METHOD AND SYSTEM FOR 5-BIT ENCODING OF COMPLETE ARABIC-FARSI LANGUAGES

METHODE ET SYSTEME DE CODAGE A 5 BITS DES CARACTERES DE TOUTES LES LANGUES ARABES ET PERSANES

English Abstract

ABSTRACT OF THE DISCLOSURE
A method and apparatus for coding, transmitting,
receiving, and displaying remotely or locally, all
Arabic-Farsi characters or letters, basic arithmetic
signs, numerals, punctuation marks and diacritical
marks, as well as teleprinter operation commands in 5-
bit standard Baudot codes. An Arabic-Farsi teleprinter
similar in operation to the English teleprinter and
compatible with the International exchange systems is
provided without eliminating any letter forms. The
teleprinter operation (the ability to compress the
data into 5-bit characters) is based upon two basic
criteria of the Arabic-Farsi languages, namely: 1)
the form (start, middle, end or independent) of a
character can be known if the preceding character and
the following characters are known; and (2) there are
six characters that are identical except for the
presence or absence of a dot. The digital logic
circuits make use of the above criteria to permit the
encoding and thus the transmitting and receiving of
complete Arabic-Farsi alphabets. A particular grouping
of characters with level coder for each grouping, as
well as a keyboard arrangement and control circuits
for more versatile use of input/output devices are
also disclosed.

Claims

The embodiments of the invention in which an exclusive
property or privilege is claimed are defined as follows:
1. A teleprinter system for Arabic-Farsi languages
comprising;
means for generating a succession of 5-bit codes
each representing an Arabic character of the Arabic-Farsi
language or one of a plurality of standard teleprinter
characters including teleprinter numerals, punctuation, and
command characters, without regard to the form of the Arabic
characters;
means for inserting one of at least three 5-bit
codes into the succession of 5-bit character codes to identify
at least one subsequent character code as being in one of
at least three predetermined groups of characters associated
with the inserted one of the three 5-bit codes;
means for receiving and storing the 5-bit code
for at least two successive characters;
means responsive to the stored 5-bit codes for
classifying each received character as one of a plurality
of predetermined character types;
means for generating a second code identifying
each stored 5-bit code representing an Arabic character as
one of four possible Arabic character forms in response to the
classified type of the character immediately preceding and
immediately following the first stored of the characters;
means for displaying in its proper form and
position each Arabic character represented by a stored 5-bit
code in response to said second code and the stored 5-bit
code, including means for displaying successive characters in
the same position in response to an indication from the
classifying means that a character to be displayed is of
a character type for which the display position is not to
change.
37

2. The teleprinter system of claim 1 wherein
said classifying means conprises means for specifying each
received character as a character of the type that joins
to the preceding or following character in a given word,
a character of the type that does not join the following
character in a given word but does join the preceding char-
acter, a character of the type that does not join either
the preceding or following character, a character of the
type that does not cause the movement of the display to
another space, or a character of the type that specifies
a teleprinter operation command.
3. The teleprinter system of claim 1 wherein
the six Arabic letters ( <IMG> ) are
coded as a 5-bit code specifying a dot followed by a 5-bit
code specifying the corresponding one of the forms ( <IMG>
<IMG> ).
4. The teleprinter system of claim 3 wherein
said classifying means comprises means for specifying each
received character as a character of the type that joins to
the preceding or following character in a given word, a char-
acter of the type that does not join the following character
in a given word but does join the preceding character, a
character of the type that does not join either the preceding
or following characters, a character of the type that does not
cause the movement of the display to another space, or a
character of the type that specifics a teleprinter operation
command.
38

5. The teleprinter system of claim 4 wherein
said displaying means comprises means for generating an 8-
bit code specifying the stored 5-bit code as one of a plurality
of possible characters including all Arabic characters and
forms thereof, standard teleprinter operation commands,
punctuation, numerals and diacritical marks, wherein the
possible characters and character forms number over one
hundred.
6. The teleprinter system of claim 1 wherein
said displaying means comprises means for gnerating an 8-
bit code specifying the stored 5-bit code as one of a plurality
of possible characters including all Arabic characters and
forms thereof, standard teleprinter operation commands,
punctuation, numerals and diacritical marks, wherein the
possible characters and character forms number over one
hundred.
39

7. An adapter for effecting the decoding of Arabic-
Farsi languages that have been encoded as a succession of
Arabic characters together with other characters including
numerals, arithmetic signs, punctuation, operation commands
and diacritical marks as a sequence of 5-bit digital character
codes without regard to the form of the Arabic characters
of the language comprising:
means for inserting one of at least three 5-bit
codes into the sequence of 5-bit digital character codes to
identify at least one subsequent character code as being in
one of at least three predetermined groups of characters
associated with the inserted one of the three 5-bit codes;
means for momentarily storing each 5-bit digital
character code in a sequence that forms the Arabic-Farsi
language;
means responsive to the stored 5-bit digital
character code for classifying each character as one of a
plurality of predetermined character types as a function of
the 5-bit digital character code preceding and following the
stored 5-bit digital character code in said sequence;
means responsive to said classifying means for
generating a second digital code specifying the form of each
Arabic character represented by the stored 5-bit digital
character code; and,
means for generating a third digital code of
greater than 5-bits specifying the character represented by
the stored 5-bit digital character code, including the form
of said character specified by said second digital code, in
response to said stored 5-bit digital character code, said
second digital signal and said inserted one of the at least
three 5-bit codes.

8. A method of teleprinting Arabic-Farsi languages
comprising the steps of:
generating a succession of 5-bit character codes
wherein each 5-bit character code represents an Arabic
character of the language or one of a plurality of characters
including punctuation, numerals, commands, and diacritical
marks, the 5-bit code of each Arabic character representing
the character without regard to its form;
inserting one of at least three 5-bit codes into
the succession of 5-bit character
codes to identify at least one subsequent character code as
being in one of at least three predetermined groups of
characters associated with the inserted one of the at least
three character codes;
transmitting the succession of 5-bit character
codes and the inserted 5-bit codes to a remote location;
receiving the succession of 5-bit character and
inserted codes at the remote location and momentarily storing
each received 5-bit code;
classifying the character represented by the
stored 5-bit code as one of a plurality of possible character
types;
for each 5-bit code representing an Arabic
character, generating a form code specifying the form of the
Arabic character represented by the 5-bit code as a function
of the type of character immediately preceding and following
the stored 5-bit character code; and
displaying the received succession of 5-bit codes
as Arabic characters in their proper forms and positions and
as other characters specified by the 5-bit codes in response
to both the 5-bit codes and the form codes, successive
characters being displayed in the same position in response
to an indication that a received character is of a predeter-
mined type for which the display position is not to change.
41

