home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Share Gallery 1
/
share_gal_1.zip
/
share_gal_1
/
GE
/
GE003C.ZIP
/
SECTION4.DOC
< prev
next >
Wrap
Text File
|
1990-07-08
|
10KB
|
184 lines
IV. Family History System Files and Datasets
The following discussion is provided for the interested user, but an
understanding of it is not necessary to make use of the system.
A. FHS Family File
The file used in the Family History System to store information about
individuals and their family relationships consists of three datasets
containing variable format `records' which are `linked' together into
a `hierarchic' logical file structure. Pictorially, a portion of the
logical file structure is represented by the diagram appearing in
Appendix E, at the end of this document.
In this diagram, each of the `boxes' represents a record type in one
of the system's datasets. In some cases, (for example child records,
marriage records, address records) a `logical record' may include
multiple records of information of that type. The number of multiple
records is essentially unlimited...there can be no more than 9999
occurrences of the name records or 32000 occurrences of the other
types of records. Records containing information of the various types
are `logically connected' by a system of program maintained `pointer'
fields. Though this structure may appear a bit more complex than the
more common `flat' file which has a single fixed record type, the
advantage in using the hierarchic structure is that there is no need
to `reserve' space in a dataset (other than a 2 character pointer
field) for types of information which may or may not be recorded for
an individual at some time in the future.
The only information that the user needs to provide to the system to
begin retrieving information in one of these extended logical records
is that required for locating the name record of the individual about
whom the information has been recorded. The basic `record key' that
the system requires is the `ID number' for the individual. This is a
number which is assigned to each individual by the system at the time
his/her Name is initially entered into the system. These numbers are
assigned sequentially (ID #216 is assigned to the 216th person entered
into the system). The ID numbers assigned two individuals need not
indicate any blood relationship between the individuals. These ID
numbers may (optionally) appear in reports produced by the system and
may also be determined by `searching' the files for other information
(such as surname, given name or birth date) using the search option of
the file maintenance program.
While the ID numbers of 2 individuals need not indicate any blood
relationship between them, these numbers are used to define the only
type of relationship that you are required to establish between indi-
viduals. The relationship of father or mother to a child is specified
by recording the ID numbers of the parents in the Name record for the
child. All other blood relationships between individuals are deter-
mined from the complete system of parent-child relationships recorded
in this way.
Perhaps it should be noted that the system provides for recording
additional information (e.g. educational, occupational, military and
27
health information) in record types not shown in the structure
described above. Further, address and comment records may also be
created under each of these record types.
It should also be noted that the system provides for extensive `com-
menting' with the comments or remarks being logically connected to the
appropriate record type. Care should be taken in the use of this
feature of the system because of the limited number of records in the
"...OTH" dataset of miscellaneous info. It is not intended for
writing a person's life history or for extended anecdotes, but rather
for recording notes concerning sources of, or explanations for, infor-
mation found in the record.
In the cases where multiple record types are permitted within a
logical record, the records are automatically maintained in some
`natural' order. For all but comment records, this is based upon a
date stored in the record. Children are ordered by increasing birth-
date; marriage records by decreasing date of marriage, and address
records by decreasing beginning date of residency. This has the
unfortunate consequence that unknown birth, marriage or beginning
dates (recorded as 00/00/0000) may result in information appearing out
of chronological sequence in the file and in reports. The order is
automatically adjusted as the correct information is determined and
recorded. You may consider entering a `best guess' for unknown dates,
making note of the fact in comments for the record type.
This discussion of the Family History System data files will conclude
with a brief description of the datasets which comprise the file.
There are three such datasets:
1. Name dataset - each record is 100 characters long;
information stored includes surname, given name, date and
time of birth and death, sex, and father and mother ID
numbers. Up to 3500 name records may be stored on a single
DS/DD diskette, or 9999 records may be stored on a hard disk.
2. Address dataset - each record is 108 characters long;
information stored includes: beginning and ending dates, 2
address lines, city, state, zip or country, and (numeric)
phone number. Up to 3500 address records may be stored on
a single DS/DD diskette, or up to 32000 may be stored on a
hard disk.
3. Miscellaneous INFO dataset - each record in this dataset is
50 characters long. Record types include:
a. spouse, family or marriage record;
b. Birth/Marriage/Death location record;
c. Comment record (1 record for each line of comments);
d. Occupation, Education, Military and Health records.
Up to 7000 records may be stored on a DS/DD diskette; up to
32000 may be stored on a hard disk.
28
B. Report Program WORK FILES
There are several types of "work" datasets that are used in the Family
History System. These are semi-temporary files that are used to
store information derived from certain processes so that those some-
times lengthy processes do not have to be repeated too frequently.
Among these work datasets are the RELATIONSHIP Work Files, the SELECT
Work file and the INDEX File.
A RELATIONSHIP WORK FILE must be created prior to producing the
ancestor, descendant or relative reports or the ancestor charts.
Although there appear to be three types of relationship files:
Ancestor, Descendant and Relative, in fact these all share a common
format...the only differences being in the number of generations of
ancestors or descendants that are searched in order to build the work
file. In building an Ancestor work file, no search is performed for
descendants at all. In building a Descendant work file, no search is
performed for ancestors. While in creating a Relative work file, the
program first searches for all ancestors and then all descendants of
the ancestors. Actually, you may specify the maximum number of
generations (of ancestors and descendants) that are to be included in
the work dataset. This is done to permit you to exercise some control
over the extent to which lines of ancestry or descendancy will be
followed in producing the report. A relationship work file may be
reused (without recreating it) if no new or changed relationships have
been recorded in the family file that was used to create the workfile.
Another type of work file is the SELECT file used by the Family Group
Report program and the export/import program. (Among the FHS
extensions, this work file is also used in the Search/Select/LIST
program and the Summary report program.) This file is essentially an
ID sequenced "check list" of individuals that have been chosen for
processing. It permits the system to isolate the sometimes complex
process of selecting ID #'s within certain programs but to pass the
results of the selection process to other programs in the system.
The Family History System INDEX File is one which contains a complete
but re-ordered sequence of ID #'s. The reordering could be accom-
plished by SORTing the ID#'s based upon the contents of certain
fields, for instance in Surname, Given Name sequence or in birthdate
sequence. The reordering could also represent the sequence that indi-
viduals identified in a relationship work file would be listed in an
ancestor, descendant or relative report. This last type index is most
likely to be used for printing family group reports in "relationship
sequence". The option for creating an INDEX file is one of the
"extensions" provided to registered users of the system.
Another type "work" file introduced in the March 1990 update is the
REFERENCE work file, which contains notes concerning the ID's that
appear in a particular report and the location of the reference within
that report. This work file is used by a program in the extended
system to print report indexes. These indexes may be printed for
ancestor, descendant, relative and family group reports as well as
sets of ancestor charts, ancestor maps and descendant charts.
29