org.eclipse.jdt.core.search
Class SearchPattern

java.lang.Object
  org.eclipse.jdt.internal.core.search.matching.InternalSearchPattern
      org.eclipse.jdt.core.search.SearchPattern

Direct Known Subclasses:

JavaSearchPattern, OrPattern

public abstract class SearchPattern
extends InternalSearchPattern

A search pattern defines how search results are found. Use SearchPattern.createPattern to create a search pattern.

Search patterns are used during the search phase to decode index entries that were added during the indexing phase (see SearchDocument.addIndexEntry(char[], char[])). When an index is queried, the index categories and keys to consider are retrieved from the search pattern using getIndexCategories() and getIndexKey(), as well as the match rule (see getMatchRule()). A blank pattern is then created (see getBlankPattern()). This blank pattern is used as a record as follows. For each index entry in the given index categories and that starts with the given key, the blank pattern is fed using decodeIndexKey(char[]). The original pattern is then asked if it matches the decoded key using matchesDecodedKey(SearchPattern). If it matches, a search document is created for this index entry using SearchParticipant.getDocument(String).

This class is intended to be sub-classed by clients. A default behavior is provided for each of the methods above, that clients can override if they wish.

Since:

3.0

See Also:

createPattern(org.eclipse.jdt.core.IJavaElement, int), createPattern(String, int, int, int)

Field Summary

static int	R_CAMELCASE_MATCH Match rule: The search pattern contains a Camel Case expression.
static int	R_CAMELCASE_SAME_PART_COUNT_MATCH Match rule: The search pattern contains a Camel Case expression with a strict expected number of parts.
static int	R_CASE_SENSITIVE Match rule: The search pattern matches the search result only if cases are the same.
static int	R_EQUIVALENT_MATCH Match rule: The search pattern matches search results as raw/parameterized types/methods with equivalent type parameters.
static int	R_ERASURE_MATCH Match rule: The search pattern matches search results as raw/parameterized types/methods with same erasure.
static int	R_EXACT_MATCH Match rule: The search pattern matches exactly the search result, that is, the source of the search result equals the search pattern.
static int	R_FULL_MATCH Match rule: The search pattern matches exactly the search result, that is, the source of the search result equals the search pattern.
static int	R_PATTERN_MATCH Match rule: The search pattern contains one or more wild cards ('*' or '?').
static int	R_PREFIX_MATCH Match rule: The search pattern is a prefix of the search result.
static int	R_REGEXP_MATCH Match rule: The search pattern contains a regular expression.

Constructor Summary

SearchPattern(int matchRule)
Creates a search pattern with the rule to apply for matching index keys.

Method Summary

static boolean	camelCaseMatch(java.lang.String pattern, int patternStart, int patternEnd, java.lang.String name, int nameStart, int nameEnd) Answers true if a sub-pattern matches the sub-part of the given name using CamelCase rules, or false otherwise.
static boolean	camelCaseMatch(java.lang.String pattern, int patternStart, int patternEnd, java.lang.String name, int nameStart, int nameEnd, boolean samePartCount) Answers true if a sub-pattern matches the sub-part of the given name using CamelCase rules, or false otherwise.
static boolean	camelCaseMatch(java.lang.String pattern, java.lang.String name) Answers true if the pattern matches the given name using CamelCase rules, or false otherwise.
static boolean	camelCaseMatch(java.lang.String pattern, java.lang.String name, boolean samePartCount) Answers true if the pattern matches the given name using CamelCase rules, or false otherwise.
static SearchPattern	createAndPattern(SearchPattern leftPattern, SearchPattern rightPattern) Deprecated. Unfortunately, this functionality is not fully supported yet (see "https://bugs.eclipse.org/bugs/show_bug.cgi?id=142044" for more details). This might be done in a further version...
static SearchPattern	createOrPattern(SearchPattern leftPattern, SearchPattern rightPattern) Returns a search pattern that combines the given two patterns into an "or" pattern.
static SearchPattern	createPattern(IJavaElement element, int limitTo) Returns a search pattern based on a given Java element.
static SearchPattern	createPattern(IJavaElement element, int limitTo, int matchRule) Returns a search pattern based on a given Java element.
static SearchPattern	createPattern(java.lang.String stringPattern, int searchFor, int limitTo, int matchRule) Returns a search pattern based on a given string pattern.
void	decodeIndexKey(char[] key) Decode the given index key in this pattern.
abstract SearchPattern	getBlankPattern() Returns a blank pattern that can be used as a record to decode an index key.
char[][]	getIndexCategories() Returns an array of index categories to consider for this index query.
char[]	getIndexKey() Returns a key to find in relevant index categories, if null then all index entries are matched.
int	getMatchRule() Returns the rule to apply for matching index keys.
boolean	matchesDecodedKey(SearchPattern decodedPattern) Returns whether this pattern matches the given pattern (representing a decoded index key).
boolean	matchesName(char[] pattern, char[] name) Returns whether the given name matches the given pattern.
java.lang.String	toString()
static int	validateMatchRule(java.lang.String stringPattern, int matchRule) Validate compatibility between given string pattern and match rule.

