Package com.ibm.icu.text
Class NFRule
- java.lang.Object
-
- com.ibm.icu.text.NFRule
-
final class NFRule extends java.lang.Object
A class representing a single rule in a RuleBasedNumberFormat. A rule inserts its text into the result string and then passes control to its substitutions, which do the same thing.
-
-
Field Summary
Fields Modifier and Type Field Description private long
baseValue
The rule's base valueprivate char
decimalPoint
If this is a fraction rule, this is the decimal point from DecimalFormatSymbols to match.(package private) static int
DEFAULT_RULE
Special base value used to identify a default ruleprivate short
exponent
The rule's exponent (the radix raised to the power of the exponent equals the rule's divisor)(package private) RuleBasedNumberFormat
formatter
The RuleBasedNumberFormat that owns this rule(package private) static int
IMPROPER_FRACTION_RULE
Special base value used to identify an improper fraction (x.x) rule(package private) static int
INFINITY_RULE
Special base value used to identify an infinity rule(package private) static int
NAN_RULE
Special base value used to identify a not a number rule(package private) static int
NEGATIVE_NUMBER_RULE
Special base value used to identify a negative-number rule(package private) static int
PROPER_FRACTION_RULE
Special base value used to identify a proper fraction (0.x) ruleprivate int
radix
The rule's radix (the radix to the power of the exponent equals the rule's divisor)private static java.lang.String[]
RULE_PREFIXES
private PluralFormat
rulePatternFormat
The rule's plural format when defined.private java.lang.String
ruleText
The rule's rule text.private NFSubstitution
sub1
The rule's first substitution (the one with the lower offset into the rule text)private NFSubstitution
sub2
The rule's second substitution (the one with the higher offset into the rule text)(package private) static java.lang.Long
ZERO
-
Constructor Summary
Constructors Constructor Description NFRule(RuleBasedNumberFormat formatter, java.lang.String ruleText)
Nominal constructor for NFRule.
-
Method Summary
All Methods Static Methods Instance Methods Concrete Methods Modifier and Type Method Description private boolean
allIgnorable(java.lang.String str)
Checks to see whether a string consists entirely of ignorable characters.void
doFormat(double number, java.lang.StringBuilder toInsertInto, int pos, int recursionCount)
Formats the number, and inserts the resulting text into toInsertInto.void
doFormat(long number, java.lang.StringBuilder toInsertInto, int pos, int recursionCount)
Formats the number, and inserts the resulting text into toInsertInto.java.lang.Number
doParse(java.lang.String text, java.text.ParsePosition parsePosition, boolean isFractionRule, double upperBound, int nonNumericalExecutedRuleMask)
Attempts to parse the string with this rule.boolean
equals(java.lang.Object that)
Tests two rules for equality.private short
expectedExponent()
This calculates the rule's exponent based on its radix and base value.private NFSubstitution
extractSubstitution(NFRuleSet owner, NFRule predecessor)
Searches the rule's rule text for the first substitution token, creates a substitution based on it, and removes the token from the rule's rule text.private void
extractSubstitutions(NFRuleSet owner, java.lang.String ruleText, NFRule predecessor)
Searches the rule's rule text for the substitution tokens, creates the substitutions, and removes the substitution tokens from the rule's rule text.private int[]
findText(java.lang.String str, java.lang.String key, PluralFormat pluralFormatKey, int startingAt)
Searches a string for another string.long
getBaseValue()
Returns the rule's base valuechar
getDecimalPoint()
Returns the rule's base valuelong
getDivisor()
Returns the rule's divisor (the value that controls the behavior of its substitutions)int
hashCode()
(package private) boolean
hasModulusSubstitution()
Internal function used by the rounding code in MultiplierSubstitution.private static int
indexOfAnyRulePrefix(java.lang.String ruleText)
Searches the rule's rule text for any of the specified strings.static void
makeRules(java.lang.String description, NFRuleSet owner, NFRule predecessor, RuleBasedNumberFormat ownersOwner, java.util.List<NFRule> returnList)
Creates one or more rules based on the description passed in.private java.lang.Number
matchToDelimiter(java.lang.String text, int startPos, double baseVal, java.lang.String delimiter, PluralFormat pluralFormatDelimiter, java.text.ParsePosition pp, NFSubstitution sub, double upperBound, int nonNumericalExecutedRuleMask)
Used by parse() to match a substitution and any following text.private java.lang.String
parseRuleDescriptor(java.lang.String description)
This function parses the rule's rule descriptor (i.e., the base value and/or other tokens that precede the rule's rule text in the description) and sets the rule's base value, radix, and exponent according to the descriptor.(package private) static long
power(long base, short exponent)
This is an equivalent to Math.pow that accurately works on 64-bit numbersprivate int
prefixLength(java.lang.String str, java.lang.String prefix)
Used by stripPrefix() to match characters.(package private) void
setBaseValue(long newBaseValue)
Sets the rule's base value, and causes the radix and exponent to be recalculated.void
setDecimalFormatSymbols(DecimalFormatSymbols newSymbols)
boolean
shouldRollBack(long number)
Used by the owning rule set to determine whether to invoke the rollback rule (i.e., whether this rule or the one that precedes it in the rule set's list should be used to format the number)private java.lang.String
stripPrefix(java.lang.String text, java.lang.String prefix, java.text.ParsePosition pp)
This function is used by parse() to match the text being parsed against a possible prefix string.java.lang.String
toString()
Returns a textual representation of the rule.