9. A method of storing and retrieving data in an
Arabic-Farsi language wherein the data are represented by
a sequence of 5-bit digital words with each word representing
a character of the data, the method comprising the steps of:
assigning each character of the data to one of at
least three predetermined groups and inserting into the
sequence of 5-bit digital words one of at least three 5-bit
codes identifying at least one subsequent 5-bit word as being
in one of the at least three predetermined groups;
storing the data as a sequence of 5-bit digital
words each representing a character of the data, including
Arabic characters, without regard to the form of the Arabic
characters;
retrieving the 5-bit digital words in sequence
from storage and determining the form of each Arabic character
represented by a 5-bit word as a function of the identify
of the Arabic character and of the 5-bit word preceding
and following the Arabic character; and
displaying the data including the Arabic
characters in their proper forms and positions with some
positions being the same for two successive characters in
response to the retrieved 5-bit digital words and the determined
form of the Arabic character.
42

10. An electronic typewriting system for Arabic-
Farsi languages comprising:
a. a keyboard having a plurality of keys each
corresponding to at least one of a single form of an
Arabic character of the Arabic-Farsi language, a numeral,
punctuation and command character and means for generating
a succession of multi-bit character codes in response to
actuation of the keys, wherein each character code represents
one of an Arabic character of the Arabic-Farsi language,
numerals, punctuation and command character and the character code
for each Arabic character is independent of the form thereof;
b. an Arabic adapter for converting the form
independent character codes into the proper language form
for display comprising
i. means dividing the characters into one of at
least three groups and for inserting one of at least three
multi-bit level codes into the succession of character codes
to identify at least one associated character code as
representing characters of one of the three groups of
characters;
ii. means for receiving and storing the codes for
at least two successive characters;
iii. means responsive to the stored codes for
classifying each received character as one of a plurality of
predetermined character types; and
iv. means for generating a form code for each
received Arabic character identifying same as one of four
Arabic character forms in response to the classification of the
character immediately preceding and immediately following
the Arabic character under consideration; and
43

c. display means responsive to the received codes
and the form code for each of a succession of characters
for displaying the last entered character in a first display
without regard to form and for displaying each Arabic character
in its proper form in successive positions in a second display
including means for displaying the successive characters in
the same position in response to an indication from the means
for classifying that a received character is of a predetermined
type for which the display position is not to change.
44

11. The system of claim 10, wherein the six
Arabic letters having a dot are coded as a character code
preceded by a predetermined one of the first, second and
third level codes.
12. The system of claim 11, wherein
the classifying means comprises means for specifying each
received character as a character of the type that joins
to the preceding or following character in a given word, a
character of the type that does not join the following
character in a given word but does join the preceding
character, a character of the type that does not join either
the preceding or following character, a character of the
type that does not cause the movement of the display to
another space, or a character of the type that specifies a
typewriter command.
13. The system of claim 12, wherein said adapter
comprises means for generating an 8-bit code specifying
the stored character code as one of a plurality of possible
characters including all Arabic characters and forms thereof,
standard operation commands, punctuation, numerals and
diacritical marks, wherein the possible characters and character
forms number over one hundred
14. The system of claim 10, wherein said adapter
comprises means for generating an 8-bit code specifying the
stored character code as one of a plurality of possible
characters including all Arabic characters and forms thereof,
standard operation commands, punctuation, numerals and
diacritical marks, wherein the possible characters and character
forms number over one hundred.

15. The system according to claim 10, wherein
the character and level codes comprise 5 bit digital words.
16. The system according to claim 10 or claim 15,
wherein the means for receiving the multi-bit codes comprises
means for alternatively receiving multi-bit codes from a
remote code generating means and wherein the adapter further
comprises means for remotely transmitting the multi-bit
codes to be received by the receiving means.
17. The system according to claim 10 wherein
the first display comprises an LED display and the second
display comprises a printer.
18. A method of word processing Arabic-Farsi
languages comprising the steps of:
a. generating a succession of multi-bit character
codes each representing one of an Arabic character of the
Arabic Farsi language and a plurality of other characters
including numerals, punctuation and command characters, wherein
the character code for each Arabic character is independent
of the form thereof by providing a keyboard having a plurality
of keys each corresponding to at least one of a single form of
an Arabic character and the plurality of other characters;
b. converting the form independent character codes
into the proper language form for display by
i. dividing the characters into one of at least
three groups and inserting one of at least three multi-bit
level codes into the succession of character codes to identify
at least one associated character code as representing
characters of one of the three groups of characters,
46

ii. receiving and storing the codes for at least
two successive characters;
iii. classifying each received character represented
by the stored codes as one of a plurality of predetermined
character types, and
iv. generating a form code for each received
Arabic character identifying same as one of four Arabic
character forms in response to the classification of the
character immediately preceding and immediately following
the Arabic character under consideration; and
c. displaying the last entered character in a first
display without regard to form and displaying each Arabic
character in its proper form in a second display in response
to the received character codes and the form code for each of
a succession of characters in successive positions by
displaying the successive characters in the same position in
response to an indication that a received character is of a
predetermined type for which the display position is not to
change.
19. The method of claim 18, wherein the six
Arabic letters having a dot are coded by inserting a
predetermined one of the first, second and third level codes
before the character code therefor.
20. The method of claim 19 wherein
the step of classifying comprises specifying each received
character as a character of the type that joins to the pre-
ceding or following character in a given word, a character of
the type that does not join the following character in a given
word but does join the preceding character, a character of the
type that does not join either the preceding or following
character, a character of the type that does not cause the
movement of the display to another space, or a character of the
type that specifies an operation command.
47

21. The method of claim 20, wherein the step of
converting comprises generating an 8-bit code specifying the
stored character code as one of a plurality of possible characters
including all Arabic characters and forms thereof, standard
operation commands, punctuation, numerals and diacritical
marks, wherein the possible characters and character forms
number over one hundred.
22. The method of claim 18 wherein the step of
converting comprises generating an 8-bit code specifying the
stored character code as one of a plurality of possible
characters including all Arabic characters and forms thereof,
standard operation commands, punctuation, numerals and
diacritical marks, wherein the possible characters and
character forms number over one hundred.
23. The method according to claim 18, wherein
the character and level codes comprises 5 bit digital words.
24. The method according to claim 18 or claim 23,
wherein the step of receiving the multi-bit codes comprises
alternatively receiving multi-bit codes from a remote code
generating means and wherein the step of converting further
comprises remotely transmitting the multi-bit level and
character codes.
25. The method according to claim 18, wherein
the first display comprises an LED display and the second
display comprises a printer.
48

26. The system of claim 10, wherein the
classifying means comprises means for specifying each
received character as a character of the type that joins
to the preceding or following character in a given word,
a character of the type that does not join the following
character in a given word but does join the preceding
character, a character of the type that does not join either
the preceding or following character, a character of the
type that does not cause the movement of the display to
another space, or a character of the type that specifies
a typewriter command.
27. The method of claim 18 wherein the step of
classifying comprises specifying each received character
as a character of the type that joins to the preceding or
following character in a given word, a character of the
type that does not join the following character in a given
word, but does join the preceding character, a character of
the type that does not join either the preceding or following
character, a character of the type that does not cause the
movement of the display to another space, or a character of
the type that specifies an operation command.
49

