zint/docs/manual.pmd

% Zint Barcode Generator and Zint Barcode Studio User Manual
% Version 2.13.0.9
% October 2024

# 1. Introduction

The Zint project aims to provide a complete cross-platform open source barcode
generating solution. The package currently consists of a Qt-based GUI, a CLI
command line executable and a library with an API to allow developers access to
the capabilities of Zint. It is hoped that Zint provides a solution which is
flexible enough for professional users while at the same time takes care of as
much of the processing as possible to allow easy translation from input data to
barcode image.

The library which forms the main component of the Zint project is currently able
to encode data in over 50 barcode symbologies (types of barcode), for each of
which it is possible to translate that data from either UTF-8 (Unicode) or a raw
8-bit data stream. The image can be rendered as a

- Windows Bitmap (BMP),
- Enhanced Metafile Format (EMF),
- Encapsulated PostScript (EPS),
- Graphics Interchange Format (GIF),
- ZSoft Paintbrush (PCX) image,
- Portable Network Graphic (PNG) image,
- Tagged Image File Format (TIF), or a
- Scalable Vector Graphic (SVG).

Many options are available for setting the characteristics of the output image
including the size and colour of the image, the amount of error correction used
in the symbol and the orientation of the image.

## 1.1 Glossary

Some of the words and phrases used in this document are specific to barcoding,
and so a brief explanation is given to help understanding:

symbol

:   A symbol is an image which encodes data according to one of the standards.
    This encompasses barcodes (linear symbols) as well as any of the other
    methods of representing data used in this program.

symbology

:   A method of encoding data to create a certain type of symbol.

linear

:   A linear or one-dimensional symbol is one which consists of bars and spaces,
    and is what most people associate with the term 'barcode'. Examples include
    Code 128.

stacked

:   A stacked symbol consists of multiple linear symbols placed one above
    another and which together hold the message, usually alongside some error
    correction data. Examples include PDF417.

matrix

:   A matrix symbol is one based on a (usually square) grid of elements called
    modules. Examples include Data Matrix, but MaxiCode and DotCode are also
    considered matrix symbologies.

composite

:   A composite symbology is one which is made up of elements which are both
    linear and stacked. Those currently supported are made up of a linear
    'primary' message above which is printed a stacked component based on the
    PDF417 symbology. These symbols also have a separator which separates the
    linear and the stacked components. The stacked component is most often
    referred to as the 2D (two-dimensional) component.

X-dimension

:   The X-dimension of a symbol is the size (usually the width) of the smallest
    element. For a linear symbology this is the width of the smallest bar. For
    matrix symbologies it is the width of the smallest module (usually a
    square). Barcode widths and heights are expressed in X-dimensions. Most
    linear symbologies can have their height varied whereas most matrix
    symbologies have a fixed width-to-height ratio where the height is
    determined by the width.

GS1 data

:   This is a structured way of representing information which consists of
    'chunks' of data, each of which starts with an Application Identifier (AI).
    The AI identifies what type of information is being encoded.

Reader Initialisation (Programming)

:   Some symbologies allow a special character to be included which can be
    detected by the scanning equipment as signifying that the data is used to
    program or change settings in that equipment. This data is usually not
    passed on to the software which handles normal input data. This feature
    should only be used if you are familiar with the programming codes relevant
    to your scanner.

ECI

:   The Extended Channel Interpretations (ECI) mechanism allows for
    multi-language data to be encoded in symbols which would usually support
    only Latin-1 (ISO/IEC 8859-1 plus ASCII) characters. This can be useful, for
    example, if you need to encode Cyrillic characters, but should be used with
    caution as not all scanners support this method.

Two other concepts that are important are raster and vector.

raster

:   A low level bitmap representation of an image. BMP, GIF, PCX, PNG and TIF
    are raster file formats.

vector

:   A high level command- or data-based representation of an image. EMF, EPS and
    SVG are vector file formats. They require renderers to turn them into
    bitmaps.


# 2. Installing Zint

## 2.1 Linux

The easiest way to configure compilation is to take advantage of the CMake
utilities. You will need to install CMake and `libpng-dev` first. For instance
on `apt` systems:

```bash
sudo apt install git cmake build-essential libpng-dev
```

If you want to take advantage of Zint Barcode Studio you will also need to have
Qt and its component `"Desktop gcc 64-bit"` installed, as well as `mesa`. For
details see `"README.linux"` in the project root directory.

Once you have fulfilled these requirements unzip the source code tarball or
clone the latest source

```bash
git clone https://git.code.sf.net/p/zint/code zint
```

and follow these steps in the top directory:

```bash
mkdir build
cd build
cmake ..
make
sudo make install
```

The CLI command line program can be accessed by typing

```bash
zint [options]
```

The GUI can be accessed by typing

```bash
zint-qt
```

To test that the installation has been successful a shell script is included in
the `"frontend"` sub-directory. To run the test type

```bash
./test.sh
```

This should create numerous files in the sub-directory `"frontend/test_sh_out"`
showing the many modes of operation which are available from Zint.

## 2.2 BSD

The latest Zint CLI, `libzint` library and GUI can be installed from the `zint`
package on FreeBSD:

```bash
su
pkg install zint
exit
```

and on OpenBSD (where the GUI is in a separate `zint-gui` package):

```bash
su
pkg_add zint zint-gui
exit
```

To build from source (including for NetBSD) see `"README.bsd"` in the project
root directory.

## 2.3 Microsoft Windows

For Microsoft Windows, Zint is distributed as a binary executable. Simply
download the ZIP file, then right-click on the ZIP file and `"Extract All"`. A
new folder will be created within which are two binary files:

* `qtZint.exe` - Zint Barcode Studio
* `zint.exe` - Command Line Interface

For fresh releases you will get a warning message from Microsoft Defender
SmartScreen that this is an 'unrecognised app'. This happens because Zint is
a free and open-source software project with no advertising and hence no income,
meaning we are not able to afford the $664 per year to have the application
digitally signed by Microsoft.

To build Zint on Windows from source, see `"win32/README"`.

## 2.4 Apple macOS

The latest Zint CLI and `libzint` can be installed using Homebrew.[^1] To
install Homebrew input the following line into the macOS terminal

```bash
/bin/bash -c "$(curl -fsSL \
    https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
```

Once Homebrew is installed use the following command to install the CLI and
library

```bash
brew install zint
```

To build from source (and install the GUI) see `"README.macos"` in the project
root directory.

[^1]: See the Homebrew website [https://brew.sh](https://brew.sh).

## 2.5 Zint Tcl Backend

The Tcl backend in the `"backend_tcl"` sub-directory may be built using the
provided TEA (Tcl Extension Architecture) build on Linux, Windows, macOS and
Android. For Windows, an MSVC6 makefile is also available. See [Annex C. Tcl
Backend Binding] for further details.


# 3. Using Zint Barcode Studio

Zint Barcode Studio is the graphical user interface for Zint. If you are
starting from a command line interface you can start the GUI by typing

```bash
zint-qt
```

or on Windows

```bash
qtZint.exe
```

See the note in section [2.3 Microsoft Windows] about Microsoft Defender
SmartScreen.

Below is a brief guide to Zint Barcode Studio.

## 3.1 Main Window and Data Tab

![Zint Barcode Studio on startup - main window with Data
tab](images/gui_main.png){.win}

This is the main window of Zint Barcode Studio. The top of the window shows a
preview of the barcode that the current settings would create. These settings
can be changed using the controls below. The text box in the `"Data to Encode"`
groupbox on this first Data tab allows you to enter the data to be encoded. When
you are happy with your settings you can use the `"Save..."` button to save the
resulting image to a file.

The `"Symbology"` drop-down box gives access to all of the symbologies supported
by Zint shown in alphabetical order. The text box to its right can filter the
drop-down to only show matching symbologies. For instance typing `"mail"` will
only show barcodes in the drop-down whose names contain the word `"mail"`. Each
word entered will match. So typing `"mail post"` will show barcodes whose names
contain `"mail"` or `"post"` (or both).

The ellipsis button `"..."` to the right of the data text box invokes the Data
Dialog - see [3.7 Data Dialog] for details. The delete button
![delete](images/gui_delete.png){.btn} next to it will clear the data text box
and the ECI (Extended Channel Interpretations) drop-down if set.

To set the barcode as a Programming Initialisation symbol click the
`"Reader Init"` checkbox. The `"1234.."` button to its right invokes the
Sequence Dialog - see [3.8 Sequence Dialog]. The zap button
![zap](images/gui_zap.png){.btn} will clear all data and reset all settings for
the barcode to defaults.

The `"BMP"` and `"SVG"` buttons at the bottom will copy the image to the
clipboard in BMP format and SVG format respectively. Further copy-to-clipboard
formats are available by clicking the `"Menu"` button, along with
`"CLI Equivalent..."`, `"Save As..."`, `"Factory Reset..."`, `"Help"`,
`"About..."` and `"Quit"` options. Most of the options are also available in a
context menu by right-clicking the preview.

![Zint Barcode Studio main menu (left) and context menu
(right)](images/gui_menus.png){.win}

## 3.2 GS1 Composite Groupbox

![Zint Barcode Studio encoding GS1 Composite
data](images/gui_composite.png){.win}

In the middle of the Data tab is an area for creating composite symbologies
which appears when the currently selected symbology is supported by the GS1
Composite symbology standard. GS1 data can then be entered with square brackets
used to separate Application Identifier (AI) information from data as shown
here. For details, see [6.3 GS1 Composite Symbols (ISO 24723)].

## 3.3 Additional ECI/Data Segments Groupbox

![Zint Barcode Studio encoding multiple segments](images/gui_segs.png){.win}

For symbologies that support ECIs (Extended Channel Interpretations) the middle
of the Data tab is an area for entering additional data segments with their own
ECIs. Up to 4 segments (including the main `"Data to Encode"` as segment 0) may
be specified. See [4.16 Multiple Segments] for details.

## 3.4 Symbology-specific Groupbox

![Zint Barcode Studio showing Code 2 of 5 Interleaved
settings](images/gui_c25inter.png){.win}

Many symbologies have extra options to change the content, format and appearance
of the symbol generated. For those with few additional options (and no support
for GS1 data or ECIs), the middle of the Data tab is an area for setting those
options.

Here is shown the check digit options for an Interleaved Code 2 of 5 symbol (see
[6.1.2.4 Interleaved Code 2 of 5 (ISO 16390)]).

Symbologies with more than a few options (or support for GS1 data or ECIs) have
a second Symbology-specific tab, shown next.

## 3.5 Symbology-specific Tab

![Zint Barcode Studio showing Aztec Code options](images/gui_aztec.png){.win}

A second tab appears for those symbologies with more than a few extra options.

Here is shown the options available for an Aztec Code symbol.

You can adjust its size or error correction level (see [6.6.8 Aztec Code (ISO
24778)]), select how its data is to be treated (see [4.11 Input Modes]), and set
it as part of a Structured Append sequence of symbols (see [4.17 Structured
Append]).

## 3.6 Appearance Tab

![Zint Barcode Studio showing Appearance tab
options](images/gui_appearance.png){.win}

The Appearance tab can be used to adjust the dimensions and other properties of
the symbol.

The `"Height"` value affects the height of symbologies which do not have a fixed
width-to-height ratio, i.e. those other than matrix symbologies. For such
symbologies the `"Automatic Height"` checkbox will be enabled - uncheck this to
manually adjust the height. The `"Compliant Height"` checkbox applies to
symbologies that define a standard height - see [4.4 Adjusting Height].

Boundary bars can be added with the `"Border Type"` drop-down and their size
adjusted with `"Border Width"`, and whitespace can be adjusted both horizontally
(first spinbox) and vertically (second spinbox), and also through the
`"Quiet Zones"` checkbox if standard quiet zones are defined for the symbology.

The size of the saved image can be specified with `"Printing Scale"`, and also
by clicking the ![scaling](images/gui_scaling.png){.btn} icon to invoke the Set
Printing Scale Dialog - see [4.9 Adjusting Image Size (X-dimension)] for further
details.

![Adjusting the Print Size](images/gui_set_printing_scale.png){.pop}

The foreground and background colours can be set either using the text boxes
which accept `"RRGGBBAA"` hexadecimal values and `"C,M,Y,K"` decimal percentage
values, or by clicking the foreground eye ![eye](images/gui_black_eye.png){.btn}
and background eye ![eye](images/gui_white_eye.png){.btn} buttons which invoke a
colour picker.

![The colour picker tool](images/gui_colour.png){.pop}

(Note that to change the colours visually, the luminence slider, the long narrow
column on the right, must be adjusted.) The color picker only deals in RGB(A),
and will overwrite any CMYK values with RGB(A) values once `"OK"` is selected.

Back in the Appearance tab, the colours can be reset to black-on-white using the
`"Reset"` button, and exchanged one for the other using the swap
![swap](images/gui_swap.png){.btn} button next to it.

## 3.7 Data Dialog

![Entering longer text input](images/gui_data_dialog.png){.pop}

Clicking on the ellipsis `"..."` button next to the `"Data to Encode"` text box
in the Data tab opens a larger window which can be used to enter longer strings
of text. You can also use this window to load data from a file.

The dialog is also available for additional ECI/Data segments by clicking the
ellipsis button to the right of their data text boxes.

Note that if your data contains line feeds (`LF`) then the data will be split
into separate lines in the dialog box. On saving the data back to the main text
box any separate lines in the data will be escaped as `'\n'` and the
`"Parse Escapes"` checkbox will be set. This only affects line feeds, not
carriage returns (`CR`) or `CR+LF` pairs, and behaves the same on both Windows
and Unix. (For details on escape sequences, see [4.1 Inputting Data].)

## 3.8 Sequence Dialog

![Creating a sequence of barcode symbols](images/gui_sequence.png){.pop}

Clicking on the sequence button (labelled `"1234.."`) in the Data tab opens the
Sequence Dialog. This allows you to create multiple barcode images by entering a
sequence of data inputs in the right hand panel. Sequences can also be
automatically generated by entering parameters on the left hand side or by
importing the data from a file. Zint will generate a separate barcode image for
each line of text in the right hand panel. The format field determines the
format of the automatically generated sequence where characters have the
meanings as given below:

| Character          | Effect                  |
|:-------------------|:------------------------|
|`$`                 | Insert leading zeroes   |
|`#`                 | Insert leading spaces   |
|`*`                 | Insert leading asterisks|
|Any other character | Interpreted literally   |

Table: {#tbl:sequence_format_characters tag=": Sequence Format Characters"}

Once you're happy with the Sequence Data, click the `"Export..."` button to
bring up the Export Dialog, discussed next.

## 3.9 Export Dialog

![Setting filenames for an exported sequence of barcode
symbols](images/gui_export.png){.pop}

The Export Dialog invoked by pressing the `"Export..."` button in the Sequence
Dialog sets the parameters for exporting the sequence of barcode images. Here
you can set the output directory, the format of the output filenames and what
their image type will be. Note that the symbology, colour and other formatting
information are taken from the main window.

## 3.10 CLI Equivalent Dialog

![CLI Equivalent Dialog](images/gui_cli_equivalent.png){.pop}

The CLI Equivalent Dialog can be invoked from the main menu or the context menu
and displays the CLI command that will reproduce the barcode as currently
configured in the GUI. Press the `"Copy"` button to copy the command to the
clipboard, which can then be pasted into the command line.


# 4. Using the Command Line

This section describes how to encode data using the command line frontend (CLI)
program. The examples given are for the Unix platform, but the same options are
available for Windows - just remember to include the executable file extension
if `".EXE"` is not in your `PATHEXT` environment variable, i.e.:

```bash
zint.exe -d "This Text"
```

For compatibility with Windows the examples use double quotes to delimit data,
though on Unix single quotes are generally preferable as they stop the shell
from processing any characters such as backslash or dollar. A single quote
itself is dealt with by terminating the single-quoted text, backslashing the
single quote, and then continuing:

```bash
zint -d 'Text containing a single quote '\'' in the middle'
```

Some examples use backslash (`\`) to continue commands onto the next line. For
Windows, use caret (`^`) instead.

Certain options that take values have short names as well as long ones, namely
`-b` (`--barcode`), `-d` (`--data`), `-i` (`--input`), `-o` (`--output`) and
`-w` (`--whitesp`). For these a space should be used to separate the short name
from its value, to avoid ambiguity. For long names a space or an equals sign may
be used. For instance:

```bash
zint -d "This Text"
zint --data="This Text"
zint --data "This Text"
```

The examples use a space separator for short option names, and an equals sign
for long option names.

## 4.1 Inputting Data

The data to encode can be entered at the command line using the `-d` or `--data`
option, for example

```bash
zint -d "This Text"
```

This will encode the text `"This Text"`. Zint will use the default symbology,
Code 128, and output to the default file `"out.png"` in the current directory.
Alternatively, if `libpng` was not present when Zint was built, the default
output file will be `"out.gif"`.

The data input to the Zint CLI is assumed to be encoded in UTF-8 (Unicode)
format (Zint will correctly handle UTF-8 data on Windows). If you are encoding
characters beyond the 7-bit ASCII set using a scheme other than UTF-8 then you
will need to set the appropriate input options as shown in [4.11 Input Modes]
below.

Non-printing characters can be entered on the command line using backslash (`\`)
as an escape character in combination with the `--esc` switch. Permissible
sequences are shown in the table below.

---------------------------------------------------------------------------
Escape      ASCII       Name    Interpretation
Sequence    Equivalent
----------  ----------  -----   -------------------------------------------
`\0`        0x00        `NUL`   Null character

`\E`        0x04        `EOT`   End of Transmission

`\a`        0x07        `BEL`   Bell

`\b`        0x08        `BS`    Backspace

`\t`        0x09        `HT`    Horizontal Tab

`\n`        0x0A        `LF`    Line Feed

`\v`        0x0B        `VT`    Vertical Tab

`\f`        0x0C        `FF`    Form Feed

`\r`        0x0D        `CR`    Carriage Return

`\e`        0x1B        `ESC`   Escape

`\G`        0x1D        `GS`    Group Separator

`\R`        0x1E        `RS`    Record Separator

`\\`        0x5C        `\`     Backslash

`\dNNN`     NNN                 Any 8-bit character where NNN is decimal
                                (000-255)

`\oNNN`     0oNNN               Any 8-bit character where NNN is octal
                                (000-377)

`\xNN`      0xNN                Any 8-bit character where NN is hexadecimal
                                (00-FF)

`\uNNNN`                        Any 16-bit Unicode BMP[^2] character where
                                NNNN is hexadecimal (0000-FFFF)

`\UNNNNNN`                      Any 21-bit Unicode character where NNNNNN
                                is hexadecimal (000000-10FFFF)
---------------------------------------------------------------------------

Table: {#tbl:escape_sequences tag=": Escape Sequences"}

[^2]: In Unicode contexts, BMP stands for Basic Multilingual Plane, the plane 0
codeset from U+0000 to U+D7FF and U+E000 to U+FFFF (i.e. excluding surrogates).
Not to be confused with the Windows Bitmap file format BMP!

(Special escape sequences are available for Code 128 only to manually switch
Code Sets and insert special FNC1 characters - see [6.1.10.1 Standard Code 128
(ISO 15417)] for details.)

Input data can be read directly from file using the `-i` or `--input` switch as
shown below. The input file is assumed to be UTF-8 formatted unless an
alternative mode is selected. This option replaces the use of the `-d` switch.

```bash
zint -i somefile.txt
```

To read from stdin specify a single hyphen `"-"` as the input file.

Note that except when batch processing (see [4.12 Batch Processing] below), the
file (or stdin) should not end with a newline (`LF` on Unix, `CR+LF` on Windows)
unless you want the newline to be encoded in the symbol.

