/*****************************************************************************\
* Copyright (c) Athena Dev Teams - Licensed under GNU GPL *
* For more information, see LICENCE in the main folder *
* *
* This file is separated in two sections: *
* (1) public typedefs, enums, unions, structures and defines *
* (2) public functions *
* *
* <B>Notes on the release system:</B> *
* Whenever an entry is removed from the database both the key and the *
* data are requested to be released. *
* At least one entry is removed when replacing an entry, removing an *
* entry, clearing the database or destroying the database. *
* What is actually released is defined by the release function, the *
* functions of the database only ask for the key and/or data to be *
* released. *
* *
* TODO: *
* - create an enum for the data (with int, unsigned int and void *) *
* - create a custom database allocator *
* - see what functions need or should be added to the database interface *
* *
* HISTORY: *
* 2007/11/09 - Added an iterator to the database.
* 2.1 (Athena build #???#) - Portability fix *
* - Fixed the portability of casting to union and added the functions *
* {@link DBMap#ensure(DBMap,DBKey,DBCreateData,...)} and *
* {@link DBMap#clear(DBMap,DBApply,...)}. *
* 2.0 (Athena build 4859) - Transition version *
* - Almost everything recoded with a strategy similar to objects, *
* database structure is maintained. *
* 1.0 (up to Athena build 4706) *
* - Previous database system. *
* *
* @version 2.1 (Athena build #???#) - Portability fix *
* @author (Athena build 4859) Flavio @ Amazon Project *
* @author (up to Athena build 4706) Athena Dev Teams *
* @encoding US-ASCII *
* @see common#db.c *
\*****************************************************************************/
#ifndef _DB_H_
#define _DB_H_
#include "../common/cbasetypes.h"
#include <stdarg.h>
/*****************************************************************************\
* (1) Section with public typedefs, enums, unions, structures and defines. *
* DB_MANUAL_CAST_TO_UNION - Define when the compiler doesn't allow casting *
* to unions. *
* DBRelease - Enumeration of release options. *
* DBType - Enumeration of database types. *
* DBOptions - Bitfield enumeration of database options. *
* DBKey - Union of used key types. *
* DBApply - Format of functions applyed to the databases. *
* DBMatcher - Format of matchers used in DBMap::getall. *
* DBComparator - Format of the comparators used by the databases. *
* DBHasher - Format of the hashers used by the databases. *
* DBReleaser - Format of the releasers used by the databases. *
* DBIterator - Database iterator. *
* DBMap - Database interface. *
\*****************************************************************************/
/**
* Define this to enable the functions that cast to unions.
* Required when the compiler doesn't support casting to unions.
* NOTE: It is recommened that the conditional tests to determine if this
* should be defined be located in the configure script or a header file
* specific for compatibility and portability issues.
* @public
* @see #db_i2key(int)
* @see #db_ui2key(unsigned int)
* @see #db_str2key(unsigned char *)
*/
//#define DB_MANUAL_CAST_TO_UNION
/**
* Bitfield with what should be released by the releaser function (if the
* function supports it).
* @public
* @see #DBReleaser
* @see #db_custom_release(DBRelease)
*/
typedef enum DBRelease {
DB_RELEASE_NOTHING = 0,
DB_RELEASE_KEY = 1,
DB_RELEASE_DATA = 2,
DB_RELEASE_BOTH = 3
} DBRelease;
/**
* Supported types of database.
* See {@link #db_fix_options(DBType,DBOptions)} for restrictions of the
* types of databases.
* @param DB_INT Uses int's for keys
* @param DB_UINT Uses unsigned int's for keys
* @param DB_STRING Uses strings for keys.
* @param DB_ISTRING Uses case insensitive strings for keys.
* @public
* @see #DBOptions
* @see #DBKey
* @see #db_fix_options(DBType,DBOptions)
* @see #db_default_cmp(DBType)
* @see #db_default_hash(DBType)
* @see #db_default_release(DBType,DBOptions)
* @see #db_alloc(const char *,int,DBType,DBOptions,unsigned short)
*/
typedef enum DBType {
DB_INT,
DB_UINT,
DB_STRING,
DB_ISTRING
} DBType;
/**
* Bitfield of options that define the behaviour of the database.
* See {@link #db_fix_options(DBType,DBOptions)} for restrictions of the
* types of databases.
* @param DB_OPT_BASE Base options: does not duplicate keys, releases nothing
* and does not allow NULL keys or NULL data.
* @param DB_OPT_DUP_KEY Duplicates the keys internally. If DB_OPT_RELEASE_KEY
* is defined, the real key is freed as soon as the entry is added.
* @param DB_OPT_RELEASE_KEY Releases the key.
* @param DB_OPT_RELEASE_DATA Releases the data whenever an entry is removed
* from the database.
* WARNING: for funtions that return the data (like DBMap::remove),
* a dangling pointer will be returned.
* @param DB_OPT_RELEASE_BOTH Releases both key and data.
* @param DB_OPT_ALLOW_NULL_KEY Allow NULL keys in the database.
* @param DB_OPT_ALLOW_NULL_DATA Allow NULL data in the database.
* @public
* @see #db_fix_options(DBType,DBOptions)
* @see #db_default_release(DBType,DBOptions)
* @see #db_alloc(const char *,int,DBType,DBOptions,unsigned short)
*/
typedef enum DBOptions {
DB_OPT_BASE = 0,
DB_OPT_DUP_KEY = 1,
DB_OPT_RELEASE_KEY = 2,
DB_OPT_RELEASE_DATA = 4,
DB_OPT_RELEASE_BOTH = 6,
DB_OPT_ALLOW_NULL_KEY = 8,
DB_OPT_ALLOW_NULL_DATA = 16,
} DBOptions;
/**
* Union of key types used by the database.
* @param i Type of key for DB_INT databases
* @param ui Type of key for DB_UINT databases
* @param str Type of key for DB_STRING and DB_ISTRING databases
* @public
* @see #DBType
* @see DBMap#get
* @see DBMap#put
* @see DBMap#remove
*/
typedef union DBKey {
int i;
unsigned int ui;
const char *str;
} DBKey;
/**
* Format of funtions that create the data for the key when the entry doesn't
* exist in the database yet.
