NOTE: This document will explain how to encode SICI codes as SISAC bar codes. How and what the SICI code itself represents is outside the scope of this document. For more information on how a SICI code is formed, please refer to http://sunsite.berkeley.edu/SICI/version2.html which provides a very complete overveiw of the subject.

Note that SISAC bar code is actually encoded using Code 128. That means that SISAC isn't really a symbology, but like UCC/EAN-128 is a format for expressing data within an existing symbology. Once the data has been converted from SICI to SISAC, Code 128 is used to actually encode the bar code.

1. An 8-digit ISSN number (with a hyphen separating the first and last four characters of the ISSN number).
2. A 3-digit date format code which indicates the format of the data that follows.
3. A variable number of characters that indicate the date of the serial item. The format of these characters and the length of this field depend on the date format code from #2 above.
4. A 4-digit enumeration data format code.
5. A variable number of characters that contain the optional enumeration data specified in #4 above.

The SISAC symbol is capable of encoding two dates when the serial item includes a range of dates. For example, if the issue data is May/June 2001 then this would best be expressed with Format Code 904, which has a first date format of yyyym (20005) and a second date format of just m (6). In this case, it is assumed the year for the second date is the same as the year for the first date.

Note that the table below shows the formatting for Format Codes 900 through 949. Any code in this range also includes enumeration data. If a symbol does not include enumeration data then you must add 50 to the Format Code. Thus code 908 represents the format yyyymd with enumeration data whereas 958 represents the format yyyymd without enumeration data.

NOTE: Format Code 900 means there is no date information but that there is enumeration data.

In addition to the standard months 01-12, the following "month" codes are defined by SISAC:

21: Spring	31: First Quarter
22: Summer	32: Second Quarter
23: Fall	33: Third Quarter
24: Winter	34: Fourth Quarter

Additionally, it is known that if there is a second set of enumeration fields that the first field of the second set is either non-existant, is of the same length as the first field of the first set, or is one character longer. That is to say, if the first field of the first set was 4 characters long, the first field of the second set is either 0 (non-existant), 4 or 5 characters longer. In other words, if the first field of the second set of data exists, its length can never be less than the length of the first field of the first set. This allows the encoding mechanism to eliminate impossible conditions which allows for a more compact code.

The "Format" column in the table below indicates the length of each possible enumeration field where "X" is the length of the first enumeration field and Y is the length of the first enumeration field + 1. Note that the following offsets should be added to this table based on the length of the first enumeration field of the first set of data:

Enumeration Field 1 Length=1: Add 0 (range 1-765)
Enumeration Field 1 Length=2: Add 765 (range 766-1530)
Enumeration Field 1 Length=3: Add 1530 (range 1531-2295)
Enumeration Field 1 Length=4: Add 2295 (range 2296-3060)
Enumeration Field 1 Length=5: Add 3060 (range 3061-3825)
Enumeration Field 1 Length=6: Add 3825 (range 3826-4400)

For example, code 0021 has a format of X110X220. X , in this case, is 1 since it is in the rang of 1-765, so this means the first enumeration field is 1 character long, as are the second and third. The fourth enumeration field is non-existant. The second set of enumeration data has its first field also of length 1, the second and third fields of length 2, and the fourth field is non-existant.

Likewise, code 0787 would be used if the length of the first enumeration field of the first set of data was "2" (since 787 is in the range of 766-1530, we subtract 765 and get a format code of 22) and the format were X110Y110. In this case X=2 so this would mean the first field of the first set of enumeration data has a length of 2, the second and third fields have a length of 1 and the fourth field is non-existant. Further, the first field of the second set of enumeration data would have a length of 3 (since Y represents a length one longer than that of the first enumeration field), a second and third field length of 1, and the fourth field would again be non-existant.

Note that this field is always encoded in the SISAC symbol as a 4-digit code. If the length of the code is less than 4 characters, it is padded with 0's to the left. Thus the code 24 would be encoded as 0024.

1. START CODE. Since ISSN numbers begin with numeric codes, the Code 128 bar code will virtually always start with a “Start C” start code which shifts into double-density numeric mode. In fact, SISAC bar codes will remain in Code 128 Character Set “C” throughout the entire bar code if there are no alphabetic characters in the code. This allows SISAC to take advantage of the double-density mode that Code 128 offers for purely-numeric data.
2. VERSION NUMBER. Note that the second-to-last character in a SICI code, after the semi-colon and before the hyphen, is a version number. This indicates with what version of the SICI spec that code was generated. This character is included in the bar code. Version 1 is encoded as “01”, version 2 would be encoded as “02”, etc. Currently, the specification is at version “2”.
3. CHECK CODE. Note that the last digit in a SICI code is a check digit. This is the last character in the code right after the last hyphen. This check code is not encoded in a SISAC bar code. Rather, the SICI check digit is dropped and the standard Code 128 check character is added.