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, wait, wait, wait

Field Detail

R_EXACT_MATCH

public static final int R_EXACT_MATCH

Match rule: The search pattern matches exactly the search result, that is, the source of the search result equals the search pattern.

See Also:

Constant Field Values

R_PREFIX_MATCH

public static final int R_PREFIX_MATCH

Match rule: The search pattern is a prefix of the search result.

See Also:

Constant Field Values

R_PATTERN_MATCH

public static final int R_PATTERN_MATCH

Match rule: The search pattern contains one or more wild cards ('*' or '?'). A '*' wild-card can replace 0 or more characters in the search result. A '?' wild-card replaces exactly 1 character in the search result.

See Also:

Constant Field Values

R_REGEXP_MATCH

public static final int R_REGEXP_MATCH

Match rule: The search pattern contains a regular expression.

See Also:

Constant Field Values

R_CASE_SENSITIVE

public static final int R_CASE_SENSITIVE

Match rule: The search pattern matches the search result only if cases are the same. Can be combined to previous rules, e.g. R_EXACT_MATCH | R_CASE_SENSITIVE

See Also:

Constant Field Values

R_ERASURE_MATCH

public static final int R_ERASURE_MATCH

Match rule: The search pattern matches search results as raw/parameterized types/methods with same erasure. This mode has no effect on other java elements search.
Type search example:

• pattern: List<Exception>
• match: List<Object>

Method search example:

• declaration: <T>foo(T t)
• pattern: <Exception>foo(new Exception())
• match: <Object>foo(new Object())

Can be combined to all other match rules, e.g. R_CASE_SENSITIVE | R_ERASURE_MATCH This rule is not activated by default, so raw types or parameterized types with same erasure will not be found for pattern List<String>, Note that with this pattern, the match selection will be only on the erasure even for parameterized types.

Since:

3.1

See Also:

Constant Field Values

R_EQUIVALENT_MATCH

public static final int R_EQUIVALENT_MATCH

Match rule: The search pattern matches search results as raw/parameterized types/methods with equivalent type parameters. This mode has no effect on other java elements search.
Type search example:

• pattern: List<Exception>
• match:
  • List<? extends Throwable>
  • List<? super RuntimeException>
  • List<?>

Method search example:

• declaration: <T>foo(T t)
• pattern: <Exception>foo(new Exception())
• match:
  • <? extends Throwable>foo(new Exception())
  • <? super RuntimeException>foo(new Exception())
  • foo(new Exception())

Can be combined to all other match rules, e.g. R_CASE_SENSITIVE | R_EQUIVALENT_MATCH This rule is not activated by default, so raw types or equivalent parameterized types will not be found for pattern List<String>, This mode is overridden by R_ERASURE_MATCH as erasure matches obviously include equivalent ones. That means that pattern with rule set to R_EQUIVALENT_MATCH | R_ERASURE_MATCH will return same results than rule only set with R_ERASURE_MATCH.

Since:

3.1

See Also:

Constant Field Values

R_FULL_MATCH

public static final int R_FULL_MATCH

Match rule: The search pattern matches exactly the search result, that is, the source of the search result equals the search pattern.

Since:

3.1

See Also:

Constant Field Values

R_CAMELCASE_MATCH

public static final int R_CAMELCASE_MATCH

Match rule: The search pattern contains a Camel Case expression.

Examples:

• 'NPE' type string pattern will match 'NullPointerException' and 'NoPermissionException' types,
• 'NuPoEx' type string pattern will only match 'NullPointerException' type.

This rule is not intended to be combined with any other match rule. In case of other match rule flags are combined with this one, then match rule validation will return a modified rule in order to perform a better appropriate search request (see validateMatchRule(String, int) for more details).

Since:

3.2

See Also:

for a detailed explanation of Camel Case matching., Constant Field Values

R_CAMELCASE_SAME_PART_COUNT_MATCH

public static final int R_CAMELCASE_SAME_PART_COUNT_MATCH

Match rule: The search pattern contains a Camel Case expression with a strict expected number of parts.
Examples:

• 'HM' type string pattern will match 'HashMap' and 'HtmlMapper' types, but not 'HashMapEntry'
• 'HMap' type string pattern will still match previous 'HashMap' and 'HtmlMapper' types, but not 'HighMagnitude'

This rule is not intended to be combined with any other match rule. In case of other match rule flags are combined with this one, then match rule validation will return a modified rule in order to perform a better appropriate search request (see validateMatchRule(String, int) for more details).

Since:

3.4

See Also:

for a detailed explanation of Camel Case matching.