* @param key Key of the database entry
* @param args Extra arguments of the funtion
* @return Data identified by the key to be put in the database
* @public
* @see DBMap#vensure
* @see DBMap#ensure
*/
typedef void* (*DBCreateData)(DBKey key, va_list args);
/**
* Format of functions to be applyed to an unspecified quantity of entries of
* a database.
* Any function that applies this function to the database will return the sum
* of values returned by this function.
* @param key Key of the database entry
* @param data Data of the database entry
* @param args Extra arguments of the funtion
* @return Value to be added up by the funtion that is applying this
* @public
* @see DBMap#vforeach
* @see DBMap#foreach
* @see DBMap#vdestroy
* @see DBMap#destroy
*/
typedef int (*DBApply)(DBKey key, void* data, va_list args);
/**
* Format of functions that match database entries.
* The purpose of the match depends on the function that is calling the matcher.
* Returns 0 if it is a match, another number otherwise.
* @param key Key of the database entry
* @param data Data of the database entry
* @param args Extra arguments of the function
* @return 0 if a match, another number otherwise
* @public
* @see DBMap#getall
*/
typedef int (*DBMatcher)(DBKey key, void* data, va_list args);
/**
* Format of the comparators used internally by the database system.
* Compares key1 to key2.
* <code>maxlen</code> is the maximum number of character used in DB_STRING and
* DB_ISTRING databases. If 0, the maximum number of maxlen is used (64K).
* Returns 0 is equal, negative if lower and positive is higher.
* @param key1 Key being compared
* @param key2 Key we are comparing to
* @param maxlen Maximum number of characters used in DB_STRING and DB_ISTRING
* databases.
* @return 0 if equal, negative if lower and positive if higher
* @public
* @see #db_default_cmp(DBType)
*/
typedef int (*DBComparator)(DBKey key1, DBKey key2, unsigned short maxlen);
/**
* Format of the hashers used internally by the database system.
* Creates the hash of the key.
* <code>maxlen</code> is the maximum number of character used in DB_STRING and
* DB_ISTRING databases. If 0, the maximum number of maxlen is used (64K).
* @param key Key being hashed
* @param maxlen Maximum number of characters used in DB_STRING and DB_ISTRING
* databases.
* @return Hash of the key
* @public
* @see #db_default_hash(DBType)
*/
typedef unsigned int (*DBHasher)(DBKey key, unsigned short maxlen);
/**
* Format of the releaser used by the database system.
* Releases nothing, the key, the data or both.
* All standard releasers use aFree to release.
* @param key Key of the database entry
* @param data Data of the database entry
* @param which What is being requested to be released
* @public
* @see #DBRelease
* @see #db_default_releaser(DBType,DBOptions)
* @see #db_custom_release(DBRelease)
*/
typedef void (*DBReleaser)(DBKey key, void* data, DBRelease which);
typedef struct DBIterator DBIterator;
typedef struct DBMap DBMap;
/**
* Database iterator.
* Supports forward iteration, backward iteration and removing entries from the database.
* The iterator is initially positioned before the first entry of the database.
* While the iterator exists the database is locked internally, so invoke
* {@link DBIterator#destroy} as soon as possible.
* @public
* @see #DBMap
*/
struct DBIterator
{
/**
* Fetches the first entry in the database.
* Returns the data of the entry.
* Puts the key in out_key, if out_key is not NULL.
* @param self Iterator
* @param out_key Key of the entry
* @return Data of the entry
* @protected
*/
void* (*first)(DBIterator* self, DBKey* out_key);
/**
* Fetches the last entry in the database.
* Returns the data of the entry.
* Puts the key in out_key, if out_key is not NULL.
* @param self Iterator
* @param out_key Key of the entry
* @return Data of the entry
* @protected
*/
void* (*last)(DBIterator* self, DBKey* out_key);
/**
* Fetches the next entry in the database.
* Returns the data of the entry.
* Puts the key in out_key, if out_key is not NULL.
* @param self Iterator
* @param out_key Key of the entry
* @return Data of the entry
* @protected
*/
void* (*next)(DBIterator* self, DBKey* out_key);
/**
* Fetches the previous entry in the database.
* Returns the data of the entry.
* Puts the key in out_key, if out_key is not NULL.
* @param self Iterator
* @param out_key Key of the entry
* @return Data of the entry
* @protected
*/
void* (*prev)(DBIterator* self, DBKey* out_key);
/**
* Returns true if the fetched entry exists.
* The databases entries might have NULL data, so use this to to test if
* the iterator is done.
* @param self Iterator
* @return true is the entry exists
* @protected
*/
bool (*exists)(DBIterator* self);
/**
* Removes the current entry from the database.
* NOTE: {@link DBIterator#exists} will return false until another entry
* is fethed
* Returns the data of the entry.
* @param self Iterator
* @return The data of the entry or NULL if not found
* @protected
* @see DBMap#remove
*/
void* (*remove)(DBIterator* self);
/**
* Destroys this iterator and unlocks the database.
* @param self Iterator
* @protected
*/
void (*destroy)(DBIterator* self);
};
/**
* Public interface of a database. Only contains funtions.
* All the functions take the interface as the first argument.
* @public
* @see #db_alloc(const char*,int,DBType,DBOptions,unsigned short)
*/
struct DBMap {
/**
* Returns a new iterator for this database.
* The iterator keeps the database locked until it is destroyed.
* The database will keep functioning normally but will only free internal
* memory when unlocked, so destroy the iterator as soon as possible.
* @param self Database
* @return New iterator
* @protected
*/
DBIterator* (*iterator)(DBMap* self);
/**
* Returns true if the entry exists.
* @param self Database
* @param key Key that identifies the entry
* @return true is the entry exists
* @protected
*/
bool (*exists)(DBMap* self, DBKey key);
/**
* Get the data of the entry identifid by the key.
* @param self Database
* @param key Key that identifies the entry
* @return Data of the entry or NULL if not found
* @protected
*/
void* (*get)(DBMap* self, DBKey key);
/**
* Just calls {@link DBMap#vgetall}.
