michael@0: /*
michael@0: **********************************************************************
michael@0: * Copyright (c) 2003-2011, International Business Machines
michael@0: * Corporation and others.  All Rights Reserved.
michael@0: **********************************************************************
michael@0: * Author: Alan Liu
michael@0: * Created: September 24 2003
michael@0: * Since: ICU 2.8
michael@0: **********************************************************************
michael@0: */
michael@0: #include "ruleiter.h"
michael@0: #include "unicode/parsepos.h"
michael@0: #include "unicode/symtable.h"
michael@0: #include "unicode/unistr.h"
michael@0: #include "unicode/utf16.h"
michael@0: #include "patternprops.h"
michael@0: 
michael@0: /* \U87654321 or \ud800\udc00 */
michael@0: #define MAX_U_NOTATION_LEN 12
michael@0: 
michael@0: U_NAMESPACE_BEGIN
michael@0: 
michael@0: RuleCharacterIterator::RuleCharacterIterator(const UnicodeString& theText, const SymbolTable* theSym,
michael@0:                       ParsePosition& thePos) :
michael@0:     text(theText),
michael@0:     pos(thePos),
michael@0:     sym(theSym),
michael@0:     buf(0),
michael@0:     bufPos(0)
michael@0: {}
michael@0: 
michael@0: UBool RuleCharacterIterator::atEnd() const {
michael@0:     return buf == 0 && pos.getIndex() == text.length();
michael@0: }
michael@0: 
michael@0: UChar32 RuleCharacterIterator::next(int32_t options, UBool& isEscaped, UErrorCode& ec) {
michael@0:     if (U_FAILURE(ec)) return DONE;
michael@0: 
michael@0:     UChar32 c = DONE;
michael@0:     isEscaped = FALSE;
michael@0: 
michael@0:     for (;;) {
michael@0:         c = _current();
michael@0:         _advance(U16_LENGTH(c));
michael@0: 
michael@0:         if (c == SymbolTable::SYMBOL_REF && buf == 0 &&
michael@0:             (options & PARSE_VARIABLES) != 0 && sym != 0) {
michael@0:             UnicodeString name = sym->parseReference(text, pos, text.length());
michael@0:             // If name is empty there was an isolated SYMBOL_REF;
michael@0:             // return it.  Caller must be prepared for this.
michael@0:             if (name.length() == 0) {
michael@0:                 break;
michael@0:             }
michael@0:             bufPos = 0;
michael@0:             buf = sym->lookup(name);
michael@0:             if (buf == 0) {
michael@0:                 ec = U_UNDEFINED_VARIABLE;
michael@0:                 return DONE;
michael@0:             }
michael@0:             // Handle empty variable value
michael@0:             if (buf->length() == 0) {
michael@0:                 buf = 0;
michael@0:             }
michael@0:             continue;
michael@0:         }
michael@0: 
michael@0:         if ((options & SKIP_WHITESPACE) != 0 && PatternProps::isWhiteSpace(c)) {
michael@0:             continue;
michael@0:         }
michael@0: 
michael@0:         if (c == 0x5C /*'\\'*/ && (options & PARSE_ESCAPES) != 0) {
michael@0:             UnicodeString tempEscape;
michael@0:             int32_t offset = 0;
michael@0:             c = lookahead(tempEscape, MAX_U_NOTATION_LEN).unescapeAt(offset);
michael@0:             jumpahead(offset);
michael@0:             isEscaped = TRUE;
michael@0:             if (c < 0) {
michael@0:                 ec = U_MALFORMED_UNICODE_ESCAPE;
michael@0:                 return DONE;
michael@0:             }
michael@0:         }
michael@0: 
michael@0:         break;
michael@0:     }
michael@0: 
michael@0:     return c;
michael@0: }
michael@0: 
michael@0: void RuleCharacterIterator::getPos(RuleCharacterIterator::Pos& p) const {
michael@0:     p.buf = buf;
michael@0:     p.pos = pos.getIndex();
michael@0:     p.bufPos = bufPos;
michael@0: }
michael@0: 
michael@0: void RuleCharacterIterator::setPos(const RuleCharacterIterator::Pos& p) {
michael@0:     buf = p.buf;
michael@0:     pos.setIndex(p.pos);
michael@0:     bufPos = p.bufPos;
michael@0: }
michael@0: 
michael@0: void RuleCharacterIterator::skipIgnored(int32_t options) {
michael@0:     if ((options & SKIP_WHITESPACE) != 0) {
michael@0:         for (;;) {
michael@0:             UChar32 a = _current();
michael@0:             if (!PatternProps::isWhiteSpace(a)) break;
michael@0:             _advance(U16_LENGTH(a));
michael@0:         }
michael@0:     }
michael@0: }
michael@0: 
michael@0: UnicodeString& RuleCharacterIterator::lookahead(UnicodeString& result, int32_t maxLookAhead) const {
michael@0:     if (maxLookAhead < 0) {
michael@0:         maxLookAhead = 0x7FFFFFFF;
michael@0:     }
michael@0:     if (buf != 0) {
michael@0:         buf->extract(bufPos, maxLookAhead, result);
michael@0:     } else {
michael@0:         text.extract(pos.getIndex(), maxLookAhead, result);
michael@0:     }
michael@0:     return result;
michael@0: }
michael@0: 
michael@0: void RuleCharacterIterator::jumpahead(int32_t count) {
michael@0:     _advance(count);
michael@0: }
michael@0: 
michael@0: /*
michael@0: UnicodeString& RuleCharacterIterator::toString(UnicodeString& result) const {
michael@0:     int32_t b = pos.getIndex();
michael@0:     text.extract(0, b, result);
michael@0:     return result.append((UChar) 0x7C).append(text, b, 0x7FFFFFFF); // Insert '|' at index
michael@0: }
michael@0: */
michael@0: 
michael@0: UChar32 RuleCharacterIterator::_current() const {
michael@0:     if (buf != 0) {
michael@0:         return buf->char32At(bufPos);
michael@0:     } else {
michael@0:         int i = pos.getIndex();
michael@0:         return (i < text.length()) ? text.char32At(i) : (UChar32)DONE;
michael@0:     }
michael@0: }
michael@0: 
michael@0: void RuleCharacterIterator::_advance(int32_t count) {
michael@0:     if (buf != 0) {
michael@0:         bufPos += count;
michael@0:         if (bufPos == buf->length()) {
michael@0:             buf = 0;
michael@0:         }
michael@0:     } else {
michael@0:         pos.setIndex(pos.getIndex() + count);
michael@0:         if (pos.getIndex() > text.length()) {
michael@0:             pos.setIndex(text.length());
michael@0:         }
michael@0:     }
michael@0: }
michael@0: 
michael@0: U_NAMESPACE_END
michael@0: 
michael@0: //eof