, Constant Field Values

Constructor Detail

SearchPattern

public SearchPattern(int matchRule)

Creates a search pattern with the rule to apply for matching index keys. It can be exact match, prefix match, pattern match or regexp match. Rule can also be combined with a case sensitivity flag.

Parameters:

matchRule - one of following match rule

• R_EXACT_MATCH
• R_PREFIX_MATCH
• R_PATTERN_MATCH
• R_REGEXP_MATCH
• R_CAMELCASE_MATCH
• R_CAMELCASE_SAME_PART_COUNT_MATCH

which may be also combined with one of following flag:

• R_CASE_SENSITIVE
• R_ERASURE_MATCH
• R_EQUIVALENT_MATCH

For example,

• R_EXACT_MATCH | R_CASE_SENSITIVE: if an exact and case sensitive match is requested,
• R_PREFIX_MATCH if a case insensitive prefix match is requested
• R_EXACT_MATCH | R_ERASURE_MATCH: if a case insensitive and erasure match is requested.

Note that R_ERASURE_MATCH or R_EQUIVALENT_MATCH has no effect on non-generic types/methods search.

Note also that default behavior for generic types/methods search is to find exact matches.

Method Detail

camelCaseMatch

public static final boolean camelCaseMatch(java.lang.String pattern,
                                           java.lang.String name)

Answers true if the pattern matches the given name using CamelCase rules, or false otherwise. char[] CamelCase matching does NOT accept explicit wild-cards '*' and '?' and is inherently case sensitive.

CamelCase denotes the convention of writing compound names without spaces, and capitalizing every term. This function recognizes both upper and lower CamelCase, depending whether the leading character is capitalized or not. The leading part of an upper CamelCase pattern is assumed to contain a sequence of capitals which are appearing in the matching name; e.g. 'NPE' will match 'NullPointerException', but not 'NewPerfData'. A lower CamelCase pattern uses a lowercase first character. In Java, type names follow the upper CamelCase convention, whereas method or field names follow the lower CamelCase convention.

The pattern may contain lowercase characters, which will be matched in a case sensitive way. These characters must appear in sequence in the name. For instance, 'NPExcep' will match 'NullPointerException', but not 'NullPointerExCEPTION' or 'NuPoEx' will match 'NullPointerException', but not 'NoPointerException'.

Digit characters are treated in a special way. They can be used in the pattern but are not always considered as leading character. For instance, both 'UTF16DSS' and 'UTFDSS' patterns will match 'UTF16DocumentScannerSupport'.

Using this method allows matching names to have more parts than the specified pattern (see camelCaseMatch(String, String, boolean)).
For instance, 'HM' , 'HaMa' and 'HMap' patterns will match 'HashMap', 'HatMapper' and also 'HashMapEntry'.

 Examples:
 
  pattern = "NPE"
  name = NullPointerException / NoPermissionException
  result => true
 
  pattern = "NuPoEx"
  name = NullPointerException
  result => true
 
  pattern = "npe"
  name = NullPointerException
  result => false
 
  pattern = "IPL3"
  name = "IPerspectiveListener3"
  result => true
 
  pattern = "HM"
  name = "HashMapEntry"
  result => true
 
  pattern = "HMap"
  name = "HatMapper"
  result => true
 

Parameters:

pattern - the given pattern

name - the given name

Returns:

true if the pattern matches the given name, false otherwise

Since:

3.2

See Also:

for algorithm implementation

camelCaseMatch

public static final boolean camelCaseMatch(java.lang.String pattern,
                                           java.lang.String name,
                                           boolean samePartCount)

Answers true if the pattern matches the given name using CamelCase rules, or false otherwise. char[] CamelCase matching does NOT accept explicit wild-cards '*' and '?' and is inherently case sensitive.

CamelCase denotes the convention of writing compound names without spaces, and capitalizing every term. This function recognizes both upper and lower CamelCase, depending whether the leading character is capitalized or not. The leading part of an upper CamelCase pattern is assumed to contain a sequence of capitals which are appearing in the matching name; e.g. 'NPE' will match 'NullPointerException', but not 'NewPerfData'. A lower CamelCase pattern uses a lowercase first character. In Java, type names follow the upper CamelCase convention, whereas method or field names follow the lower CamelCase convention.

The pattern may contain lowercase characters, which will be matched in a case sensitive way. These characters must appear in sequence in the name. For instance, 'NPExcep' will match 'NullPointerException', but not 'NullPointerExCEPTION' or 'NuPoEx' will match 'NullPointerException', but not 'NoPointerException'.

Digit characters are treated in a special way. They can be used in the pattern but are not always considered as leading character. For instance, both 'UTF16DSS' and 'UTFDSS' patterns will match 'UTF16DocumentScannerSupport'.