* Get the data of the entries matched by <code>match</code>.
* It puts a maximum of <code>max</code> entries into <code>buf</code>.
* If <code>buf</code> is NULL, it only counts the matches.
* Returns the number of entries that matched.
* NOTE: if the value returned is greater than <code>max</code>, only the
* first <code>max</code> entries found are put into the buffer.
* @param self Database
* @param buf Buffer to put the data of the matched entries
* @param max Maximum number of data entries to be put into buf
* @param match Function that matches the database entries
* @param ... Extra arguments for match
* @return The number of entries that matched
* @protected
* @see DBMap#vgetall(DBMap*,void **,unsigned int,DBMatcher,va_list)
*/
unsigned int (*getall)(DBMap* self, void** buf, unsigned int max, DBMatcher match, ...);
/**
* Get the data of the entries matched by <code>match</code>.
* It puts a maximum of <code>max</code> entries into <code>buf</code>.
* If <code>buf</code> is NULL, it only counts the matches.
* Returns the number of entries that matched.
* NOTE: if the value returned is greater than <code>max</code>, only the
* first <code>max</code> entries found are put into the buffer.
* @param self Database
* @param buf Buffer to put the data of the matched entries
* @param max Maximum number of data entries to be put into buf
* @param match Function that matches the database entries
* @param ... Extra arguments for match
* @return The number of entries that matched
* @protected
* @see DBMap#getall(DBMap*,void **,unsigned int,DBMatcher,...)
*/
unsigned int (*vgetall)(DBMap* self, void** buf, unsigned int max, DBMatcher match, va_list args);
/**
* Just calls {@link DBMap#vensure}.
* Get the data of the entry identified by the key.
* If the entry does not exist, an entry is added with the data returned by
* <code>create</code>.
* @param self Database
* @param key Key that identifies the entry
* @param create Function used to create the data if the entry doesn't exist
* @param ... Extra arguments for create
* @return Data of the entry
* @protected
* @see DBMap#vensure(DBMap*,DBKey,DBCreateData,va_list)
*/
void* (*ensure)(DBMap* self, DBKey key, DBCreateData create, ...);
/**
* Get the data of the entry identified by the key.
* If the entry does not exist, an entry is added with the data returned by
* <code>create</code>.
* @param self Database
* @param key Key that identifies the entry
* @param create Function used to create the data if the entry doesn't exist
* @param args Extra arguments for create
* @return Data of the entry
* @protected
* @see DBMap#ensure(DBMap*,DBKey,DBCreateData,...)
*/
void* (*vensure)(DBMap* self, DBKey key, DBCreateData create, va_list args);
/**
* Put the data identified by the key in the database.
* Returns the previous data if the entry exists or NULL.
* NOTE: Uses the new key, the old one is released.
* @param self Database
* @param key Key that identifies the data
* @param data Data to be put in the database
* @return The previous data if the entry exists or NULL
* @protected
*/
void* (*put)(DBMap* self, DBKey key, void* data);
/**
* Remove an entry from the database.
* Returns the data of the entry.
* NOTE: The key (of the database) is released.
* @param self Database
* @param key Key that identifies the entry
* @return The data of the entry or NULL if not found
* @protected
*/
void* (*remove)(DBMap* self, DBKey key);
/**
* Just calls {@link DBMap#vforeach}.
* Apply <code>func</code> to every entry in the database.
* Returns the sum of values returned by func.
* @param self Database
* @param func Function to be applyed
* @param ... Extra arguments for func
* @return Sum of the values returned by func
* @protected
* @see DBMap#vforeach(DBMap*,DBApply,va_list)
*/
int (*foreach)(DBMap* self, DBApply func, ...);
/**
* Apply <code>func</code> to every entry in the database.
* Returns the sum of values returned by func.
* @param self Database
* @param func Function to be applyed
* @param args Extra arguments for func
* @return Sum of the values returned by func
* @protected
* @see DBMap#foreach(DBMap*,DBApply,...)
*/
int (*vforeach)(DBMap* self, DBApply func, va_list args);
/**
* Just calls {@link DBMap#vclear}.
* Removes all entries from the database.
* Before deleting an entry, func is applyed to it.
* Releases the key and the data.
* Returns the sum of values returned by func, if it exists.
* @param self Database
* @param func Function to be applyed to every entry before deleting
* @param ... Extra arguments for func
* @return Sum of values returned by func
* @protected
* @see DBMap#vclear(DBMap*,DBApply,va_list)
*/
int (*clear)(DBMap* self, DBApply func, ...);
/**
* Removes all entries from the database.
* Before deleting an entry, func is applyed to it.
* Releases the key and the data.
* Returns the sum of values returned by func, if it exists.
* @param self Database
* @param func Function to be applyed to every entry before deleting
* @param args Extra arguments for func
* @return Sum of values returned by func
* @protected
* @see DBMap#clear(DBMap*,DBApply,...)
*/
int (*vclear)(DBMap* self, DBApply func, va_list args);
/**
* Just calls {@link DBMap#vdestroy}.
* Finalize the database, feeing all the memory it uses.
* Before deleting an entry, func is applyed to it.
* Releases the key and the data.
* Returns the sum of values returned by func, if it exists.
* NOTE: This locks the database globally. Any attempt to insert or remove
* a database entry will give an error and be aborted (except for clearing).
* @param self Database
* @param func Function to be applyed to every entry before deleting
* @param ... Extra arguments for func
* @return Sum of values returned by func
* @protected
* @see DBMap#vdestroy(DBMap*,DBApply,va_list)
*/
int (*destroy)(DBMap* self, DBApply func, ...);
/**
* Finalize the database, feeing all the memory it uses.
* Before deleting an entry, func is applyed to it.
* Returns the sum of values returned by func, if it exists.
* NOTE: This locks the database globally. Any attempt to insert or remove
* a database entry will give an error and be aborted (except for clearing).
* @param self Database
* @param func Function to be applyed to every entry before deleting
* @param args Extra arguments for func
* @return Sum of values returned by func
* @protected
* @see DBMap#destroy(DBMap*,DBApply,...)