Description

~.~L2~
B~C~GRO~ND OF THE INVENTIGN
Field of Invention
This invention relates to a method and devices to
be used in Arabic-Farsi teleprinters, typewriters,
typesetting control, compu-ter input/ou-tput terminals,
and displays. In addition the devices and method-may
be applied to similar terminals which may combine
Arabic with other languages.
State of the Prior Art
Arabic scripts used for languages such as Arabic,
Persian and Urdu (Arabic-Farsi languages) generally ,~
contain many more characters and character forms than
are found in Roman script used for English, French,
etc. Acco~dingly, coding techniques developed for
transmitting, receiving, typesetting, and the like in
connection with languages based upon Roman scripts may
not be directly applicable ~or use in encoding and
decoding of languages employing Arabic scripts.
A prime example of a codin~ technique that is
used for transmission of the English language is the
5-bit Baudot code used in teleprinting throughout the
world on the International exchan~e system. This 5-
bit code can accommodate Roman script since only 26
letters or characters are involved and all 26 letters
plus 10 numbers and various punctuations, symbols and
functional keys can be accommodated by the Baudot
code. On the contrary, it has been thought that the
--2--

5-bit Baudot code cannot accommodate -the 60 or more
characters and character forms that might be required
to provide for the transmission of good quality Axabic-
Farsi languages by teleprintex. Accordingly, various
compromises have been suggested as well as various
coding techniques that require more than 5 bits and
thus are not compatible with the existing In-ternational
exchange requirements.
One solution offered by ~. S. Chaudhry in U.S.
Patent No. 3,998,310 does not provide all the character
forms, does not take into consideration the require-
ments for numerals, arithmetic signs, punctuation, and
diacritical marks and expands coding requirements so
as to be incompatible with e~isting teleprinter systems.
Chaudhry reduces the number of letters on a keyboard by
dividing Arabic letters into two -Eorms, short form and
full form, ignoring the other forms described hereinafter.
Charac-ters having both full and short forms are stored
in short form when followed by another character and
in full form when followed by space. Chaudhry also
expands the coding re~uirements by using a 6-bit code
with a seventh bit for "checking". Although it is
suggested that other codes may be used, there is no
disclosure of a system that provides Eor transmission
and reception of complete Arabic-Farsi languages over
standard teleprinter systems.
Hanson U.S. Patent No. 3,513,968 discloses a
typesetting con-trol system in which 6-bit signals
representing Arabic characters and space units are
--3--
.

~z~
stored ln a first shift register and successively
decoded to classify the data into one of three classes
for storage in a second shift register. A second
decoder determines the form of the character from the
character classification immedia-tely preceding and
following the given character. The latter informa-tion,
and the character form are used to address a memory to
select a character in its desired form.
Hyder U.S. Patent No. 3,938,099 discloses a
printing system in which Arabic characters are coded
using 8 bits and 11 bits. An analyzer is provided to
analyze the concatenation properties applicable to
each character using Boolean equations based on
knowledge of the variables of the preceding and follow-
ing characters. Thls information from the analy~er
combined with the character representation code and
the composite code is then converted into a code
suitable for driving output means.
Other approaches have been undertaken to reduce
the number o~ required characters on machines such as
teleprinters by omitting some Arablc character forms
and deleting the arithmetic signs and punctuation
marks so that the remaining number of characters and
operations can be coded in the standard 5 bit binary
Baudot coding. Another approach has been -to use the
English (i.e., Latin or Roman) alphabet to transmit
Arabic on English teleprinters.
None of the above approaches solves the problem
of transmitting good quality Arabic plus the numerals,

arithmetic signs, etc., over the International exchange
ne-tworks which use Telex and Gentex Exchange systems
and utilize s-tandardized 5-bit binary Baudot coding.
The elimination of characters greatly diminishes the
quality of the Arabic language transmission and much
of the expression may be lost or at leas-t may be
difficult to read.
To achieve desired quality levels by past approaches
has required many more than 5 binary bits for encoding
the Arabic charac-ters. As a result, considerably more
computer storage is required when Arabic script rather
than Roman script languages are used in conjunction
with computer sys-tems. Furthermore, the -transmission
energy requirement of a given message is reduced as
the number of bits per character is reduced so such
xeduction i-s very desirable.

.1)61
BRIEF SUM~ARY AND OBJECTS OF THE INVENTION
It is accordingly an object of the present
invention to provide a novel me-thod and system that
overcomes the foregoing problems of the prior art.
It is another object of the present invention to
provide a novel method and system for high quality
reproduction of languages that use ~rabic characters
wherein digital encoding and decoding is employed and
each character is represented by and may be transmitted
using no more than 5 binary bits.
It is yet another object of the present invention
to provide a novel method and system for teleprinting
Arabic-Farsi languages using existing International
exchange networks includiny Telex a:nd Gentex Exchange
systems wit~h a minimum of additional equipment and no ~.
change in the code now employed.
Still a further object of the invention is to
provide a novel method and system for encoding and
decoding the Arabic-Farsi lanyuages for transmission
or storage wherein the input device may be used for -
diverse purposes and is not tied to a sole function -
such as that of teleprinting.
Yet a further objec-t of the present invention is
to provide a method and system for telecommunication
wherein messages may be transmitted and received on
different languages employing different character
forms.
`~
-6-

The apparatus and the method according -to this
invention enable the user to transmit and receive up
to four forms per letter of the Arabic-Farsi languages
plus -the numerals 0-9, various tel.eprinter commands,
the basic arithmetical signs, and a selected number
of punctuation and diacritical marks. The transmitted
and received code uses the standardized 5-bit binary
Baudot coding. Hence the International Telex, and
Gentex networks may be used to transmit complete Arabic-
Farsi texts without compromising the quality of the
language. Savings are obtained in the required number
of code words and bits for the message and in computer
storage requirements for the Arabic~Farsi texts.
In accordance with one embodiment of the inven~
tion, various characteristics of Arabic-Farsi languages
are used to~provide for the complete reproduction of
all Arabic characters as well as required numerals,
punctuation marks, etc., required for complete tele-
printing of Arabic-Farsi languages using standard ;~
5-bit coding. The language characteristics used include:
1. Although there may be more than 60 characters and
character forms (or variations) in the Arabic-Farsi
languages, there are 28 basic letters or characters
in the Arabic-Farsi languages, some of these characters
take different forms depending on the character preceding
it and the character following it and the used calli-
graphy style. Hence only one code word for each Arabic
character is required to be transmitted if a logic is
--7--

implemented at the receiver printer or display to
select the required form and command the printer or
output display, or device accordingly.
2. Certain Arabic letters are otherwise identical with
the exception of a dot above each letter in a first
group. Hence each letter in the first group can be
recognized if a code is received for the dot followed
by a code for the corresponding letter. Thus the
required code words can further be reduced by five words.
3~ Arabic letters, numerals, punctuation marks,
arithmetical signs, diacritical marks including the dot
above selected letters, and teleprinter operational
commands can be classified into th~e following types:
Type A; Those characters that join to the
following character in a given word
and join to the preceding character.
Type s: Those characters that do not join to
the following letter in a given
word but join the preceding character.
Type C: Those characters that do not join
to the preceding or to the following
characters. These include nu~erals,
arithmetical signs, and punctuation
marks.

