_TYPE_ARG_TYPE, typeIndex, length, 0, errorCode);
}
// look for an argument style (pattern)
if(c==u_rightCurlyBrace) {
if(argType!=UMSGPAT_ARG_TYPE_SIMPLE) {
setParseError(parseError, nameIndex); // No style field for complex argument.
errorCode=U_PATTERN_SYNTAX_ERROR;
return 0;
}
} else /* ',' */ {
++index;
if(argType==UMSGPAT_ARG_TYPE_SIMPLE) {
index=parseSimpleStyle(index, parseError, errorCode);
} else if(argType==UMSGPAT_ARG_TYPE_CHOICE) {
index=parseChoiceStyle(index, nestingLevel, parseError, errorCode);
} else {
index=parsePluralOrSelectStyle(argType, index, nestingLevel, parseError, errorCode);
}
}
}
// Argument parsing stopped on the '}'.
addLimitPart(argStart, UMSGPAT_PART_TYPE_ARG_LIMIT, index, 1, argType, errorCode);
return index+1;
}