*/
int (*vdestroy)(DBMap* self, DBApply func, va_list args);
/**
* Return the size of the database (number of items in the database).
* @param self Database
* @return Size of the database
* @protected
*/
unsigned int (*size)(DBMap* self);
/**
* Return the type of the database.
* @param self Database
* @return Type of the database
* @protected
*/
DBType (*type)(DBMap* self);
/**
* Return the options of the database.
* @param self Database
* @return Options of the database
* @protected
*/
DBOptions (*options)(DBMap* self);
};
//For easy access to the common functions.
#ifdef DB_MANUAL_CAST_TO_UNION
# define i2key db_i2key
# define ui2key db_ui2key
# define str2key db_str2key
#else /* not DB_MANUAL_CAST_TO_UNION */
# define i2key(k) ((DBKey)(int)(k))
# define ui2key(k) ((DBKey)(unsigned int)(k))
# define str2key(k) ((DBKey)(const char *)(k))
#endif /* not DB_MANUAL_CAST_TO_UNION */
#define db_exists(db,k) ( (db)->exists((db),(k)) )
#define idb_exists(db,k) ( (db)->exists((db),i2key(k)) )
#define uidb_exists(db,k) ( (db)->exists((db),ui2key(k)) )
#define strdb_exists(db,k) ( (db)->exists((db),str2key(k)) )
#define db_get(db,k) ( (db)->get((db),(k)) )
#define idb_get(db,k) ( (db)->get((db),i2key(k)) )
#define uidb_get(db,k) ( (db)->get((db),ui2key(k)) )
#define strdb_get(db,k) ( (db)->get((db),str2key(k)) )
#define db_put(db,k,d) ( (db)->put((db),(k),(d)) )
#define idb_put(db,k,d) ( (db)->put((db),i2key(k),(d)) )
#define uidb_put(db,k,d) ( (db)->put((db),ui2key(k),(d)) )
#define strdb_put(db,k,d) ( (db)->put((db),str2key(k),(d)) )
#define db_remove(db,k) ( (db)->remove((db),(k)) )
#define idb_remove(db,k) ( (db)->remove((db),i2key(k)) )
#define uidb_remove(db,k) ( (db)->remove((db),ui2key(k)) )
#define strdb_remove(db,k) ( (db)->remove((db),str2key(k)) )
//These are discarding the possible vargs you could send to the function, so those
//that require vargs must not use these defines.
#define db_ensure(db,k,f) ( (db)->ensure((db),(k),(f)) )
#define idb_ensure(db,k,f) ( (db)->ensure((db),i2key(k),(f)) )
#define uidb_ensure(db,k,f) ( (db)->ensure((db),ui2key(k),(f)) )
#define strdb_ensure(db,k,f) ( (db)->ensure((db),str2key(k),(f)) )
// Database creation and destruction macros
#define idb_alloc(opt) db_alloc(__FILE__,__LINE__,DB_INT,(opt),sizeof(int))
#define uidb_alloc(opt) db_alloc(__FILE__,__LINE__,DB_UINT,(opt),sizeof(unsigned int))
#define strdb_alloc(opt,maxlen) db_alloc(__FILE__,__LINE__,DB_STRING,(opt),(maxlen))
#define stridb_alloc(opt,maxlen) db_alloc(__FILE__,__LINE__,DB_ISTRING,(opt),(maxlen))
#define db_destroy(db) ( (db)->destroy((db),NULL) )
// Other macros
#define db_clear(db) ( (db)->clear(db,NULL) )
#define db_iterator(db) ( (db)->iterator(db) )
#define dbi_first(dbi) ( (dbi)->first(dbi,NULL) )
#define dbi_last(dbi) ( (dbi)->last(dbi,NULL) )
#define dbi_next(dbi) ( (dbi)->next(dbi,NULL) )
#define dbi_prev(dbi) ( (dbi)->prev(dbi,NULL) )
#define dbi_exists(dbi) ( (dbi)->exists(dbi) )
#define dbi_destroy(dbi) ( (dbi)->destroy(dbi) )
/*****************************************************************************\
* (2) Section with public functions. *
* db_fix_options - Fix the options for a type of database. *
* db_default_cmp - Get the default comparator for a type of database. *
* db_default_hash - Get the default hasher for a type of database. *
* db_default_release - Get the default releaser for a type of database *
* with the fixed options. *
* db_custom_release - Get the releaser that behaves as specified. *
* db_alloc - Allocate a new database. *
* db_i2key - Manual cast from 'int' to 'DBKey'. *
* db_ui2key - Manual cast from 'unsigned int' to 'DBKey'. *
* db_str2key - Manual cast from 'unsigned char *' to 'DBKey'. *
* db_init - Initialise the database system. *
* db_final - Finalise the database system. *
\*****************************************************************************/
/**
* Returns the fixed options according to the database type.
* Sets required options and unsets unsupported options.
* For numeric databases DB_OPT_DUP_KEY and DB_OPT_RELEASE_KEY are unset.
* @param type Type of the database
* @param options Original options of the database
* @return Fixed options of the database
* @private
* @see #DBType
* @see #DBOptions
* @see #db_default_release(DBType,DBOptions)
*/
DBOptions db_fix_options(DBType type, DBOptions options);
/**
* Returns the default comparator for the type of database.
* @param type Type of database
* @return Comparator for the type of database or NULL if unknown database
* @public
* @see #DBType
* @see #DBComparator
*/
DBComparator db_default_cmp(DBType type);
/**
* Returns the default hasher for the specified type of database.
* @param type Type of database
* @return Hasher of the type of database or NULL if unknown database
* @public
* @see #DBType
* @see #DBHasher
*/
DBHasher db_default_hash(DBType type);
/**
* Returns the default releaser for the specified type of database with the
* specified options.
* NOTE: the options are fixed by {@link #db_fix_options(DBType,DBOptions)}
* before choosing the releaser
* @param type Type of database
* @param options Options of the database
* @return Default releaser for the type of database with the fixed options
* @public
* @see #DBType
* @see #DBOptions
* @see #DBReleaser
* @see #db_fix_options(DBType,DBOptions)
* @see #db_custom_release(DBRelease)
*/
DBReleaser db_default_release(DBType type, DBOptions options);
/**
* Returns the releaser that behaves as <code>which</code> specifies.