~-~z~
Type ~: Those characters that do not cause
the carriage or printing cylinder,
or display to move to the next space
such as diacrltical marks, and the
upper case and lower case signals. .~
~ype 0: Those teleprinter operational ~:
commands such as "Who are you?",
Here is, sell, Carriage return, and
Line feed.
4. The diacritical marks fall above or below the corres- '
ponding le-tter the same as the dot above the letters.
When diacritical marks are printed they do not cause
the carriage feed or the printiny cylinder, or ball or
C~T display to advance to the next space and do not
affect the choice of the letter form. Alsor the trans-
rnission of the teleprinter commands such as change from
upper to lower case and.vice-versa are not printed and
do not cause carriage feed, or display space movement.
Accordlng to another embodiment of the invention,
the characteristics of the Arabic-Farsi languages ,
specified in paragraphs 1, 3 and fi above are used to
decrease the total numbexs of code words required to
encoda a complete ~rabic-Farsi language. Moreover,
additional code words are made available for encoding
by sacrificing at least three code words as level
indicators. These level indicator code words, when
present in a sequence of character code words, specify .;
that the immediately following code words in a sequence ::
are of a predetermined group of characters. In this
manner, a 5-bit code can be used to encode up to 93

charac-ters (96 minus the 3 level code words), and a
complete Arabic-Farsi lanyuaye as well as o-ther charac-
ters such as Telex control characters, numerals, etc.,
can be encoded.
In addition to the foreyoing, one aspect of the
invention involves a technique for reducing the number
of keys on an Arabic-Farsi typewriter or teletypewriter
keyboard by providing only one form of a multiple-form
character on the keyboard. An Arabic adapter in accordance
with the foregoin~ principles is employed to determine
the true form of the character and, despite the fewer
number of keys, the complete Arabic-Farsi language can
be produced. It will be appreciatec1 that this arrange-
ment greatly simplifies the keyboard and simplifies the
task of the machine operator thereby facilitating both
the operation of the machine and the training required.
Another aspect of the invention involves the
provision of con-txols that permit an Arabic-Farsi
keyboard controlled prin-ting device to be used both as
a teletypewriter and a typewriter. ~lso provision is
made for using both Latin and Arablc-Farsi devices with
peripheral units such as a tape punch.
Using the above characteristics, this invention
provides apparatus and a method to code the complete
Arabic alphabet, the numerals, the basic arithmetic
signs, and the selected punctuations, and diacritical
marks, plus the teleprinter operational commands in
5-bit binary Baudot codes. Apparatus is provided to
interface with the printer, or display so that all the
--10--

required Arabic letter Eorms can be indicated and printed
or displayed accordinyly.
The ~oregoiny objects and advantages of the invention
will become apparent to one skilled in the art to which
the invention pertains from the following detailed
description when read in conjunction with the appended
drawings.

~c~
BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 is a functional hlock diagram of a tele-
printer sys-tem operable in accordance with the present
invention to transmit and receive Arabic-Farsi languages
using standard 5 bit codes;
Figure 2 is a functional block diagram illustrating
the Arabic adap-ter circuit of Figure 1 in greater detail;
Figure 3 is a functional block diagram illustra-ting
in greater de-tail a coder that automa-tically causes -the
generation o~ a dot code and a charac-ter code when
certain characters are commanded by a user;
Figure ~ is a pictorial representation of one fo:rm
of teleprin-ter keyboard that may be utilized in conjunc-
tion with -the present invention for transmission and
reception of Arabic; and
Figure 5 is a pictorial representation of a variation
of the keyboard of Figure 4 with the English (~oman)
characters also appearing on the keys as they are shown
. in Table III herein.
Figure 6 is a table indicating the groups of
classification of the Arabic-Farsi characters in the
teleprinter or telex operations in order to illustrate
one form of the character type classification according
to the Figure 1 embodiment of the invention;

z~
Figure 7 is a table summa.rizing the rules for
determining the letter form according to the present
invention;
Figure 8 is a table illustrating an example of ;~
saudot coding of Arabic Farsi characters in accordance
with one embodiment of the invention;
Figures 9a and 9b indicate the two groups o:E
Arabic-Farsi characters that are similar except for a :
dot above each character in one group that does not ~ ;
appear in the other; ~ -
..~'
Figure 10 is a table illustrating another example
of Baudot coding oE Arabic-Farsi characters in accordance
with the present invention wherein three level codes ~
are used and the characters are placed into three ;.
groups for coding;
Figure ll is a functional block diagram of an - :
embodiment of the present inven-tion compatible with up ;,'.
to three groups of characters;
: ~;
~ Figure 12 is a plan view schematically illustrating
an abbreviated keyboard input device in accordance with :
the present invention; and ~ `.
Figure 13 is a functional block diagram of an
embodiment of an Arabic-Farsi printing system employing
-13~
~. ~' , ',

6~
an abbreviated keyboard input device as shown in
Figure 12.
-14- ~:
.
: '

6~
DETAILED DESCRIPTION OF ~ PREFERRED EMBODIMENT
Arabic letters or characters are baslcally 28 in
number but some le-tters may have as many as four diEferent
forms depending upon their position in relation to other
characters. As indicated in Figure 8, there are charac-
ters which may take four forms, while others may take
three forms, others two forms, and others only one form.
The form of the character is decided upon in accordance
with the logic and classification se-t forth in Figures
6 and 7.
In accordance with the preferred embodiment of the
invention, Arabic letters, numerals, arithmetic signs,
punctuations, diacri-tic marks (including the dot above
selec-ted letters), and the teleprinter operational
commands a~e classified into the five types of teleprinter
characters Ar B~ C~ D, and O previously defined and in
Figure 6.
The Arabic-Farsi letter Eorms may be one of four
possibilities: the start form, the middle form, the
end form, and the independent form. The form of a ~ -
letter is logically determined in the preferred embodi-
ment according to the rules in Figure 7 where the (+)
sign means "or."
Figure l illustrates a teleprinter system in
accordance with the present invention which utilizes
the foregoing criteria to transmit and receive Arabic-
Farsi languages using standard 5-bit coding tachniques.
Referring to Figure l, a keyboard ll is connected by a
15-
.. ~ :

