Working with the Database

The Neuros has several on-disk databases that are used by both the device during normal operation and the host computer during synchronization. The primary databases are:

audio: Audio files stored on the Neuros
pcaudio: Audio files stored on the host computer
unidedhisi: HiSi clips that have not been identified yet. These clips should be fingerprinted and looked up on the HiSi server during synchronization.
idedhisi: HiSi clips that have been identified. If the fingerprint is successfully located on the HiSi server, the database record corresponding to the clip should be removed from unidedhisi and put into this database along with the metadata returned from the server.
failedhisi: HiSi clips that could not be identified. If the lookup fails, the record from the HiSi clip should be moved from the unidedhisi database to this database.

Each database can be thought as a collection of records with a fixed number (greater than or equal to 1) of fields. The first field is the primary field. Next are zero or more fields called access keys, which are fields whose contents are indexed. All the records that contain a particular value in an access key field can be quickly looked up. (Example: finding all the songs with the genre "Rock") Finally, the access keys are followed by zero or more extra info fields. These fields contain data that does not need to be indexed, like filenames or file sizes. Some fields may also contain a collection of values, called a "bag". This is used in the audio database to allow one file to be in multiple playlists at once (i.e. have multiple values in its playlist field). Note that every database is required to have a special null record.

Design

The database structure described above is implemented recursively using a root database (audio, unidedhisi, etc.) with a child database (artist, genre, etc.) for each access key. When a record is added to the root, the actual contents of each access key field are replaced with a pointer to the record in the child db containing the value in its primary field. Of course, if the value doesn't already exist in the child database, it must be added. The following diagram shows an example record from the audio database:

Notice that null values are possible for access keys; just use a pointer to the null record in the child database.

File Format

Each root database is stored in a directory whose name is the same as the database name. Inside that directory are two files holding the contents of the root database, a MDB file and a SAI file. Child databases have a MDB and SAI file as well as a PAI file used for reverse lookups. The name of each of these files will be the name of the database with either a ".mdb", ".sai", or ".pai" extension. All the files for the root and child databases are stored in the same directory. The file layout for the audio database is:

audio/albums.mdb - Album child database
audio/albums.pai
audio/albums.sai
audio/artist.mdb - Artist child database
audio/artist.pai
audio/artist.sai
audio/audio.mdb - Root database
audio/audio.sai
audio/genre.mdb - Genre child database
audio/genre.pai
audio/genre.sai
audio/playlist.mdb - Playlist child database
audio/playlist.pai
audio/playlist.sai
audio/recordings.mdb - Recordings child database
audio/recordings.pai
audio/recordings.sai

All database files are treated as a sequence of 16-bit words. A "pointer" is a word offset in a file. So a pointer with value 0 refers to bytes 0 and 1 in the file, and a pointer with value 22 refers to bytes 44 and 45. Null-terminated strings are padded with an extra null byte if they do not end on a word boundary and are terminated with a null word (0x0000). Integers (including pointers) and bitfields are always stored in big-endian format.

Standard Database Field Definitions

In the following sections, the fields for each of the standard databases are defined. The following types are used to describe the extra info fields:

sz - Null-terminated string
uint32 - Unsigned 32-bit integer

Note that the primary field is always a null-terminated string, and the access keys are pointers to records in child databases which have only a primary field (also a null-terminated string).

audio

#	Name	Type	Description
0