Category : Science and Education
Archive   : ASCIIRF2.ZIP
Filename : ASCIIREF.DOC

ASCIIREF.EXE
Version 2.0

Copyright (C) 1989 Peter M. LePoer



A user-configurable utility to reformat online bibliographic
citations into the ASCII delimited format.
-----------------------------------------------------------------

CONTENTS OF THIS ZIP FILE:

ASCIIREF.EXE the main program
ASCIIREF.DOC this file
MICROREC.TXT example file of captured citations
MICROREC.NEW example file of reformatted citations
ABST0002.ABS-
ABST0010.ABS abstracts for reformatted citations
ABLOAD.PRG dBASE IV program to load abstracts
READ.ME registration information

PURPOSE OF THE PROGRAM:

ASCIIREF was created by a professional librarian to
facilitate the use of popular relational-type databases
for bibliographic purposes. The growing use of online
bibliographic databases for research purposes has
spurred the development of PC-based bibliographic
databases to store and manipulate local files of
research material. While there are a number of good
products for this purpose, - both generic text
databases and databases specifically for bibliographic
information, - these are specialized tools and lack
the programmability and flexibility of a
relational-type database such as dBASE, rBASE, Paradox,
etc. In addition, many researchers may already have
one of these relational-type systems and might like to
use it for bibliographic purposes.

It was to facilitate this use that ASCIIREF was
developed. ASCIIREF will take just about any file of
citations captured during an online session and convert
it into a format that most relational-type databases
can directly import.

Perhaps the most important feature of this program is
that it is USER CONFIGURABLE - it is not limited to any
one bibliographic database format, but is configurable
to just about any system in which the records meet
certain minimum requirements (see below).

NOTE:

This document will assume that you know how to define a
database compatible with the data you wish to import,
and how to import an ASCII (comma) delimited file into
your particular system.

REQUIREMENTS:

An IBM PC or compatible with 256k and one floppy drive
A file of citations, in straight ASCII format
Each citation can have up to 12 fields
Each field must be labeled
Each citation must begin with the same field label
Each field can contain up to 256 characters
There can be one lengthy "abstract" type field per
record

OUTPUT:

A file with each citation on one line, stripped of
field labels, with each field surrounded by quotes, and
the fields separated by commas. Multiply-occurring
fields, such as the AU field in the accompanying file
(MICROREC.TXT) are condensed into one field, with
occurrences separated by semicolons. If the citations
contain abstracts, the abstracts are removed from the
citations and each is placed in a separate,
consecutively numbered, file with names such as
ABST0001.ABS, ABST0002.ABS, etc. The name of the
abstract for the particular citation replaces the
abstract in the comma delimited record. This
facilitates adding the records to dBASE IV with the
"APPEND MEMO" command ABLOAD.PRG (see below).

PROCEDURE:

There are 4 options on the main menu: Input
Parameters, Begin Processing, View File and Quit.

The first thing you will need to do is to input the
parameters of the file you wish to reformat. You will
need to input this information:

The input file name
The output file name
The number of fields per record
The width of the field labels
The column in which data begins after the field
labels
Whether or not the citations contain abstracts,
and the name of the abstract field if they do
Whether continuation lines wrap to column 1
The field labels

Once this has been entered you will be returned to the
main menu, and can begin processing the file.

Be sure to enter all of the field labels that can occur
in a record, even if they don't occur in each record.
The only field that MUST be in each record is the first
one. All other empty fields will be indicated by two
quotes with nothing in between ("") in the output file.

The field label processing is case sensitive, so be
sure to enter the labels in the case in which they
appear in the data (usually upper case).

Enclosed is a sample file, MICROREC.TXT, which contains
10 citations captured during a MEDLINE (the National
Library of Medicine database) search session.

The appropriate parameters for this file would be:

INPUT FILE: MICROREC.TXT
OUTPUT FILE: (WHATEVER YOU WANT)
NUMBER OF FIELDS: 5
FIELD LABEL WIDTH: 2
BEGINNING DATA COL: 7
ABSTRACTS? Y
ABSTRACT FIELD NAME:AB
WRAP TO COL 1? N
FIELD LABELS: UI,AU,TI,AB,SO

The included file MICROREC.NEW contains the results of
running ASCIIREF with these parameters, and
ABST0002.ABS - ABST0010.ABS contain the abstracts for
these cites.

The View File option on the menu allows you to look at
any ASCII file. This is helpful in determining the
correct parameters to enter, and in checking the
results of the reformatting process.


DEFINING THE DATABASE

SAMPLE DATABASE DEFINITION FOR MEDLINE RECORDS


Field Field Name Type Width Dec Index

1 UI Character 8 N
2 AUTHOR Character 128 N
3 TITLE Character 128 N
4 ABNAME Character 12 N
5 SO Character 128 N
6 ABSTRACT Memo 10 N

** Total ** 415


Above is a sample definition for a database of citations
with abstracts downloaded from MEDLINE. The most important
thing to note about this definition is the presence of two
fields related to the abstract. One, ABNAME, includes only
the 12-character name of the file containing the abstract
for that record, given by ASCIIREF. The other, ABSTRACT, is
the memo field which will actually contain the abstract
data. The fields must be in the same order as they appear
in the original records, with the ABNAME field substituting
for the original abstract field, and the memo field added at
the end. Neither of these fields would be necessary, of
course, if abstracts were not included in the records.

The field lengths for the AUTHOR, TITLE, and SOURCE fields
are approximations. They may be defined shorter or longer
as needed. The AUTHOR field is defined as 128 characters in
order to store multiple author names, since ASCIIREF
reformats multiple author names into one field.



ABLOAD.PRG


ABLOAD is run from the dBASE dot prompt by issuing the
command DO ABLOAD. The program first prompts for the
necessary information regarding file and database names.
ABLOAD.PRG then appends the records in the delimited file,
goes back to the first of these records, looks in the field
containing the name of the abstract, and if there is a name
in that field, uses the APPEND MEMO command to import the
appropriate abstract. It then goes on to the next record,
looping until the end of the file. When it is through, a
complete database of citations with abstracts is available.

NOTE THAT THIS PROGRAM ONLY WORKS WITH DBASE IV, AS IT USES
THE NEW APPEND MEMO COMMAND.


SYSTEMS TESTED:

This program has been successfully tested on these systems:

MEDLINE (National Library of Medicine)
BRS Colleague (MEDLINE offered through BRS)
BRS PSYCLIT (Psychological literature database on
BRS)
ERIC (Educational literature on DIALOG -must
specify TAG in printing format)
PaperChase (Compuserve MEDLINE - use NLM
printing format)
SilverPlatter (MEDLINE on compact disk)

I would be happy to learn of any successful or unsuccessful
tests on other databases.

-----------------------------------------------------------------
Please read the accompanying READ.ME file for registration
information. Questions, comments, bugs, etc. can be communicated to:

Peter M. LePoer
481 Cliffside Drive
Columbus, OH 43202

EXEC-PC (Peter LePoer)
Compuserve (72037,3021)