~LZ~L~6~
line 12 to a conventional 5-bit Baudot coder 13. The
keyboard 11 may be a standard English keyboard arranged
with the Arabic letters, the numerals, the arithmetic
signs, and the selected punctuations and diacritical
marks plus the teleprinter commands as shown in Figure 8
hereinafter by way o example. The coder 13 codes the
characters into 5-bit binary Baudot codes. Figure 8
also provides an example oE a Baudot 5-bit binary coding
arrangements for the keyboard characters.
The 5-bi-t coder 13 is connected by a line 14 to a
conventional memory or tape punch 15 which is controlled
by a suitable memory or tape punch control 16 by way
of line 129. The 5-bit coder 13 is also connected to a
conventional modem 111 by a line 13 and the modem 111
is connected to a conventional transmitter 117 and to
a conventiQnal receiver 113. The transmitter and
receiver are controlled by a conventional call control
circui-t 119 by a line 116 as illustrated.
The modem 111 and -the memory or tape punch 15 are
interconnec-ted as schematically indicated by the lines
126 and 128. The memory or tape punch 15, the 5-bit
coder 13 and the modem 111 are also connected as is
schematically illustrated by the respective lines 17,
18 and 110 to a switch 125. The switch 125 selectively
connects the three units 13, 15, and 111 either to a ~-
conventional English or other language printing or `
display unit 124 or to an Arabic adapter 120 hereinafter
described in greater detail. The Arabic adapter 120
is connected to an Arabic printing or display unit 122
-16-

~Lf~
such as a conventional CRT display or a conventional
Arabic typewriter.
The keyboard 11, the 5-bit coder 13, the memory or
tape punch 15, the memory or tape punch control 16, the
modem 111, the receiver 113, the transmitter 117 and
the call control 119, as well as the English or other
language printing or display unit 124 together make up
a standard teleprinter unit of the type commercially
available ~or English language or other Roman character
based language transmission. The one difference in
this system is that the keyboard 11 is provided with
Arabic letters, numerals, arithmetic signs, selected
punctuations, diacritical marks and in one embodiment
third level coding, as well as the teleprinter commands
shown, for example, in Figure 8. Moreover, the switch
125 would ordinarily be unnecessary :in a one language
system.
The teleprinter of Figure 1 may operate in a transmit
or receive mode, or in a purely local mode in which mode
data is neither transmitted nor received. In local mode,
the transmitter 117 is disabled so that data entered by
way of the keyboard 11 is not transmitted. The data is,
however, coded by the 5-bit coder 13 to form 5-bit
. ~.
Baudot codes. These 5-bit codes are supplied either to
the printing or display unit 124 or the Arabic adapter
120 depending upon the position of the switch 125.
Assuming that the system is set up for Arabic
operation and the user is using the Arabic characters
on the keyboard 11, the switch 125 will be in the position
-17-

z~
illustrated. The keys depressed on the keyboard 11
result in a 5-bit code for each depressed key and this
5-bit code is supplied to the Arabic adapter 120. The
Arabic adapter translates -the 5-bit codes into 8-bit
codes by addinc3 2 bits to indicate the proper form oE
the character and 1 bit to indicate whether the character
is upper or lower case. The additional 2 bits indicating
the form of the character are arrived at by utili2ing
the previously described characteristics of the Arabic-
Farsi languages.
If the teleprinter is operating in the receive
mode with the switch in the illustrated position, the
data received on input lead 114 by the receiver 113
is supplied by the modem 111 to the memory or tape punch
15 and the Arabic adapter 120. Depending upon the state
of the memory or tape punch control 16 the incoming
data may be stored by the memory or tape punch 15 in
a conventional manner. The data supplied to the Arabic
adapter 120 is translated in-to the 8-bit signal previously
described and causes the printlng or display unit 122 to
reproduce the proper Arabic characters.
If the teleprinter of Figure 1 is operating in the
transmit mode with the switch 125 in the illustrated
Arabic position, the transmitter 117 is enabled and the
5-bit codes from the coder 13 are supplied both to the
Arabic adapter and through the modem 111 to the transmitter
117. The modem may alternatively receive 5-bit codes
from the memory or tape punch 15 by a line 128 as in
typical conventional teleprinter systems. Figure 8
-18-

provides an example of 5-bit binary saudot coding oE
Arabic as compared to English character coding on a
standard teleprinter ]~eyboard. The codes are 32 in
number but many more than 32 characters can be encoded
because the ~eys can be operated either in upper or
lower case. Also, the characters listed in slots 33~38
do not have separate codes but are made up of a composite
of the dot code (00000) followed by the corresponding
character code. It will be appreciated from Figure 8
that all of the ~rabic characters and character forms are
provided on the keyboard in addition to the numerals,
the arithmetic signs, the selected punctuations and
diacritical marks, and the teleprinter commands of a
standard teleprinter. The code for the characters
listed as 33-3~ may thus be formed by first depressing
the key (and thus generating the code~ Eor the dot and
then depressing the key for the character or by depressing
only one key and automatically generating both codes as
is described hereinafter. ;;
It can be seen from Figure 8 -that those characters
having more than one form are provided only one key
position on the keyboard and one corresponding 5-bit
code. Thus, one 5-bit code represents a character that
may have up to four forms with nothing in the 5-bit
code itself to indicate the form of the characters. The
receiving end of the system (i.e., a remote receiver or
a local printer) must therefore determine the form of the
character from the foregoing criteria.
-19-

The Arabic character in the lower case keyboard
position labelled number 1 ln Figure 8, for example,
corresponds to the English Q (see Figure 5 for example).
The Arabic character in the number 1 position in the
table of Figure 8 and the English Q are therefore encoded
with the same code, i.e., 10111. Accordingly, if the
Arabic character in the number one position in the table
of Figure 8 is depressed on the keyboard of Figures 1 and
5 the coder 13 will produce the 10111 code. If the
switch 125 is in the illustrated position, one of the
two forms of the Arabic--Farsi character in the number :
one position in Figure ~ will be reproduced depending
on the position of the character relative to other
characters. Similarly, if the switch 125 is in the
position connecting the coder to the printing or display
unit 12~, the letter Q will be printed. It will there-
fore be appreciated that the Arabic adapter 120 makes
the decision based upon the previously described
criteria as to what Arabic character form will be printed
despite the fact -that the 5-bit code does not directly :
carry information as to the forms of the character.
One embodiment of the Arabic adaptex 120 of Figure 1
i5 illustrated in greater detail in Figure 2. Referring
now to Fi~ure 2, it will be seen that the Arabic adapter
provides for the utilization of the standard 5-bit
saudot code in the transmission and reception of Arabic-
Farsi languages and provides the ability to print all
Arabic characters in their exact forms at the receiving
end of the transmission system. It will also be
-20-
_ _ _ _ . _ _ .. . . . . . .. .. _ _ . .. , . . , , ~,