## 4.2 Directing Output

Output can be directed to a file other than the default using the `-o` or
`--output` switch. For example:

```bash
zint -o here.png -d "This Text"
```

This draws a Code 128 barcode in the file `"here.png"`. If an Encapsulated
PostScript file is needed simply append the filename with `".eps"`, and so on
for the other supported file types:

```bash
zint -o there.eps -d "This Text"
```

The currently supported output file formats are shown in the following table.

Extension   File format
---------   ------------------------------------
bmp         Windows Bitmap
emf         Enhanced Metafile Format
eps         Encapsulated PostScript
gif         Graphics Interchange Format
pcx         ZSoft Paintbrush image
png         Portable Network Graphic
svg         Scalable Vector Graphic
tif         Tagged Image File Format
txt         Text file (see [4.19 Other Options])

Table: {#tbl:output_file_formats tag=": Output File Formats"}

The filename can contain directories and sub-directories also, which will be
created if they don't already exist:

```bash
zint -o "dir/subdir/filename.eps" -d "This Text"
```

Note that on Windows, filenames are assumed to be UTF-8 encoded.

## 4.3 Selecting Barcode Type

Selecting which type of barcode you wish to produce (i.e. which symbology to
use) can be done at the command line using the `-b` or `--barcode` switch
followed by the appropriate integer value or name in the following table. For
example to create a Data Matrix symbol you could use:

```bash
zint -b 71 -o datamatrix.png -d "Data to encode"
```

or

```bash
zint -b DATAMATRIX -o datamatrix.png -d "Data to encode"
```

Names are treated case-insensitively by the CLI, and the `BARCODE_` prefix and
any underscores are optional.

-----------------------------------------------------------------------------
Numeric  Name[^3]                  Barcode Name
Value
-------  ------------------------  ------------------------------------------
 1       `BARCODE_CODE11`          Code 11

 2`*`    `BARCODE_C25STANDARD`     Standard Code 2 of 5

 3       `BARCODE_C25INTER`        Interleaved 2 of 5

 4       `BARCODE_C25IATA`         Code 2 of 5 IATA

 6       `BARCODE_C25LOGIC`        Code 2 of 5 Data Logic

 7       `BARCODE_C25IND`          Code 2 of 5 Industrial

 8       `BARCODE_CODE39`          Code 3 of 9 (Code 39)

 9       `BARCODE_EXCODE39`        Extended Code 3 of 9 (Code 39+)

 13      `BARCODE_EANX`            EAN (EAN-2, EAN-5, EAN-8 and EAN-13)

 14      `BARCODE_EANX_CHK`        EAN + Check Digit

 16`*`   `BARCODE_GS1_128`         GS1-128 (UCC.EAN-128)

 18      `BARCODE_CODABAR`         Codabar

 20      `BARCODE_CODE128`         Code 128 (automatic Code Set switching)

 21      `BARCODE_DPLEIT`          Deutsche Post Leitcode

 22      `BARCODE_DPIDENT`         Deutsche Post Identcode

 23      `BARCODE_CODE16K`         Code 16K

 24      `BARCODE_CODE49`          Code 49

 25      `BARCODE_CODE93`          Code 93

 28      `BARCODE_FLAT`            Flattermarken

 29`*`   `BARCODE_DBAR_OMN`        GS1 DataBar Omnidirectional (including GS1
                                      DataBar Truncated)

 30`*`   `BARCODE_DBAR_LTD`        GS1 DataBar Limited

 31`*`   `BARCODE_DBAR_EXP`        GS1 DataBar Expanded

 32      `BARCODE_TELEPEN`         Telepen Alpha

 34      `BARCODE_UPCA`            UPC-A

 35      `BARCODE_UPCA_CHK`        UPC-A + Check Digit

 37      `BARCODE_UPCE`            UPC-E

 38      `BARCODE_UPCE_CHK`        UPC-E + Check Digit

 40      `BARCODE_POSTNET`         POSTNET

 47      `BARCODE_MSI_PLESSEY`     MSI Plessey

 49      `BARCODE_FIM`             FIM

 50      `BARCODE_LOGMARS`         LOGMARS

 51      `BARCODE_PHARMA`          Pharmacode One-Track

 52      `BARCODE_PZN`             PZN

 53      `BARCODE_PHARMA_TWO`      Pharmacode Two-Track

 54      `BARCODE_CEPNET`          Brazilian CEPNet

 55      `BARCODE_PDF417`          PDF417

 56`*`   `BARCODE_PDF417COMP`      Compact PDF417 (Truncated PDF417)

 57      `BARCODE_MAXICODE`        MaxiCode

 58      `BARCODE_QRCODE`          QR Code

 60      `BARCODE_CODE128AB`       Code 128 (Suppress Code Set C)

 63      `BARCODE_AUSPOST`         Australia Post Standard Customer

 66      `BARCODE_AUSREPLY`        Australia Post Reply Paid

 67      `BARCODE_AUSROUTE`        Australia Post Routing

 68      `BARCODE_AUSDIRECT`       Australia Post Redirection

 69      `BARCODE_ISBNX`           ISBN (EAN-13 with verification stage)

 70      `BARCODE_RM4SCC`          Royal Mail 4-State Customer Code (RM4SCC)

 71      `BARCODE_DATAMATRIX`      Data Matrix (ECC200)

 72      `BARCODE_EAN14`           EAN-14

 73      `BARCODE_VIN`             Vehicle Identification Number

 74      `BARCODE_CODABLOCKF`      Codablock-F

 75      `BARCODE_NVE18`           NVE-18 (SSCC-18)

 76      `BARCODE_JAPANPOST`       Japanese Postal Code

 77      `BARCODE_KOREAPOST`       Korea Post

 79`*`   `BARCODE_DBAR_STK`        GS1 DataBar Stacked

 80`*`   `BARCODE_DBAR_OMNSTK`     GS1 DataBar Stacked Omnidirectional

 81`*`   `BARCODE_DBAR_EXPSTK`     GS1 DataBar Expanded Stacked

 82      `BARCODE_PLANET`          PLANET

 84      `BARCODE_MICROPDF417`     MicroPDF417

 85`*`   `BARCODE_USPS_IMAIL`      USPS Intelligent Mail (OneCode)

 86      `BARCODE_PLESSEY`         UK Plessey

 87      `BARCODE_TELEPEN_NUM`     Telepen Numeric

 89      `BARCODE_ITF14`           ITF-14

 90      `BARCODE_KIX`             Dutch Post KIX Code

 92      `BARCODE_AZTEC`           Aztec Code

 93      `BARCODE_DAFT`            DAFT Code

 96      `BARCODE_DPD`             DPD Code

 97      `BARCODE_MICROQR`         Micro QR Code

 98      `BARCODE_HIBC_128`        HIBC Code 128

 99      `BARCODE_HIBC_39`         HIBC Code 39

 102     `BARCODE_HIBC_DM`         HIBC Data Matrix ECC200

 104     `BARCODE_HIBC_QR`         HIBC QR Code

 106     `BARCODE_HIBC_PDF`        HIBC PDF417

 108     `BARCODE_HIBC_MICPDF`     HIBC MicroPDF417

 110     `BARCODE_HIBC_BLOCKF`     HIBC Codablock-F

 112     `BARCODE_HIBC_AZTEC`      HIBC Aztec Code

 115     `BARCODE_DOTCODE`         DotCode

 116     `BARCODE_HANXIN`          Han Xin (Chinese Sensible) Code

 119     `BARCODE_MAILMARK_2D`     Royal Mail 2D Mailmark (CMDM) (Data
                                   Matrix)

 121     `BARCODE_MAILMARK_4S`     Royal Mail 4-State Mailmark

 128     `BARCODE_AZRUNE`          Aztec Runes

 129     `BARCODE_CODE32`          Code 32

 130     `BARCODE_EANX_CC`         GS1 Composite Symbol with EAN linear
                                      component

 131`*`  `BARCODE_GS1_128_CC`      GS1 Composite Symbol with GS1-128 linear
                                      component

 132`*`  `BARCODE_DBAR_OMN_CC`     GS1 Composite Symbol with GS1 DataBar
                                      Omnidirectional linear component

 133`*`  `BARCODE_DBAR_LTD_CC`     GS1 Composite Symbol with GS1 DataBar
                                      Limited linear component

 134`*`  `BARCODE_DBAR_EXP_CC`     GS1 Composite Symbol with GS1 DataBar
                                      Expanded linear component

 135     `BARCODE_UPCA_CC`         GS1 Composite Symbol with UPC-A linear
                                      component

 136     `BARCODE_UPCE_CC`         GS1 Composite Symbol with UPC-E linear
                                      component

 137`*`  `BARCODE_DBAR_STK_CC`     GS1 Composite Symbol with GS1 DataBar
                                      Stacked component

 138`*`  `BARCODE_DBAR_OMNSTK_CC`  GS1 Composite Symbol with GS1 DataBar
                                      Stacked Omnidirectional component

 139`*`  `BARCODE_DBAR_EXPSTK_CC`  GS1 Composite Symbol with GS1 DataBar
                                      Expanded Stacked component

 140     `BARCODE_CHANNEL`         Channel Code

 141     `BARCODE_CODEONE`         Code One

 142     `BARCODE_GRIDMATRIX`      Grid Matrix

 143     `BARCODE_UPNQR`           UPNQR (Univerzalnega Plačilnega Naloga QR)

 144     `BARCODE_ULTRA`           Ultracode

 145     `BARCODE_RMQR`            Rectangular Micro QR Code (rMQR)

 146     `BARCODE_BC412`           IBM BC412 (SEMI T1-95)
-----------------------------------------------------------------------------

Table: {#tbl:barcode_types tag=": Barcode Types (Symbologies)"}

[^3]: The symbologies marked with an asterisk (`*`) in Table
{@tbl:barcode_types} above used different names in Zint before version 2.9.0.
For example, symbology 29 used the name `BARCODE_RSS14`. These names are now
deprecated but are still recognised by Zint and will continue to be supported in
future versions.

## 4.4 Adjusting Height

The height of a symbol (except those with a fixed width-to-height ratio) can be
adjusted using the `--height` switch. For example:

```bash
zint --height=100 -d "This Text"
```

This specifies a symbol height of 100 times the X-dimension of the symbol.

The default height of most linear barcodes is 50.0X, but this can be changed for
barcodes whose specifications give a standard height by using the switch
`--compliantheight`. For instance

```bash
zint -b LOGMARS -d "This Text" --compliantheight
```

will produce a barcode of height 45.455X instead of the normal default of 50.0X.
The flag also causes Zint to return a warning if a non-compliant height is
given:

```bash
zint -b LOGMARS -d "This Text" --compliantheight --height=6.2
Warning 247: Height not compliant with standards
```

Another switch is `--heightperrow`, which can be useful for symbologies that
have a variable number of linear rows, namely Codablock-F, Code 16K, Code 49,
GS1 DataBar Expanded Stacked, MicroPDF417 and PDF417, as it changes the
treatment of the height value from overall height to per-row height, allowing
you to specify a consistent height for each linear row without having to know
how many there are. For instance

```bash
zint -b PDF417 -d "This Text" --height=4 --heightperrow
```

![`zint -b PDF417 -d "This Text" --height=4
--heightperrow`](images/pdf417_heightperrow.svg){.lin}

will produce a barcode of height 32X, with each of the 8 rows 4X high.

## 4.5 Adjusting Whitespace

The amount of horizontal whitespace to the left and right of the generated
barcode can be altered using the `-w` or `--whitesp` switch, in integral
multiples of the X-dimension. For example:

```bash
zint -w 10 -d "This Text"
```

This specifies a whitespace width of 10 times the X-dimension of the symbol both
to the left and to the right of the barcode.

The amount of vertical whitespace above and below the barcode can be altered
using the `--vwhitesp` switch, in integral multiples of the X-dimension. For
example for 3 times the X-dimension:

```bash
zint --vwhitesp=3 -d "This Text"
```

Note that the whitespace at the bottom appears below the text, if any.

Horizontal and vertical whitespace can of course be used together:

```bash
zint -b DATAMATRIX --whitesp=1 --vwhitesp=1 -d "This Text"
```

A `--quietzones` option is also available which adds quiet zones compliant with
the symbology's specification. This is in addition to any whitespace specified
with the `--whitesp` or `--vwhitesp` switches.

Note that Codablock-F, Code 16K, Code 49, ITF-14, EAN-2 to EAN-13, ISBN, UPC-A
and UPC-E have compliant quiet zones added by default. This can be disabled with
the option `--noquietzones`.

## 4.6 Adding Boundary Bars and Boxes

Zint allows the symbol to be bound with 'boundary bars' (also known as 'bearer
bars') using the option `--bind`. These bars help to prevent misreading of the
symbol by corrupting a scan if the scanning beam strays off the top or bottom of
the symbol. Zint can also put a border right around the symbol and its
horizontal whitespace with the `--box` option.

The width of the boundary bars or box borders, in integral multiples of the
X-dimension, must be specified using the `--border` switch. For example:

```bash
zint --box --border=10 -w 10 -d "This Text"
```

![`zint --border=10 --box -d "This Text" -w 10`](images/code128_box.svg){.lin}

gives a box with a width 10 times the X-dimension of the symbol. Note that when
specifying a box, horizontal whitespace is usually required in order to create a
quiet zone between the barcode and the sides of the box. To add a boundary bar
to the top only use `--bindtop`.

For linear symbols, horizontal boundary bars appear tight against the barcode,
inside any vertical whitespace (or text). For matrix symbols, however, where
they are decorative rather than functional, boundary bars appear outside any
whitespace.

![`zint -b QRCODE --border=1 --box -d "This Text"
--quietzones`](images/qrcode_box.svg){.i2d}

Codablock-F, Code 16K and Code 49 always have boundary bars, and default to
particular horizontal whitespace values. Special considerations apply to ITF-14
and DPD - see [6.1.2.6 ITF-14] and [6.1.10.7 DPD Code] for those symbologies.

## 4.7 Using Colour

The default colours of a symbol are a black symbol on a white background. Zint
allows you to change this. The `-r` or `--reverse` switch allows the default
colours to be inverted so that a white symbol is shown on a black background
(known as "reflectance reversal" or "reversed reflectance"). For example the
command

```bash
zint -r -d "This Text"
```

gives an inverted Code 128 symbol. This is not practical for most symbologies
but white-on-black is allowed by the Aztec Code, Data Matrix, DotCode, Han Xin
Code, Grid Matrix and QR Code symbology specifications.

For more specific needs the foreground (ink) and background (paper) colours can
be specified using the `--fg` and `--bg` options followed by a number in
`"RRGGBB"` hexadecimal notation (the same system used in HTML) or in `"C,M,Y,K"`
decimal percentages format (the latter normally used with the `--cmyk` option -
see below). For example the command

```bash
zint --fg=00FF00 -d "This Text"
```

alters the symbol to a bright green.

![`zint -d "This Text" --fg=00FF00`](images/code128_green.svg){.lin}

Zint also supports RGBA colour information for those output file formats which
support alpha channels (currently only GIF, PCX, PNG, SVG and TIF, with GIF
supporting either a background or foreground alpha but not both) in a
`"RRGGBBAA"` format. For example:

```bash
zint --fg=00ff0055 -d "This Text"
```

![`zint -d "This Text" --fg=00FF0055`](images/code128_green_alpha.svg){.lin}

will produce a semi-transparent green foreground with a standard (white)
background. Note that transparency is treated differently by raster and vector
(SVG) output formats, as for vector output the background will "shine through" a
transparent foreground. For instance

```bash
zint --bg=ff0000 --fg=ffffff00 ...
```

will give different results for PNG and SVG. Experimentation is advised!

In addition the `--nobackground` option will remove the background from all
output formats except BMP.[^4]

The `--cmyk` option is specific to output in Encapsulated PostScript (EPS) and
TIF, and selects the CMYK colour space. Custom colours should then usually be
given in the comma-separated `"C,M,Y,K"` format, where `C`, `M`, `Y` and `K` are
expressed as decimal percentage values from 0 to 100. RGB values may still be
used, in which case they will be converted formulaically to CMYK approximations.

[^4]: The background is omitted for vector outputs EMF, EPS and SVG when
`--nobackground` is given. For raster outputs GIF, PCX, PNG and TIF, the
background's alpha channel is set to zero (fully transparent).

## 4.8 Rotating the Symbol

The symbol can be rotated through four orientations using the `--rotate` option
followed by the angle of rotation as shown below.

```
--rotate=0 (default)
--rotate=90
--rotate=180
--rotate=270
```

![`zint -d "This Text" --rotate=90`](images/code128_rotate90.svg){.lin}

\clearpage

## 4.9 Adjusting Image Size (X-dimension)

The size of the image can be altered using the `--scale` option, which sets the
X-dimension. The default scale is 1.0.

The scale is multiplied by 2 (with the exception of MaxiCode) before being
applied to the X-dimension. For MaxiCode, it is multiplied by 10 for raster
output, by 40 for EMF vector output, and by 2 otherwise (non-EMF vector output).

For non-MaxiCode raster output, the default scale of 1 results in an X-dimension
of 2 pixels. For example for non-MaxiCode PNG images a scale of 5 will increase
the X-dimension to 10 pixels. For MaxiCode, see [4.9.3 MaxiCode Raster Scaling]
below.

Scales for non-MaxiCode raster output should be given in increments of 0.5, i.e.
0.5, 1, 1.5, 2, 2.5, 3, 3.5, etc., to avoid the X-dimension varying across the
symbol due to interpolation. 0.5 increments are also faster to render.

The minimum scale for non-MaxiCode raster output in non-dotty mode is 0.5,
giving a minimum X-dimension of 1 pixel. For MaxiCode, it is 0.2. The minimum
scale for raster output in dotty mode is 1 (see [4.15 Working with Dots]). For
raster output, text will not be printed for scales less than 1.

The minimum scale for vector output is 0.1, giving a minimum X-dimension of 0.2
(or for MaxiCode EMF output, 4). The maximum scale for both raster and vector is
200.

To summarize the more intricate details:

-----------------------------------------------------------------
MaxiCode?  Output            Multiplier  Min. Scale    Min. Scale
                                         (non-dotty)   (dotty)
---------  ----------------  ----------  ------------  ----------
No         Raster            2           0.5           1

No         Vector            2           0.1           0.1

Yes        Raster            10          0.2           N/A

Yes        Vector (non-EMF)  2           0.1           N/A

Yes        EMF               40          0.1           N/A
-----------------------------------------------------------------

Table: {#tbl:scaling_multiplers tag=": Scaling Multipliers and Minima"}

### 4.9.1 Scaling by X-dimension and Resolution

An alternative way to specify the scale, which takes the above details into
account, is to specify measurable units using the `--scalexdimdp` option, which
has the format

```
--scalexdimdp=X[,R]
```

where `X` is the X-dimension (in mm by default) and `R` is the resolution (in
dpmm, dots per mm, by default). `R` is optional, and defaults to 12 dpmm, and
`X` may be zero, in which case it uses a symbology-specific default. The units
may be given in inches for `X` by appending `"in"`, and in dpi (dots per inch)
for `R` by appending `"dpi"`. For example

```bash
zint -d "1234" --scalexdimdp=0.013in,300dpi
```

Explicit metric units may also be given by appending `"mm"` and `"dpmm"` as
appropriate, and may be mixed with U.S. units:

```bash
zint -d "1234" --scalexdimdp=0.33mm,300dpi
```

### 4.9.2 Scaling Example

The GS1 General Specifications Section 5.2.6.6 'Symbol dimensions at nominal
size' gives an example of an EAN-13 barcode using the X-dimension of 0.33mm. To
print that example as a PNG at 12 dpmm, the approximate equivalent of 300 dpi
(`dpi = dpmm * 25.4`), specify a scale of 2, since `0.33 * 12 = 3.96` pixels, or
4 pixels rounding to the nearest pixel:

```bash
zint -b EANX -d "501234567890" --compliantheight --scale=2
```

This will result in output of 37.29mm x 25.56mm (WxH) at 12 dpmm. The same
result can be achieved using the `--scalexdimdp` option with

```bash
zint -b EANX -d "501234567890" --compliantheight --scalexdimdp=0
```

as 0.33mm is the default X-dimension for EAN, and 12 dpmm the default
resolution.

