class KSycocaDict

Public Methods

•  KSycocaDict ()
•  KSycocaDict (QDataStream *str, int offset)
•  ~KSycocaDict ()
• void  add (const QString &key, KSycocaEntry *payload)
• int  find_string (const QString &key )
• uint  count ()
• void  clear ()
• void  save (QDataStream &str)

Protected Methods

• Q_UINT32  hashKey ( const QString &)

Detailed Description

Hash table implementation for the sycoca database file

Create an empty dict, for building the database

Create a dict from an existing database

Adds a 'payload' to the dictionary with key 'key'.

'payload' should have a valid offset by the time the dictionary gets saved.

Looks up an entry identified by 'key'.

If 0 is returned, no matching entry exists. Otherwise, the offset of the entry is returned.

NOTE: It is not guaranteed that this entry is indeed the one you were looking for. After loading the entry you should check that it indeed matches the search key. If it doesn't then no matching entry exists.

The number of entries in the dictionary.

Only valid when building the database.

Reset the dictionary.

Only valid when building the database.

Save the dictionary to the stream A reasonable fast hash algorithm will be created.

Typically this will find 90% of the entries directly. Average hash table size: nrOfItems * 20 bytes. Average duplicate list size: nrOfItms * avgKeyLength / 5.

Unknown keys have an average 20% chance to give a false hit. (That's why your program should check the result)

Example: Assume 1000 items with an average key length of 60 bytes.

Approx. 900 items will hash directly to the right entry. Approx. 100 items require a lookup in the duplicate list.

The hash table size will be approx. 20Kb. The duplicate list size will be approx. 12Kb.

[protected]