-
-
-
Field Detail
-
NEGATIVE_NUMBER_RULE
static final int NEGATIVE_NUMBER_RULE
Special base value used to identify a negative-number rule- See Also:
- Constant Field Values
-
IMPROPER_FRACTION_RULE
static final int IMPROPER_FRACTION_RULE
Special base value used to identify an improper fraction (x.x) rule- See Also:
- Constant Field Values
-
PROPER_FRACTION_RULE
static final int PROPER_FRACTION_RULE
Special base value used to identify a proper fraction (0.x) rule- See Also:
- Constant Field Values
-
DEFAULT_RULE
static final int DEFAULT_RULE
Special base value used to identify a default rule- See Also:
- Constant Field Values
-
INFINITY_RULE
static final int INFINITY_RULE
Special base value used to identify an infinity rule- See Also:
- Constant Field Values
-
NAN_RULE
static final int NAN_RULE
Special base value used to identify a not a number rule- See Also:
- Constant Field Values
-
ZERO
static final java.lang.Long ZERO
-
baseValue
private long baseValue
The rule's base value
-
radix
private int radix
The rule's radix (the radix to the power of the exponent equals the rule's divisor)
-
exponent
private short exponent
The rule's exponent (the radix raised to the power of the exponent equals the rule's divisor)
-
decimalPoint
private char decimalPoint
If this is a fraction rule, this is the decimal point from DecimalFormatSymbols to match.
-
ruleText
private java.lang.String ruleText
The rule's rule text. When formatting a number, the rule's text is inserted into the result string, and then the text from any substitutions is inserted into the result string
-
rulePatternFormat
private PluralFormat rulePatternFormat
The rule's plural format when defined. This is not a substitution because it only works on the current baseValue. It's normally not used due to the overhead.
-
sub1
private NFSubstitution sub1
The rule's first substitution (the one with the lower offset into the rule text)
-
sub2
private NFSubstitution sub2
The rule's second substitution (the one with the higher offset into the rule text)
-
formatter
final RuleBasedNumberFormat formatter
The RuleBasedNumberFormat that owns this rule
-
RULE_PREFIXES
private static final java.lang.String[] RULE_PREFIXES
-
-
Constructor Detail
-
NFRule
public NFRule(RuleBasedNumberFormat formatter, java.lang.String ruleText)
Nominal constructor for NFRule. Most of the work of constructing an NFRule is actually performed by makeRules().
-
-
Method Detail
-
makeRules
public static void makeRules(java.lang.String description, NFRuleSet owner, NFRule predecessor, RuleBasedNumberFormat ownersOwner, java.util.List<NFRule> returnList)
Creates one or more rules based on the description passed in.- Parameters:
description
- The description of the rule(s).owner
- The rule set containing the new rule(s).predecessor
- The rule that precedes the new one(s) in "owner"'s rule listownersOwner
- The RuleBasedNumberFormat that owns the rule set that owns the new rule(s)returnList
- One or more instances of NFRule are added and returned here
-
parseRuleDescriptor
private java.lang.String parseRuleDescriptor(java.lang.String description)
This function parses the rule's rule descriptor (i.e., the base value and/or other tokens that precede the rule's rule text in the description) and sets the rule's base value, radix, and exponent according to the descriptor. (If the description doesn't include a rule descriptor, then this function sets everything to default values and the rule set sets the rule's real base value).- Parameters:
description
- The rule's description- Returns:
- If "description" included a rule descriptor, this is "description" with the descriptor and any trailing whitespace stripped off. Otherwise; it's "descriptor" unchanged.