### 4.9.3 MaxiCode Raster Scaling

For MaxiCode symbols, which use hexagons, the scale for raster output is
multiplied by 10 before being applied. The 0.5 increment recommended for normal
raster output does not apply.

The minimum scale is 0.2, so the minimum X-dimension is 2 pixels. However scales
below 0.5 are not recommended and may produce symbols that are not within the
following size ranges.

MaxiCode symbols have fixed size ranges of 24.82mm to 27.93mm in width, and
23.71mm to 26.69mm in height, excluding quiet zones. The default X-dimension is
0.88mm. For example, to output at the default X-dimension at 600 dpi specify:

```bash
zint -b MAXICODE -d "MaxiCode (19 chars)" --scalexdimdp=0,600dpi
```

## 4.10 Human Readable Text (HRT) Options

For linear barcodes the text present in the output image can be removed by using
the `--notext` option. Note also that for raster output text will not be printed
for scales less than 1 (see [4.9 Adjusting Image Size (X-dimension)]).

Text can be set to bold using the `--bold` option, or a smaller font can be
substituted using the `--small` option. The `--bold` and `--small` options can
be used together if required, but only for vector output.

![`zint --bold -d "This Text" --small`](images/code128_small_bold.svg){.lin}

The gap between the barcode and the text can be adjusted using the `--textgap`
option, where the gap is given in X-dimensions, and may be negative (minimum
-5.0X, maximum 10.0X). The default gap is 1X. Note that a very small gap may
cause accented texts to overlap with the barcode:

![`zint -d "Áccent" --textgap=0.1`](images/code128_textgap.svg){.lin}

For SVG output, the font preferred by Zint (monospaced "OCR-B" for EAN/UPC,
"Arimo" - a proportional sans-serif font metrically compatible with "Arial" -
for all others) can be embedded in the file for portability using the
`--embedfont` option:

![`zint -d "Áccent" --embedfont`](images/code128_embedfont.svg){.lin}

## 4.11 Input Modes

### 4.11.1 Unicode, Data, and GS1 Modes

By default all CLI input data is assumed to be encoded in UTF-8 format. Many
barcode symbologies encode data using the Latin-1 (ISO/IEC 8859-1 plus ASCII)
character set, so input is converted from UTF-8 to Latin-1 before being put in
the symbol. In addition QR Code and its variants and Han Xin Code can by default
encode Japanese (Kanji) or Chinese (Hanzi) characters which are also converted
from UTF-8.

There are two exceptions to the Latin-1 default: Grid Matrix, whose default
character set is GB 2312 (Chinese); and UPNQR, whose default character set is
Latin-2 (ISO/IEC 8859-2 plus ASCII).

Symbology      Default character sets    Alternate if input not Latin-1
-------------  ------------------------  ------------------------------
Aztec Code     Latin-1                   None
Codablock-F    Latin-1                   None
Code 128       Latin-1                   None
Code 16K       Latin-1                   None
Code One       Latin-1                   None
Data Matrix    Latin-1                   None
DotCode        Latin-1                   None
Grid Matrix    GB 2312 (includes ASCII)  N/A
Han Xin        Latin-1                   GB 18030 (includes ASCII)
MaxiCode       Latin-1                   None
MicroPDF417    Latin-1                   None
Micro QR Code  Latin-1                   Shift JIS (includes ASCII[^5])
PDF417         Latin-1                   None
QR Code        Latin-1                   Shift JIS (see above)
rMQR           Latin-1                   Shift JIS (see above)
Ultracode      Latin-1                   None
UPNQR          Latin-2                   N/A
All others     ASCII                     N/A

Table: {#tbl:default_character_sets tag=": Default Character Sets"}

[^5]: Shift JIS (JIS X 0201 Roman) re-maps two ASCII characters: backslash (`\`)
to the yen sign (¥), and tilde (`~`) to overline (U+203E).

If Zint encounters characters which can not be encoded using the default
character encoding then it will take advantage of the ECI (Extended Channel
Interpretations) mechanism to encode the data if the symbology supports it - see
[4.11.2 Input Modes and ECI] below.

GS1 data can be encoded in a number of symbologies. Application Identifiers
(AIs) should be enclosed in `[square brackets]` followed by the data to be
encoded (see [6.1.10.3 GS1-128]). To encode GS1 data use the `--gs1` option.
GS1 mode is assumed (and doesn't need to be set) for GS1-128, EAN-14, GS1
DataBar and GS1 Composite symbologies but is also available for Aztec Code, Code
16K, Code 49, Code One, Data Matrix, DotCode, QR Code and Ultracode.

Health Industry Barcode (HIBC) data may also be encoded in the symbologies Aztec
Code, Codablock-F, Code 128, Code 39, Data Matrix, MicroPDF417, PDF417 and QR
Code. Within this mode, the leading `'+'` and the check character are
automatically added by Zint, conforming to HIBC Labeler Identification Code
(HIBC LIC). For HIBC Provider Applications Standard (HIBC PAS), preface the data
with a slash `'/'`.

The `--binary` option encodes the input data as given. Automatic code page
translation to an ECI page is disabled, and no validation of the data's encoding
takes place. This may be used for raw binary or binary encrypted data. This
switch plays together with the built-in ECI logic and examples may be found
below.

The `--fullmultibyte` option uses the multibyte modes of QR Code, Micro QR Code,
Rectangular Micro QR Code, Han Xin Code and Grid Matrix for non-ASCII data,
maximizing density. This is achieved by using compression designed for
Kanji/Hanzi characters; however some decoders take blocks which are encoded this
way and interpret them as Kanji/Hanzi characters, thus causing data corruption.
Symbols encoded with this option should be checked against decoders before they
are used. The popular open-source ZXing decoder is known to exhibit this
behaviour.

### 4.11.2 Input Modes and ECI

If your data contains characters that are not in the default character set, you
may encode it using an ECI-aware symbology and an ECI value from Table
{@tbl:eci_codes} below. The ECI information is added to your code symbol as
prefix data. The symbologies that support ECI are

------------  ------------  ------------
Aztec Code    Grid Matrix   PDF417
Code One      Han Xin Code  QR Code
Data Matrix   MaxiCode      rMQR
DotCode       MicroPDF417   Ultracode
------------  ------------  ------------

Table: {#tbl:eci_aware_symbologies tag=": ECI-Aware Symbologies"}

Be aware that not all barcode readers support ECI mode, so this can sometimes
lead to unreadable barcodes. If you are using characters beyond those supported
by the default character set then you should check that the resulting barcode
can be understood by your target barcode reader.

The ECI value may be specified with the `--eci` switch, followed by the value in
the column `"ECI Code"` in the table below. The input data should be UTF-8
formatted. Zint automatically translates the data into the target encoding.

ECI Code  Character Encoding Scheme (ISO/IEC 8859 schemes include ASCII)
--------  --------------------------------------------------------------
3         ISO/IEC 8859-1 - Latin alphabet No. 1
4         ISO/IEC 8859-2 - Latin alphabet No. 2
5         ISO/IEC 8859-3 - Latin alphabet No. 3
6         ISO/IEC 8859-4 - Latin alphabet No. 4
7         ISO/IEC 8859-5 - Latin/Cyrillic alphabet
8         ISO/IEC 8859-6 - Latin/Arabic alphabet
9         ISO/IEC 8859-7 - Latin/Greek alphabet
10        ISO/IEC 8859-8 - Latin/Hebrew alphabet
11        ISO/IEC 8859-9 - Latin alphabet No. 5 (Turkish)
12        ISO/IEC 8859-10 - Latin alphabet No. 6 (Nordic)
13        ISO/IEC 8859-11 - Latin/Thai alphabet
15        ISO/IEC 8859-13 - Latin alphabet No. 7 (Baltic)
16        ISO/IEC 8859-14 - Latin alphabet No. 8 (Celtic)
17        ISO/IEC 8859-15 - Latin alphabet No. 9
18        ISO/IEC 8859-16 - Latin alphabet No. 10
20        Shift JIS (JIS X 0208 and JIS X 0201)
21        Windows 1250 - Latin 2 (Central Europe)
22        Windows 1251 - Cyrillic
23        Windows 1252 - Latin 1
24        Windows 1256 - Arabic
25        UTF-16BE (High order byte first)
26        UTF-8
27        ASCII (ISO/IEC 646 IRV)
28        Big5 (Taiwan) Chinese Character Set
29        GB 2312 (PRC) Chinese Character Set
30        Korean Character Set EUC-KR (KS X 1001:2002)
31        GBK Chinese Character Set
32        GB 18030 Chinese Character Set
33        UTF-16LE (Low order byte first)
34        UTF-32BE (High order bytes first)
35        UTF-32LE (Low order bytes first)
170       ISO/IEC 646 Invariant[^6]
899       8-bit binary data

Table: {#tbl:eci_codes tag=": ECI Codes"}

[^6]: ISO/IEC 646 Invariant is a subset of ASCII with 12 characters undefined:
`#`, `$`, `@`, `[`, `\`, `]`, `^`, `` ` ``, `{`, `|`, `}`, `~`.

An ECI value of 0 does not encode any ECI information in the code symbol (unless
the data contains non-default character set characters). In this case, the
default character set applies (see Table @tbl:default_character_sets above).

If no ECI is specified or a value of 0 is given, and the data does contain
characters other than in the default character set, then Zint will automatically
insert the appropriate single-byte ECI if possible (ECIs 3 to 24, excluding ECI
20), or failing that ECI 26 (UTF-8). A warning will be generated. This mechanism
is not applied if the `--binary` option is given.

Multiple ECIs can be specified using the `--segN` options - see [4.16 Multiple
Segments].

Note: the `--eci=3` specification should only be used for special purposes.
Using this parameter, the ECI information is explicitly added to the symbol.
Nevertheless, for ECI Code 3, this is not usually required, as this is the
default encoding for most barcodes, which is also active without any ECI
information.

#### 4.11.2.1 Input Modes and ECI Example 1

The Euro sign U+20AC can be encoded in ISO/IEC 8859-15. The Euro sign has the
ISO/IEC 8859-15 codepoint hex `"A4"`. It is encoded in UTF-8 as the hex
sequence: `"E2 82 AC"`. Those 3 bytes are contained in the file
`"utf8euro.txt"`. This command will generate the corresponding code:

```bash
zint -b 71 --scale=10 --eci=17 -i utf8euro.txt
```

This is equivalent to the commands (using the `--esc` switch):

```bash
zint -b 71 --scale=10 --eci=17 --esc -d "\xE2\x82\xAC"

zint -b 71 --scale=10 --eci=17 --esc -d "\u20AC"
```

and to the command:

```bash
zint -b 71 --scale=10 --eci=17 -d "€"
```

![`zint -b DATAMATRIX --eci=17 -d "€"`](images/datamatrix_euro.svg){.i2d}

#### 4.11.2.2 Input Modes and ECI Example 2

The Chinese character with the Unicode codepoint U+5E38 can be encoded in Big5
encoding. The Big5 representation of this character is the two hex bytes:
`"B1 60"` (contained in the file `"big5char.txt"`). The generation command for
Data Matrix is:

```bash
zint -b 71 --scale=10 --eci=28 --binary -i big5char.txt
```

This is equivalent to the command (using the `--esc` switch):

```bash
zint -b 71 --scale=10 --eci=28 --binary --esc -d "\xB1\x60"
```

and to the commands (no `--binary` switch so conversion occurs):

```bash
zint -b 71 --scale=10 --eci=28 --esc -d "\xE5\xB8\xB8"

zint -b 71 --scale=10 --eci=28 --esc -d "\u5E38"

zint -b 71 --scale=10 --eci=28 -d "常"
```

![`zint -b DATAMATRIX --eci=28 -d "\u5E38"
--esc`](images/datamatrix_big5.svg){.i2d}

#### 4.11.2.3 Input Modes and ECI Example 3

Some decoders (in particular mobile app ones) for QR Code assume UTF-8 encoding
by default and do not support ECI. In this case supply UTF-8 data and use the
`--binary` switch so that the data will be encoded as UTF-8 without conversion:

```bash
zint -b 58 --binary -d "UTF-8 data"
```

![`zint -b QRCODE --binary -d "\xE2\x82\xAC\xE5\xB8\xB8"
--esc`](images/qrcode_binary_utf8.svg){.i2d}

## 4.12 Batch Processing

Data can be batch processed by reading from a text file and producing a
separate barcode image for each line of text in that file. To do this use the
`--batch` switch together with `-i` to select the input file from which to read
data. For example

```bash
zint -b EANX --batch -i ean13nos.txt
```

where `"ean13nos.txt"` contains a list of EAN-13 numbers (GTINs), each on its
own line. Zint will automatically detect the end of a line of text (in either
Unix or Windows formatted text files) and produce a symbol each time it finds
this.

Input files should end with a line feed character - if this is not present then
Zint will not encode the last line of text, and will warn you that there is a
problem.

By default Zint will output numbered filenames starting with `00001.png`,
`00002.png` etc. To change this behaviour specify the `-o` option using special
characters in the output filename as shown in the table below:

Input Character      Interpretation
---------------      ------------------------------------------
`~`                  Insert a number or 0
`#`                  Insert a number or space
`@`                  Insert a number or `*` (or `+` on Windows)
Any other            Insert literally

Table: {#tbl:batch_filename_formatting tag=": Batch Filename Formatting"}

For instance

```bash
zint -b EANX --batch -i ean13nos.txt -o file~~~.svg
```

The following table shows some examples to clarify this method:

Input                Filenames Generated
-----------------    ---------------------------------------------------------
`-o file~~~.svg`     `"file001.svg"`, `"file002.svg"`, `"file003.svg"`
`-o @@@@bar.png`     `"***1.png"`, `"***2.png"`, `"***3.png"` (except Windows)
`-o @@@@bar.png`     `"+++1.png"`, `"+++2.png"`, `"+++3.png"` (on Windows)
`-o my~~~bar.eps`    `"my001bar.eps"`, `"my002bar.eps"`, `"my003bar.eps"`
`-o t#es~t~.png`     `"t es0t1.png"`, `"t es0t2.png"`, `"t es0t3.png"`

Table: {#tbl:batch_filename_examples tag=": Batch Filename Examples"}

The special characters can span directories also, which is useful when creating
a large number of barcodes:

Input                    Filenames Generated
--------------------     ------------------------------------------------------
`-o dir~/file~~~.svg`    `"dir0/file001.svg"`, `"dir0/file002.svg"`, ...
                         , `"dir0/file999.svg"`, `"dir1/file000.svg"`, ...

Table: {#tbl:batch_dir_examples tag=": Batch Directory Examples"}

For an alternative method of naming output files see the `--mirror` option in
[4.14 Automatic Filenames] below.

## 4.13 Direct Output to stdout

The finished image files can be output directly to stdout for use as part of a
pipe by using the `--direct` option. By default `--direct` will output data as a
PNG image (or GIF image if `libpng` is not present), but this can be altered by
supplementing the `--direct` option with a `--filetype` option followed by the
suffix of the file type required. For example:

```bash
zint -b 84 --direct --filetype=pcx -d "Data to encode"
```

This command will output the symbol as a PCX file to stdout. For the supported
output file formats see Table {@tbl:output_file_formats}.

* * *
CAUTION: Outputting binary files to the command shell without catching that data
in a pipe can have unpredictable results. Use with care!

* * *

## 4.14 Automatic Filenames

The `--mirror` option instructs Zint to use the data to be encoded as an
indicator of the filename to be used. This is particularly useful if you are
processing batch data. For example the input data `"1234567"` will result in a
file named `"1234567.png"`.

There are restrictions, however, on what characters can be stored in a filename,
so the filename may vary from the data if the data includes non-printable
characters, for example, and may be shortened if the data input is long.

To set the output file format use the `--filetype` option as detailed above in
[4.13 Direct Output to stdout]. To output to a specific directory use the `-o`
option giving the name of the directory (any filename will be ignored, unless
`--filetype` is not specified, in which case the filename's extension will be
used).

## 4.15 Working with Dots

Matrix codes can be rendered as a series of dots or circles rather than the
normal squares by using the `--dotty` option. This option is only available for
matrix symbologies, and is automatically selected for DotCode. The size of the
dots can be adjusted using the `--dotsize` option followed by the diameter of
the dot, where that diameter is in X-dimensions. The minimum dot size is 0.01,
the maximum is 20. The default size is 0.8.

The default and minimum scale for raster output in dotty mode is 1.

![`zint -b CODEONE -d "123456789012345678" --dotty
--vers=9`](images/codeone_s_dotty.svg){.dotty}

## 4.16 Multiple Segments

If you need to specify different ECIs for different sections of the input data,
the `--seg1` to `--seg9` options can be used. Each option is of the form
`--segN=ECI,data` where `ECI` is the ECI code (see Table {@tbl:eci_codes}) and
`data` is the data to which this applies. This is in addition to the ECI and
data specified using the `--eci` and `-d` options which must still be present
and which in effect constitute segment 0. For instance

```bash
zint -b AZTEC_CODE --eci=9 -d "Κείμενο" --seg1=7,"Текст" --seg2=20,"文章"
```

specifies 3 segments: segment 0 with ECI 9 (Greek), segment 1 with ECI 7
(Cyrillic), and segment 2 with ECI 20 (Shift JIS). Segments must be consecutive.

Naturally the symbology must be ECI-aware (see Table
{@tbl:eci_aware_symbologies}).

![`zint -b AZTEC --eci=9 -d "Κείμενο" --seg1=7,"Текст"
--seg2=20,"文章"`](images/aztec_segs.svg){.i2d}

ECIs of zero may be given, in which case Zint will automatically determine an
ECI if necessary, as described in section [4.11.2 Input Modes and ECI].

Multiple segments are not currently supported for use with GS1 data.

## 4.17 Structured Append

Structured Append is a method of splitting data among several symbols so that
they form a sequence that can be scanned and re-assembled in the correct order
on reading, and is available for Aztec Code, Code One, Data Matrix, DotCode,
Grid Matrix, MaxiCode, MicroPDF417, PDF417, QR Code and Ultracode.

The `--structapp` option marks a symbol as part of a Structured Append sequence,
and has the format

```
--structapp=I,C[,ID]
```

![`zint -b DATAMATRIX -d "2nd of 3"
--structapp="2,3,5006"`](images/datamatrix_structapp.svg){.i2d}

where `I` is the index (position) of the symbol in the Structured Append
sequence, `C` is the count or total number of symbols in the sequence, and `ID`
is an optional identifier (not available for Code One, DotCode or MaxiCode) that
is the same for all symbols belonging to the same sequence. The index is 1-based
and goes from 1 to count. Count must be 2 or more. See the individual
symbologies for further details.

## 4.18 Help Options

There are three help options which give information about how to use the command
line. The `-h` or `--help` option will display a list of all of the valid
options available, and also gives the exact version of the software (the version
by itself can be displayed with `-v` or `--version`).

The `-t` or `--types` option gives the table of symbologies along with the
symbol ID numbers and names.

The `-e` or `--ecinos` option gives a list of the ECI codes.

## 4.19 Other Options

Zint can output a representation of the symbol data as a set of hexadecimal
values if asked to output to a text file (`"*.txt"`) or if given the option
`--filetype=txt`. This can be used for test and diagnostic purposes.

Additional options are available which are specific to certain symbologies.
These may, for example, control the amount of error correction data or the size
of the symbol. These options are discussed in section [6. Types of Symbology] of
this guide.