* @param which Defines what the releaser releases
* @return Releaser for the specified release options
* @public
* @see #DBRelease
* @see #DBReleaser
* @see #db_default_release(DBType,DBOptions)
*/
DBReleaser db_custom_release(DBRelease which);
/**
* Allocate a new database of the specified type.
* It uses the default comparator, hasher and releaser of the specified
* database type and fixed options.
* NOTE: the options are fixed by {@link #db_fix_options(DBType,DBOptions)}
* before creating the database.
* @param file File where the database is being allocated
* @param line Line of the file where the database is being allocated
* @param type Type of database
* @param options Options of the database
* @param maxlen Maximum length of the string to be used as key in string
* databases
* @return The interface of the database
* @public
* @see #DBType
* @see #DBMap
* @see #db_default_cmp(DBType)
* @see #db_default_hash(DBType)
* @see #db_default_release(DBType,DBOptions)
* @see #db_fix_options(DBType,DBOptions)
*/
DBMap* db_alloc(const char *file, int line, DBType type, DBOptions options, unsigned short maxlen);
#ifdef DB_MANUAL_CAST_TO_UNION
/**
* Manual cast from 'int' to the union DBKey.
* Created for compilers that don't support casting to unions.
* @param key Key to be casted
* @return The key as a DBKey union
* @public
* @see #DB_MANUAL_CAST_TO_UNION
*/
DBKey db_i2key(int key);
/**
* Manual cast from 'unsigned int' to the union DBKey.
* Created for compilers that don't support casting to unions.
* @param key Key to be casted
* @return The key as a DBKey union
* @public
* @see #DB_MANUAL_CAST_TO_UNION
*/
DBKey db_ui2key(unsigned int key);
/**
* Manual cast from 'unsigned char *' to the union DBKey.
* Created for compilers that don't support casting to unions.
* @param key Key to be casted
* @return The key as a DBKey union
* @public
* @see #DB_MANUAL_CAST_TO_UNION
*/
DBKey db_str2key(const char *key);
#endif /* DB_MANUAL_CAST_TO_UNION */
/**
* Initialize the database system.
* @public
* @see #db_final(void)
*/
void db_init(void);
/**
* Finalize the database system.
* Frees the memory used by the block reusage system.
* @public
* @see #db_init(void)
*/
void db_final(void);
// Link DB System - From jAthena
struct linkdb_node {
struct linkdb_node *next;
struct linkdb_node *prev;
void *key;
void *data;
};
typedef void (*LinkDBFunc)(void* key, void* data, va_list args);
void linkdb_insert ( struct linkdb_node** head, void *key, void* data); // �d�����l�����Ȃ�
void linkdb_replace( struct linkdb_node** head, void *key, void* data); // �d�����l������
void* linkdb_search ( struct linkdb_node** head, void *key);
void* linkdb_erase ( struct linkdb_node** head, void *key);
void linkdb_final ( struct linkdb_node** head );
void linkdb_foreach( struct linkdb_node** head, LinkDBFunc func, ... );
/// Finds an entry in an array.
/// ex: ARR_FIND(0, size, i, list[i] == target);
///
/// @param __start Starting index (ex: 0)
/// @param __end End index (ex: size of the array)
/// @param __var Index variable
/// @param __cmp Expression that returns true when the target entry is found
#define ARR_FIND(__start, __end, __var, __cmp) \
do{ \
for( (__var) = (__start); (__var) < (__end); ++(__var) ) \
if( __cmp ) \
break; \
}while(0)
/// Moves an entry of the array.
/// Use ARR_MOVERIGHT/ARR_MOVELEFT if __from and __to are direct numbers.
/// ex: ARR_MOVE(i, 0, list, int);// move index i to index 0
///
///
/// @param __from Initial index of the entry
/// @param __to Target index of the entry
/// @param __arr Array
/// @param __type Type of entry
#define ARR_MOVE(__from, __to, __arr, __type) \
do{ \
if( (__from) != (__to) ) \
{ \
__type __backup__; \
memmove(&__backup__, (__arr)+(__from), sizeof(__type)); \
if( (__from) < (__to) ) \
memmove((__arr)+(__from), (__arr)+(__from)+1, ((__to)-(__from))*sizeof(__type)); \
else if( (__from) > (__to) ) \
memmove((__arr)+(__to)+1, (__arr)+(__to), ((__from)-(__to))*sizeof(__type)); \
memmove((__arr)+(__to), &__backup__, sizeof(__type)); \
} \
}while(0)
/// Moves an entry of the array to the right.
/// ex: ARR_MOVERIGHT(1, 4, list, int);// move index 1 to index 4
///
/// @param __from Initial index of the entry
/// @param __to Target index of the entry
/// @param __arr Array
/// @param __type Type of entry
#define ARR_MOVERIGHT(__from, __to, __arr, __type) \
do{ \
__type __backup__; \
memmove(&__backup__, (__arr)+(__from), sizeof(__type)); \
memmove((__arr)+(__from), (__arr)+(__from)+1, ((__to)-(__from))*sizeof(__type)); \
memmove((__arr)+(__to), &__backup__, sizeof(__type)); \
}while(0)
/// Moves an entry of the array to the left.
/// ex: ARR_MOVELEFT(3, 0, list, int);// move index 3 to index 0
///
/// @param __from Initial index of the entry
/// @param __end Target index of the entry
/// @param __arr Array
/// @param __type Type of entry
#define ARR_MOVELEFT(__from, __to, __arr, __type) \
do{ \
__type __backup__; \
memmove(&__backup__, (__arr)+(__from), sizeof(__type)); \
memmove((__arr)+(__to)+1, (__arr)+(__to), ((__from)-(__to))*sizeof(__type)); \
memmove((__arr)+(__to), &__backup__, sizeof(__type)); \
}while(0)
/////////////////////////////////////////////////////////////////////
// Vector library based on defines. (dynamic array)
// uses aMalloc, aRealloc, aFree
/// Declares an anonymous vector struct.
