Home | History | Annotate | Download | only in common
      1 /*
      2 **********************************************************************
      3 * Copyright (c) 2003-2007, International Business Machines
      4 * Corporation and others.  All Rights Reserved.
      5 **********************************************************************
      6 * Author: Alan Liu
      7 * Created: September 24 2003
      8 * Since: ICU 2.8
      9 **********************************************************************
     10 */
     11 #ifndef _RULEITER_H_
     12 #define _RULEITER_H_
     13 
     14 #include "unicode/uobject.h"
     15 
     16 U_NAMESPACE_BEGIN
     17 
     18 class UnicodeString;
     19 class ParsePosition;
     20 class SymbolTable;
     21 
     22 /**
     23  * An iterator that returns 32-bit code points.  This class is deliberately
     24  * <em>not</em> related to any of the ICU character iterator classes
     25  * in order to minimize complexity.
     26  * @author Alan Liu
     27  * @since ICU 2.8
     28  */
     29 class RuleCharacterIterator : public UMemory {
     30 
     31     // TODO: Ideas for later.  (Do not implement if not needed, lest the
     32     // code coverage numbers go down due to unused methods.)
     33     // 1. Add a copy constructor, operator==() method.
     34     // 2. Rather than return DONE, throw an exception if the end
     35     // is reached -- this is an alternate usage model, probably not useful.
     36 
     37 private:
     38     /**
     39      * Text being iterated.
     40      */
     41     const UnicodeString& text;
     42 
     43     /**
     44      * Position of iterator.
     45      */
     46     ParsePosition& pos;
     47 
     48     /**
     49      * Symbol table used to parse and dereference variables.  May be 0.
     50      */
     51     const SymbolTable* sym;
     52 
     53     /**
     54      * Current variable expansion, or 0 if none.
     55      */
     56     const UnicodeString* buf;
     57 
     58     /**
     59      * Position within buf.  Meaningless if buf == 0.
     60      */
     61     int32_t bufPos;
     62 
     63 public:
     64     /**
     65      * Value returned when there are no more characters to iterate.
     66      */
     67     enum { DONE = -1 };
     68 
     69     /**
     70      * Bitmask option to enable parsing of variable names.  If (options &
     71      * PARSE_VARIABLES) != 0, then an embedded variable will be expanded to
     72      * its value.  Variables are parsed using the SymbolTable API.
     73      */
     74     enum { PARSE_VARIABLES = 1 };
     75 
     76     /**
     77      * Bitmask option to enable parsing of escape sequences.  If (options &
     78      * PARSE_ESCAPES) != 0, then an embedded escape sequence will be expanded
     79      * to its value.  Escapes are parsed using Utility.unescapeAt().
     80      */
     81     enum { PARSE_ESCAPES   = 2 };
     82 
     83     /**
     84      * Bitmask option to enable skipping of whitespace.  If (options &
     85      * SKIP_WHITESPACE) != 0, then whitespace characters will be silently
     86      * skipped, as if they were not present in the input.  Whitespace
     87      * characters are defined by UCharacterProperty.isRuleWhiteSpace().
     88      */
     89     enum { SKIP_WHITESPACE = 4 };
     90 
     91     /**
     92      * Constructs an iterator over the given text, starting at the given
     93      * position.
     94      * @param text the text to be iterated
     95      * @param sym the symbol table, or null if there is none.  If sym is null,
     96      * then variables will not be deferenced, even if the PARSE_VARIABLES
     97      * option is set.
     98      * @param pos upon input, the index of the next character to return.  If a
     99      * variable has been dereferenced, then pos will <em>not</em> increment as
    100      * characters of the variable value are iterated.
    101      */
    102     RuleCharacterIterator(const UnicodeString& text, const SymbolTable* sym,
    103                           ParsePosition& pos);
    104 
    105     /**
    106      * Returns true if this iterator has no more characters to return.
    107      */
    108     UBool atEnd() const;
    109 
    110     /**
    111      * Returns the next character using the given options, or DONE if there
    112      * are no more characters, and advance the position to the next
    113      * character.
    114      * @param options one or more of the following options, bitwise-OR-ed
    115      * together: PARSE_VARIABLES, PARSE_ESCAPES, SKIP_WHITESPACE.