# 5. Using the API

Zint has been written using the C language and has an API for use with C/C++
language programs. A Qt interface (see [Annex B. Qt Backend QZint]) is available
in the `"backend_qt"` sub-directory, and a Tcl interface is available in the
`"backend_tcl"` sub-directory (see [Annex C. Tcl Backend Binding]).

The `libzint` API has been designed to be very similar to that used by the GNU
Barcode package. This allows easy migration from GNU Barcode to Zint. Zint,
however, uses none of the same function names or option names as GNU Barcode.
This allows you to use both packages in your application without conflict if you
wish.

## 5.1 Creating and Deleting Symbols

The symbols manipulated by Zint are held in a `zint_symbol` structure defined in
`"zint.h"`. These symbol structures are created with the `ZBarcode_Create()`
function and deleted using the `ZBarcode_Delete()` function. For example the
following code creates and then deletes a symbol:

```c
#include <zint.h>
#include <stdio.h>
int main()
{
    struct zint_symbol *my_symbol;
    my_symbol = ZBarcode_Create();
    if (my_symbol != NULL) {
        printf("Symbol successfully created!\n");
        ZBarcode_Delete(my_symbol);
    }
    return 0;
}
```

When compiling this code it will need to be linked with the `libzint` library
using the `-lzint` option:

```bash
gcc -o simple simple.c -lzint
```

## 5.2 Encoding and Saving to File

To encode data in a barcode use the `ZBarcode_Encode()` function. To write the
symbol to a file use the `ZBarcode_Print()` function. For example the following
code takes a string from the command line and outputs a Code 128 symbol to a PNG
file named `"out.png"` (or a GIF file `"out.gif"` if `libpng` is not present) in
the current working directory:

```c
#include <zint.h>
int main(int argc, char **argv)
{
    struct zint_symbol *my_symbol;
    my_symbol = ZBarcode_Create();
    ZBarcode_Encode(my_symbol, argv[1], 0);
    ZBarcode_Print(my_symbol, 0);
    ZBarcode_Delete(my_symbol);
    return 0;
}
```

This can also be done in one stage using the `ZBarcode_Encode_and_Print()`
function as shown in the next example:

```c
#include <zint.h>
int main(int argc, char **argv)
{
    struct zint_symbol *my_symbol;
    my_symbol = ZBarcode_Create();
    ZBarcode_Encode_and_Print(my_symbol, argv[1], 0, 0);
    ZBarcode_Delete(my_symbol);
    return 0;
}
```

Note that when using the API, the input data is assumed to be 8-bit binary
unless the `input_mode` member of the `zint_symbol` structure is set - see [5.11
Setting the Input Mode] for details.

## 5.3 Encoding and Printing Functions in Depth

The functions for encoding and printing barcodes are defined as:

```c
int ZBarcode_Encode(struct zint_symbol *symbol,
      const unsigned char *source, int length);

int ZBarcode_Encode_File(struct zint_symbol *symbol,
      const char *filename);

int ZBarcode_Print(struct zint_symbol *symbol, int rotate_angle);

int ZBarcode_Encode_and_Print(struct zint_symbol *symbol,
      const unsigned char *source, int length, int rotate_angle);

int ZBarcode_Encode_File_and_Print(struct zint_symbol *symbol,
      const char *filename, int rotate_angle);
```

In these definitions `length` can be used to set the length of the input string.
This allows the encoding of `NUL` (ASCII 0) characters in those symbologies
which allow this. A value of 0 (or less than 0) will disable this usage and Zint
will encode data up to the first `NUL` character in the input string, which must
be present.

The `rotate_angle` value can be used to rotate the image when outputting. Valid
values are 0, 90, 180 and 270.

The `ZBarcode_Encode_File()` and `ZBarcode_Encode_File_and_Print()` functions
can be used to encode data read directly from a text file where the filename is
given in the `NUL`-terminated `filename` string. The special filename `"-"`
(single hyphen) can be used to read from stdin. Note that on Windows, filenames
are assumed to be UTF-8 encoded.

If printing more than one barcode, the `zint_symbol` structure may be re-used by
calling the `ZBarcode_Clear()` function after each barcode to free any output
buffers allocated. The `zint_symbol` input members must be reset. To fully
restore `zint_symbol` to its default state, call `ZBarcode_Reset()` instead.

## 5.4 Buffering Symbols in Memory (raster)

In addition to saving barcode images to file Zint allows you to access a
representation of the resulting bitmap image in memory. The following functions
allow you to do this:

```c
int ZBarcode_Buffer(struct zint_symbol *symbol, int rotate_angle);

int ZBarcode_Encode_and_Buffer(struct zint_symbol *symbol,
      const unsigned char *source, int length, int rotate_angle);

int ZBarcode_Encode_File_and_Buffer(struct zint_symbol *symbol,
      const char *filename, int rotate_angle);
```

The arguments here are the same as above, and rotation and colour options can be
used with the buffer functions in the same way as when saving to a file. The
difference is that instead of saving the image to a file it is placed in a byte
(`unsigned char`) array pointed to by the `bitmap` member, with `bitmap_width`
set to the number of columns and `bitmap_height` set to the number of rows.

The RGB channels are split into 3 consecutive red, green, blue bytes per pixel,
and there are `bitmap_width` pixels per row and `bitmap_height` rows, so the
total size of the `bitmap` array is `3 * bitmap_width * bitmap_height`.

If the background and/or foreground are RGBA then the byte array `alphamap` will
also be set, with a single alpha value for each pixel. Its total size will be
`bitmap_width * bitmap_height`.

The pixel data can be extracted from the array (or arrays) by the method shown
in the example below, where `render_rgb()` and `render_rgba()` are assumed to be
functions for drawing an RGB and RGBA pixel on the screen implemented by the
client application:

```c
int row, col, i = 0, j = 0;
int red, blue, green, alpha;

for (row = 0; row < my_symbol->bitmap_height; row++) {
     for (col = 0; col < my_symbol->bitmap_width; col++) {
          red = (int) my_symbol->bitmap[i];
          green = (int) my_symbol->bitmap[i + 1];
          blue = (int) my_symbol->bitmap[i + 2];
          if (my_symbol->alphamap) {
              alpha = (int) my_symbol->alphamap[j];
              render_rgba(row, col, red, green, blue, alpha);
              j++;
          } else {
              render_rgb(row, col, red, green, blue);
          }
          i += 3;
     }
}
```

Where speed is important, the buffer can be returned instead in a more compact
intermediate form using the output option `OUT_BUFFER_INTERMEDIATE`. Here each
byte is an ASCII value: `'1'` for foreground colour and `'0'` for background
colour, except for Ultracode, which also uses colour codes: `'W'` for white,
`'C'` for cyan, `'B'` for blue, `'M'` for magenta, `'R'` for red, `'Y'` for
yellow, `'G'` for green, and `'K'` for black. Alpha values are not reported
(`alphamap` will always be `NULL`). The loop for accessing the data is then:

```c
int row, col, i = 0;

for (row = 0; row < my_symbol->bitmap_height; row++) {
     for (col = 0; col < my_symbol->bitmap_width; col++) {
          render_pixel(row, col, my_symbol->bitmap[i]);
          i++;
     }
}
```

## 5.5 Buffering Symbols in Memory (vector)

Symbols can also be saved to memory in a vector representation as well as a
bitmap one. The following functions, exactly analogous to the ones above, allow
you to do this:

```c
int ZBarcode_Buffer_Vector(struct zint_symbol *symbol, int rotate_angle);

int ZBarcode_Encode_and_Buffer_Vector(struct zint_symbol *symbol,
      const unsigned char *source, int length, int rotate_angle);

int ZBarcode_Encode_File_and_Buffer_Vector(struct zint_symbol *symbol,
      const char *filename, int rotate_angle);
```

Here the `vector` member is set to point to a `zint_vector` header structure
which contains pointers to lists of structures representing the various elements
of the barcode: rectangles, hexagons, strings and circles. To draw the barcode,
each of the element types is iterated in turn, and using the information stored
is drawn by a rendering system. For instance, to draw a barcode using a
rendering system with `prepare_canvas()`, `draw_rect()`, `draw_hexagon()`,
`draw_string()`, and `draw_circle()` routines available:

```c
struct zint_vector_rect *rect;
struct zint_vector_hexagon *hex;
struct zint_vector_string *string;
struct zint_vector_circle *circle;

prepare_canvas(my_symbol->vector->width, my_symbol->vector->height,
               my_symbol->scale, my_symbol->fgcolour, my_symbol->bgcolour,
               rotate_angle);

for (rect = my_symbol->vector->rectangles; rect; rect = rect->next) {
    draw_rect(rect->x, rect->y, rect->width, rect->height,
              rect->colour);
}
for (hex = my_symbol->vector->hexagons; hex; hex = hex->next) {
    draw_hexagon(hex->x, hex->y, hex->diameter, hex->rotation);
}
for (string = my_symbol->vector->strings; string; string = string->next) {
    draw_string(string->x, string->y, string->fsize,
                string->rotation, string->halign,
                string->text, string->length);
}
for (circle = my_symbol->vector->circles; circle; circle = circle->next) {
    draw_circle(circle->x, circle->y, circle->diameter, circle->width);
}
```

## 5.6 Buffering Symbols in Memory (memfile)

Symbols can also be stored as "in-memory" file buffers by giving the
`BARCODE_MEMORY_FILE` option to the `output_options` member, which saves the
print output to member `memfile` instead of to the output file `outfile`. The
length of the buffer is given in `memfile_size`. For instance:

```c
#include <zint.h>
#include <stdio.h>
#include <string.h>
int main(int argc, char **argv)
{
    struct zint_symbol *my_symbol;
    my_symbol = ZBarcode_Create();
	my_symbol->output_options |= BARCODE_MEMORY_FILE;
	/* Only the extension is used, to determine output format */
	strcpy(my_symbol->outfile, "mem.svg");
    ZBarcode_Encode_and_Print(my_symbol, argv[1], 0, 0);
	/* `my_symbol->memfile` now contains the SVG output */
	fwrite(my_symbol->memfile, 1, my_symbol->memfile_size, stdout);
    ZBarcode_Delete(my_symbol);
    return 0;
}

```

will print the SVG output to `stdout` (the file "mem.svg" is not created). This
is particularly useful for the textual formats EPS and SVG,[^7] allowing the
output to be manipulated and processed by the client.

[^7]: BARCODE_MEMORY_FILE textual formats EPS and SVG will have Unix newlines
(LF) on both Windows and Unix, i.e. not CR+LF on Windows.

## 5.7 Setting Options

So far our application is not very useful unless we plan to only make Code 128
symbols and we don't mind that they only save to `"out.png"` (or to memory, as
above). As with the CLI program, of course, these options can be altered. The
way this is done is by altering the contents of the `zint_symbol` structure
between the creation and encoding stages. The `zint_symbol` structure consists
of the following members:

-----------------------------------------------------------------------------
Member Name          Type        Meaning                    Default Value
-------------------  ----------  -------------------------  -----------------
`symbology`          integer     Symbol to use - see [5.9   `BARCODE_CODE128`
                                 Specifying a Symbology].

`height`             float       Symbol height in            Symbol dependent
                                 X-dimensions, excluding
                                 fixed width-to-height
                                 symbols.[^8]

`scale`              float       Scale factor for            1.0
                                 adjusting size of image
                                 (sets X-dimension).

`whitespace_width`   integer     Horizontal whitespace       0
                                 width in X-dimensions.

`whitespace_height`  integer     Vertical whitespace         0
                                 height in X-dimensions.

`border_width`       integer     Border width in             0
                                 X-dimensions.

`output_options`     integer     Set various output          0 (none)
                                 parameters - see [5.10
                                 Adjusting Output
                                 Options].

`fgcolour`           character   Foreground (ink)            `"000000"`
                     string      colour as RGB/RGBA
                                 hexadecimal string or
                                 `"C,M,Y,K"` decimal
                                 percentages string, with
                                 a terminating `NUL`.

`bgcolour`           character   Background (paper)          `"ffffff"`
                     string      colour as RGB/RGBA
                                 hexadecimal string or
                                 `"C,M,Y,K"` decimal
                                 percentages string, with
                                 a terminating `NUL`.

`fgcolor`            pointer     Points to fgcolour
                                 allowing alternate
                                 spelling.

`bgcolor`            pointer     Points to bgcolour
                                 allowing alternate
                                 spelling.

`outfile`            character   Contains the name of the    `"out.png"`
                     string      file to output a
                                 resulting barcode symbol
                                 to. Must end in `.png`,
                                 `.gif`, `.bmp`, `.emf`,
                                 `.eps`, `.pcx`, `.svg`,
                                 `.tif` or `.txt` followed
                                 by a terminating
                                 `NUL`.[^9]

`primary`            character   Primary message data for    `""` (empty)
                     string      more complex symbols,
                                 with a terminating `NUL`.

`option_1`           integer     Symbol specific options.    -1

`option_2`           integer     Symbol specific options.    0

`option_3`           integer     Symbol specific options.    0

`show_hrt`           integer     Set to 0 to hide Human      1
                                 Readable Text (HRT).

`input_mode`         integer     Set encoding of input       `DATA_MODE`
                                 data - see [5.11 Setting
                                 the Input Mode].

`eci`                integer     Extended Channel            0 (none)
                                 Interpretation code.

`dpmm`               float       Resolution of output in     0 (none)
                                 dots per mm (BMP, EMF,
                                 PCX, PNG and TIF only).

`dot_size`           float       Diameter of dots used in    0.8
                                 dotty mode (in
                                 X-dimensions).

`text_gap`           float       Gap between barcode and     1.0
                                 text (HRT) in
                                 X-dimensions.

`guard_descent`      float       Height of guard bar         5.0
                                 descent (EAN/UPC only) in
                                 X-dimensions.

`structapp`          Structured  Mark a symbol as part of    count 0
                     Append      a sequence of symbols.      (disabled)
                     structure

`debug`              integer     Debugging flags.            0

`warn_level`         integer     Affects error/warning       `WARN_DEFAULT`
                                 value returned by Zint
                                 API - see [5.8 Handling
                                 Errors].

`text`               unsigned    Human Readable Text,        `""` (empty)
                     character   which usually consists of   (output only)
                     string      input data plus one more
                                 check digit. Uses UTF-8
                                 formatting, with a
                                 terminating `NUL`.

`rows`               integer     Number of rows used by      (output only)
                                 the symbol.

`width`              integer     Width of the generated      (output only)
                                 symbol.

`encoded_data`       array of    Representation of the       (output only)
                     unsigned    encoded data.
                     character
                     arrays

`row_height`         array of    Heights of each row.        (output only)
                     floats

`errtxt`             character   Error message in the        (output only)
                     string      event that an error
                                 occurred, with a
                                 terminating `NUL` - see
                                 [5.8 Handling Errors].

`bitmap`             pointer to  Pointer to stored bitmap    (output only)
                     unsigned    image - see [5.4
                     character   Buffering Symbols in
                     array       Memory (raster)].

`bitmap_width`       integer     Width of stored bitmap      (output only)
                                 image (in pixels) - see
                                 `bitmap` member.

`bitmap_height`      integer     Height of stored bitmap     (output only)
                                 image (in pixels) - see
                                 `bitmap` member.

`alphamap`           pointer to  Pointer to array            (output only)
                     unsigned    representing alpha
                     character   channel of stored bitmap
                     array       image (or `NULL` if no
                                 alpha channel used) - see
                                 `bitmap` member.

`vector`             pointer to  Pointer to vector header    (output only)
                     vector      containing pointers to
                     structure   vector elements - see
                                 [5.5 Buffering Symbols
                                 in Memory (vector)].

`memfile`            pointer to  Pointer to in-memory        (output only)
                     unsigned    file buffer if
                     character   `BARCODE_MEMORY_FILE`
                     array       set in `output_options`
                                 - see [5.6 Buffering
                                 Symbols in Memory
                                 (memfile)].

`memfile_size`       integer     Length of in-memory file    (output only)
                                 buffer.
-----------------------------------------------------------------------------

Table: API Structure `zint_symbol` {#tbl:api_structure_zint_symbol tag="$ $"}

[^8]: The `height` value is ignored for Aztec (including HIBC and Aztec Rune),
Code One, Data Matrix (including HIBC), DotCode, Grid Matrix, Han Xin, MaxiCode,
QR Code (including HIBC, Micro QR, rMQR and UPNQR), and Ultracode - all of which
have a fixed width-to-height ratio (or, in the case of Code One, a fixed
height).

[^9]: For Windows, `outfile` is assumed to be UTF-8 encoded.

To alter these values use the syntax shown in the example below. This code has
the same result as the previous example except the output is now taller and
plotted in green.

```c
#include <zint.h>
#include <string.h>
int main(int argc, char **argv)
{
    struct zint_symbol *my_symbol;
    my_symbol = ZBarcode_Create();
    strcpy(my_symbol->fgcolour, "00ff00");
    my_symbol->height = 400.0f;
    ZBarcode_Encode_and_Print(my_symbol, argv[1], 0, 0);
    ZBarcode_Delete(my_symbol);
    return 0;
}
```

Note that background removal for all outputs except BMP can be achieved by
setting the background alpha to `"00"` where the values for R, G and B will be
ignored:

```c
    strcpy(my_symbol->bgcolour, "55555500");
```

This is what the CLI option `--nobackground` does - see [4.7 Using Colour].

## 5.8 Handling Errors

If errors occur during encoding a non-zero integer value is passed back to the
calling application. In addition the `errtxt` member is set to a message
detailing the nature of the error. The errors generated by Zint are:

------------------------------------------------------------------------------
Return Value                   Meaning
-----------------------------  -----------------------------------------------
`ZINT_WARN_HRT_TRUNCATED`      The Human Readable Text returned in `text` was
                               truncated (maximum 199 bytes).

`ZINT_WARN_INVALID_OPTION`     One of the values in `zint_struct` was set
                               incorrectly but Zint has made a guess at what
                               it should have been and generated a barcode
                               accordingly.

`ZINT_WARN_USES_ECI`           Zint has automatically inserted an ECI
                               character. The symbol may not be readable with
                               some readers.

`ZINT_WARN_NONCOMPLIANT`       The symbol was created but is not compliant
                               with certain standards set in its specification
                               (e.g. height, GS1 AI data lengths).

`ZINT_ERROR`                   Marks the divide between warnings and errors.
                               For return values greater than or equal to this
                               no symbol (or only an incomplete symbol) is
                               generated.

`ZINT_ERROR_TOO_LONG`          The input data is too long or too short for the
                               selected symbology. No symbol has been
                               generated.

`ZINT_ERROR_INVALID_DATA`      The data to be encoded includes characters
                               which are not permitted by the selected
                               symbology (e.g. alphabetic characters in an EAN
                               symbol). No symbol has been generated.

`ZINT_ERROR_INVALID_CHECK`     Data with an incorrect check digit has been
                               entered. No symbol has been generated.

`ZINT_ERROR_INVALID_OPTION`    One of the values in `zint_struct` was set
                               incorrectly and Zint was unable (or unwilling)
                               to guess what it should have been. No symbol
                               has been generated.

`ZINT_ERROR_ENCODING_PROBLEM`  A problem has occurred during encoding of the
                               data. This should never happen. Please contact
                               the developer if you encounter this error.

`ZINT_ERROR_FILE_ACCESS`       Zint was unable to open the requested output
                               file. This is usually a file permissions
                               problem.

`ZINT_ERROR_MEMORY`            Zint ran out of memory. This should only be a
                               problem with legacy systems.

`ZINT_ERROR_FILE_WRITE`        Zint failed to write all contents to the
                               requested output file. This should only occur
                               if the output device becomes full.

`ZINT_ERROR_USES_ECI`          Returned if `warn_level` set to `WARN_FAIL_ALL`
                               and `ZINT_WARN_USES_ECI` occurs.

