Code 128 is a very high-density barcode symbology. It is used for alphanumeric or numeric-only barcodes. It can encode all 128 characters of ASCII and, by use of an extension character (FNC4), the Latin-1 characters defined in ISO/IEC 8859-1.
GS1-128 (formerly known as UCC/EAN-128) is a subset of Code 128 and is used extensively world wide in shipping and packaging industries as a product identification code for the container and pallet levels in the supply chain. The symbology was formerly defined as ISO/IEC 15417:2007.
- 1 Specification
- 2 References
- 3 External links
A Code 128 barcode will have six sections:
- Quiet Zone
- Start Character
- Encoded Data
- Check Character
- Stop Character
- Quiet Zone
Code 128 includes 108 symbols: 103 data symbols, 3 start symbols, and 2 stop symbols. The symbols consist of three black bars and three white spaces. The stop pattern is composed of two overlapped symbols and has four bars. The stop pattern permits bidirectional scanning. When the stop pattern is read left-to-right (the usual case), the stop symbol is recognized. When the stop pattern is read right-to-left, the reverse stop symbol is recognized. A scanner seeing the reverse stop symbol then knows it must read the symbol in reverse. To represent all 128 ASCII values, there are three code sets (A, B, C), which can be mixed within a single barcode (by using codes 98 and 99 in code sets A and B, 100 in code sets A and C and 101 in code sets B and C to switch between them):
- 128A (Code Set A) - ASCII characters 00 to 95 (0-9, A-Z and control codes), special characters, and FNC 1-4
- 128B (Code Set B) - ASCII characters 32 to 127 (0-9, A-Z, a-z), special characters, and FNC 1-4
- 128C (Code Set C) - 00-99 (encodes each two digits with one code) and FNC1
The quiet zone should be at least ten times the width of the narrowest bar/space element. It is mandatory at the left and right side of the barcode. The quiet zone must not be less than 6.4mm wide.
Start/stop and encoded data
Each character in the barcode symbol is composed of three bars and three spaces. (The stop adds an additional extra bar of length 2.) Each bar or space is 1, 2, 3 or 4 units wide, the sum of the widths of bars must be even, the sum of the widths of the spaces must be odd, and total 11 units per character. For instance, encoding the ASCII value 0 can be viewed as 10011101100, where a 1 (one) is a bar and a 0 is a space. A combination which contains a single 1 would be the thinnest line in the bar code. A combination including three 1 (111) in sequence indicates a bar three times as thick as a single 1 bar.
The code uses 107 of the 108 possible patterns that fulfill the required criteria, with the pattern 211133 deliberately remaining unused. This is done to enhance omnidirectional scanning and help prevent misreads (if 211133 represented a legitimate symbol it could be mistaken for the stop code 2331112 when scanned backwards).
Check digit calculation
The check digit is a Modulo 103 checksum. It is calculated by summing the start code 'value' to the products of each character's 'value' multiplied by its position in the barcode string. The start character and first encoded value is in position 1. The sum of the start code value and the products is then divided by 103. The remainder of the division is the check digit's 'value' which is then converted into a character (following the instructions given below) and appended to the end of the barcode.
For example, in the following table, the code 128 variant A checksum value is calculated for the alphanumeric string PJJ123C
|Position||Code||Value||Position x Value|
|1||Start Code A||103||103|
|Remainder &||Check Sum Value||54|
Calculating check digit with multiple variants
As Code 128 allows multiple variants, as well as switching between variants within a single barcode, it is important to remember that the absolute Code 128 value of a character is completely independent of its value within a given variant. For instance the Variant C value "33" and the Variant B value "A" are both considered to be a Code 128 value of 33, and the check digit would be computed based on the value of 33 times the character's position within the barcode.
Using FNC4 to encode high (128-255) characters
Function 4 (FNC4) can be used to encode all the LATIN-1 (ISO-8859-1) characters in a Code 128 barcode. The feature is not widely supported. When a single FNC4 is present in a string, the following character will be converted to ASCII as usual (character sets A and B) and then have 128 added to the ASCII value. (If the following symbol is a shift, then a second symbol will be used to obtain the character.) If two FNC4s are used consecutively then all following characters will be treated as such, up to the end of the string or another pair of FNC4s. Between the double FNC4s, a single FNC4 will be used to denote that the following character will be standard ASCII.