    116      * @param isEscaped output parameter set to TRUE if the character
    117      * was escaped
    118      * @param ec input-output error code.  An error will only be set by
    119      * this routing if options includes PARSE_VARIABLES and an unknown
    120      * variable name is seen, or if options includes PARSE_ESCAPES and
    121      * an invalid escape sequence is seen.
    122      * @return the current 32-bit code point, or DONE
    123      */
    124     UChar32 next(int32_t options, UBool& isEscaped, UErrorCode& ec);
    125 
    126     /**
    127      * Returns true if this iterator is currently within a variable expansion.
    128      */
    129     inline UBool inVariable() const;
    130 
    131     /**
    132      * An opaque object representing the position of a RuleCharacterIterator.
    133      */
    134     struct Pos : public UMemory {
    135     private:
    136         const UnicodeString* buf;
    137         int32_t pos;
    138         int32_t bufPos;
    139         friend class RuleCharacterIterator;
    140     };
    141 
    142     /**
    143      * Sets an object which, when later passed to setPos(), will
    144      * restore this iterator's position.  Usage idiom:
    145      *
    146      * RuleCharacterIterator iterator = ...;
    147      * RuleCharacterIterator::Pos pos;
    148      * iterator.getPos(pos);
    149      * for (;;) {
    150      *   iterator.getPos(pos);
    151      *   int c = iterator.next(...);
    152      *   ...
    153      * }
    154      * iterator.setPos(pos);
    155      *
    156      * @param p a position object to be set to this iterator's
    157      * current position.
    158      */
    159     void getPos(Pos& p) const;
    160 
    161     /**
    162      * Restores this iterator to the position it had when getPos()
    163      * set the given object.
    164      * @param p a position object previously set by getPos()
    165      */
    166     void setPos(const Pos& p);
    167 
    168     /**
    169      * Skips ahead past any ignored characters, as indicated by the given
    170      * options.  This is useful in conjunction with the lookahead() method.
    171      *
    172      * Currently, this only has an effect for SKIP_WHITESPACE.
    173      * @param options one or more of the following options, bitwise-OR-ed
    174      * together: PARSE_VARIABLES, PARSE_ESCAPES, SKIP_WHITESPACE.
    175      */
    176     void skipIgnored(int32_t options);
    177 
    178     /**
    179      * Returns a string containing the remainder of the characters to be
    180      * returned by this iterator, without any option processing.  If the
    181      * iterator is currently within a variable expansion, this will only
    182      * extend to the end of the variable expansion.  This method is provided
    183      * so that iterators may interoperate with string-based APIs.  The typical
    184      * sequence of calls is to call skipIgnored(), then call lookahead(), then
    185      * parse the string returned by lookahead(), then call jumpahead() to
    186      * resynchronize the iterator.
    187      * @param result a string to receive the characters to be returned
    188      * by future calls to next()
    189      * @param maxLookAhead The maximum to copy into the result.
    190      * @return a reference to result
    191      */
    192     UnicodeString& lookahead(UnicodeString& result, int32_t maxLookAhead = -1) const;
    193 
    194     /**
    195      * Advances the position by the given number of 16-bit code units.
    196      * This is useful in conjunction with the lookahead() method.
    197      * @param count the number of 16-bit code units to jump over
    198      */
    199     void jumpahead(int32_t count);
    200 
    201     /**
    202      * Returns a string representation of this object, consisting of the
    203      * characters being iterated, with a '|' marking the current position.
    204      * Position within an expanded variable is <em>not</em> indicated.
    205      * @param result output parameter to receive a string
    206      * representation of this object
    207      */
    208 //    UnicodeString& toString(UnicodeString& result) const;
    209 
    210 private:
    211     /**
    212      * Returns the current 32-bit code point without parsing escapes, parsing
    213      * variables, or skipping whitespace.
    214      * @return the current 32-bit code point
    215      */
    216     UChar32 _current() const;
    217 
    218     /**
    219      * Advances the position by the given amount.
    220      * @param count the number of 16-bit code units to advance past
    221      */
    222     void _advance(int32_t count);
    223 };
    224 
    225 inline UBool RuleCharacterIterator::inVariable() const {
    226     return buf != 0;
    227 }
    228 
    229 U_NAMESPACE_END
    230 
    231 #endif // _RULEITER_H_
    232 //eof
    233