`ZINT_ERROR_NONCOMPLIANT`      Returned if `warn_level` set to `WARN_FAIL_ALL`
                               and `ZINT_WARN_NONCOMPLIANT` occurs.

`ZINT_ERROR_HRT_TRUNCATED`     Returned if `warn_level` set to `WARN_FAIL_ALL`
                               and `ZINT_WARN_HRT_TRUNCATED` occurs.
------------------------------------------------------------------------------

Table: {#tbl:api_warnings_errors tag=": API Warning and Error Return Values"}

To catch errors use an integer variable as shown in the code below:

```c
#include <zint.h>
#include <stdio.h>
#include <string.h>
int main(int argc, char **argv)
{
    struct zint_symbol *my_symbol;
    int error;
    my_symbol = ZBarcode_Create();
    /* Set invalid foreground colour */
    strcpy(my_symbol->fgcolour, "nonsense");
    error = ZBarcode_Encode_and_Print(my_symbol, argv[1], 0, 0);
    if (error != 0) {
        /* Some warning or error occurred */
        printf("%s\n", my_symbol->errtxt);
        if (error >= ZINT_ERROR) {
            /* Stop now */
            ZBarcode_Delete(my_symbol);
            return 1;
        }
    }
    /* Otherwise carry on with the rest of the application */
    ZBarcode_Delete(my_symbol);
    return 0;
}
```

This code will exit with the appropriate message:

```
Error 881: Malformed foreground RGB colour 'nonsense' (hexadecimal only)
```

To treat all warnings as errors, set `symbol->warn_level` to `WARN_FAIL_ALL`.

## 5.9 Specifying a Symbology

Symbologies can be specified by number or by name as shown in the Table
{@tbl:barcode_types}. For example

```c
symbol->symbology = BARCODE_LOGMARS;
```

means the same as

```c
symbol->symbology = 50;
```

## 5.10 Adjusting Output Options

The `output_options` member can be used to adjust various aspects of the output
file. To select more than one option from the table below simply `OR` them
together when adjusting this value:

```c
my_symbol->output_options |= BARCODE_BIND | READER_INIT;
```

------------------------------------------------------------------------------
Value                      Effect
-------------------------  ---------------------------------------------------
 0                         No options selected.

`BARCODE_BIND_TOP`         Boundary bar above the symbol only.[^10]

`BARCODE_BIND`             Boundary bars above and below the symbol and
                           between rows if stacking multiple symbols.[^11]

`BARCODE_BOX`              Add a box surrounding the symbol and whitespace.

`BARCODE_STDOUT`           Output the file to stdout.

`READER_INIT`              Create as a Reader Initialisation (Programming)
                           symbol.

`SMALL_TEXT`               Use a smaller font for the Human Readable Text.

`BOLD_TEXT`                Embolden the Human Readable Text.

`CMYK_COLOUR`              Select the CMYK colour space option for
                           Encapsulated PostScript and TIF files.

`BARCODE_DOTTY_MODE`       Plot a matrix symbol using dots rather than
                           squares.

`GS1_GS_SEPARATOR`         Use GS (Group Separator) instead of FNC1 as GS1
                           separator (Data Matrix only).

`OUT_BUFFER_INTERMEDIATE`  Return the bitmap buffer as ASCII values instead of
                           separate colour channels - see [5.4 Buffering
                           Symbols in Memory (raster)].

`BARCODE_QUIET_ZONES`      Add compliant quiet zones (additional to any
                           specified whitespace).[^12]

`BARCODE_NO_QUIET_ZONES`   Disable quiet zones, notably those with defaults.

`COMPLIANT_HEIGHT`         Warn if height specified not compliant, or use
                           standard height (if any) as default.

`EANUPC_GUARD_WHITESPACE`  Add quiet zone indicators ("<" and/or ">") to HRT
                           whitespace (EAN/UPC).

`EMBED_VECTOR_FONT`        Embed font in vector output - currently available
                           for SVG output only.

`BARCODE_MEMORY_FILE`      Write output to in-memory buffer `symbol->memfile`
                           instead of to `outfile` file.
------------------------------------------------------------------------------

Table: API `output_options` Values {#tbl:api_output_options tag="$ $"}

[^10]: The `BARCODE_BIND_TOP` flag is set by default for DPD - see [6.1.10.7 DPD
Code].

[^11]: The `BARCODE_BIND` flag is always set for Codablock-F, Code 16K and Code
49. Special considerations apply to ITF-14 - see [6.1.2.6 ITF-14].

[^12]: Codablock-F, Code 16K, Code 49, EAN-2 to EAN-13, ISBN, ITF-14, UPC-A and
UPC-E have compliant quiet zones added by default.

## 5.11 Setting the Input Mode

The way in which the input data is encoded can be set using the `input_mode`
member. Valid values are shown in the table below.

------------------------------------------------------------------------------
Value               Effect
------------------  ----------------------------------------------------------
`DATA_MODE`         Uses full 8-bit range interpreted as binary data.

`UNICODE_MODE`      Uses UTF-8 input.

`GS1_MODE`          Encodes GS1 data using FNC1 characters.

                    _The above are exclusive, the following optional and
                    OR-ed._

`ESCAPE_MODE`       Process input data for escape sequences.

`GS1PARENS_MODE`    Parentheses (round brackets) used in GS1 data instead of
                    square brackets to delimit Application Identifiers
                    (parentheses must not otherwise occur in the data).

`GS1NOCHECK_MODE`   Do not check GS1 data for validity, i.e. suppress checks
                    for valid AIs and data lengths. Invalid characters (e.g.
                    control characters, extended ASCII characters) are still
                    checked for.

`HEIGHTPERROW_MODE` Interpret the `height` member as per-row rather than as
                    overall height.

`FAST_MODE`         Use faster if less optimal encodation or other shortcuts
                    if available (affects `DATAMATRIX`, `MICROPDF417`,
                    `PDF417`, `QRCODE` and `UPNQR` only).

`EXTRA_ESCAPE_MODE` Process special symbology-specific escape sequences
                    (`CODE128` only).

------------------------------------------------------------------------------

Table: API `input_mode` Values {#tbl:api_input_mode tag="$ $"}

The default mode is `DATA_MODE`. (Note that this differs from the default for
the CLI and GUI, which is `UNICODE_MODE`.)

`DATA_MODE`, `UNICODE_MODE` and `GS1_MODE` are mutually exclusive, whereas
`ESCAPE_MODE`, `GS1PARENS_MODE`, `GS1NOCHECK_MODE`, `HEIGHTPERROW_MODE`,
`FAST_MODE` and `EXTRA_ESCAPE_MODE` are optional. So, for example, you can set

```c
my_symbol->input_mode = UNICODE_MODE | ESCAPE_MODE;
```

or

```c
my_symbol->input_mode = GS1_MODE | GS1PARENS_MODE | GS1NOCHECK_MODE;
```

whereas

```c
my_symbol->input_mode = DATA_MODE | GS1_MODE;
```

is not valid.

Permissible escape sequences (`ESCAPE_MODE`) are listed in Table
{@tbl:escape_sequences}, and the special Code 128-only `EXTRA_ESCAPE_MODE`
escape sequences are given in [6.1.10.1 Standard Code 128 (ISO 15417)]. An
example of `GS1PARENS_MODE` usage is given in section [6.1.10.3 GS1-128].

`GS1NOCHECK_MODE` is for use with legacy systems that have data that does not
conform to the current GS1 standard. Printable ASCII input is still checked for,
as is the validity of GS1 data specified without AIs (e.g. linear data for GS1
DataBar Omnidirectional/Limited/etc.). Also checked is GS1 DataBar Expanded and
GS1 Composite input that is not in the GS1 encodable character set 82 (see GS1
General Specifications Figure 7.11.1 'GS1 AI encodable character set 82'),
otherwise encodation would fail.

For `HEIGHTPERROW_MODE`, see `--heightperrow` in section [4.4 Adjusting Height].
The `height` member should be set to the desired per-row value on input (it will
be set to the overall height on output).

`FAST_MODE` causes a less optimal encodation scheme to be used for Data Matrix,
MicroPDF417 and PDF417. For QR Code and UPNQR, it affects Zint's automatic mask
selection - see [6.6.3 QR Code (ISO 18004)] for details.

## 5.12 Multiple Segments

For input data requiring multiple ECIs, the following functions may be used:

```c
int ZBarcode_Encode_Segs(struct zint_symbol *symbol,
      const struct zint_seg segs[], const int seg_count);

int ZBarcode_Encode_Segs_and_Print(struct zint_symbol *symbol,
      const struct zint_seg segs[], const int seg_count, int rotate_angle);

int ZBarcode_Encode_Segs_and_Buffer(struct zint_symbol *symbol,
      const struct zint_seg segs[], const int seg_count, int rotate_angle);

int ZBarcode_Encode_Segs_and_Buffer_Vector(struct zint_symbol *symbol,
      const struct zint_seg segs[], const int seg_count, int rotate_angle);
```

These are direct analogues of the previously mentioned `ZBarcode_Encode()`,
`ZBarcode_Encode_and_Print()`, `ZBarcode_Encode_and_Buffer()` and
`ZBarcode_Encode_and_Buffer_Vector()` respectively, where instead of a pair
consisting of `"source, length"`, a pair consisting of `"segs, seg_count"` is
given, with `segs` being an array of `struct zint_seg` segments and `seg_count`
being the number of elements it contains. The zint_seg structure is of the form:

```c
struct zint_seg {
    unsigned char *source; /* Data to encode */
    int length;            /* Length of `source`. If 0, `source` must be
                              NUL-terminated */
    int eci;               /* Extended Channel Interpretation */
};
```

The symbology must support ECIs (see Table {@tbl:eci_aware_symbologies}). For
example:

```c
#include <zint.h>
int main(int argc, char **argv)
{
    struct zint_seg segs[] = {
        { "Κείμενο", 0, 9 },
        { "Текст", 0, 7 },
        { "文章", 0, 20 }
    };
    struct zint_symbol *my_symbol;
    my_symbol = ZBarcode_Create();
    my_symbol->symbology = BARCODE_AZTEC;
    my_symbol->input_mode = UNICODE_MODE;
    ZBarcode_Encode_Segs(my_symbol, segs, 3);
    ZBarcode_Print(my_symbol, 0);
    ZBarcode_Delete(my_symbol);
    return 0;
}
```

A maximum of 256 segments may be specified. Use of multiple segments with GS1
data is not currently supported.

## 5.13 Scaling Helpers

To help with scaling the output, the following three function are available:

```c
float ZBarcode_Default_Xdim(int symbol_id);

float ZBarcode_Scale_From_XdimDp(int symbol_id, float x_dim_mm, float dpmm,
        const char *filetype) {

float ZBarcode_XdimDP_From_Scale(int symbol_id, float scale,
        float x_dim_mm_or_dpmm, const char *filetype);
```

The first `ZBarcode_Default_Xdim()` returns the default X-dimension suggested by
Zint for symbology `symbol_id`.

The second `ZBarcode_Scale_From_XdimDp()` returns the scale to use to output to
a file of type `filetype` with X-dimension `x_dim_mm` at `dpmm` dots per mm. The
given X-dimension must be non-zero and less than or equal to 10mm, however
`dpmm` may be zero and defaults to 12 dpmm, and `filetype` may be NULL or empty
in which case a GIF filetype is assumed. For raster output (BMP/GIF/PCX/PNG/TIF)
the scale is rounded to half-integer increments.

For example:

```c
/* Royal Mail 4-State Customer Code */
my_symbol->symbology = BARCODE_RM4SCC;
my_symbol->dpmm = 600.0f / 25.4f; /* 600 dpi */
my_symbol->scale = ZBarcode_Scale_From_XdimDp(
                        my_symbol->symbology,
                        ZBarcode_Default_Xdim(my_symbol->symbology),
                        my_symbol->dpmm, "PNG"); /* Returns 7.5 */
```

The third function `ZBarcode_XdimDP_From_Scale()` is the "reverse" of
`ZBarcode_Scale_From_XdimDp()`, returning the X-dimension (in mm) or the dot
density (in dpmm) given a scale `scale`. Both `scale` and `x_dim_mm_or_dpmm`
must be non-zero. The returned value is bound to the maximum value of dpmm
(1000), so must be further bound to 10 on return if the X-dimension is sought.

Note that the X-dimension to use is application dependent, and varies not only
due to the symbology, resolution and filetype but also due to the type of
scanner used, the intended scanning distance, and what media ("substrates") the
barcode appears on.

## 5.14 Verifying Symbology Availability

An additional function available in the API is:

```c
int ZBarcode_ValidID(int symbol_id);
```

which allows you to check whether a given symbology is available, returning a
non-zero value if so. For example:

```c
if (ZBarcode_ValidID(BARCODE_PDF417) != 0) {
    printf("PDF417 available\n");
} else {
    printf("PDF417 not available\n");
}
```

Another function that may be useful is:

```c
int ZBarcode_BarcodeName(int symbol_id, char name[32]);
```

which copies the name of a symbology into the supplied `name` buffer, which
should be 32 characters in length. The name is `NUL`-terminated, and zero is
returned on success. For instance:

```c
char name[32];
if (ZBarcode_BarcodeName(BARCODE_PDF417, name) == 0) {
    printf("%s\n", name);
}
```

will print `BARCODE_PDF417`.

## 5.15 Checking Symbology Capabilities

It can be useful for frontend programs to know the capabilities of a symbology.
This can be determined using another additional function:

```c
unsigned int ZBarcode_Cap(int symbol_id, unsigned int cap_flag);
```

by `OR`-ing the flags below in the `cap_flag` argument and checking the return
to see which are set.

------------------------------------------------------------------------------
Value                       Meaning
-------------------------   --------------------------------------------------
`ZINT_CAP_HRT`              Can the symbology print Human Readable Text?

`ZINT_CAP_STACKABLE`        Is the symbology stackable?

`ZINT_CAP_EANUPC`[^13]      Is the symbology EAN/UPC?

`ZINT_CAP_COMPOSITE`        Does the symbology support composite data? (see
                            [6.3 GS1 Composite Symbols (ISO 24723)] below)

`ZINT_CAP_ECI`              Does the symbology support Extended Channel
                            Interpretations?

`ZINT_CAP_GS1`              Does the symbology support GS1 data?

`ZINT_CAP_DOTTY`            Can the symbology be outputted as dots?

`ZINT_CAP_QUIET_ZONES`      Does the symbology have default quiet zones?

`ZINT_CAP_FIXED_RATIO`      Does the symbology have a fixed width-to-height
                            (aspect) ratio?

`ZINT_CAP_READER_INIT`      Does the symbology support Reader Initialisation?

`ZINT_CAP_FULL_MULTIBYTE`   Is the `ZINT_FULL_MULTIBYTE` option applicable?

`ZINT_CAP_MASK`             Is mask selection applicable?

`ZINT_CAP_STRUCTAPP`        Does the symbology support Structured Append?

`ZINT_CAP_COMPLIANT_HEIGHT` Does the symbology have a compliant height
                            defined?
------------------------------------------------------------------------------

Table: {#tbl:api_cap tag=": API Capability Flags"}

[^13]: `ZINT_CAP_EANUPC` was previously named `ZINT_CAP_EXTENDABLE`, which is
still recognised.

For example:

```c
unsigned int cap;
cap = ZBarcode_Cap(BARCODE_PDF417, ZINT_CAP_HRT | ZINT_CAP_ECI);
if (cap & ZINT_CAP_HRT) {
    printf("PDF417 supports HRT\n");
} else {
    printf("PDF417 does not support HRT\n");
}
if (cap & ZINT_CAP_ECI) {
    printf("PDF417 supports ECI\n");
} else {
    printf("PDF417 does not support ECI\n");
}
```

## 5.16 Zint Version

Whether the Zint library linked to was built with PNG support may be determined
with:

```c
int ZBarcode_NoPng();
```

which returns 1 if no PNG support is available, else zero.

Lastly, the version of the Zint library linked to is returned by:

```c
int ZBarcode_Version();
```

The version parts are separated by hundreds. For instance, version `"2.9.1"` is
returned as `"20901"`.


# 6. Types of Symbology

## 6.1 One-Dimensional Symbols

One-dimensional or linear symbols are what most people associate with the term
barcode. They consist of a number of bars and a number of spaces of differing
widths.

### 6.1.1 Code 11

![`zint -b CODE11 -d "9212320967"`](images/code11.svg){.lin}

Developed by Intermec in 1977, Code 11 is similar to Code 2 of 5 Matrix and is
primarily used in telecommunications. The symbol can encode data consisting of
the digits 0-9 and the dash character (`-`) up to a maximum of 140 characters.
Two modulo-11 check digits are added by default. To add just one check digit,
set `--vers=1` (API `option_2 = 1`). To add no check digits, set `--vers=2`
(API `option_2 = 2`).

### 6.1.2 Code 2 of 5

Code 2 of 5 is a family of one-dimensional symbols, 8 of which are supported by
Zint. Note that the names given to these standards alters from one source to
another so you should take care to ensure that you have the right barcode type
before using these standards.

#### 6.1.2.1 Standard Code 2 of 5

![`zint -b C25STANDARD -d "9212320967"`](images/c25standard.svg){.lin}

Also known as Code 2 of 5 Matrix, this is a self-checking code used in
industrial applications and photo development. Standard Code 2 of 5 will encode
numeric input (digits 0-9) up to a maximum of 112 digits. No check digit is
added by default. To add a check digit, set `--vers=1` (API `option_2 = 1`). To
add a check digit but not show it in the Human Readable Text, set `--vers=2`
(API `option_2 = 2`).

#### 6.1.2.2 IATA Code 2 of 5

![`zint -b C25IATA -d "9212320967"`](images/c25iata.svg){.lin}

Used for baggage handling in the air-transport industry by the International Air
Transport Agency, this self-checking code will encode numeric input (digits 0-9)
up to a maximum of 80 digits. No check digit is added by default, but can be set
the same as for [6.1.2.1 Standard Code 2 of 5].

#### 6.1.2.3 Industrial Code 2 of 5

![`zint -b C25IND -d "9212320967"`](images/c25ind.svg){.lin}

Industrial Code 2 of 5 can encode numeric input (digits 0-9) up to a maximum of
79 digits. No check digit is added by default, but can be set the same as for
[6.1.2.1 Standard Code 2 of 5].

#### 6.1.2.4 Interleaved Code 2 of 5 (ISO 16390)

![`zint -b C25INTER --compliantheight -d
"9212320967"`](images/c25inter.svg){.lin}

This self-checking symbology encodes pairs of numbers, and so can only encode an
even number of digits (0-9). If an odd number of digits is entered a leading
zero is added by Zint. A maximum of 62 pairs (124 digits) can be encoded. No
check digit is added by default, but can be set the same as for [6.1.2.1
Standard Code 2 of 5].

#### 6.1.2.5 Code 2 of 5 Data Logic

![`zint -b C25LOGIC -d "9212320967"`](images/c25logic.svg){.lin}

Data Logic does not include a check digit by default and can encode numeric
input (digits 0-9) up to a maximum of 113 digits. Check digit options are the
same as for [6.1.2.1 Standard Code 2 of 5].

#### 6.1.2.6 ITF-14

![`zint -b ITF14 --compliantheight -d "9212320967145"`](images/itf14.svg){.lin}

ITF-14, also known as UPC Shipping Container Symbol or Case Code, is based on
Interleaved Code 2 of 5 and requires a 13-digit numeric input (digits 0-9). One
modulo-10 check digit is added by Zint.

If no border option is specified Zint defaults to adding a bounding box with a
border width of 5. This behaviour can be overridden by using the `--bind` option
(API `output_options |= BARCODE_BIND`). Similarly the border width can be
overridden using `--border` (API `border_width`). If a symbol with no border is
required this can be achieved by explicitly setting the border type to box (or
bind or bindtop) and leaving the border width 0.