[Question: please clarify if switching to Code C will terminate the effect of previous double FNC4s. For example, the code is currently in Code A, and there are 2 consecutive FNC4s which bring the context to latin submode of Code A, then it's switched to Code C and then switched back to Code A again (or Code B), question is, is it still in latin submode? If the above statement is true, the answer will be yes. please someone confirm this. Thanks!]
The answer is YES, still in latin submode.
I implement this in my barcode generator http://www.generatorkodowkreskowych.pl/en/code-128/, create barcodes and read with barcode reader Argox AS-8000.
Text to encode: óóóó1234óóabózz
[Start B] [FNC 4] [FNC 4] 83 83 83 83 [Code C] 12 34 [Code B] 83 83 [FNC 4] 65 [FNC 4] 66 83 [FNC 4] [FNC 4] 90 90 02
||begin using charset B|
||go into latin mode|
||goto charset C - digits 1234|
||back to charset B - letters óó|
||terminate latin mode|
Bar code widths
Code128 specifies a combination of 6 bars and spaces for each character except the Stop character, which uses 7. Thus, each character begins with a bar and ends with a space (with the exception of the stop character, which ends in a bar). The following tables detail the widths associated with each bar and space for each character. The width of each bar or space can be 1, 2, 3 or 4 units. Using the example above, an 'A' would be depicted with the pattern 10100011000, or as 111323 in the tables below.
|Value||128A||128B||128C||ASCII Code (Common/Barcodesoft)||Character (Common/Barcodesoft)||Bar/Space Pattern||Bar/Space Weights|
|0||space||space||00||0032 or 0212 / 252||Space or Ô / ü||11011001100||212222|
|95||US||DEL||95||200 / 240||È / ð||10111101000||114113|
|96||FNC 3||FNC 3||96||201 / 241||É / ñ||10111100010||114311|
|97||FNC 2||FNC 2||97||202 / 242||Ê / ò||11110101000||411113|
|98||Shift B||Shift A||98||203 / 243||Ë / ó||11110100010||411311|
|99||Code C||Code C||99||204 / 244||Ì / ô||10111011110||113141|
|100||Code B||FNC 4||Code B||205 / 245||Í / õ||10111101110||114131|
|101||FNC 4||Code A||Code A||206 / 246||Î / ö||11101011110||311141|
|102||FNC 1||FNC 1||FNC 1||207 / 247||Ï / ÷||11110101110||411131|
|103||Start Code A||208 / 248||Ð / ø||11010000100||211412|
|104||Start Code B||209 / 249||Ñ / ù||11010010000||211214|
|105||Start Code C||210 / 250||Ò / ú||11010011100||211232|
|106||Stop (7 bars/spaces)||211 / 251||Ó / û||1100011101011||2331112|
used to detect reading right to left
The "Code A", "Code B" and "Code C" symbols cause all future symbols to be interpreted according to the corresponding subcode. The "Shift" symbol switches a single following symbol's interpretation between subcodes A and B.
The accoding ASCII char depends on the actual used barcode-font. Especially the ASCII char of value 0 and of value 95 and above can be defined differently in the font that is installed.
The FNCx codes are used for special purposes. FNC1 at the beginning of a bar code indicates that it begins with a 2- 3- or 4-digit application identifier assigned by the Uniform Code Council, which explains the following digits. For example, application identifier 421 indicates that an ISO 3166-1 numeric country code and ship-to postal code follows. For example, the U.S. ZIP code for the White House would generally be printed as "(421) 840 20500", but would actually be coded as "
[Start C] [FNC1] 42 18 40 20 50 [Code A] 0 [Check symbol 92] [Stop]"
Check Digit Calculation for above Zip Code Example
|Value||Weight||Weight x Value|
|1740||Mod 103 =||92|
For the end user, Code 128 barcodes may be generated by either an outside application to create an image of the barcode, or by a font-based barcode solution. Either solution requires the use of an application or an application add in to calculate the check digit and create the barcode.