CamelCase can be restricted to match only the same count of parts. When this restriction is specified the given pattern and the given name must have exactly the same number of parts (i.e. the same number of uppercase characters).
For instance, 'HM' , 'HaMa' and 'HMap' patterns will match 'HashMap' and 'HatMapper' but not 'HashMapEntry'.

 Examples:
 
  pattern = "NPE"
  name = NullPointerException / NoPermissionException
  result => true
 
  pattern = "NuPoEx"
  name = NullPointerException
  result => true
 
  pattern = "npe"
  name = NullPointerException
  result => false
 
  pattern = "IPL3"
  name = "IPerspectiveListener3"
  result => true
 
  pattern = "HM"
  name = "HashMapEntry"
  result => (samePartCount == false)
 

Parameters:

pattern - the given pattern

name - the given name

samePartCount - flag telling whether the pattern and the name should have the same count of parts or not.
For example:

• 'HM' type string pattern will match 'HashMap' and 'HtmlMapper' types, but not 'HashMapEntry'
• 'HMap' type string pattern will still match previous 'HashMap' and 'HtmlMapper' types, but not 'HighMagnitude'

Returns:

true if the pattern matches the given name, false otherwise

Since:

3.4

See Also:

for algorithm implementation

camelCaseMatch

public static final boolean camelCaseMatch(java.lang.String pattern,
                                           int patternStart,
                                           int patternEnd,
                                           java.lang.String name,
                                           int nameStart,
                                           int nameEnd)

Answers true if a sub-pattern matches the sub-part of the given name using CamelCase rules, or false otherwise. char[] CamelCase matching does NOT accept explicit wild-cards '*' and '?' and is inherently case sensitive. Can match only subset of name/pattern, considering end positions as non-inclusive. The sub-pattern is defined by the patternStart and patternEnd positions.

CamelCase denotes the convention of writing compound names without spaces, and capitalizing every term. This function recognizes both upper and lower CamelCase, depending whether the leading character is capitalized or not. The leading part of an upper CamelCase pattern is assumed to contain a sequence of capitals which are appearing in the matching name; e.g. 'NPE' will match 'NullPointerException', but not 'NewPerfData'. A lower CamelCase pattern uses a lowercase first character. In Java, type names follow the upper CamelCase convention, whereas method or field names follow the lower CamelCase convention.

The pattern may contain lowercase characters, which will be matched in a case sensitive way. These characters must appear in sequence in the name. For instance, 'NPExcep' will match 'NullPointerException', but not 'NullPointerExCEPTION' or 'NuPoEx' will match 'NullPointerException', but not 'NoPointerException'.

Digit characters are treated in a special way. They can be used in the pattern but are not always considered as leading character. For instance, both 'UTF16DSS' and 'UTFDSS' patterns will match 'UTF16DocumentScannerSupport'.

Digit characters are treated in a special way. They can be used in the pattern but are not always considered as leading character. For instance, both 'UTF16DSS' and 'UTFDSS' patterns will match 'UTF16DocumentScannerSupport'.

Using this method allows matching names to have more parts than the specified pattern (see camelCaseMatch(String, int, int, String, int, int, boolean)).
For instance, 'HM' , 'HaMa' and 'HMap' patterns will match 'HashMap', 'HatMapper' and also 'HashMapEntry'.

Examples:

 
  pattern = "NPE"
  patternStart = 0
  patternEnd = 3
  name = NullPointerException
  nameStart = 0
  nameEnd = 20
  result => true
 
  pattern = "NPE"
  patternStart = 0
  patternEnd = 3
  name = NoPermissionException
  nameStart = 0
  nameEnd = 21
  result => true
 
  pattern = "NuPoEx"
  patternStart = 0
  patternEnd = 6
  name = NullPointerException
  nameStart = 0
  nameEnd = 20
  result => true
 
  pattern = "NuPoEx"
  patternStart = 0
  patternEnd = 6
  name = NoPermissionException
  nameStart = 0
  nameEnd = 21
  result => false
 
  pattern = "npe"
  patternStart = 0
  patternEnd = 3
  name = NullPointerException
  nameStart = 0
  nameEnd = 20
  result => false
 
  pattern = "IPL3"
  patternStart = 0
  patternEnd = 3
  name = "IPerspectiveListener3"
  nameStart = 0
  nameEnd = 21
  result => true
 
  pattern = "HM"
  patternStart = 0
  patternEnd = 2
  name = "HashMapEntry"
  nameStart = 0
  nameEnd = 12
  result => true
 
  pattern = "HMap"
  patternStart = 0
  patternEnd = 4
  name = "HatMapper"
  nameStart = 0
  nameEnd = 9
  result => true
 

Parameters:

pattern - the given pattern

patternStart - the start index of the pattern, inclusive

patternEnd - the end index of the pattern, exclusive

name - the given name

nameStart - the start index of the name, inclusive

nameEnd - the end index of the name, exclusive

Returns:

true if a sub-pattern matches the sub-part of the given name, false otherwise

Since:

3.2

camelCaseMatch

public static final boolean camelCaseMatch(java.lang.String pattern,
                                           int patternStart,
                                           int patternEnd,
                                           java.lang.String name,
                                           int nameStart,
                                           int nameEnd,
                                           boolean samePartCount)

Answers true if a sub-pattern matches the sub-part of the given name using CamelCase rules, or false otherwise. char[] CamelCase matching does NOT accept explicit wild-cards '*' and '?' and is inherently case sensitive. Can match only subset of name/pattern, considering end positions as non-inclusive. The sub-pattern is defined by the patternStart and patternEnd positions.