///
/// @param __type Type of data
#define VECTOR_DECL(__type) \
struct { \
size_t _max_; \
size_t _len_; \
__type* _data_; \
}
/// Declares a named vector struct.
///
/// @param __name Structure name
/// @param __type Type of data
#define VECTOR_STRUCT_DECL(__name,__type) \
struct __name { \
size_t _max_; \
size_t _len_; \
__type* _data_; \
}
/// Declares and initializes an anonymous vector variable.
///
/// @param __type Type of data
/// @param __var Variable name
#define VECTOR_VAR(__type,__var) \
VECTOR_DECL(__type) __var = {0,0,NULL}
/// Declares and initializes a named vector variable.
///
/// @param __name Structure name
/// @param __var Variable name
#define VECTOR_STRUCT_VAR(__name,__var) \
struct __name __var = {0,0,NULL}
/// Initializes a vector.
///
/// @param __vec Vector
#define VECTOR_INIT(__vec) \
memset(&(__vec), 0, sizeof(__vec))
/// Returns the internal array of values.
///
/// @param __vec Vector
/// @return Array of values
#define VECTOR_DATA(__vec) \
( (__vec)._data_ )
/// Returns the length of the vector.
///
/// @param __vec Vector
/// @return Length
#define VECTOR_LENGTH(__vec) \
( (__vec)._len_ )
/// Returns the capacity of the vector.
///
/// @param __vec Vector
/// @return Capacity
#define VECTOR_CAPACITY(__vec) \
( (__vec)._max_ )
/// Returns the value at the target index.
/// Assumes the index exists.
///
/// @param __vec Vector
/// @param __idx Index
/// @return Value
#define VECTOR_INDEX(__vec,__idx) \
( VECTOR_DATA(__vec)[__idx] )
/// Returns the first value of the vector.
/// Assumes the array is not empty.
///
/// @param __vec Vector
/// @return First value
#define VECTOR_FIRST(__vec) \
( VECTOR_INDEX(__vec,0) )
/// Returns the last value of the vector.
/// Assumes the array is not empty.
///
/// @param __vec Vector
/// @return Last value
#define VECTOR_LAST(__vec) \
( VECTOR_INDEX(__vec,VECTOR_LENGTH(__vec)-1) )
/// Resizes the vector.
/// Excess values are discarded, new positions are zeroed.
///
/// @param __vec Vector
/// @param __n Size
#define VECTOR_RESIZE(__vec,__n) \
do{ \
if( (__n) > VECTOR_CAPACITY(__vec) ) \
{ /* increase size */ \
if( VECTOR_CAPACITY(__vec) == 0 ) VECTOR_DATA(__vec) = aMalloc((__n)*sizeof(VECTOR_FIRST(__vec))); /* allocate new */ \
else VECTOR_DATA(__vec) = aRealloc(VECTOR_DATA(__vec),(__n)*sizeof(VECTOR_FIRST(__vec))); /* reallocate */ \
memset(VECTOR_DATA(__vec)+VECTOR_LENGTH(__vec), 0, (VECTOR_CAPACITY(__vec)-VECTOR_LENGTH(__vec))*sizeof(VECTOR_FIRST(__vec))); /* clear new data */ \
VECTOR_CAPACITY(__vec) = (__n); /* update capacity */ \
} \
else if( (__n) == 0 && VECTOR_CAPACITY(__vec) ) \
{ /* clear vector */ \
aFree(VECTOR_DATA(__vec)); VECTOR_DATA(__vec) = NULL; /* free data */ \
VECTOR_CAPACITY(__vec) = 0; /* clear capacity */ \
VECTOR_LENGTH(__vec) = 0; /* clear length */ \
} \
else if( (__n) < VECTOR_CAPACITY(__vec) ) \
{ /* reduce size */ \
VECTOR_DATA(__vec) = aRealloc(VECTOR_DATA(__vec),(__n)*sizeof(VECTOR_FIRST(__vec))); /* reallocate */ \
VECTOR_CAPACITY(__vec) = (__n); /* update capacity */ \
if( VECTOR_LENGTH(__vec) > (__n) ) VECTOR_LENGTH(__vec) = (__n); /* update length */ \
} \
}while(0)
/// Ensures that the array has the target number of empty positions.
/// Increases the capacity in multiples of __step.
///
/// @param __vec Vector
/// @param __n Empty positions
/// @param __step Increase
#define VECTOR_ENSURE(__vec,__n,__step) \
do{ \
size_t _empty_ = VECTOR_CAPACITY(__vec)-VECTOR_LENGTH(__vec); \
while( (__n) > _empty_ ) _empty_ += (__step); \
if( _empty_ != VECTOR_CAPACITY(__vec)-VECTOR_LENGTH(__vec) ) VECTOR_RESIZE(__vec,_empty_+VECTOR_LENGTH(__vec)); \
}while(0)
/// Inserts a zeroed value in the target index.
/// Assumes the index is valid and there is enough capacity.
///
/// @param __vec Vector
/// @param __idx Index
#define VECTOR_INSERTZEROED(__vec,__idx) \
do{ \
if( (__idx) < VECTOR_LENGTH(__vec) ) /* move data */ \
memmove(&VECTOR_INDEX(__vec,(__idx)+1),&VECTOR_INDEX(__vec,__idx),(VECTOR_LENGTH(__vec)-(__idx))*sizeof(VECTOR_FIRST(__vec))); \
memset(&VECTOR_INDEX(__vec,__idx), 0, sizeof(VECTOR_INDEX(__vec,__idx))); /* set zeroed value */ \
++VECTOR_LENGTH(__vec); /* increase length */ \
}while(0)
/// Inserts a value in the target index. (using the '=' operator)
/// Assumes the index is valid and there is enough capacity.
///
/// @param __vec Vector
/// @param __idx Index
/// @param __val Value
#define VECTOR_INSERT(__vec,__idx,__val) \
do{ \
if( (__idx) < VECTOR_LENGTH(__vec) ) /* move data */ \
memmove(&VECTOR_INDEX(__vec,(__idx)+1),&VECTOR_INDEX(__vec,__idx),(VECTOR_LENGTH(__vec)-(__idx))*sizeof(VECTOR_FIRST(__vec))); \
VECTOR_INDEX(__vec,__idx) = (__val); /* set value */ \
++VECTOR_LENGTH(__vec); /* increase length */ \
}while(0)
/// Inserts a value in the target index. (using memcpy)
/// Assumes the index is valid and there is enough capacity.