-
extractSubstitutions
private void extractSubstitutions(NFRuleSet owner, java.lang.String ruleText, NFRule predecessor)
Searches the rule's rule text for the substitution tokens, creates the substitutions, and removes the substitution tokens from the rule's rule text.- Parameters:
owner
- The rule set containing this rulepredecessor
- The rule preceding this one in "owners" rule listruleText
- The rule text
-
extractSubstitution
private NFSubstitution extractSubstitution(NFRuleSet owner, NFRule predecessor)
Searches the rule's rule text for the first substitution token, creates a substitution based on it, and removes the token from the rule's rule text.- Parameters:
owner
- The rule set containing this rulepredecessor
- The rule preceding this one in the rule set's rule list- Returns:
- The newly-created substitution. This is never null; if the rule text doesn't contain any substitution tokens, this will be a NullSubstitution.
-
setBaseValue
final void setBaseValue(long newBaseValue)
Sets the rule's base value, and causes the radix and exponent to be recalculated. This is used during construction when we don't know the rule's base value until after it's been constructed. It should not be used at any other time.- Parameters:
newBaseValue
- The new base value for the rule.
-
expectedExponent
private short expectedExponent()
This calculates the rule's exponent based on its radix and base value. This will be the highest power the radix can be raised to and still produce a result less than or equal to the base value.
-
indexOfAnyRulePrefix
private static int indexOfAnyRulePrefix(java.lang.String ruleText)
Searches the rule's rule text for any of the specified strings.- Returns:
- The index of the first match in the rule's rule text (i.e., the first substring in the rule's rule text that matches _any_ of the strings in "strings"). If none of the strings in "strings" is found in the rule's rule text, returns -1.
-
equals
public boolean equals(java.lang.Object that)
Tests two rules for equality.- Overrides:
equals
in classjava.lang.Object
- Parameters:
that
- The rule to compare this one against- Returns:
- True if the two rules are functionally equivalent
-
hashCode
public int hashCode()
- Overrides:
hashCode
in classjava.lang.Object
-
toString
public java.lang.String toString()
Returns a textual representation of the rule. This won't necessarily be the same as the description that this rule was created with, but it will produce the same result.- Overrides:
toString
in classjava.lang.Object
- Returns:
- A textual description of the rule
-
getDecimalPoint
public final char getDecimalPoint()
Returns the rule's base value- Returns:
- The rule's base value
-
getBaseValue
public final long getBaseValue()
Returns the rule's base value- Returns:
- The rule's base value
-
getDivisor
public long getDivisor()
Returns the rule's divisor (the value that controls the behavior of its substitutions)- Returns:
- The rule's divisor
-
hasModulusSubstitution
boolean hasModulusSubstitution()
Internal function used by the rounding code in MultiplierSubstitution.
-
doFormat
public void doFormat(long number, java.lang.StringBuilder toInsertInto, int pos, int recursionCount)
Formats the number, and inserts the resulting text into toInsertInto.- Parameters:
number
- The number being formattedtoInsertInto
- The string where the resultant text should be insertedpos
- The position in toInsertInto where the resultant text should be inserted
-
doFormat
public void doFormat(double number, java.lang.StringBuilder toInsertInto, int pos, int recursionCount)
Formats the number, and inserts the resulting text into toInsertInto.- Parameters:
number
- The number being formattedtoInsertInto
- The string where the resultant text should be insertedpos
- The position in toInsertInto where the resultant text should be inserted
-
power
static long power(long base, short exponent)
This is an equivalent to Math.pow that accurately works on 64-bit numbers- Parameters:
base
- The baseexponent
- The exponent- Returns:
- radix ** exponent
- See Also:
Math.pow(double, double)
-
shouldRollBack
public boolean shouldRollBack(long number)
Used by the owning rule set to determine whether to invoke the rollback rule (i.e., whether this rule or the one that precedes it in the rule set's list should be used to format the number)- Parameters:
number
- The number being formatted- Returns:
- True if the rule set should use the rule that precedes this one in its list; false if it should use this rule
-
doParse
public java.lang.Number doParse(java.lang.String text, java.text.ParsePosition parsePosition, boolean isFractionRule, double upperBound, int nonNumericalExecutedRuleMask)
Attempts to parse the string with this rule.- Parameters:
text
- The string being parsedparsePosition
- On entry, the value is ignored and assumed to be 0. On exit, this has been updated with the position of the first character not consumed by matching the text against this rule (if this rule doesn't match the text at all, the parse position if left unchanged (presumably at 0) and the function returns Long.valueOf(0)).isFractionRule
- True if this rule is contained within a fraction rule set. This is only used if the rule has no substitutions.- Returns:
- If this rule matched the text, this is the rule's base value combined appropriately with the results of parsing the substitutions. If nothing matched, this is Long.valueOf(0) and the parse position is left unchanged. The result will be an instance of Long if the result is an integer and Double otherwise. The result is never null.