~Z~6~
appreciated that the technique used to accomplish this
result in the circuit of Figure 2 includes (A) identifi-
cation of a sequence of characters as upper or lower
case, (s) identification of each character by type
including whether -this character is printed with or
without carria~e feed and whether carriage feed occurs
without printing, and (C) utilization of the information
of (A) and (B) above in conjunction with delay so that
a form for each character is identified at the time of
pri.nting. It will be seen that because of the delay the
printed character is, in general, one character behind
the last charac-ter received.
Referring now to Figure 2, the ~rabic adapter
includes two position selec-tor switches 25 and 28 which
select coded character information and character
indicator info:rmation (a strobe signal IND that acts as
a timing signal for the received information), respec-
tively, from eithe.r a local keyboard or mernory, or from
a transmission system. For example, in one position of
the switches 25 and 28, a 5-bit saudot code and a valid
character indicator signal (the strobe) will be accepted
fron~ the transmission system modem 111 of Figure 1. In
the other position of the switches 25 and 28, the
character and character indicator code will be accepted
from the keyboard coder 13 of Figure 1.
The selected character is supplied as a 5-bit
signal along line 26 to a conventional 5-bit parallel
in/parallel out shift register 210. The output signai
from the register 210 is supplied to a second identical
-21-
. . . ~

~q ~DA# ~L~
L
shift register 220, to an upper and lower case xecognizing
circuit 218 and to the address inpu-t terminal of a read
only memory (ROM) 231.
The valid character indicator signal selected by
the switch 28 is supplied along line 29 to a conventional
dela~ circuit 213 such as a flip flop and to one input
terminal of a gate 216. The output signal from the delay
circuit 213 is applied over line 212 to the shift input
terminal of the xegister 210 and to a second conventional
delay circuit 214. The signal from the second delay
circuit 214 is supplied to a third conventional delay
circuit 222 and to one input terminal of a conventional
three input terminal logic gate 235. The output signal
from the delay circuit 222 is supplied to the clock
input terminal of a register 227 and to one input terminal
of a loyic gate 247.
The change output signal from the upper and lower :
case recognizer 218 indicating that a change from upper
to lower case or vice-versa has occurred is supplied to
one input terminal of the ga-te 235, to an inverting
(negative logic) input terminal of the gate 216, and
to one input terminal of each of three conventional
logic gates 243, 245, and 247 (e.g., AND gates). The
output signal *rom the logic gate 216 is supplied along
line 217 to the clock or shift input terminal of the
register 220. The output signals from the logic gates
243, 245, and 247 are supplied to the printer 122 of
Figure 1 as the respective indicator (IND), carriage
feed (CARRFEED) and print (PRINT) signals.
-22-
, . . . .. .. ~ ., ., _ _ . .,, , ,, _ - - - . , , . ,. ., _. _.
'

,he read only memory 22~ receives an 8-bit address
signal ~the delayed character plus the upper/lower case
STATE plus a 2-bit signal MODE specifying the character
form) and supplied an ~-bit charaçter code to a conven-
tional 8-bit register 227. The output signal C~AR
from the register 227 is -the code identifying which
character form is to be printed.
The read only memory 231 receives 6 bits of informa-
tion, including the last received 5-bit character code
and the current upper/lower case STATE, and provides
four bits of information specifying the type of character
received (TYPE), whether or not the carriage should be
moved (C~RRIAGi~) and whether or not -the character should
be printed ( PRINT)o The T~PE signal is a 2-bit code
supplied to both a register 236 and a logic circuit
241. The type of charac-te:r may be type A, B, C, or O
as was previously described (the type D being excluded
since it is a non-carriage character). The CARRIAGE
signal is a l-bit signal specifying whether or not a
movement of the carriage is specified by the current
character. The PRINT signal is a l-bit signal specifying
whether or not -the character is to be printed (e.g.,
type O characters will not be printed~
The TYPE signal is applied over line 232 to the
data input terminals of two stages of a conventional
four bit parallel in/parallel out shift register 236.
The output signals from the first two stages are applied
via line 238 to the input terminals of the other two
stages of the register 236, and the output signals from
-23-
.. . . _ .... . . . . i , .. .. .. _ _ . , ~ , _

these latter two stages are applied as the PRECEDE signal
to a conventional logic circuit 241. The TYPE signal
is supplied to two other input terminals of the logic
circuit ~41 as the FOLLOW signal. The logic circuit 241
may be any conventional logic circuit ~e.g., a plurality
of AND, OR, NAND, or NOR gates) connected in a conven-
tional manner to solve the eguations of Table II. The
resulting Mode signal thereby indicates by 2 bits one
of the four possible forms previously discussed.
In operation, switches 25 and 28 enable selection
of the 5-bit Baudot charac-ters plus a valid character
indicator from either a local keyboard or from a trans-
mission system. The transmission system will be of the
standard 5-bit Baudot type.
The character indicator pulse is conventionally
provided in a teleprinter system to indlcate the presence ~ ;
oE a character. This pulse :is delayed by the respective
delay circuits 213, 214, and 222 to provide ~or a
controlled order of sequence of events described below.
The upper~lower case recognizer 218 examines the
last received character in register 210 and provides
two output signals. The current state output signal
STATE indicates that all characters are either upper or
lower case depending upon the ~inary state so indicated
and until the state is unchanged. For example, a binary
ONE on line 229 might indicate upper case while a ZERO
might indicate lower case. The CHANGE signal on 219
indicates a state change when the last received character
was an upper or lower case indicator character~ If the
-24-

~10~
last received character was an upper or lower case
indicator (i.e., the character indicating the upper or
lower case key as shown in Figure ~ has been depressed)
then the only activity on the next character is the
change of the state 229 and loading of register 210,
while the CHANGE signal on lead 219 inhibits gate 216
and prohibits loadiny of register 220.
If -the last received character is not an upper or
lower case indicator, then the next received character
indicator pulse on lead 29 is passed by gate 216 and
causes the previous character in regis-ter 210 to be
transferred to register 220.
Next, aftex the delay Tl the latest character received
over line 26 is stored in register 210 in response to the
delayed pulse from circuit 213. This latest 5-bit
character along with the l-bit STATF, signal produces a
6-bit address for the read only memory (ROM) 231. This
ROM stores a 4-bit word for each address. Two bits
identify the form as O, A, B, or C as indicated in
Figure 7, one bit indicates if carriage feed is associated
with this character and one bit indicates i~ printing is
associated with the character. For example, a space
does not involve printing while adding a dot or diacritical
mark to a character involves printing but does not involve -
carriage feed.
Next a~ter additional delay T2, the TYPE data in
register 236 is advanced by the signal from logic gate
235. If the character in register 210 is an upper or
lower case indicator, or if carriage feed is not
-25-

