Functions | |
int | kdbGetKey (KDB *handle, Key *dest) |
int | kdbSetKey (KDB *handle, const Key *key) |
int | kdbGetString (KDB *handle, const char *keyname, char *returned, size_t maxSize) |
int | kdbSetString (KDB *handle, const char *keyname, const char *value) |
int | kdbRemove (KDB *handle, const char *keyname) |
ssize_t | kdbGetByName (KDB *handle, KeySet *returned, const char *name, option_t options) |
To use them:
#include <kdb.h>
These methods are higher level. They use kdbOpen(), kdbClose(), kdbGet() and kdbSet() methods to do their job, and don't have to be reimplemented for a different backend.
These functions avoid limitations through not implemented capabilities. This will of course cost some effort, so read through the description carefully and decide if it is appropriate for your problem.
Binding writers don't have to implement these functions, use features of the binding language instead. But you can use these functions as ideas what high level methods may be useful.
Don't use writing single keys in a loop, prefer always writing out a keyset!
ssize_t kdbGetByName | ( | KDB * | handle, | |
KeySet * | returned, | |||
const char * | name, | |||
option_t | options | |||
) |
This method is similar kdbGet() but the path is given by a string.
When it is not possible to make a key out of that string -1 is returned .
When parentName starts with / cascading will be used and both keys from user and system will be fetched.
A typically app with about 3000 keys may have this line:
KDB *handle = kdbOpen(); KeySet *myConfig = (4096, KS_END); ssize_t ret = kdbGetByName (handle, myConfig, "/sw/app/current", 0); // check ret and work with keyset myConfig ksDel (myConfig); kdbClose (handle); *
myConfig will be loaded with keys from system/sw/app/current but also user/sw/app/current.
When one of these kdbGet() fails -1 will be returned, but the other kdbGet() will be tried too.
handle | contains internal information of opened key database | |
name | the name where to get the keys below | |
returned | the (pre-initialized) KeySet returned with all keys found | |
options | ORed options to control approaches Unlike to kdbGet() is KDB_O_POP set per default. |
returned
-1 on failure
-1 when name
is no valid key
-1 on NULL pointer
int kdbGetKey | ( | KDB * | handle, | |
Key * | dest | |||
) |
Fully retrieves the passed key
from the backend storage.
The backend will try to get the key, identified through its name.
It uses kdbGet() for retrieving the key and copies the found data to dest.
While kdbGetKey() is perfect for a simple get of a specific key, kdbGet() and kdbGetByName() gives you more control over the keyset.
handle | contains internal information of opened key database | |
dest | a pointer to a Key that has a name set |
-1 on failure
-1 on NULL pointer
commandGet() code in KDB :: Low Level Methods command for usage example
kdbGet() and kdbGetByName() to have more control over KeySet :: Class Methods and options
int kdbGetString | ( | KDB * | handle, | |
const char * | keyname, | |||
char * | returned, | |||
size_t | maxSize | |||
) |
A high-level method to get a key value, by key name.
This method gets a backend from any backend with kdbGetKey() and extracts the string and store it into returned. It only works with string keys.
This method gives you the direct relation between a keyname and the value, without any kdb specific structures. Use it when you just want some values out of the kdb namespace.
You need to know the maximum string length of the object. That could be the case when you e.g. save a path which is limited with MAX_PATH.
KDB *handle = kdbOpen(); char buffer [MAX_PATH]; if (kdbGetString(handle, "user/key/to/get/pathname", buffer, sizeof(buffer)) == -1) { // handle error cases } else { printf ("The keys value is %s\n", buffer); } kdbClose(handle);
handle | contains internal information of opened key database | |
keyname | the name of the key to receive the value | |
returned | a buffer to put the key value | |
maxSize | the size of the buffer |
-1 on failure
-1 on NULL pointers
-1 if maxSize is 0 or larger than SSIZE_MAX
kdbGetKey(), keySetKey() to work with Keys
kdbGet() and kdbGetByName() for full access to KDB Backends :: Internal Helper for Elektra datastructures
int kdbRemove | ( | KDB * | handle, | |
const char * | keyname | |||
) |
Remove a key by its name from the backend storage.
With kdbSetString() its only possible to set a key with an empty string. To really remove a key in a highlevel way you can use this method.
handle | contains internal information of opened key database | |
keyname | the name of the key to be removed |
-1 on failure
-1 on NULL pointers
commandRemove() code in KDB :: Low Level Methods command for usage example
int kdbSetKey | ( | KDB * | handle, | |
const Key * | key | |||
) |
Sets key
in the backend storage.
While kdbSetKey() is perfect for a simple get of a specific key, kdbGet() and kdbGetByName() gives you more control over the keyset.
handle | contains internal information of opened key database | |
key | Key to set |
-1 on failure
-1 on NULL pointer
kdbSet() for more control over KeySet :: Class Methods and options
commandSet() code in KDB :: Low Level Methods command for usage example
int kdbSetString | ( | KDB * | handle, | |
const char * | keyname, | |||
const char * | value | |||
) |
A high-level method to set a value to a key, by key name.
It will check if key exists first, and keep its metadata. So you'll not loose the previous key comment.
This will set a text key. So if the key was previously a binary it will be retyped as string.
handle | contains internal information of opened key database | |
keyname | the name of the key to receive the value | |
value | the value to be set |
-1 on NULL pointers
-1 on failure