next up previous contents index
Next: The Header and Data Up: Internal CLASS Format Previous: Contents of one observation   Contents   Index

File organization

Data is stored in a direct access file. The record length is 512 bytes. Blocks may contain either ovservations or indexing information. Each observation begins on a logical block limit. The first block of the file contains a file description in the following format:


      length     content 

(bytes)
4 file identification code
4 next block available for writing on the file.
4 number of index entries per index extension.
4 nex: number of index extensions (max. 251)
4*nex first block number of each index extension
The list of index extension block numbers extends onto the second block of the file.

The file identification code is a 4 character code used to indicate the file type (single versus multiple) and the internal storage format of the machine which has written the data. Table [*] summarizes the possible codes. This enables the reading machine to convert the data to its internal format.


Table: CLASS file identification codes depending on the file type.
  Vax IEEE EEEI
MULTIPLE '1 ' '1A ' '1B '
SINGLE '9 ' '9A ' '9B '


Each file contains its own index. Each index entry has a length of 32 words. Thus 4 index entries fill a 512-byte block. Several (by default 64) index entries are grouped in an index extension, of by default 16 consecutive blocks. The addresses of the first block of each extension are stored in the first block. During writing, observations are written at the end of the file, and indexed in the current index extension. When the latter is full, a new index extension is created at the end of the file.

The contents of each index entry may depend on the application. for CLASS spectral line and continuum observations, it contains the following:


      name        length     content 

(words)
XBLOC 1 observation block number
XNUM 1 observation sequence number
XVER 1 observation version number
XSOURC 3 source name (12 char)
XLINE 3 telescope name (12 char)
XDOBS 1 observation date
XDRED 1 reduction date
XOFF1 1 first offset (real, radians)
XOFF2 1 second offset (real, radians)
XTYPE 1 coordinate system ('EQ'', 'GA', 'HO')
XKIND 1 Kind of observation (0: spectral, 1: continuum, )
XQUAL 1 Quality (0-9)
XSCAN 1 Scan number
XPOSA 1 Position Angle
XFRONT 2 (8 char) Front-end ID (PROPOSED)
XBACK 2 (8 char) Back-end ID (PROPOSED)
XPROC 2 (8 char) Procedure ID (PROPOSED)
XPROJ 2 (8 char) Project ID (PROPOSED)
(unused) 6 for future extensions


next up previous contents index
Next: The Header and Data Up: Internal CLASS Format Previous: Contents of one observation   Contents   Index
Gildas manager 2014-07-01