![`zint -b ITF14 --box --compliantheight -d
"9212320967145"`](images/itf14_border0.svg){.lin}

#### 6.1.2.7 Deutsche Post Leitcode

![`zint -b DPLEIT -d "9212320967145"`](images/dpleit.svg){.lin}

Leitcode is based on Interleaved Code 2 of 5 and is used by Deutsche Post for
routing purposes. Leitcode requires a 13-digit numerical input to which Zint
adds a check digit.

#### 6.1.2.8 Deutsche Post Identcode

![`zint -b DPIDENT -d "91232096712"`](images/dpident.svg){.lin}

Identcode is based on Interleaved Code 2 of 5 and is used by Deutsche Post for
identification purposes. Identcode requires an 11-digit numerical input to which
Zint adds a check digit.

\clearpage

### 6.1.3 UPC (Universal Product Code) (ISO 15420)

#### 6.1.3.1 UPC Version A

![`zint -b UPCA --compliantheight -d "72527270270"`](images/upca.svg){.upcean}

UPC-A is used in the United States for retail applications. The symbol requires
an 11-digit article number. The check digit is calculated by Zint. In addition
EAN-2 and EAN-5 add-on symbols can be added using the + character. For example,
to draw a UPC-A symbol with the data 72527270270 with an EAN-5 add-on showing
the data 12345 use the command:

```bash
zint -b UPCA -d "72527270270+12345"
```

or using the API encode a data string with the + character included:

```c
my_symbol->symbology = BARCODE_UPCA;
error = ZBarcode_Encode_and_Print(my_symbol, "72527270270+12345", 0, 0);
```

![`zint -b UPCA --compliantheight -d
"72527270270+12345"`](images/upca_5.svg){.upcean}

If your input data already includes the check digit symbology `BARCODE_UPCA_CHK`
(35) can be used which takes a 12-digit input and validates the check digit
before encoding.

A quiet zone indicator can be added to the HRT by setting `--guardwhitespace`
(API `output_options |= EANUPC_GUARD_WHITESPACE`). For UPC, this is only
relevant when there is add-on:

```bash
zint -b UPCA -d "72527270270+12345" --guardwhitespace
```

or using the API:

```c
my_symbol->symbology = BARCODE_UPCA;
my_symbol->output_options |= EANUPC_GUARD_WHITESPACE;
error = ZBarcode_Encode_and_Print(my_symbol, "72527270270+12345", 0, 0);
```

![`zint -b UPCA --compliantheight -d "72527270270+12345"
--guardwhitespace`](images/upca_5_gws.svg){.upcean}

You can adjust the gap between the main symbol and an add-on in integral
multiples of the X-dimension by setting `--addongap` (API `option_2`) to a value
between 9 (default) and 12. The height in X-dimensions that the guard bars
descend below the main bars can be adjusted by setting `--guarddescent` (API
`guard_descent`) to a value between 0.0 and 20.0 (default 5.0).

#### 6.1.3.2 UPC Version E

![`zint -b UPCE --compliantheight -d "1123456"`](images/upce.svg){.upcean}

UPC-E is a zero-compressed version of UPC-A developed for smaller packages. The
code requires a 6-digit article number (digits 0-9). The check digit is
calculated by Zint. EAN-2 and EAN-5 add-on symbols can be added using the +
character as with UPC-A. In addition Zint also supports Number System 1 encoding
by entering a 7-digit article number starting with the digit 1. For example:

```bash
zint -b UPCE -d "1123456"
```

or

```c
my_symbol->symbology = BARCODE_UPCE;
error = ZBarcode_Encode_and_Print(my_symbol, "1123456", 0, 0);
```

If your input data already includes the check digit symbology `BARCODE_UPCE_CHK`
(38) can be used which takes a 7 or 8-digit input and validates the check digit
before encoding.

As with UPC-A, a quiet zone indicator can be added when there is an add-on by
setting `--guardwhitespace` (API `output_options |= EANUPC_GUARD_WHITESPACE`):

```bash
zint -b UPCE -d "1123456+12" --guardwhitespace
```

![`zint -b UPCE --compliantheight -d "1123456+12"
--guardwhitespace`](images/upce_2_gws.svg){.upcean}

You can adjust the gap between the main symbol and an add-on in integral
multiples of the X-dimension by setting `--addongap` (API `option_2`) to a value
between 7 (default) and 12. The height in X-dimensions that the guard bars
descend below the main bars can be adjusted by setting `--guarddescent` (API
`guard_descent`) to a value between 0.0 and 20.0 (default 5.0).

### 6.1.4 EAN (European Article Number) (ISO 15420)

#### 6.1.4.1 EAN-2, EAN-5, EAN-8 and EAN-13

![`zint -b EANX --compliantheight -d
"4512345678906"`](images/eanx13.svg){.upcean}

The EAN system is used in retail across Europe and includes standards for EAN-2,
EAN-5, EAN-8 and EAN-13 which encode 2, 5, 7 or 12-digit numbers respectively.
Zint will decide which symbology to use depending on the length of the input
data. In addition EAN-2 and EAN-5 add-on symbols can be added to EAN-8 and
EAN-13 symbols using the + character as with UPC symbols. For example:

```bash
zint -b EANX -d "54321"
```

![`zint -b EANX --compliantheight -d "54321"`](images/eanx5.svg){.upcean}

will encode a stand-alone EAN-5, whereas

```bash
zint -b EANX -d "7432365+54321"
```

will encode an EAN-8 symbol with an EAN-5 add-on. As before these results can be
achieved using the API:

```c
my_symbol->symbology = BARCODE_EANX;

error = ZBarcode_Encode_and_Print(my_symbol, "54321", 0, 0);

error = ZBarcode_Encode_and_Print(my_symbol, "7432365+54321", 0, 0);
```

![`zint -b EANX --compliantheight -d
"7432365+54321"`](images/eanx8_5.svg){.upcean}

All of the EAN symbols include check digits which are added by Zint.

If you are encoding an EAN-8 or EAN-13 symbol and your data already includes the
check digit then you can use symbology `BARCODE_EANX_CHK` (14) which takes an 8
or 13-digit input and validates the check digit before encoding.

Options to add quiet zone indicators and to adjust the add-on gap and the guard
bar descent height are the same as for [6.1.3.2 UPC Version E]. For instance:

```bash
zint -b EANX_CHK -d "74323654" --guardwhitespace
```

![`zint -b EANX_CHK --compliantheight -d "74323654"`
--guardwhitespace](images/eanx8_gws.svg){.upcean}

#### 6.1.4.2 SBN, ISBN and ISBN-13

![`zint -b ISBNX --compliantheight -d
"9789295055124"`](images/isbnx.svg){.upcean}

EAN-13 symbols (also known as Bookland EAN-13) can also be produced from 9-digit
SBN, 10-digit ISBN or 13-digit ISBN-13 data. The relevant check digit needs to
be present in the input data and will be verified before the symbol is
generated.

As with EAN-13, a quiet zone indicator can be added using `--guardwhitespace`:

![`zint -b ISBNX --compliantheight -d "9789295055124"
--guardwhitespace`](images/isbnx_gws.svg){.upcean}

EAN-2 and EAN-5 add-on symbols can be added using the + character, and there are
options to adjust the add-on gap and the guard bar descent height - see [6.1.3.2
UPC Version E].

### 6.1.5 Plessey

#### 6.1.5.1 UK Plessey

![`zint -b PLESSEY -d "C64"`](images/plessey.svg){.lin}

Also known as Plessey Code, this symbology was developed by the Plessey Company
Ltd. in the UK. The symbol can encode data consisting of digits (0-9) or letters
A-F up to a maximum of 67 characters and includes a hidden CRC check digit.

#### 6.1.5.2 MSI Plessey

![`zint -b MSI_PLESSEY -d "6502" --vers=2`](images/msi_plessey.svg){.lin}

Based on Plessey and developed by MSI Data Corporation, MSI Plessey can encode
numeric (digits 0-9) input of up to 92 digits. It has a range of check digit
options that are selectable by setting `--vers` (API `option_2`), shown in the
table below:

Value   Check Digits
-----   ---------------------------
0       None
1       Modulo-10 (Luhn)
2       Modulo-10 & Modulo-10
3       Modulo-11 (IBM)
4       Modulo-11 (IBM) & Modulo-10
5       Modulo-11 (NCR)
6       Modulo-11 (NCR) & Modulo-10

Table: {#tbl:msi_plessey_check_digits tag=": MSI Plessey Check Digit Options"}

To not show the check digit or digits in the Human Readable Text, add 10 to the
`--vers` value. For example `--vers=12` (API `option_2 = 12`) will add two
hidden modulo-10 check digits.

### 6.1.6 Telepen

#### 6.1.6.1 Telepen Alpha

![`zint -b TELEPEN --compliantheight -d "Z80"`](images/telepen.svg){.lin}

Telepen Alpha was developed by SB Electronic Systems Limited and can encode
ASCII text input, up to a maximum of 69 characters. Telepen includes a
hidden modulo-127 check digit, added by Zint.

#### 6.1.6.2 Telepen Numeric

![`zint -b TELEPEN_NUM --compliantheight -d
"466X33"`](images/telepen_num.svg){.lin}

Telepen Numeric allows compression of numeric data into a Telepen symbol. Data
can consist of pairs of numbers or pairs consisting of a numerical digit
followed an X character. For example: 466333 and 466X33 are valid codes whereas
46X333 is not (the digit pair `"X3"` is not valid). Up to 136 digits can be
encoded. Telepen Numeric includes a hidden modulo-127 check digit which is added
by Zint.

### 6.1.7 Code 39

#### 6.1.7.1 Standard Code 39 (ISO 16388)

![`zint -b CODE39 --compliantheight -d "1A" --vers=1`](images/code39.svg){.lin}

Standard Code 39 was developed in 1974 by Intermec. Input data can be up to 86
characters in length and can include the characters 0-9, A-Z, dash (`-`), full
stop (`.`), space, asterisk (`*`), dollar (`$`), slash (`/`), plus (`+`) and
percent (`%`). The standard does not require a check digit but a modulo-43 check
digit can be added if desired by setting `--vers=1` (API `option_2 = 1`). To add
a check digit but not show it in the Human Readable Text, set `--vers=2` (API
`option_2 = 2`).

\clearpage

#### 6.1.7.2 Extended Code 39

![`zint -b EXCODE39 --compliantheight -d
"123.45$@fd"`](images/excode39.svg){.lin}

Also known as Code 39e and Code39+, this symbology expands on Standard Code 39
to provide support for the full 7-bit ASCII character set. The check digit
options are the same as for [6.1.7.1 Standard Code 39 (ISO 16388)].

#### 6.1.7.3 Code 93

![`zint -b CODE93 --compliantheight -d "C93"`](images/code93.svg){.lin}

A variation of Extended Code 39, Code 93 also supports full ASCII text,
accepting up to 123 characters. Two check characters are added by Zint. By
default these check characters are not shown in the Human Readable Text, but may
be shown by setting `--vers=1` (API `option_2 = 1`).

#### 6.1.7.4 PZN (Pharmazentralnummer)

![`zint -b PZN --compliantheight -d "2758089"`](images/pzn.svg){.lin}

PZN is a Code 39 based symbology used by the pharmaceutical industry in Germany.
PZN encodes a 7-digit number to which Zint will add a modulo-11 check digit
(PZN8). Input less than 7 digits will be zero-filled. An 8-digit input can be
supplied in which case Zint will validate the check digit.

To encode a PZN7 (obsolete since 2013) instead set `--vers=1` (API
`option_2 = 1`) and supply up to 7 digits. As with PZN8, a modulo-11 check digit
will be added or if 7 digits supplied the check digit validated.

#### 6.1.7.5 LOGMARS

![`zint -b LOGMARS --compliantheight -d "12345/ABCDE"
--vers=1`](images/logmars.svg){.lin}

LOGMARS (Logistics Applications of Automated Marking and Reading Symbols) is a
variation of the Code 39 symbology used by the U.S. Department of Defense.
LOGMARS encodes the same character set as [6.1.7.1 Standard Code 39 (ISO
16388)], and the check digit options are also the same. Input is restricted to
a maximum of 30 characters.

#### 6.1.7.6 Code 32

![`zint -b CODE32 --compliantheight -d "14352312"`](images/code32.svg){.lin}

A variation of Code 39 used by the Italian Ministry of Health ("Ministero della
Sanità") for encoding identifiers on pharmaceutical products. This symbology
requires a numeric input up to 8 digits in length. A check digit is added by
Zint.

#### 6.1.7.7 HIBC Code 39

![`zint -b HIBC_39 --compliantheight -d "14352312"`](images/hibc_39.svg){.lin}

This variant adds a leading `'+'` character and a trailing modulo-49 check digit
to a standard Code 39 symbol as required by the Health Industry Barcode
standards.

#### 6.1.7.8 Vehicle Identification Number (VIN)

![`zint -b VIN -d "2FTPX28L0XCA15511" --vers=1`](images/vin.svg){.lin}

A variation of Code 39 that for vehicle identification numbers used in North
America (first character `'1'` to `'5'`) has a check character verification
stage. A 17 character input (0-9, and A-Z excluding `'I'`, `'O'` and `'Q'`) is
required. An invisible Import character prefix `'I'` can be added by setting
`--vers=1` (API `option_2 = 1`).

### 6.1.8 Codabar (EN 798)

![`zint -b CODABAR --compliantheight -d "A37859B"`](images/codabar.svg){.lin}

Also known as NW-7, Monarch, ABC Codabar, USD-4, Ames Code and Code 27, this
symbology was developed in 1972 by Monarch Marketing Systems for retail
purposes. The American Blood Commission adopted Codabar in 1977 as the standard
symbology for blood identification. Codabar can encode up to 103 characters
starting and ending with the letters A-D and containing between these letters
the numbers 0-9, dash (`-`), dollar (`$`), colon (`:`), slash (`/`), full stop
(`.`) or plus (`+`). No check character is generated by default, but a modulo-16
one can be added by setting `--vers=1` (API `option_2 = 1`). To have the check
character appear in the Human Readable Text, set `--vers=2` (API
`option_2 = 2`).

### 6.1.9 Pharmacode

![`zint -b PHARMA --compliantheight -d "130170"`](images/pharma.svg){.lin}

Developed by Laetus, Pharmacode is used for the identification of
pharmaceuticals. The symbology is able to encode whole numbers between 3 and
131070.

### 6.1.10 Code 128

#### 6.1.10.1 Standard Code 128 (ISO 15417)

![`zint -b CODE128 --bind -d "130170X178"`](images/code128.svg){.lin}

One of the most ubiquitous one-dimensional barcode symbologies, Code 128 was
developed in 1981 by Computer Identics. This symbology supports full ASCII text
and uses a three-Code Set system to compress the data into a smaller symbol.
Zint automatically switches between Code Sets A, B and C (but see following) and
adds a hidden modulo-103 check digit.

Manual switching of Code Sets is possible using the `--extraesc` option (API
`input_mode |= EXTRA_ESCAPE_MODE`) and the Code 128-specific escapes `\^A`,
`\^B`, `\^C`. For instance the following will force switching to Code Set B for
the data `"5678"` (normally Code Set C would be used throughout):

```bash
zint -b CODE128 -d "1234\^B5678" --extraesc
```

The manually selected Code Set will apply until the next Code Set escape
sequence, with the exception that data that cannot be represented in that Code
Set will be switched as appropriate. If the data contains a special code
sequence, it can be escaped by doubling the caret (`^`). For instance

```bash
zint -b CODE128 -d "\^AABC\^^BDEF" --extraesc
```

will encode the data `"ABC\^BDEF"` in Code Set A.

There is also the extra escape `\^1`, which will encode a special Function Code
1 character (FNC1) anywhere you chose in the data, for instance

```bash
zint -b CODE128 -d "A\^1BC\^1DEF" --extraesc
```

Code 128 is the default barcode symbology used by Zint. In addition Zint
supports the encoding of ISO/IEC 8859-1 (non-English) characters in Code 128
symbols. The ISO/IEC 8859-1 character set is shown in Annex [A.2 Latin Alphabet
No. 1 (ISO/IEC 8859-1)].

Zint can encode a maximum of 99 symbol characters, which allows for e.g. 198
all-numeric characters.

#### 6.1.10.2 Code 128 Suppress Code Set C (Code Sets A and B only)

![`zint -b CODE128AB -d "130170X178"`](images/code128ab.svg){.lin}

It is sometimes advantageous to stop Code 128 from using Code Set C which
compresses numerical data. The `BARCODE_CODE128AB`[^14] variant (symbology 60)
suppresses Code Set C in favour of Code Sets A and B.

Note that the special extra escapes mentioned above are not available for this
variant (nor for any other).

[^14]: `BARCODE_CODE128AB` previously used the name `BARCODE_CODE128B`, which is
still recognised.

#### 6.1.10.3 GS1-128

![`zint -b GS1_128 --compliantheight -d
"[01]98898765432106[3202]012345[15]991231"`](images/gs1_128.svg){.lin}

A variation of Code 128 previously known as UCC/EAN-128, this symbology is
defined by the GS1 General Specifications. Application Identifiers (AIs) should
be entered using [square bracket] notation. These will be converted to
parentheses (round brackets) for the Human Readable Text. This will allow round
brackets to be used in the data strings to be encoded.

For compatibility with data entry in other systems, if the data does not include
round brackets, the option `--gs1parens` (API `input_mode |= GS1PARENS_MODE`)
may be used to signal that AIs are encased in round brackets instead of square
ones.

Fixed length data should be entered at the appropriate length for correct
encoding. GS1-128 does not support extended ASCII (ISO/IEC 8859-1) characters.
Check digits for GTIN data AI (01) are not generated and need to be included in
the input data. The following is an example of a valid GS1-128 input:

```bash
zint -b 16 -d "[01]98898765432106[3202]012345[15]991231"
```

or using the `--gs1parens` option:

```bash
zint -b 16 --gs1parens -d "(01)98898765432106(3202)012345(15)991231"
```

#### 6.1.10.4 EAN-14

![`zint -b EAN14 --compliantheight -d "9889876543210"`](images/ean14.svg){.lin}

A shorter version of GS1-128 which encodes GTIN data only. A 13-digit number is
required. The GTIN check digit and HRT-only AI "(01)" are added by Zint.

#### 6.1.10.5 NVE-18 (SSCC-18)

![`zint -b NVE18 --compliantheight -d
"37612345000001003"`](images/nve18.svg){.lin}

A variation of Code 128 the 'Nummer der Versandeinheit' standard, also known as
SSCC-18 (Serial Shipping Container Code), includes both a visible modulo-10 and
a hidden modulo-103 check digit. NVE-18 requires a 17-digit numerical input.
Check digits and HRT-only AI "(00)" are added by Zint.

#### 6.1.10.6 HIBC Code 128

![`zint -b HIBC_128 -d "A123BJC5D6E71"`](images/hibc_128.svg){.lin}

This option adds a leading `'+'` character and a trailing modulo-49 check digit
to a standard Code 128 symbol as required by the Health Industry Barcode
standards.

#### 6.1.10.7 DPD Code

![`zint -b DPD --compliantheight -d
"000393206219912345678101040"`](images/dpd.svg){.lin}

Another variation of Code 128 as used by DPD (Deutscher Paketdienst). Requires a
27 or 28 character input. For 28 character input, the first character is an
identification tag (Barcode ID), which should usually be `"%"` (ASCII 37). If 27
characters are supplied, `"%"` will be prefixed by Zint (except if marked as a
"relabel", see below). The rest of the 27-character input must be alphanumeric,
and is of the form:

-----------------------------------------------------------------------
Destination Post   Tracking Number     Service     Destination Country
Code                                   Code        Code
-----------------  ------------------  ----------  --------------------
PPPPPPP            TTTTTTTTTTTTTT      SSS         CCC
(7 alphanumerics)  (14 alphanumerics)  (3 digits)  (3-digit ISO 3166-1)
-----------------------------------------------------------------------

Table: {#tbl:dpd_input_fields tag=": DPD Input Fields"}

A warning will be generated if the Service Code, the Destination Country Code,
or the last 10 characters of the Tracking Number are non-numeric.