CamelCase denotes the convention of writing compound names without spaces, and capitalizing every term. This function recognizes both upper and lower CamelCase, depending whether the leading character is capitalized or not. The leading part of an upper CamelCase pattern is assumed to contain a sequence of capitals which are appearing in the matching name; e.g. 'NPE' will match 'NullPointerException', but not 'NewPerfData'. A lower CamelCase pattern uses a lowercase first character. In Java, type names follow the upper CamelCase convention, whereas method or field names follow the lower CamelCase convention.

The pattern may contain lowercase characters, which will be matched in a case sensitive way. These characters must appear in sequence in the name. For instance, 'NPExcep' will match 'NullPointerException', but not 'NullPointerExCEPTION' or 'NuPoEx' will match 'NullPointerException', but not 'NoPointerException'.

Digit characters are treated in a special way. They can be used in the pattern but are not always considered as leading character. For instance, both 'UTF16DSS' and 'UTFDSS' patterns will match 'UTF16DocumentScannerSupport'.

CamelCase can be restricted to match only the same count of parts. When this restriction is specified the given pattern and the given name must have exactly the same number of parts (i.e. the same number of uppercase characters).
For instance, 'HM' , 'HaMa' and 'HMap' patterns will match 'HashMap' and 'HatMapper' but not 'HashMapEntry'.

Examples:

 
  pattern = "NPE"
  patternStart = 0
  patternEnd = 3
  name = NullPointerException
  nameStart = 0
  nameEnd = 20
  result => true
 
  pattern = "NPE"
  patternStart = 0
  patternEnd = 3
  name = NoPermissionException
  nameStart = 0
  nameEnd = 21
  result => true
 
  pattern = "NuPoEx"
  patternStart = 0
  patternEnd = 6
  name = NullPointerException
  nameStart = 0
  nameEnd = 20
  result => true
 
  pattern = "NuPoEx"
  patternStart = 0
  patternEnd = 6
  name = NoPermissionException
  nameStart = 0
  nameEnd = 21
  result => false
 
  pattern = "npe"
  patternStart = 0
  patternEnd = 3
  name = NullPointerException
  nameStart = 0
  nameEnd = 20
  result => false
 
  pattern = "IPL3"
  patternStart = 0
  patternEnd = 3
  name = "IPerspectiveListener3"
  nameStart = 0
  nameEnd = 21
  result => true
 
  pattern = "HM"
  patternStart = 0
  patternEnd = 2
  name = "HashMapEntry"
  nameStart = 0
  nameEnd = 12
  result => (samePartCount == false)
 

Parameters:

pattern - the given pattern

patternStart - the start index of the pattern, inclusive

patternEnd - the end index of the pattern, exclusive

name - the given name

nameStart - the start index of the name, inclusive

nameEnd - the end index of the name, exclusive

samePartCount - flag telling whether the pattern and the name should have the same count of parts or not.
For example:

• 'HM' type string pattern will match 'HashMap' and 'HtmlMapper' types, but not 'HashMapEntry'
• 'HMap' type string pattern will still match previous 'HashMap' and 'HtmlMapper' types, but not 'HighMagnitude'

Returns:

true if a sub-pattern matches the sub-part of the given name, false otherwise

Since:

3.4

See Also:

from which algorithm implementation has been entirely copied.

createAndPattern

public static SearchPattern createAndPattern(SearchPattern leftPattern,
                                             SearchPattern rightPattern)

Deprecated. Unfortunately, this functionality is not fully supported yet (see "https://bugs.eclipse.org/bugs/show_bug.cgi?id=142044" for more details). This might be done in a further version...

Returns a search pattern that combines the given two patterns into an "and" pattern. The search result will match both the left pattern and the right pattern.

Parameters:

leftPattern - the left pattern

rightPattern - the right pattern

Returns:

an "and" pattern

createOrPattern

public static SearchPattern createOrPattern(SearchPattern leftPattern,
                                            SearchPattern rightPattern)

Returns a search pattern that combines the given two patterns into an "or" pattern. The search result will match either the left pattern or the right pattern.