///
/// @param __vec Vector
/// @param __idx Index
/// @param __val Value
#define VECTOR_INSERTCOPY(__vec,__idx,__val) \
VECTOR_INSERTARRAY(__vec,__idx,&(__val),1)
/// Inserts the values of the array in the target index. (using memcpy)
/// Assumes the index is valid and there is enough capacity.
///
/// @param __vec Vector
/// @param __idx Index
/// @param __pval Array of values
/// @param __n Number of values
#define VECTOR_INSERTARRAY(__vec,__idx,__pval,__n) \
do{ \
if( (__idx) < VECTOR_LENGTH(__vec) ) /* move data */ \
memmove(&VECTOR_INDEX(__vec,(__idx)+(__n)),&VECTOR_INDEX(__vec,__idx),(VECTOR_LENGTH(__vec)-(__idx))*sizeof(VECTOR_FIRST(__vec))); \
memcpy(&VECTOR_INDEX(__vec,__idx), (__pval), (__n)*sizeof(VECTOR_FIRST(__vec))); /* set values */ \
VECTOR_LENGTH(__vec) += (__n); /* increase length */ \
}while(0)
/// Inserts a zeroed value in the end of the vector.
/// Assumes there is enough capacity.
///
/// @param __vec Vector
#define VECTOR_PUSHZEROED(__vec) \
do{ \
memset(&VECTOR_INDEX(__vec,VECTOR_LENGTH(__vec)), 0, sizeof(VECTOR_INDEX(__vec,VECTOR_LENGTH(__vec)))); /* set zeroed value */ \
++VECTOR_LENGTH(__vec); /* increase length */ \
}while(0)
/// Inserts a value in the end of the vector. (using the '=' operator)
/// Assumes there is enough capacity.
///
/// @param __vec Vector
/// @param __val Value
#define VECTOR_PUSH(__vec,__val) \
do{ \
VECTOR_INDEX(__vec,VECTOR_LENGTH(__vec)) = (__val); /* set value */ \
++VECTOR_LENGTH(__vec); /* increase length */ \
}while(0)
/// Inserts a value in the end of the vector. (using memcpy)
/// Assumes there is enough capacity.
///
/// @param __vec Vector
/// @param __val Value
#define VECTOR_PUSHCOPY(__vec,__val) \
VECTOR_PUSHARRAY(__vec,&(__val),1)
/// Inserts the values of the array in the end of the vector. (using memcpy)
/// Assumes there is enough capacity.
///
/// @param __vec Vector
/// @param __pval Array of values
/// @param __n Number of values
#define VECTOR_PUSHARRAY(__vec,__pval,__n) \
do{ \
memcpy(&VECTOR_INDEX(__vec,VECTOR_LENGTH(__vec)), (__pval), (__n)*sizeof(VECTOR_FIRST(__vec))); /* set values */ \
VECTOR_LENGTH(__vec) += (__n); /* increase length */ \
}while(0)
/// Removes and returns the last value of the vector.
/// Assumes the array is not empty.
///
/// @param __vec Vector
/// @return Removed value
#define VECTOR_POP(__vec) \
( VECTOR_INDEX(__vec,--VECTOR_LENGTH(__vec)) )
/// Removes the last N values of the vector and returns the value of the last pop.
/// Assumes there are enough values.
///
/// @param __vec Vector
/// @param __n Number of pops
/// @return Last removed value
#define VECTOR_POPN(__vec,__n) \
( VECTOR_INDEX(__vec,(VECTOR_LENGTH(__vec)-=(__n))) )
/// Removes the target index from the vector.
/// Assumes the index is valid and there are enough values.
///
/// @param __vec Vector
/// @param __idx Index
#define VECTOR_ERASE(__vec,__idx) \
VECTOR_ERASEN(__vec,__idx,1)
/// Removes N values from the target index of the vector.
/// Assumes the index is valid and there are enough values.
///
/// @param __vec Vector
/// @param __idx Index
/// @param __n Number of values
#define VECTOR_ERASEN(__vec,__idx,__n) \
do{ \
if( (__idx) < VECTOR_LENGTH(__vec)-(__n) ) /* move data */ \
memmove(&VECTOR_INDEX(__vec,__idx),&VECTOR_INDEX(__vec,(__idx)+(__n)),(VECTOR_LENGTH(__vec)-((__idx)+(__n)))*sizeof(VECTOR_FIRST(__vec))); \
VECTOR_LENGTH(__vec) -= (__n); /* decrease length */ \
}while(0)
/// Clears the vector, freeing allocated data.
///
/// @param __vec Vector
#define VECTOR_CLEAR(__vec) \
do{ \
if( VECTOR_CAPACITY(__vec) ) \
{ \
aFree(VECTOR_DATA(__vec)); VECTOR_DATA(__vec) = NULL; /* clear allocated array */ \
VECTOR_CAPACITY(__vec) = 0; /* clear capacity */ \
VECTOR_LENGTH(__vec) = 0; /* clear length */ \
} \
}while(0)
/////////////////////////////////////////////////////////////////////
// Binary heap library based on defines. (uses the vector defines above)
// uses aMalloc, aRealloc, aFree
/// Declares an anonymous binary heap struct.
///
/// @param __type Type of data
#define BHEAP_DECL(__type) VECTOR_DECL(__type)
/// Declares a named binary heap struct.
///
/// @param __name Structure name
/// @param __type Type of data
#define BHEAP_STRUCT_DECL(__name,__type) VECTOR_STRUCT_DECL(__name,__type)
/// Declares and initializes an anonymous binary heap variable.
///
/// @param __type Type of data
/// @param __var Variable name
#define BHEAP_VAR(__type,__var) VECTOR_VAR(__type,__var)
/// Declares and initializes a named binary heap variable.