~L~2~
associated with this character including -the U/L case
state as indicated by the CARRIAGE signal on line 233,
then the data in register 236 is not advanced. When
the data is advanced then the 2 binary bits of -the TYPE
signal are stored in register 236 and appear on leads
238 while the previous data on leads 238 is simultaneously
advanced to appear on leads 240 of register 236. The
resulting 4 bits applied Erom the register 236 to
loyic element 241 produces a ~ODE output signal. The
MODE output identifies the character form as being either
of the start, middle, end or independent form as
indicated in Figure 7. The 5-bit character code from
register 220 -together with the character form signal
MODE and the upper/lower case state signal thus form an
address that selects the indicator for the proper character
form from the appropriate Memory location of the ROM 224
for printing.
The final operations are the loading of register
227 after an additional delay T3 and an outputting of a
character indicator to the printer via lead 248. If
the last received character was an upper or lower case
indicator then the character indicator signal IND,
carriage feed signal CARRFEED, and print commands PRINT
to the printer are all inhibited. Depending upon the
type of prin-ter being used, the independent carriage
feed and print commands may not be necessary since this
information is also inherently contained in the 8-bit
character 228 being fed to the printer.
-26-

The final selection of up to an 8-bit (some systems
may require only 7 bits) character data for the printer
is accomplished in ROM 224. The 8-bit address -to this
~OM is composed of 5 bits 221 of-the originally received
saudot character, one on bit 229 indicating upper or
lower case, and 2-bits 242 indicating mode as was
previously described.
As was previously men-tioned, there are six Arabic
letters or characters that are formed identically to
six other but quite different charac-ters except that the
latter six include a dot over the character. ~ith the
keyboard discussed in connection with Figure 8 it was
suggested that -these letters with dots could be encoded
for subseguent decodiny by providing a "dot" key on the
keyboard, which key could be depressed before depressing
any of the.six "non-dotted" characters already on the
keyboard to -transform these to the "dot" characters. As
an alternative, the character with the dot itself can
be placed on the keyboard as shown in Figure 4 and a
circuit such as that shown in Figure 3 can be used to
automatically generate the dot code plus the code of the
corresponding characters whenever these keyboard characters
are depressed. It will, of course, be appreciated that
this requires no additional code words but merely
simplifies the operation of the keyboard.
Figure 3 illustrates a circuit that causes the
transmission of a dot code ~00000) followed by the code
of the corresponding letter when any of the letters with
the dot above is selected. For example if the third
-27-

o~
character from the left in Figure 9a is selected, two
character codes will be transmitted; namely (00000) for
the dot followed by (10101) Eor the character that
corresponds to the dotted character (the third character
from the left in Figure 9a).
The six input characters, the first of which is
iden~ified by the numeral 301, are the six Arabic letters
without the dot and they are processed ln the S-bit
keyboard coder 305 as previously descri~ed. The same
six letter forms (e~cept for having the dot) are
identified by the number 302. Through six OR gates
303 each letter activation on -the keyboard of one of the
six letters with or without the dot en-ters the coder
305 identically~ In addition, a six input terminal
OR gate 304 provides an output signal on lead 31~ in
response to the entry of one of the six "dotted" letters.
The character indicator pulse on line 318 from the coder
305 is combined through AND gates 3:L0 and 311 with the
signal from the OR gate 304 and its inverted form,
respectively. The signals ~rom -the AND gates 310 and
311 and the signal from the coder 305 are then used
~ith a conventional delay circui-t 312, OR gate 308, AND
gate 306, and OR gate 313 to produce the following responses
and outputs.
If a key 301 for the letter without the dot is
activated then the AND gate 306 is enabled to cause the
5-bit character to appear at the output terminal 307
during the occurrence of a single character indicator on
lead 317.
-28-

If a key 302 for a letter with a dot is activated,
then two sequential charaeter indieator pulses oeeur
on ou-tput lead 317 separated by the -time delay 312 which
will be in the order of 10 to 30 milliseeonds to enable
separation but to prevent operator ac-tivation of another
key before the double eharaeter out is completed. During
the first character indicator pulse CHAR IND, the AND
gate 306 is disabled so that -the outpu-t on lead 307
is the all ZERO code for the dot. During the second
eharacter indicator pulse, the OR gate 308 is enabled so
that the code for the charaeter without the dot 301 whieh
is on lead 306 out of the eoder now appears at -the output
307.
In the Flyure 1 embodiment of the invention using
the eharae-ter grouping and eoding oE Figure 8, a dot
eode is prQvided to aeeount for the differenee between
the eharacters o-f Figures 9a and 9b. As is illustrated
in the table of Figure 10, the eharaeters are elassified
into three yroups and a level code is provided for eaeh
group. In -this manner it will be appreeiated that as
many as 93 character and con-trol eodes are available.
Aeeordingly, all the information shown in Figure 10,
as well as additional information if necessary or
desirable, can be eneoded for transmission or storage.
It ean be seen from Figure 10 that two of the level
codes are the lower and upper case codes 11111 and 11011,
respeetively. The all zeros code (00000) may be used
as the third level code.
-29-

With such a three-level system, the circuit o~
Figures 1 and 2 may be utilized with a slight modifica-
tion to the upper/lower case recognizer circuit 218 and
the read only memories so that the third level code can
be recognized and utilized to address the appropriate
characters in memory, as was discussed previously. Of
course, such functions as were previously mentioned can
be implemented using standard microprocessor and memory
chips (integrated circuits) in order to provide added
flexibility.
Alternatively, a circuit such as that illustrated
in Figure 11 may be utilized. Referring to Figure 11,
a keyboard 44 such as that illustrated in Figure 4 (or
a standard Arabic typewriter keyboard with additional
telex operation keys) supplies an ASCII coded character
signal CHAR to an input memory 402, and a strobe signal
CHAR IND is supplied to an input logic circuit 404.
The CHAR signal from the ~eyboard is an 8-bit ASCII
signal when using commercially available equipment, so
the input memory 402 utilizes this code and a 2-bit
code LVL from the input logic circuit 404 to address
appropriate 5-bit Baudot codes CHA~' in memory. Also,
there is a 3-bit identification code ID stored with the
5-bit code in memory and this 3-bit code is supplied to
the input logic circuit 404.
The CHAR' signal is supplied to conventional "first-
in, first-out" (FIFO) circuit 406, which is controlled
by IN and O~T signals from the logic circuit 404. The
FIFO conventionally "smooths" the transmission rate of
-30-
....--

the CHAR' signal so that this rate is relatively uniform
and does not exceed the capacity of a telex or a printer
(e.g., does not exceed 75 bauds).
The 5-bit character signal C~AR' is supplied from
the FIFO to a conven-tional "universal asynchronous
receiver-transmitter" ~UART) 408 which in turn supplies
this signal for transmission as the TS signal to con-
ventional interface circuits (STUNT CIRCUITS) ~10 that
control transmission and reception of signals at acceptable
levels and n appropriate forms. Control slgnals for
the STUNT circuits 410 are supplied from a signal control
logic circuit 412 which receives control signals LOCAL
and OFF-LINE from the keyboard or other control signal
source~ Thus, with the system in telex mode (not in
LOCAL and not OFF-LINE) the input information from the
keyboard 40.0 is directed out of the STUNT circuits to
the TTY and to the UART for simultaneous transmission
and printing. With the system in LOCAL mode, the TS
signal will be routed to the printer (as the RS signal)
by the STUNT circuits and also to the routing switch
(not shown)~ The routing switch may route the information
signals to a paper punch or other type of memory so that
the keyboard data may be recorded as it is printed. Also,
the system may be placed OFF-LINE so that any incoming ;-
data received by the STUNT circuits will be applied to
the routing switch for recording and later printing.
With continued reference to Figure 11, the RS signal
from the UART is applied to a printing control circuit
similar to the input circuit in the sense that the level
-31-