Zint formats the Human Readable Text as specified by DPD, leaving out the
identication tag, and adds a modulo-36 check character to the text (not to the
barcode itself), thus:

`PPPP PPP TTTT TTTT TTTT TT SSS CCC D`

By default a top boundary bar is added, with default width 3X. The width can be
overridden using `--border` (API `border_width`). For a symbol with no top
boundary bar, explicitly set the border type to bindtop (or bind or box) and
leave the border width 0.

A DPD Code can be marked as a "relabel" by specifying `--vers=1` (API
`option_2 = 1`), which omits the identification tag and prints the barcode at
half height. In this case, an input of 27 alphanumeric characters is required.

\clearpage

#### 6.1.10.8 UPU S10

![`zint -b UPU_S10 --compliantheight -d
"EE876543216CA"`](images/upu_s10.svg){.lin}

The Universal Postal Union S10 variant of Code 128 encodes 13 characters in the
format `"SSNNNNNNNNXCC"`, where `"SS"` is a two-character alphabetic service
indicator, `"NNNNNNNN"` is an 8-digit serial number, `"X"` is a modulo-11 check
digit, and `"CC"` is a two-character ISO 3166-1 country code.

The check digit may be omitted in which case Zint will add it. Warnings will be
generated if the service indicator is non-standard or the country code is not
ISO 3361-1.

### 6.1.11 GS1 DataBar (ISO 24724)

Previously known as RSS (Reduced Spaced Symbology), these symbols are due to
replace GS1-128 symbols in accordance with the GS1 General Specifications. If a
GS1 DataBar symbol is to be printed with a 2D component as specified in ISO/IEC
24723 set `--mode=2` (API `option_1 = 2`). See [6.3 GS1 Composite Symbols (ISO
24723)] to find out how to generate DataBar symbols with 2D components.

#### 6.1.11.1 GS1 DataBar Omnidirectional and GS1 DataBar Truncated

![`zint -b DBAR_OMN --compliantheight -d
"0950110153001"`](images/dbar_omn.svg){.lin}

Previously known as RSS-14 this standard encodes a 13-digit item code. A check
digit and HRT-only Application Identifier of "(01)" are added by Zint. (A
14-digit code that appends the check digit may be given, in which case the check
digit will be verified.)

GS1 DataBar Omnidirectional symbols should have a height of 33 or greater. To
produce a GS1 DataBar Truncated symbol set the symbol height to a value between
13 and 32. Truncated symbols may not be scannable by omnidirectional scanners.

![`zint -b DBAR_OMN -d "0950110153001"
--height=13`](images/dbar_truncated.svg){.lin}

#### 6.1.11.2 GS1 DataBar Limited

![`zint -b DBAR_LTD --compliantheight -d
"0950110153001"`](images/dbar_ltd.svg){.lin}

Previously known as RSS Limited this standard encodes a 13-digit item code and
can be used in the same way as GS1 DataBar Omnidirectional above. GS1 DataBar
Limited, however, is limited to data starting with digits 0 and 1 (i.e. numbers
in the range 0 to 1999999999999). As with GS1 DataBar Omnidirectional a check
digit and HRT-only Application Identifier of "(01)" are added by Zint, and a
14-digit code may be given in which case the check digit will be verified.

#### 6.1.11.3 GS1 DataBar Expanded

![`zint -b DBAR_EXP --compliantheight -d
"[01]98898765432106[3202]012345[15]991231"`](images/dbar_exp.svg){.lin}

Previously known as RSS Expanded this is a variable length symbology capable of
encoding data from a number of AIs in a single symbol. AIs should be encased in
[square brackets] in the input data, which will be converted to parentheses
(round brackets) before being included in the Human Readable Text attached to
the symbol. This method allows the inclusion of parentheses in the data to be
encoded. If the data does not include parentheses, the AIs may alternatively be
encased in parentheses using the `--gs1parens` switch. See [6.1.10.3 GS1-128].

GTIN data AI (01) should also include the check digit data as this is not
calculated by Zint when this symbology is encoded. Fixed length data should be
entered at the appropriate length for correct encoding. The following is an
example of a valid GS1 DataBar Expanded input:

```bash
zint -b 31 -d "[01]98898765432106[3202]012345[15]991231"
```

### 6.1.12 Korea Post Barcode

![`zint -b KOREAPOST -d "923457"`](images/koreapost.svg){.lin}

The Korean Postal Barcode is used to encode a 6-digit number and includes one
check digit.

### 6.1.13 Channel Code

![`zint -b CHANNEL -d "453678" --compliantheight`](images/channel.svg){.lin}

A highly compressed symbol for numeric data. The number of channels in the
symbol can be between 3 and 8 and this can be specified by setting the value of
the `--vers` option (API `option_2`). It can also be determined by the length of
the input data: e.g. a three character input string generates a 4 channel code
by default.

The maximum values permitted depend on the number of channels used as shown in
the table below:

| Channels | Minimum Value | Maximum Value
|:---------|:--------------|:-------------
| 3        | 00            | 26
| 4        | 000           | 292
| 5        | 0000          | 3493
| 6        | 00000         | 44072
| 7        | 000000        | 576688
| 8        | 0000000       | 7742862

Table: {#tbl:channel_maxima tag=": Channel Value Ranges"}

### 6.1.14 BC412 (SEMI T1-95)

![`zint -b BC412 -d "AQ45670" --compliantheight`](images/bc412.svg){.lin}

Designed by IBM for marking silicon wafers, each BC412 character is represented
by 4 bars of a single size, interleaved with 4 spaces of varying sizes that
total 8 (hence 4 bars in 12). Zint implements the SEMI T1-95 standard, where
input must be alphanumeric, excluding the letter `O`, and must be from 7 to 18
characters in length. A single check character is added by Zint, appearing in
the 2nd character position. Lowercase input is automatically made uppercase.

\clearpage

## 6.2 Stacked Symbologies

### 6.2.1 Basic Symbol Stacking

An early innovation to get more information into a symbol, used primarily in the
vehicle industry, is to simply stack one-dimensional codes on top of each other.
This can be achieved at the command prompt by giving more than one set of input
data. For example

```bash
zint -d "This" -d "That"
```

will draw two Code 128 symbols, one on top of the other. The same result can be
achieved using the API by executing the `ZBarcode_Encode()` function more than
once on a symbol. For example:

```c
my_symbol->symbology = BARCODE_CODE128;

error = ZBarcode_Encode(my_symbol, "This", 0);

error = ZBarcode_Encode(my_symbol, "That", 0);

error = ZBarcode_Print(my_symbol);
```

![`zint -d "This" -d "That"`](images/code128_stacked.svg){.lin}

Note that the Human Readable Text will be that of the last data, so it's best to
use the option `--notext` (API `show_hrt = 0`).

The stacked barcode rows can be separated by row separator bars by specifying
`--bind` (API `output_options |= BARCODE_BIND`). The height of the row separator
bars in integral multiples of the X-dimension (minimum and default 1, maximum 4)
can be set by `--separator` (API `option_3`):

```bash
zint --bind --notext --separator=2 -d "This" -d "That"
```

![`zint --notext --bind --separator=2 -d "This" -d
"That"`](images/code128_stacked_sep2.svg){.lin}

A more sophisticated method is to use some type of line indexing which indicates
to the barcode reader which order the stacked symbols should be read in. This is
demonstrated by the symbologies below.

### 6.2.2 Codablock-F

![`zint -b CODABLOCKF -d "CODABLOCK F Symbology"
--rows=3`](images/codablockf.svg){.lin}

This is a stacked symbology based on Code 128 which can encode Latin-1 data up
to a maximum length of 2726 symbol characters, meaning for instance up to 2726
all ASCII characters, or 5452 all numeric, or up to 1363 all extended ASCII
(ISO/IEC 8859-1).

The width of the Codablock-F symbol can be set using the `--cols` option (API
`option_2`), to a value between 9 and 67. The height (number of rows) can be set
using the `--rows` option (API `option_1`), with a maximum of 44. Zint does not
currently support encoding of GS1 data in Codablock-F symbols.

A separate symbology ID (`BARCODE_HIBC_BLOCKF`) can be used to encode Health
Industry Barcode (HIBC) data which adds a leading `'+'` character and a
modulo-49 check digit to the encoded data.

### 6.2.3 Code 16K (EN 12323)

![`zint -b CODE16K --compliantheight -d
"ab0123456789"`](images/code16k.svg){.lin}

Code 16K uses a Code 128 based system which can stack up to 16 rows in a block.
This gives a maximum data capacity of 77 characters or 154 numerical digits and
includes two modulo-107 check digits. Code 16K also supports ISO/IEC 8859-1
character encoding in the same manner as Code 128. GS1 data encoding is also
supported. The minimum number of rows to use can be set using the `--rows`
option (API `option_1`), with values from 2 to 16.

### 6.2.4 PDF417 (ISO 15438)

![`zint -b PDF417 -d "PDF417"`](images/pdf417.svg){.lin}

Heavily used in the parcel industry, the PDF417 symbology can encode a vast
amount of data into a small space. Zint supports encoding up to the ISO standard
maximum symbol size of 925 codewords which (at error correction level 0) allows
a maximum data size of 1850 text characters, or 2710 digits.

The width of the generated PDF417 symbol can be specified at the command line
using the `--cols` switch (API `option_2`) followed by a number between 1 and
30, the number of rows using the `--rows` switch (API `option_3`) followed by a
number between 3 and 90, and the amount of error correction information can be
specified by using the `--secure` switch (API `option_1`) followed by a number
between 0 and 8 where the number of codewords used for error correction is
determined by `2^(value + 1)`. The default level of error correction is
determined by the amount of data being encoded.

This symbology uses Latin-1 character encoding by default but also supports the
ECI encoding mechanism. A separate symbology ID (`BARCODE_HIBC_PDF`) can be used
to encode Health Industry Barcode (HIBC) data.

For a faster but less optimal encoding, the `--fast` option (API `input_mode |=
FAST_MODE`) may be used.

PDF417 supports Structured Append of up to 99,999 symbols and an optional
numeric ID of up to 30 digits, which can be set by using the `--structapp`
option (see [4.17 Structured Append]) (API `structapp`). The ID consists of up
to 10 triplets, each ranging from `"000"` to `"899"`. For instance `"123456789"`
would be a valid ID of 3 triplets. However `"123456900"` would not, as the last
triplet `"900"` exceeds `"899"`. The triplets are 0-filled, for instance
`"1234"` becomes `"123004"`. If an ID is not given, no ID is encoded.

### 6.2.5 Compact PDF417 (ISO 15438)

![`zint -b PDF417COMP -d "PDF417"`](images/pdf417comp.svg){.lin}

Previously known as Truncated PDF417, Compact PDF417 omits some per-row overhead
to produce a narrower but less robust symbol. Options are the same as for PDF417
above.

### 6.2.6 MicroPDF417 (ISO 24728)

![`zint -b MICROPDF417 -d "12345678"`](images/micropdf417.svg){.lin}

A variation of the PDF417 standard, MicroPDF417 is intended for applications
where symbol size needs to be kept to a minimum. 34 predefined symbol sizes are
available with 1 - 4 columns and 4 - 44 rows. The maximum amount a MicroPDF417
symbol can hold is 250 alphanumeric characters or 366 digits. The amount of
error correction used is dependent on symbol size. The number of columns used
can be determined using the `--cols` switch (API `option_2`) as with PDF417.

This symbology uses Latin-1 character encoding by default but also supports the
ECI encoding mechanism. A separate symbology ID (`BARCODE_HIBC_MICPDF`) can be
used to encode Health Industry Barcode (HIBC) data. MicroPDF417 supports
`FAST_MODE` and Structured Append the same as PDF417, for which see details.

### 6.2.7 GS1 DataBar Stacked (ISO 24724)

#### 6.2.7.1 GS1 DataBar Stacked

![`zint -b DBAR_STK --compliantheight -d
"9889876543210"`](images/dbar_stk.svg){.lin}

A stacked variation of the GS1 DataBar Truncated symbol requiring the same input
(see [6.1.11.1 GS1 DataBar Omnidirectional and GS1 DataBar Truncated]), this
symbol is the same as the following GS1 DataBar Stacked Omnidirectional symbol
except that its height is reduced and its central separator is a single row,
making it suitable for small items when omnidirectional scanning is not
required. It can be generated with a two-dimensional component to make a
composite symbol.

#### 6.2.7.2 GS1 DataBar Stacked Omnidirectional

![`zint -b DBAR_OMNSTK --compliantheight -d
"9889876543210"`](images/dbar_omnstk.svg){.lin}

A stacked variation of the GS1 DataBar Omnidirectional symbol requiring the same
input (see [6.1.11.1 GS1 DataBar Omnidirectional and GS1 DataBar Truncated]).
The data is encoded in two rows of bars with a central 3-row separator. This
symbol can be generated with a two-dimensional component to make a composite
symbol.

#### 6.2.7.3 GS1 DataBar Expanded Stacked

![`zint -b DBAR_EXPSTK --compliantheight -d
"[01]98898765432106[3202]012345[15]991231"`](images/dbar_expstk.svg){.lin}

A stacked variation of the GS1 DataBar Expanded symbol for smaller packages.
Input is the same as for GS1 DataBar Expanded (see [6.1.11.3 GS1 DataBar
Expanded]). In addition the width of the symbol can be altered using the
`--cols` switch (API `option_2`). In this case the number of columns (values 1
to 11) relates to the number of character pairs on each row of the symbol.
Alternatively the `--rows` switch (API `option_3`) can be used to specify the
maximum number of rows (values 2 to 11), and the number of columns will be
adjusted accordingly. This symbol can be generated with a two-dimensional
component to make a composite symbol. For symbols with a 2D component the number
of columns must be at least 2.

### 6.2.8 Code 49

![`zint -b CODE49 --compliantheight -d
"MULTIPLE ROWS IN CODE 49"`](images/code49.svg){.lin}

Developed in 1987 at Intermec, Code 49 is a cross between UPC and Code 39. It is
one of the earliest stacked symbologies and influenced the design of Code 16K a
few years later. It supports full 7-bit ASCII input up to a maximum of 49
characters or 81 numeric digits. GS1 data encoding is also supported. The
minimum number of rows to use can be set using the `--rows` option (API
`option_1`), with values from 2 to 8.

\clearpage

## 6.3 GS1 Composite Symbols (ISO 24723)

GS1 Composite symbols employ a mixture of components to give more comprehensive
information about a product. The permissible contents of a composite symbol is
determined by the terms of the GS1 General Specifications. Composite symbols
consist of a linear component which can be an EAN, UPC, GS1-128 or GS1 DataBar
symbol, a two-dimensional (2D) component which is based on PDF417 or
MicroPDF417, and a separator pattern. The type of linear component to be used is
determined using the `-b` or `--barcode` switch (API `symbology`) as with other
encoding methods. Valid values are shown below.

---------------------------------------------------------------------------
Numeric  Name                      Barcode Name
Value
-------  ------------------------  ----------------------------------------
130      `BARCODE_EANX_CC`         GS1 Composite Symbol with EAN linear
                                     component

131      `BARCODE_GS1_128_CC`      GS1 Composite Symbol with GS1-128 linear
                                     component

132      `BARCODE_DBAR_OMN_CC`     GS1 Composite Symbol with GS1 DataBar
                                     Omnidirectional linear component

133      `BARCODE_DBAR_LTD_CC`     GS1 Composite Symbol with GS1 DataBar
                                     Limited linear component

134      `BARCODE_DBAR_EXP_CC`     GS1 Composite Symbol with GS1 DataBar
                                     Expanded linear component

135      `BARCODE_UPCA_CC`         GS1 Composite Symbol with UPC-A linear
                                     component

136      `BARCODE_UPCE_CC`         GS1 Composite Symbol with UPC-E linear
                                     component

137      `BARCODE_DBAR_STK_CC`     GS1 Composite Symbol with GS1 DataBar
                                     Stacked component

138      `BARCODE_DBAR_OMNSTK_CC`  GS1 Composite Symbol with GS1 DataBar
                                     Stacked Omnidirectional component

139      `BARCODE_DBAR_EXPSTK_CC`  GS1 Composite Symbol with GS1 DataBar
                                     Expanded Stacked component
---------------------------------------------------------------------------

Table: {#tbl:composite_symbologies tag=": GS1 Composite Symbology Values"}

The data to be encoded in the linear component of a composite symbol should be
entered into a primary string with the data for the 2D component being entered
in the normal way. To do this at the command prompt use the `--primary` switch
(API `primary`). For example:

```bash
zint -b EANX_CC --mode=1 --primary=331234567890 -d "[99]1234-abcd"
```

This creates an EAN-13 linear component with the data `"331234567890"` and a 2D
CC-A (see [below][6.3.1 CC-A]) component with the data `"(99)1234-abcd"`. The
same results can be achieved using the API as shown below:

```c
my_symbol->symbology = BARCODE_EANX_CC;

my_symbol->option_1 = 1;

strcpy(my_symbol->primary, "331234567890");

ZBarcode_Encode_and_Print(my_symbol, "[99]1234-abcd", 0, 0);
```

EAN-2 and EAN-5 add-on data can be used with EAN and UPC symbols using the +
symbol as described in sections [6.1.3 UPC (Universal Product Code) (ISO 15420)]
and [6.1.4 EAN (European Article Number) (ISO 15420)].

The 2D component of a composite symbol can use one of three systems: CC-A, CC-B
and CC-C, as described below. The 2D component type can be selected
automatically by Zint dependent on the length of the input string. Alternatively
the three methods can be accessed using the `--mode` prompt (API `option_1`)
followed by 1, 2 or 3 for CC-A, CC-B or CC-C respectively.

### 6.3.1 CC-A

![`zint -b EANX_CC --compliantheight -d "[99]1234-abcd" --mode=1
--primary=331234567890`](images/eanx_cc_a.svg){.upcean}

This system uses a variation of MicroPDF417 which is optimised to fit into a
small space. The size of the 2D component and the amount of error correction is
determined by the amount of data to be encoded and the type of linear component
which is being used. CC-A can encode up to 56 numeric digits or an alphanumeric
string of shorter length. To select CC-A use `--mode=1` (API `option_1 = 1`).

### 6.3.2 CC-B

![`zint -b EANX_CC --compliantheight -d "[99]1234-abcd" --mode=2
--primary=331234567890`](images/eanx_cc_b.svg){.upcean}

This system uses MicroPDF417 to encode the 2D component. The size of the 2D
component and the amount of error correction is determined by the amount of data
to be encoded and the type of linear component which is being used. CC-B can
encode up to 338 numeric digits or an alphanumeric string of shorter length. To
select CC-B use `--mode=2` (API `option_1 = 2`).

### 6.3.3 CC-C

![`zint -b GS1_128_CC --compliantheight -d "[99]1234-abcd" --mode=3
--primary="[01]03312345678903"`](images/gs1_128_cc_c.svg){.upcean}

This system uses PDF417 and can only be used in conjunction with a GS1-128
linear component. CC-C can encode up to 2361 numeric digits or an alphanumeric
string of shorter length. To select CC-C use `--mode=3` (API `option_1 = 3`).