Parameters:

leftPattern - the left pattern

rightPattern - the right pattern

Returns:

an "or" pattern

createPattern

public static SearchPattern createPattern(java.lang.String stringPattern,
                                          int searchFor,
                                          int limitTo,
                                          int matchRule)

Returns a search pattern based on a given string pattern. The string patterns support '*' wild-cards. The remaining parameters are used to narrow down the type of expected results.
Examples:

• search for case insensitive references to Object: createSearchPattern("Object", IJavaSearchConstants.TYPE, IJavaSearchConstants.REFERENCES, false);
• search for case sensitive references to exact Object() constructor: createSearchPattern("java.lang.Object()", IJavaSearchConstants.CONSTRUCTOR, IJavaSearchConstants.REFERENCES, true);
• search for implementers of java.lang.Runnable: createSearchPattern("java.lang.Runnable", IJavaSearchConstants.TYPE, IJavaSearchConstants.IMPLEMENTORS, true);

Parameters:

stringPattern - the given pattern

• Type patterns have the following syntax:

  [qualification '.']typeName ['<' typeArguments '>']

  Some samples:

  • java.lang.Object
  • Runnable
  • List<String>

  Type arguments can be specified to search for references to parameterized types using following syntax:

  '<' { [ '?' {'extends'|'super'} ] type ( ',' [ '?' {'extends'|'super'} ] type )* | '?' } '>'

  Note that:

  • '*' is not valid inside type arguments definition <>
  • '?' is treated as a wildcard when it is inside <> (i.e. it must be put on first position of the type argument)

• Method patterns have the following syntax:

  [declaringType '.'] ['<' typeArguments '>'] methodName ['(' parameterTypes ')'] [returnType]

  Type arguments have the same syntax as explained in the type patterns section

  Some samples:

  • java.lang.Runnable.run() void
  • main(*)
  • <String>toArray(String[])

• Constructor patterns have the following syntax:

  ['<' typeArguments '>'] [declaringQualification '.'] typeName ['(' parameterTypes ')']

  Type arguments have the same syntax as explained in the type patterns section

  Some samples:

  • java.lang.Object()
  • Test(*)
  • <Exception>Sample(Exception)

• Field patterns have the following syntax:

  [declaringType '.'] fieldName [fieldType]

  Some samples:

  • java.lang.String.serialVersionUID long
  • field*

• Package patterns have the following syntax:

  packageNameSegment {'.' packageNameSegment}

  Some samples:

  • java.lang
  • org.e*.jdt.c*e

searchFor - determines the nature of the searched elements

• IJavaSearchConstants.CLASS: only look for classes
• IJavaSearchConstants.INTERFACE: only look for interfaces
• IJavaSearchConstants.ENUM: only look for enumeration
• IJavaSearchConstants.ANNOTATION_TYPE: only look for annotation type
• IJavaSearchConstants.CLASS_AND_ENUM: only look for classes and enumerations
• IJavaSearchConstants.CLASS_AND_INTERFACE: only look for classes and interfaces
• IJavaSearchConstants.TYPE: look for all types (i.e. classes, interfaces, enum and annotation types)
• IJavaSearchConstants.FIELD: look for fields
• IJavaSearchConstants.METHOD: look for methods
• IJavaSearchConstants.CONSTRUCTOR: look for constructors
• IJavaSearchConstants.PACKAGE: look for packages

limitTo - determines the nature of the expected matches

• IJavaSearchConstants.DECLARATIONS: will search declarations matching with the corresponding element. In case the element is a method, declarations of matching methods in sub-types will also be found, allowing to find declarations of abstract methods, etc.
  Note that additional flags IJavaSearchConstants.IGNORE_DECLARING_TYPE and IJavaSearchConstants.IGNORE_RETURN_TYPE are ignored for string patterns. This is due to the fact that client may omit to define them in string pattern to have same behavior.
• IJavaSearchConstants.REFERENCES: will search references to the given element.
• IJavaSearchConstants.ALL_OCCURRENCES: will search for either declarations or references as specified above.
• IJavaSearchConstants.IMPLEMENTORS: for types, will find all types which directly implement/extend a given interface. Note that types may be only classes or only interfaces if IJavaSearchConstants.CLASS or IJavaSearchConstants.INTERFACE is respectively used instead of IJavaSearchConstants.TYPE.
• TODO (frederic) add specification for fine grain flags

matchRule - one of the following match rules

• R_EXACT_MATCH
• R_PREFIX_MATCH
• R_PATTERN_MATCH
• R_REGEXP_MATCH
• R_CAMELCASE_MATCH
• R_CAMELCASE_SAME_PART_COUNT_MATCH

, which may be also combined with one of the following flags:

• R_CASE_SENSITIVE
• R_ERASURE_MATCH
• R_EQUIVALENT_MATCH

For example,

• R_EXACT_MATCH | R_CASE_SENSITIVE: if an exact and case sensitive match is requested,
• R_PREFIX_MATCH if a case insensitive prefix match is requested
• R_EXACT_MATCH | R_ERASURE_MATCH: if a case insensitive and erasure match is requested.