Barcode length optimization using code set C
Although code set C provides code symbols between 0 and 99 providing ability to store pairs of neighboring digits in one code symbol, its usage does not necessarily save final code length compared to code sets A or B. While data "X00Y" encoded using code set A or B result in 6 code symbols (
[Start B] 56 16 16 57 56) including start and stop symbol, an attempt to encode inner "00" from given example in code set C will result in code 7 symbols long, because although one symbol is saved on encoding "00" to code symbol 0, adding of two code set switching symbols is required: first to switch into code set C, second to switch back to code set A or B (
[Start B] 56 [Code C] 00 [Code B] 57 CHECKSUM).
Analogically, applying of code set C to a sequence of digits produces code with less symbols compared to code sets A or B only under these conditions:[dubious ]
|Position of sequence of digits in data||Number of digits for switching to code set C|
|beginning of data||4+|
|end of data||4+|
|middle of data (surrounded by characters from code set A or B)||6+|
|entire data||either 2 or 4+|
Example: For data "098X1234567Y23", savings on barcode length using code set C are achieved only if it is applied to middle part of data ("123456" or "234567"). For beginning and ending part of the code, switching to code set C is not effective. Code symbols:
[Start B] 16 25 24 56 17 [Code C] 23 45 67 [Code B] 57 18 19 33, where
33 is checksum.
The description above is incomplete. The mistake is a transition from Code A or Code B to Code C is not done when there are an odd number of digits remaining. It is better to code the first digit in A/B and then switch to C. The delay avoids an extra symbol. Consider the string "...01234": a delayed switch produces
0 [Code C] 12 34 [Stop] but an early switch produces
[Code C] 01 23 [Code A] 4 [Stop].
Optimizing length of resulting barcode is important where barcode readers are used which require to detect the entire barcode image at once in order to read a barcode, such as laser scanners. The longer the barcode is, the greater distance of laser barcode reader from barcode image is needed, making reading difficult or impossible above some threshold lengths/distances.
- Apparently ISO 15417 Annex F
- Softmatic, http://www.softmatic.com/barcode-code-128.html, stating, "In principle non-ASCII characters like German umlauts (e.g. ÄÖÜ) can be encoded in a Code 128 symbol by using a special character (FNC4). However, this feature is not widely supported. Using a 2D barcode symbology like Aztec or Datamatrix with dedicated support for non-ASCII data might be a better choice."
- http://www.han-soft.com/releases/barcode1d/documents/b_code128.html stating, "If a single "FNC 4" character is used, indicates the following data character in the symbol is a extended ASCII character. A "SHIFT" character may follow the "FNC 4" character if it is necessary to change character subset for the following data character. Subsequent data characters revert to the standard ASCII character set. If two consecutive "FNC4" characters are used, all following data characters are extended ASCII characters until two further consecutive "FNC4" characters are encountered or the end of the symbol is reached. If during this sequence of extended encodation a single "FNC4" character is encountered it is used to revert to standard ASCII encodation for the next data character only. "SHIFT" and character subset characters shall have their normal effect during such a sequence."
- GS1 document gives the compression strategy.
- GS1-128 Specification Detailed List of Application Identifiers
- Barcodesoft Font mapping of Barcodesoft, which differs from the common ascii mapping (see http://ascii-code.com/).
- The 128 code Learn the Code 128 encoding algorithm with a font-based barcode solution.
- Online barcode generator Free online Barcode generator for the various barcode types.
- Free Barcode Generator for Excel Free Excel Macro and font for Multiple Barcode 128 Generation.
- ZXing Multiplatform open source barcode scanner / generator with versions available in Java (core project) and ports to ActionScript, C++, C#, ObjectiveC and Ruby.
- Python Bar Code 128 This code appears to draw boxes 1 pixel wide. It appears it was modified from a short line long line bar code which would have drawn lines. The "Black boxes" should be the same size as the "White Boxes".
- GenCode128 Free C# source code implementation of Code128. Almost all features are implemented, but is not 100% complete.
- Barcode1DTools Ruby gem Ruby source code for many 1D barcode symbologies including Code 128.
- Perl barcode generation code Perl source code for many 1D barcode symbologies including Code 128.
- Barcode::Code128 Free Perl barcode generation module.
- GOCR Free OCR with Code 128 recognition.
- Barcode4J Free Java API with implementation of Code128 and other standard barcodes.