-
stripPrefix
private java.lang.String stripPrefix(java.lang.String text, java.lang.String prefix, java.text.ParsePosition pp)
This function is used by parse() to match the text being parsed against a possible prefix string. This function matches characters from the beginning of the string being parsed to characters from the prospective prefix. If they match, pp is updated to the first character not matched, and the result is the unparsed part of the string. If they don't match, the whole string is returned, and pp is left unchanged.- Parameters:
text
- The string being parsedprefix
- The text to match againstpp
- On entry, ignored and assumed to be 0. On exit, points to the first unmatched character (assuming the whole prefix matched), or is unchanged (if the whole prefix didn't match).- Returns:
- If things match, this is the unparsed part of "text"; if they didn't match, this is "text".
-
matchToDelimiter
private java.lang.Number matchToDelimiter(java.lang.String text, int startPos, double baseVal, java.lang.String delimiter, PluralFormat pluralFormatDelimiter, java.text.ParsePosition pp, NFSubstitution sub, double upperBound, int nonNumericalExecutedRuleMask)
Used by parse() to match a substitution and any following text. "text" is searched for instances of "delimiter". For each instance of delimiter, the intervening text is tested to see whether it matches the substitution. The longest match wins.- Parameters:
text
- The string being parsedstartPos
- The position in "text" where we should start looking for "delimiter".baseVal
- A partial parse result (often the rule's base value), which is combined with the result from matching the substitutiondelimiter
- The string to search "text" for.pp
- Ignored and presumed to be 0 on entry. If there's a match, on exit this will point to the first unmatched character.sub
- If we find "delimiter" in "text", this substitution is used to match the text between the beginning of the string and the position of "delimiter." (If "delimiter" is the empty string, then this function just matches against this substitution and updates everything accordingly.)upperBound
- When matching the substitution, it will only consider rules with base values lower than this value.- Returns:
- If there's a match, this is the result of composing baseValue with the result of matching the substitution. Otherwise, this is Long.valueOf(0). It's never null. If the result is an integer, this will be an instance of Long; otherwise, it's an instance of Double.
-
prefixLength
private int prefixLength(java.lang.String str, java.lang.String prefix)
Used by stripPrefix() to match characters. If lenient parse mode is off, this just calls startsWith(). If lenient parse mode is on, this function uses CollationElementIterators to match characters in the strings (only primary-order differences are significant in determining whether there's a match).- Parameters:
str
- The string being testedprefix
- The text we're hoping to see at the beginning of "str"- Returns:
- If "prefix" is found at the beginning of "str", this is the number of characters in "str" that were matched (this isn't necessarily the same as the length of "prefix" when matching text with a collator). If there's no match, this is 0.
-
findText
private int[] findText(java.lang.String str, java.lang.String key, PluralFormat pluralFormatKey, int startingAt)
Searches a string for another string. If lenient parsing is off, this just calls indexOf(). If lenient parsing is on, this function uses CollationElementIterator to match characters, and only primary-order differences are significant in determining whether there's a match.- Parameters:
str
- The string to searchkey
- The string to search "str" forstartingAt
- The index into "str" where the search is to begin- Returns:
- A two-element array of ints. Element 0 is the position of the match, or -1 if there was no match. Element 1 is the number of characters in "str" that matched (which isn't necessarily the same as the length of "key")
-
allIgnorable
private boolean allIgnorable(java.lang.String str)
Checks to see whether a string consists entirely of ignorable characters.- Parameters:
str
- The string to test.- Returns:
- true if the string is empty of consists entirely of characters that the number formatter's collator says are ignorable at the primary-order level. false otherwise.
-
setDecimalFormatSymbols
public void setDecimalFormatSymbols(DecimalFormatSymbols newSymbols)
-
-