Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.stecf.org/conferences/adass/adassVII/reprints/swaded.ps.gz Äàòà èçìåíåíèÿ: Mon Jun 12 18:51:50 2006 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 03:11:18 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: ï ï ï ï ï ï ï ï ï ï ï ï ï ï ï ï ï ï ï ï ð ï ï ð ï ï ð ï ï ð ï ï ð ï ï ð ï

Astronomical Data Analysis Software and Systems VII
ASP Conference Series, Vol. 145, 1998
R. Albrecht, R. N. Hook and H. A. Bushouse, e
Ö Copyright 1998 Astronomical Society of the Pacific. All rights reserved.
ds.
HST Keyword Dictionary
Daryl A. Swade 1 , Lisa Gardner, Ed Hopkins, Tim Kimball, Karen
Lezon, Jim Rose 1 and Bernie Shiao
Space Telescope Science Institute, 3700 San Martin Drive, Baltimore,
MD 21218, Email: swade@stsci.edu
Abstract. STScI has undertaken a project to place all HST keyword
information in one source, the keyword database, and to provide a mecha­
nism for making this keyword information accessible to all HST users, the
keyword dictionary, which is a WWW interface to the keyword database.
1. Overview
The HST OPUS 2 data processing pipeline receives the telemetry stream from
the spacecraft and formats the data as either GEIS or standard FITS files. In
both cases header keywords are used to characterize the data. Along with form­
ing the headers for the science data, most keyword values are ingested into HST
archive catalog fields where they are accessible for archive searches and data
characterization. One goal of this project is to strive for unity in keyword defi­
nitions across all HST science instrument data. In order to establish standards
for HST keywords and to provide a central location for keyword information,
STScI has undertaken a project to place all keyword information in one source
­ the keyword database.
In addition, a Web­based interface to the keyword database, the keyword
dictionary 3 , has been developed as a mechanism for making this keyword infor­
mation accessible to all HST users.
2. Keyword Flow
The keyword database provides the capability to trace the flow of a keyword
value from source, through headers, to the archive catalog.
Keyword values for the HST science data are derived from three main
sources: the Proposal Management Database (PMDB), spacecraft telemetry
as defined in the Project Database (PDB), and calculations performed during
data processing. The PMDB contains information derived from the submitted
proposals and their scheduling process. The PDB contains mnemonic definitions
1 Computer Sciences Corporation
2 http://www.stsci.edu/software/OPUS/
3 http://archive.stsci.edu/keyword/
442

HST Keyword Dictionary 443
for downlinked telemetry values and aperture positions. A table in the keyword
database links a keyword to the source of its value.
Headers are constructed and the keyword values are populated in the OPUS
(science) data processing pipeline and the OMS (engineering and FGS) data pro­
cessing pipeline. All science data keywords are inserted into the headers during
a processing step called Generic Conversion. Some of the keyword values may
be left blank at this stage and populated in subsequent pipeline steps. For ex­
ample, the OPUS pipeline includes the instrument specific STSDAS calibration
step that will update some keyword values.
Along with forming the headers for the data, most keyword values are cat­
aloged into HST archive database fields where they are accessible for archive
searches and data characterization. Tables in the keyword database associate
the header keywords with the archive catalog field into which they are ingested.
3. Keyword Database
The keyword database is a relational database that contains all relevant informa­
tion about a keyword. It is the definitive source of keyword information for HST
data processing and archiving software systems. This information includes the
keyword source, datatype, default value, possible values, units, header comment,
and textual long description with detailed information. The keyword database
also contains all the necessary information on how to build headers from the
individual keywords, which archive catalog fields are populated by each key­
word, and information necessary to build DDL for StarView screens. A detailed
description of the keyword database schema is available through the keyword
dictionary Web site.
4. Information Contained in the Keyword Database
The following keyword related information is contained within the keyword
database fields.
. Keyword name ­ name of the keyword following FITS standards
. HST instrument ­ the instrument onboard HST to which the keyword
applies (the same keyword can be used for multiple instruments)
. Datatype ­ datatype of the keyword value: character, I2, I4, R4, R8, or L1
. Units ­ the units of the keyword value
. Default value ­ a default value for each keyword may be specified
. Possible values ­ a list of all possible values for a given keyword. This may
be an enumerated list in which case all values are stored in the database,
or a range in which case the minimum and maximum values are stored.
. Source ­ the source of the keyword value (PMDB, PDB, or calculated; see
above)