///
/// @param __name Structure name
/// @param __var Variable name
#define BHEAP_STRUCT_VAR(__name,__var) VECTOR_STRUCT_VAR(__name,__var)
/// Initializes a heap.
///
/// @param __heap Binary heap
#define BHEAP_INIT(__heap) VECTOR_INIT(__heap)
/// Returns the internal array of values.
///
/// @param __heap Binary heap
/// @return Array of values
#define BHEAP_DATA(__heap) VECTOR_DATA(__heap)
/// Returns the length of the heap.
///
/// @param __heap Binary heap
/// @return Length
#define BHEAP_LENGTH(__heap) VECTOR_LENGTH(__heap)
/// Returns the capacity of the heap.
///
/// @param __heap Binary heap
/// @return Capacity
#define BHEAP_CAPACITY(__heap) VECTOR_CAPACITY(__heap)
/// Ensures that the heap has the target number of empty positions.
/// Increases the capacity in multiples of __step.
///
/// @param __heap Binary heap
/// @param __n Empty positions
/// @param __step Increase
#define BHEAP_ENSURE(__heap,__n,__step) VECTOR_ENSURE(__heap,__n,__step)
/// Returns the top value of the heap.
/// Assumes the heap is not empty.
///
/// @param __heap Binary heap
/// @return Value at the top
#define BHEAP_PEEK(__heap) VECTOR_INDEX(__heap,0)
/// Inserts a value in the heap. (using the '=' operator)
/// Assumes there is enough capacity.
///
/// The comparator takes two values as arguments, returns:
/// - negative if the first value is on the top
/// - positive if the second value is on the top
/// - 0 if they are equal
///
/// @param __heap Binary heap
/// @param __val Value
/// @param __topcmp Comparator
#define BHEAP_PUSH(__heap,__val,__topcmp) \
do{ \
size_t _i_ = VECTOR_LENGTH(__heap); \
VECTOR_PUSH(__heap,__val); /* insert at end */ \
while( _i_ ) \
{ /* restore heap property in parents */ \
size_t _parent_ = (_i_-1)/2; \
if( __topcmp(VECTOR_INDEX(__heap,_parent_),VECTOR_INDEX(__heap,_i_)) < 0 ) \
break; /* done */ \
swap(VECTOR_INDEX(__heap,_parent_),VECTOR_INDEX(__heap,_i_)); \
_i_ = _parent_; \
} \
}while(0)
/// Removes the top value of the heap. (using the '=' operator)
/// Assumes the heap is not empty.
///
/// The comparator takes two values as arguments, returns:
/// - negative if the first value is on the top
/// - positive if the second value is on the top
/// - 0 if they are equal
///
/// @param __heap Binary heap
/// @param __topcmp Comparator
#define BHEAP_POP(__heap,__topcmp) BHEAP_POPINDEX(__heap,0,__topcmp)
/// Removes the target value of the heap. (using the '=' operator)
/// Assumes the index exists.
///
/// The comparator takes two values as arguments, returns:
/// - negative if the first value is on the top
/// - positive if the second value is on the top
/// - 0 if they are equal
///
/// @param __heap Binary heap
/// @param __idx Index
/// @param __topcmp Comparator
#define BHEAP_POPINDEX(__heap,__idx,__topcmp) \
do{ \
size_t _i_ = __idx; \
VECTOR_INDEX(__heap,__idx) = VECTOR_POP(__heap); /* put last at index */ \
while( _i_ ) \
{ /* restore heap property in parents */ \
size_t _parent_ = (_i_-1)/2; \
if( __topcmp(VECTOR_INDEX(__heap,_parent_),VECTOR_INDEX(__heap,_i_)) < 0 ) \
break; /* done */ \
swap(VECTOR_INDEX(__heap,_parent_),VECTOR_INDEX(__heap,_i_)); \
_i_ = _parent_; \
} \
while( _i_ < VECTOR_LENGTH(__heap) ) \
{ /* restore heap property in childs */ \
size_t _lchild_ = _i_*2 + 1; \
size_t _rchild_ = _i_*2 + 2; \
if( (_lchild_ >= VECTOR_LENGTH(__heap) || __topcmp(VECTOR_INDEX(__heap,_i_),VECTOR_INDEX(__heap,_lchild_)) <= 0) && \
(_rchild_ >= VECTOR_LENGTH(__heap) || __topcmp(VECTOR_INDEX(__heap,_i_),VECTOR_INDEX(__heap,_rchild_)) <= 0) ) \
break; /* done */ \
else if( _rchild_ >= VECTOR_LENGTH(__heap) || __topcmp(VECTOR_INDEX(__heap,_lchild_),VECTOR_INDEX(__heap,_rchild_)) <= 0 ) \
{ /* left child */ \
swap(VECTOR_INDEX(__heap,_i_),VECTOR_INDEX(__heap,_lchild_)); \
_i_ = _lchild_; \
} \
else \
{ /* right child */ \
swap(VECTOR_INDEX(__heap,_i_),VECTOR_INDEX(__heap,_rchild_)); \
_i_ = _rchild_; \
} \
} \
}while(0)
/// Clears the binary heap, freeing allocated data.
///
/// @param __heap Binary heap
#define BHEAP_CLEAR(__heap) VECTOR_CLEAR(__heap)
/// Generic comparator for a min-heap. (minimum value at top)
/// Returns -1 if v1 is smaller, 1 if v2 is smaller, 0 if equal.
///
/// @param v1 First value
/// @param v2 Second value
/// @return negative if v1 is top, positive if v2 is top, 0 if equal
#define BHEAP_MINTOPCMP(v1,v2) ( v1 == v2 ? 0 : v1 < v2 ? -1 : 1 )
/// Generic comparator for a max-heap. (maximum value at top)
/// Returns -1 if v1 is bigger, 1 if v2 is bigger, 0 if equal.
///
/// @param v1 First value
/// @param v2 Second value
/// @return negative if v1 is top, positive if v2 is top, 0 if equal
#define BHEAP_MAXTOPCMP(v1,v2) ( v1 == v2 ? 0 : v1 > v2 ? -1 : 1 )
#endif /* _DB_H_ */