Note that R_ERASURE_MATCH or R_EQUIVALENT_MATCH has no effect on non-generic types/methods search.

Note also that the default behavior for generic types/methods search is to find exact matches.

Returns:

a search pattern on the given string pattern, or null if the string pattern is ill-formed

createPattern

public static SearchPattern createPattern(IJavaElement element,
                                          int limitTo)

Returns a search pattern based on a given Java element. The pattern is used to trigger the appropriate search.
Note that for generic searches, the returned pattern consider R_ERASURE_MATCH matches. If other kind of generic matches (i.e. R_EXACT_MATCH or R_EQUIVALENT_MATCH) are expected, createPattern(IJavaElement, int, int) method need to be used instead with the explicit match rule specified.
The pattern can be parameterized as follows:

Parameters:

element - the Java element the search pattern is based on

limitTo - determines the nature of the expected matches

• IJavaSearchConstants.DECLARATIONS: will search declarations matching with the corresponding element. In case the element is a method, declarations of matching methods in sub-types will also be found, allowing to find declarations of abstract methods, etc. Some additional flags may be specified while searching declaration:
  • IJavaSearchConstants.IGNORE_DECLARING_TYPE: declaring type will be ignored during the search.
    For example using following test case:

                      class A { A method() { return null; } }
                      class B extends A { B method() { return null; } }
                      class C { A method() { return null; } }
                                            

    search for method declaration with this flag will return 2 matches: in A and in C
  • IJavaSearchConstants.IGNORE_RETURN_TYPE: return type will be ignored during the search.
    Using same example, search for method declaration with this flag will return 2 matches: in A and in B.
  Note that these two flags may be combined and both declaring and return types can be ignored during the search. Then, using same example, search for method declaration with these 2 flags will return 3 matches: in A, in B and in C
• IJavaSearchConstants.REFERENCES: will search references to the given element.
• IJavaSearchConstants.ALL_OCCURRENCES: will search for either declarations or references as specified above.
• IJavaSearchConstants.IMPLEMENTORS: for types, will find all types which directly implement/extend a given interface.
• TODO (frederic) add specification for fine grain flags

Returns:

a search pattern for a Java element or null if the given element is ill-formed

createPattern

public static SearchPattern createPattern(IJavaElement element,
                                          int limitTo,
                                          int matchRule)

Returns a search pattern based on a given Java element. The pattern is used to trigger the appropriate search, and can be parameterized as follows:

Parameters:

element - the Java element the search pattern is based on

limitTo - determines the nature of the expected matches

• IJavaSearchConstants.DECLARATIONS: will search declarations matching with the corresponding element. In case the element is a method, declarations of matching methods in sub-types will also be found, allowing to find declarations of abstract methods, etc. Some additional flags may be specified while searching declaration:
  • IJavaSearchConstants.IGNORE_DECLARING_TYPE: declaring type will be ignored during the search.
    For example using following test case:

                      class A { A method() { return null; } }
                      class B extends A { B method() { return null; } }
                      class C { A method() { return null; } }
                                            

    search for method declaration with this flag will return 2 matches: in A and in C
  • IJavaSearchConstants.IGNORE_RETURN_TYPE: return type will be ignored during the search.
    Using same example, search for method declaration with this flag will return 2 matches: in A and in B.
  Note that these two flags may be combined and both declaring and return types can be ignored during the search. Then, using same example, search for method declaration with these 2 flags will return 3 matches: in A, in B and in C
• IJavaSearchConstants.REFERENCES: will search references to the given element.
• IJavaSearchConstants.ALL_OCCURRENCES: will search for either declarations or references as specified above.
• IJavaSearchConstants.IMPLEMENTORS: for types, will find all types which directly implement/extend a given interface.
• TODO (frederic) add specification for fine grain flags

matchRule - one of the following match rules:

• R_EXACT_MATCH
• R_PREFIX_MATCH
• R_PATTERN_MATCH
• R_REGEXP_MATCH
• R_CAMELCASE_MATCH
• R_CAMELCASE_SAME_PART_COUNT_MATCH

, which may be also combined with one of the following flags:

• R_CASE_SENSITIVE
• R_ERASURE_MATCH
• R_EQUIVALENT_MATCH

For example,

• R_EXACT_MATCH | R_CASE_SENSITIVE: if an exact and case sensitive match is requested,
• R_PREFIX_MATCH if a case insensitive prefix match is requested
• R_EXACT_MATCH | R_ERASURE_MATCH: if a case insensitive and erasure match is requested.

Note that R_ERASURE_MATCH or R_EQUIVALENT_MATCH has no effect on non-generic types/methods search.

Note also that default behavior for generic types/methods search is to find exact matches.