444 Swade et al.
. Long description ­ an arbitrarily long text field which describes any rele­
vant information about a keyword or how its value is derived
. Header structure information ­ all the information necessary to construct
each HST data header including which keywords to include and their order
. Optional flag ­ a flag to identify a keyword as optional. Optional keywords
will not appear in the header if their value is blank.
. Short comment ­ a short (<48 character) description of the keyword for
inclusion in the headers
. Archive catalog table ­ the archive catalog table to which the keyword is
ingested
. Archive catalog field ­ the field in the archive catalog table to which the
keyword is ingested
. StarView DDL ­ information necessary to construct StarView screens
5. Keyword Database Products
The following products are currently generated from the keyword database.
. ICD­19: The HST project document that defines the headers for HST
science data files. This document now resides on­line within the keyword
dictionary.
. OPUS load files: The OPUS data processing system uses database load
files generated by the keyword database to populate tables in the opera­
tional database which define the keywords contained in the data headers,
the order those keywords appear in the headers, and the source of the
keyword values.
. OMS keyword file: In a manner similar to OPUS, the OMS engineering
data processing system uses the keyword database to define the contents
and order of keywords in observation logs.
. Archive catalog verification file: The archive catalog is verified to insure
all keywords ingested into the catalog correspond to the keyword value in
the header. This file contains archive catalog keyword/fieldname mapping.
. StarView DDL files: StarView is the interface to the HST archive catalog.
DDL files provide the database attributes defined on StarView screens.
6. Keyword Database Configuration Process
A formal configuration management procedure has been established to control
modifications to the database. There are three versions of the keyword database
active at any one time. Development of new keywords and headers is performed
in keyworddev. After development the new keywords enter a test phase and

HST Keyword Dictionary 445
the contents of keyworddev are copied to keywordtst. New development can
continue in keyworddev as the previously developed keywords are tested. After
successfully completing the test phase keywordtst is copied into the operational
keyword database. In addition to the three versions of the keyword database,
the latest version of the OPUS load files derived from the keyword database are
configured in a CMS software archive.
7. Keyword Update Process
STScI has set up a Keyword Review Team (KRT) to consider all proposed
changes to HST headers. Proposals to change a keyword or any keyword at­
tribute are submitted to the KRT coordinator. Proposals have originated from a
number of sources such as science instrument teams, software developers, system
engineers, user support personnel, etc. Once the proposed changes are clearly
documented, the proposal is forwarded to the rest of the KRT for consideration
and evaluation of any impacts. When approved the change is implemented in
the development version of the keyword database, and the revised headers enter
into the formal software test cycle.
During development of headers for new science instruments placed aboard
HST during servicing missions, the process is streamlined in that entire headers
are considered by the KRT instead of individual keywords.
8. Keyword Dictionary
The HST Keyword Dictionary Web Interface is written in Perl using the Sybperl
module. For Javascript­enabled browsers, a small window can be popped up that
allows quick searches ``on the side''. Frames were used where possible in the
initial design, but they encumber the interface (and the CGI script) somewhat;
the next generation will design them out.
An administration tool was written in Tcl/Tk. (At the time, Web security
was judged to be insu#cient for an administrative interface.) The keyword
database administrator can use this tool to load new or edit existing database
records.
9. Keyword Dictionary Capabilities
The HST keyword dictionary allows for access to keyword information through
a number of di#erent approaches. It is possible to search for a specific keyword,
browse all the keywords in alphabetical order for a specific HST instrument,
or generate schematic headers for each science file generated from HST data.
All information about each keyword contained within the keyword database
is accessible through any of these routes. It is also possible to look in the
development, test, or operational versions of the keyword database.
Since the keyword dictionary accesses the same database that provides the
information and instructions on how to construct the headers in the data pro­
cessing pipeline, this documentation can never be out of date with respect to
current data processing!