~z~
code is recoynized to assist in recognizing the received
character, the form of the character is determined and an
appropriate character is selected by meMory addressing
as was previously discussed. In this connection, the
RS sigllal is applied to a register 414 (Register 1),
the output signal of which is applied to a register
414, whose output goes to a register 416 and a register
~18.
The output signals from registers 416 and 418 are
supplied to an output memory 420 which supplies a PRINT
signal (the ASCII coded or otherwise compatible character
code) to a sui-table printing mechanism such as the tele-
typewriter solenoid drivers that drive the individual
character printing devices. The PRINT signal is also :
supplied to an output logic circuit 422 that provides the
print strobe signal PRINT STRB to the teletypewriter or
other output device. The output logic circuit also
provides a 2-bit level indicating signal to the Register
(414) and control signals that control the loading o-f
the registers 414-418.
In operation, the 7-bit character signal CHAR from
the keyboard and the LVL signal from -the input logic
circuit 404 select the proper 5-bit 13audot code from the
memory 402 for application to the FIFO. The selected
code is strobed into the FIFO by the logic signal I2~.
Also, the logic circuit 404 determines the group in which
the character belongs and causes the appropriate level
code to be accessed in the memory and placed into the
sequence of character codes.

L
The input logic circuit also determines from the ID
signal if the accessed code is tha-t of a diacritical mark.
I-f it is, it does not have any par-ticular form associated
therewith ~e.g., start, middle, end, independent) and, :
when placed into a sequence of characters it is printed
over the imrnediately preceding character. Thus if a
sequence of characters Cl, C2, C3, C4 is transmitted with
a diacritical mark DM between C~ and C3 (i.e., the sequence
Cl, C2, DM, C3, C4) the DM must be printed before the
C3 is received, DM is printed when Ll is received and
C3 is printed when C4 is received. The printer may
therefore be held up occasionally but neve.r has to print
at a faster than normal speed. Of course, the FIFO
functions as a smoothing buffer during the insertion of
the level codes and during any non-uniform insertion of
characters.so that a uniform transmi.ssion rate is achieved.
When receiving characters for printing, Register 1
contains the last received character code. The next
received character causes the contents of Register 1
to be transferred to Register 2 (or to Register 3 if
Register 1 contains a diacritical mark) and this next
received character is strobed into Register 1. The output
logic 42~ determines the form of the character and supplies
form ID signals FID to the output memory 420 along with
the S-bit character code. The form of the characters
will of course depend upon the 5-bit code, the position
of the character in a word and the level code preceding
the character (first, second, or third le~el). Whether
or not the character is a diacritical mark determines
.

the printing procedure since diacritical marks do no-t
cause a carriage feed. In this connection, a diacritical
mark will be stored in Register 3 and a timing sequence
will start so -that the time order of printing previously
discussed will occur (i.e., Cl, C2, DM, C3, Ll, C4).
In accordance with -the present invention, the Arabic ~ ;
adapters of Figures 2 or 11 may be utilized to simplify
the keyboard of an Arabic typewriter or other such
printing machine to thereby increase the operator's speed
and also simplify training. The manner in which this
may be accomplished is illustrated in Figures 12 and 13.
Referring now to Figures 12 and 13, a simplified
Arabic-Farsi keyboard 450 having only one form of each
letter is connected to the Arabic adapter 120 such as
that illustrated in Figure 2. In this connection, the
signals indicated a~ the inpu~ and output terminals of
the adapter 120 are the same as those shown in Figure 2.
The ou-tput signals ~rom the adapter 120 are supplied to
a suitable conventional printer or other suitable output
device 452.
The line 211 from -the register 210 of the Arabic
adapter (as shown in Figure 2) supplies the last entered
character code to a light emitting diode (LED) display
452 for display of the last entered character. This
LED display may be positioned in any suitable location
readily visible to the machine operator.
In operation, the machine operator is provided with a
keyboard 450 with keys for only one form of the Arabic-
Farsi characters. Accordingly, it can be seen that each
-34-

charac-ter can be typed without the necessity of shifting
from one character level to another (e.g., as with upper
to lower case shifting that is required when one key is
used for two characters). When a word is typed, the code
for each charac-ter key depressed by the operator is
generated in a conventional manner and the Arabic adapter
120 determines the form of the character by its position
in the sequence of characters as was previously described.
The character that is typed or printed is thus the proper
form of the character despite the fact that the operator
always inputs only the one form on the keyboard. In
this regard, it should be noted that the output device
is of the type discussed previously (e.g., a conventional
complete Arabic-Farsi language typewriter output) and is
capable of printing all forms of the charac-ters.
Since the Arabic adapter 120 provides an output
character only after two successive characters have been
entered by the keyboard, the LED display is provided to
display the last entered character. This display may
not be necessary for the trained operator but it may be
helpful should interruptions in the typing occur. -~-
It will be appreciated that this simplified form
of keyboard greatly reduces the amount of shifting
presently required on a standard ~rabic typewriter.
Whereas a good typist may be able to type 35 to 40 words ~ -
per minute on a standard device, it is contemplated that
speeds will almost double to a rate comparable to that
on Latin machines using the device of Figures 12 and 13.

It will also be appreciated that a standard ~rabic
keyboard with all forms of the ~rabic-Farsi characters
can be used in the event that an operator previously
trained on such a machine wishes to continue with its use.
This same system illus-trated in Figure 13 may still be
used, however, and the shifts from upper to lower case
or vice-versa need not be made or, if made, are ignored
in encoding the characters.
.
The present lnvention may be embodied in other
specific forms without departing from the spirit or
essential characteristics thereof. It should be under-
stood, for example, that the present invention is
readily usable for the storage and re-trieval of informa-
tion as shown in Figure 1 merely by operating in local
mode and recording or storing information in memory 15
as it is keyed in through the ~eyboard 11. When it is
desired to use the information, it can be retrieved and
applied through the Arabic adapter for display. The
presently disclosed exemplary embodiment is therefore to be
considered in all respects as illustrative and not res-
trictive, the scope of the invention being indicated by the
appended claims rather than by the foregoing description,
and all changes which come within the meaning and range of
equivalency of the claims are therefore intended to be
embraced therein.