Returns:

a search pattern for a Java element or null if the given element is ill-formed

Since:

3.1

decodeIndexKey

public void decodeIndexKey(char[] key)

Decode the given index key in this pattern. The decoded index key is used by matchesDecodedKey(SearchPattern) to find out if the corresponding index entry should be considered.

This method should be re-implemented in subclasses that need to decode an index key.

Parameters:

key - the given index key

getBlankPattern

public abstract SearchPattern getBlankPattern()

Returns a blank pattern that can be used as a record to decode an index key.

Implementors of this method should return a new search pattern that is going to be used to decode index keys.

Returns:

a new blank pattern

See Also:

decodeIndexKey(char[])

getIndexKey

public char[] getIndexKey()

Returns a key to find in relevant index categories, if null then all index entries are matched. The key will be matched according to some match rule. These potential matches will be further narrowed by the match locator, but precise match locating can be expensive, and index query should be as accurate as possible so as to eliminate obvious false hits.

This method should be re-implemented in subclasses that need to narrow down the index query.

Returns:

an index key from this pattern, or null if all index entries are matched.

getIndexCategories

public char[][] getIndexCategories()

Returns an array of index categories to consider for this index query. These potential matches will be further narrowed by the match locator, but precise match locating can be expensive, and index query should be as accurate as possible so as to eliminate obvious false hits.

This method should be re-implemented in subclasses that need to narrow down the index query.

Returns:

an array of index categories

getMatchRule

public final int getMatchRule()

Returns the rule to apply for matching index keys. Can be exact match, prefix match, pattern match or regexp match. Rule can also be combined with a case sensitivity flag.

Returns:

one of R_EXACT_MATCH, R_PREFIX_MATCH, R_PATTERN_MATCH, R_REGEXP_MATCH combined with R_CASE_SENSITIVE, e.g. R_EXACT_MATCH | R_CASE_SENSITIVE if an exact and case sensitive match is requested, or R_PREFIX_MATCH if a prefix non case sensitive match is requested. [TODO (frederic) I hope R_ERASURE_MATCH doesn't need to be on this list. Because it would be a breaking API change.]

matchesDecodedKey

public boolean matchesDecodedKey(SearchPattern decodedPattern)

Returns whether this pattern matches the given pattern (representing a decoded index key).

This method should be re-implemented in subclasses that need to narrow down the index query.

Parameters:

decodedPattern - a pattern representing a decoded index key

Returns:

whether this pattern matches the given pattern

matchesName

public boolean matchesName(char[] pattern,
                           char[] name)

Returns whether the given name matches the given pattern.

This method should be re-implemented in subclasses that need to define how a name matches a pattern.

Parameters:

pattern - the given pattern, or null to represent "*"

name - the given name

Returns:

whether the given name matches the given pattern

validateMatchRule

public static int validateMatchRule(java.lang.String stringPattern,
                                    int matchRule)

Validate compatibility between given string pattern and match rule.
In certain circumstances described in the table below, the returned match rule is modified in order to provide a more efficient search pattern:

1. when the R_REGEXP_MATCH flag is set, then the pattern is rejected as this kind of match is not supported yet and -1 is returned).
2. when the string pattern has no pattern characters (e.g. '*' or '?') and the pattern match flag is set (i.e. the match rule has the R_PATTERN_MATCH flag), then the pattern match flag is reset.
   Reversely, when the string pattern has pattern characters and the pattern match flag is not set, then the pattern match flag is set.
3. when the R_PATTERN_MATCH flag is set then, other R_PREFIX_MATCH, R_CAMELCASE_MATCH or R_CAMELCASE_SAME_PART_COUNT_MATCH flags are reset if they are tentatively combined.
4. when the R_CAMELCASE_MATCH flag is set, then other R_PREFIX_MATCH or R_CAMELCASE_SAME_PART_COUNT_MATCH flags are reset if they are tentatively combined.
   Reversely, if the string pattern cannot be a camel case pattern (i.e. contains invalid Java identifier characters or does not have at least two uppercase characters - one for method camel case patterns), then the CamelCase match flag is replaced with a prefix match flag.
5. when the R_CAMELCASE_SAME_PART_COUNT_MATCH flag is set, then (R_PREFIX_MATCH flag is reset if it's tentatively combined.
   Reversely, if the string pattern cannot be a camel case pattern (i.e. contains invalid Java identifier characters or does not have at least two uppercase characters - one for method camel case patterns), then the CamelCase part count match flag is reset.

Note: the rules are validated in the documented order. For example, it means that as soon as the string pattern contains one pattern character, the pattern match flag will be set and all other match flags reset: validation of rule 2) followed by rule 3)...

Parameters:

stringPattern - The string pattern

matchRule - The match rule

Returns:

Optimized valid match rule or -1 if an incompatibility was detected.

Since:

3.2

toString

public java.lang.String toString()

Overrides:

toString in class java.lang.Object

See Also:

Object.toString()