\clearpage

## 6.4 Two-Track Symbols

### 6.4.1 Two-Track Pharmacode

![`zint -b PHARMA_TWO --compliantheight -d
"29876543"`](images/pharma_two.svg){.trk}

Developed by Laetus, Pharmacode Two-Track is an alternative system to Pharmacode
One-Track (see [6.1.9 Pharmacode]) used for the identification of
pharmaceuticals. The symbology is able to encode whole numbers between 4 and
64570080.

### 6.4.2 POSTNET

![`zint -b POSTNET --compliantheight -d
"12345678901"`](images/postnet.svg){.trk}

Used by the United States Postal Service until 2009, the POSTNET barcode was
used for encoding zip-codes on mail items. POSTNET uses numerical input data and
includes a modulo-10 check digit. While Zint will encode POSTNET symbols of up
to 38 digits in length, standard lengths as used by USPS were `PostNet6`
(5-digit ZIP input), `PostNet10` (5-digit ZIP + 4-digit user data) and
`PostNet12` (5-digit ZIP + 6-digit user data), and a warning will be issued if
the input length is not one of these.

### 6.4.3 PLANET

![`zint -b PLANET --compliantheight -d
"4012345235636"`](images/planet.svg){.trk}

Used by the United States Postal Service until 2009, the PLANET (Postal Alpha
Numeric Encoding Technique) barcode was used for encoding routing data on mail
items. PLANET uses numerical input data and includes a modulo-10 check digit.
While Zint will encode PLANET symbols of up to 38 digits in length, standard
lengths used by USPS were `Planet12` (11-digit input) and `Planet14` (13-digit
input), and as with POSTNET a warning will be issued if the length is not one of
these.

### 6.4.4 Brazilian CEPNet

![`zint -b CEPNET --compliantheight -d "12345678"`](images/cepnet.svg){.trk}

Based on POSTNET, the CEPNet symbol is used by Correios, the Brazilian postal
service, to encode CEP (Código de Endereçamento Postal) numbers on mail items.
Input should consist of eight digits with the check digit being automatically
added by Zint.

\clearpage

## 6.5 4-State Postal Codes

### 6.5.1 Australia Post 4-State Symbols

#### 6.5.1.1 Customer Barcodes

![`zint -b AUSPOST --compliantheight -d "96184209"`](images/auspost.svg){.trk}

Australia Post Standard Customer Barcode, Customer Barcode 2 and Customer
Barcode 3 are 37-bar, 52-bar and 67-bar specifications respectively, developed
by Australia Post for printing Delivery Point ID (DPID) and customer information
on mail items. Valid data characters are 0-9, A-Z, a-z, space and hash (#). A
Format Control Code (FCC) is added by Zint and should not be included in the
input data. Reed-Solomon error correction data is generated by Zint. Encoding
behaviour is determined by the length of the input data according to the formula
shown in the following table.

-------------------------------------------------------------
Input     Required Input Format       Symbol   FCC   Encoding
Length                                Length         Table
------    -------------------------   ------   ---   --------
8         `99999999`                  37-bar   11    None

13        `99999999AAAAA`             52-bar   59    C

16        `9999999999999999`          52-bar   59    N

18        `99999999AAAAAAAAAA`        67-bar   62    C

23        `99999999999999999999999`   67-bar   62    N
-------------------------------------------------------------

Table: {#tbl:auspost_input_formats tag=": Australia Post Input Formats"}

#### 6.5.1.2 Reply Paid Barcode

![`zint -b AUSREPLY --compliantheight -d "12345678"`](images/ausreply.svg){.trk}

A Reply Paid version of the Australia Post 4-State Barcode (FCC 45) which
requires an 8-digit DPID input.

#### 6.5.1.3 Routing Barcode

![`zint -b AUSROUTE --compliantheight -d "34567890"`](images/ausroute.svg){.trk}

A Routing version of the Australia Post 4-State Barcode (FCC 87) which requires
an 8-digit DPID input.

#### 6.5.1.4 Redirect Barcode

![`zint -b AUSREDIRECT --compliantheight -d
"98765432"`](images/ausredirect.svg){.trk}

A Redirection version of the Australia Post 4-State Barcode (FCC 92) which
requires an 8-digit DPID input.

### 6.5.2 Dutch Post KIX Code

![`zint -b KIX --compliantheight -d "2500GG30250"`](images/kix.svg){.trk}

This symbology is used by Royal Dutch TPG Post (Netherlands) for Postal code and
automatic mail sorting. Data input can consist of numbers 0-9 and letters A-Z
and needs to be 11 characters in length. No check digit is included.

### 6.5.3 Royal Mail 4-State Customer Code (RM4SCC)

![`zint -b RM4SCC --compliantheight -d "W1J0TR01"`](images/rm4scc.svg){.trk}

The RM4SCC standard is used by the Royal Mail in the UK to encode postcode and
customer data on mail items. Data input can consist of numbers 0-9 and letters
A-Z and usually includes delivery postcode followed by house number. For example
`"W1J0TR01"` for 1 Piccadilly Circus in London. Check digit data is generated by
Zint.

### 6.5.4 Royal Mail 4-State Mailmark

![`zint -b MAILMARK_4S --compliantheight -d
"1100000000000XY11"`](images/mailmark_4s.svg){.trk}

Developed in 2014 as a replacement for RM4SCC this 4-state symbol includes Reed-
Solomon error correction. Input is a pre-formatted alphanumeric string of 22
(for Barcode C) or 26 (for Barcode L) characters, producing a symbol with 66 or
78 bars respectively. The rules for the input data are complex, as summarized in
the following table.

---------------------------------------------------------------------------
Format   Version  Class        Supply Chain ID  Item ID   Destination+DPS
         ID
-------  -------  -----------  ---------------  --------  -----------------
1 digit  1 digit  1 alphanum.  2 digits (C) or  8 digits  9 alphanumerics
(0-4)    (0-3)    (0-9A-E)     6 digits (L)               (1 of 6 patterns)
---------------------------------------------------------------------------

Table: {#tbl:mailmark_4s_input_fields
tag=": Royal Mail 4-State Mailmark Input Fields"}

The 6 Destination+DPS (Destination Post Code plus Delivery Point Suffix)
patterns are:

-----------  -----------  -----------
`FNFNLLNLS`  `FFNNLLNLS`  `FFNNNLLNL`
`FFNFNLLNL`  `FNNLLNLSS`  `FNNNLLNLS`
-----------  -----------  -----------

Table: {#tbl:mailmark_destination_dps
tag=": Royal Mail Mailmark Destination+DPS Patterns"}

where `'F'` stands for full alphabetic (A-Z), `'L'` for limited alphabetic (A-Z
less `'CIKMOV'`), `'N'` for numeric (0-9), and `'S'` for space.

Four of the permitted patterns include a number of trailing space characters -
these will be appended by Zint if not included in the input data.

For the two-dimensional Data Matrix-based version, see [6.6.2 Royal Mail 2D
Mailmark (CMDM) (Data Matrix)].

### 6.5.5 USPS Intelligent Mail

![`zint -b USPS_IMAIL --compliantheight -d
"01234567094987654321-01234"`](images/usps_imail.svg){.trk}

Also known as the OneCode barcode and used in the U.S. by the United States
Postal Service (USPS), the Intelligent Mail system replaced the POSTNET and
PLANET symbologies in 2009. Intelligent Mail is a fixed length (65-bar) symbol
which combines routing and customer information in a single symbol. Input data
consists of a 20-digit tracking code, followed by a dash (`-`), followed by a
delivery point zip-code which can be 0, 5, 9 or 11 digits in length. For example
all of the following inputs are valid data entries:

- `"01234567094987654321"`
- `"01234567094987654321-01234"`
- `"01234567094987654321-012345678"`
- `"01234567094987654321-01234567891"`

### 6.5.6 Japanese Postal Code

![`zint -b JAPANPOST --compliantheight -d
"15400233-16-4-205"`](images/japanpost.svg){.trk}

Used for address data on mail items for Japan Post. Accepted values are 0-9,
A-Z and dash (`-`). A modulo 19 check digit is added by Zint.

### 6.5.7 DAFT Code

![`zint -b DAFT -d "AAFDTTDAFADTFTTFFFDATFTADTTFFTDAFAFDTF" --height=8.494
--vers=256`](images/daft_rm4scc.svg){.trk}

This is a method for creating 4-state codes where the data encoding is provided
by an external program. Input data should consist of the letters `'D'`, `'A'`,
`'F'` and `'T'` where these refer to descender, ascender, full (ascender and
descender) and tracker (neither ascender nor descender) respectively. All other
characters are invalid. The ratio of the tracker size to full height can be
given in thousandths (permille) using the `--vers` option (API `option_2`). The
default value is 250 (25%).

For example the following

```bash
zint -b DAFT -d AAFDTTDAFADTFTTFFFDATFTADTTFFTDAFAFDTF --height=8.494 --vers=256
```

produces the same barcode (see [6.5.3 Royal Mail 4-State Customer Code
(RM4SCC)]) as

```bash
zint -b RM4SCC --compliantheight -d "W1J0TR01"
```

\clearpage

## 6.6 Matrix Symbols

### 6.6.1 Data Matrix (ISO 16022)

![`zint -b HIBC_DM -d "/ACMRN123456/V200912190833" --fast
--square`](images/hibc_dm.svg){.i2dbig}

Also known as Semacode this symbology was developed in 1989 by Acuity CiMatrix
in partnership with the U.S. DoD and NASA. The symbol can encode a large amount
of data in a small area. Data Matrix encodes characters in the Latin-1 set by
default but also supports encoding in other character sets using the ECI
mechanism. It can also encode GS1 data. The size of the generated symbol can be
adjusted using the `--vers` option (API `option_2`) as shown in the table below.
A separate symbology ID (`BARCODE_HIBC_DM`) can be used to encode Health
Industry Barcode (HIBC) data. Note that only ECC200 encoding is supported, the
older standards have now been removed from Zint.

Input  Symbol Size      Input  Symbol Size      Input  Symbol Size
-----  -----------  --  -----  -----------  --  -----  -----------
1      10 x 10          11     36 x 36          21     104 x 104
2      12 x 12          12     40 x 40          22     120 x 120
3      14 x 14          13     44 x 44          23     132 x 132
4      16 x 16          14     48 x 48          24     144 x 144
5      18 x 18          15     52 x 52          25     8 x 18
6      20 x 20          16     64 x 64          26     8 x 32
7      22 x 22          17     72 x 72          28     12 x 26
8      24 x 24          18     80 x 80          28     12 x 36
9      26 x 26          19     88 x 88          29     16 x 36
10     32 x 32          20     96 x 96          30     16 x 48

Table: {#tbl:datamatrix_sizes tag=": Data Matrix Sizes"}

The largest version 24 (144 x 144) can encode 3116 digits, around 2335
alphanumeric characters, or 1555 bytes of data.

When using automatic symbol sizes you can force Zint to use square symbols
(versions 1-24) at the command line by using the option `--square` (API
`option_3 = DM_SQUARE`).

Data Matrix Rectangular Extension (ISO/IEC 21471) codes may be generated with
the following values as before:

Input  Symbol Size      Input  Symbol Size
-----  -----------  --  -----  -----------
31     8 x 48           40     20 x 36
32     8 x 64           41     20 x 44
33     8 x 80           42     20 x 64
34     8 x 96           43     22 x 48
35     8 x 120          44     24 x 48
36     8 x 144          45     24 x 64
37     12 x 64          46     26 x 40
38     12 x 88          47     26 x 48
39     16 x 64          48     26 x 64

Table: {#tbl:dmre_sizes tag=": DMRE Sizes"}

DMRE symbol sizes may be activated in automatic size mode using the option
`--dmre` (API `option_3 = DM_DMRE`).

GS1 data may be encoded using FNC1 (default) or GS (Group Separator, ASCII 29)
as separator. Use the option `--gssep` to change to GS (API `output_options |=
GS1_GS_SEPARATOR`).

By default Zint uses a "de facto" codeword placement for symbols of size 144 x
144 (version 24). To override this and use the now clarified ISO/IEC standard
placement, use option `--dmiso144` (API `option_3 |= DM_ISO_144`).

For a faster but less optimal encoding, the `--fast` option (API `input_mode |=
FAST_MODE`) may be used.

Data Matrix supports Structured Append of up to 16 symbols and a numeric ID
(file identifications), which can be set by using the `--structapp` option (see
[4.17 Structured Append]) (API `structapp`). The ID consists of 2 numbers `ID1`
and `ID2`, each of which can range from 1 to 254, and is specified as the single
number `ID1 * 1000 + ID2`, so for instance `ID1` `"123"` and `ID2` `"234"` would
be given as `"123234"`. Note that both `ID1` and `ID2` must be non-zero, so e.g.
`"123000"` or `"000123"` would be invalid IDs. If an ID is not given it defaults
to `"001001"`.

### 6.6.2 Royal Mail 2D Mailmark (CMDM) (Data Matrix)

![`zint -b MAILMARK_2D -d "JGB 01Z999999900000001EC1A1AA1A0SN35TQ"
--vers=30`](images/mailmark_2d.svg){.i2dbig}

This variant of Data Matrix, also known as "Complex Mail Data Mark" (CMDM), was
introduced by Royal Mail along with [6.5.4 Royal Mail 4-State Mailmark], and
offers space for customer data following an initial pre-formatted 45 character
section, as summarized below.

Field Name        Length       Values
----------------  -----------  ------------------------------
UPU Country ID    4            `"JGB "`
Information Type  1            Alphanumeric
Version ID        1            `"1"`
Class             1            Alphanumeric
Supply Chain ID   7            Numeric
Item ID           8            Numeric
Destination+DPS   9            Alphanumeric (1 of 6 patterns)
Service Type      1            Numeric
RTS Post Code     7            Alphanumeric (1 of 6 patterns)
Reserved          6            Spaces
Customer Data     6, 45 or 29  Anything (Latin-1)

Table: {#tbl:mailmark_2d_input_fields
tag=": Royal Mail 2D Mailmark Input Fields"}

The 6 Destination+DPS (Destination Post Code plus Delivery Point Suffix)
patterns are the same as for the 4-state - see Table
{@tbl:mailmark_destination_dps}. The 6 RTS (Return to Sender) Post Code patterns
are the same also except without the additional DPS `'NL'`, i.e.

---------  ---------  ---------
`FNFNLLS`  `FFNNLLS`  `FFNNNLL`
`FFNFNLL`  `FNNLLSS`  `FNNNLLS`
---------  ---------  ---------

Table: {#tbl:mailmark_2d_rts
tag=": Royal Mail 2D Mailmark RTS Patterns"}

where `'F'` is full alphabetic (A-Z), `'L'` limited alphabetic (A-Z less
`'CIKMOV'`), `'N'` numeric (0-9), and `'S'` space.

Three sizes are defined, one rectangular, with varying maximum amounts of
optional customer data:

Name     Size     Customer Data  Zint Version
-------  -------  -------------  ------------
Type 7   24 x 24  6 characters   8
Type 9   32 x 32  45 characters  10
Type 29  16 x 48  29 characters  30

Table: {#tbl:mailmark_2d_sizes tag=": Royal Mail 2D Mailmark Sizes"}

Zint will automatically select a size based on the amount of customer data, or
it can be specified using the `--vers` option (API `option_2`), which takes the
Zint version number (one more than the Royal Mail Type number). Zint will prefix
the input data with `"JGB "` if it's missing, and also space-pad the input if
the customer data is absent or falls short. As with Data Matrix, the rectangular
symbol Type 29 can be excluded from automatic size selection by using the option
`--square` (API `option_3 = DM_SQUARE`).

GS1 data, the ECI mechanism, and Structured Append are not supported.

### 6.6.3 QR Code (ISO 18004)

![`zint -b QRCODE -d "QR Code Symbol" --mask=5`](images/qrcode.svg){.i2dbig}

Also known as Quick Response Code this symbology was developed by Denso. Four
levels of error correction are available using the `--secure` option (API
`option_1`) as shown in the following table.

Input  ECC Level  Error Correction Capacity  Recovery Capacity
-----  ---------  -------------------------  -----------------
1      L          Approx 20% of symbol       Approx 7%
2      M          Approx 37% of symbol       Approx 15%
3      Q          Approx 55% of symbol       Approx 25%
4      H          Approx 65% of symbol       Approx 30%

Table: {#tbl:qrcode_eccs tag=": QR Code ECC Levels"}

The size of the symbol can be specified by setting the `--vers` option (API
`option_2`) to the QR Code version required (1-40). The size of symbol generated
is shown in the table below.

Input Symbol Size    Input Symbol Size    Input Symbol Size
----- ----------- -- ----- ----------- -- ----- -----------
1     21 x 21        15    77 x 77        29    133 x 133
2     25 x 25        16    81 x 81        30    137 x 137
3     29 x 29        17    85 x 85        31    141 x 141
4     33 x 33        18    89 x 89        32    145 x 145
5     37 x 37        19    93 x 93        33    149 x 149
6     41 x 41        20    97 x 97        34    153 x 153
7     45 x 45        21    101 x 101      35    157 x 157
8     49 x 49        22    105 x 105      36    161 x 161
9     53 x 53        23    109 x 109      37    165 x 165
10    57 x 57        24    113 x 113      38    169 x 169
11    61 x 61        25    117 x 117      39    173 x 173
12    65 x 65        26    121 x 121      40    177 x 177
13    69 x 69        27    125 x 125
14    73 x 73        28    129 x 129

Table: {#tbl:qrcode_sizes tag=": QR Code Sizes"}

The maximum capacity of a QR Code symbol (version 40) is 7089 numeric digits,
4296 alphanumeric characters or 2953 bytes of data. QR Code symbols can also be
used to encode GS1 data. QR Code symbols can by default encode either characters
in the Latin-1 set or Kanji, Katakana and ASCII characters which are members of
the Shift JIS encoding scheme. In addition QR Code supports other character sets
using the ECI mechanism. Input should usually be entered as UTF-8 with
conversion to Latin-1 or Shift JIS being carried out by Zint. A separate
symbology ID (`BARCODE_HIBC_QR`) can be used to encode Health Industry Barcode
(HIBC) data.

Non-ASCII data density may be maximized by using the `--fullmultibyte` switch
(API `option_3 = ZINT_FULL_MULTIBYTE`), but check that your barcode reader
supports this before using.

QR Code has eight different masks designed to minimize unwanted patterns. The
best mask to use is selected automatically by Zint but may be manually specified
by using the `--mask` switch with values 0-7, or in the API by setting
`option_3 = (N + 1) << 8` where N is 0-7. To use with `ZINT_FULL_MULTIBYTE` set
```c
option_3 = ZINT_FULL_MULTIBYTE | (N + 1) << 8
```

The `--fast` option (API `input_mode |= FAST_MODE`) may be used when leaving
Zint to automatically select a mask to reduce the number of masks to try to four
(0, 2, 4, 7).

QR Code supports Structured Append of up to 16 symbols and a numeric ID
(parity), which can be set by using the `--structapp` option (see [4.17
Structured Append]) (API `structapp`). The parity ID ranges from 0 (default) to
255, and for full compliance should be set to the value obtained by `XOR`-ing
together each byte of the complete data forming the sequence. Currently this
calculation must be done outside of Zint.

### 6.6.4 Micro QR Code (ISO 18004)

![`zint -b MICROQR -d "01234567"`](images/microqr.svg){.i2dbig}

A miniature version of the QR Code symbol for short messages, Micro QR Code
symbols can encode either Latin-1 characters or Shift JIS characters. Input
should be entered as a UTF-8 stream with conversion to Latin-1 or Shift JIS
being carried out automatically by Zint. A preferred symbol size can be selected
by using the `--vers` option (API `option_2`), as shown in the table below. Note
that versions M1 and M2 have restrictions on what characters can be encoded.

------------------------------------------------------------------
Input   Version   Symbol Size   Allowed Characters
-----   -------   -----------   ----------------------------------
1       M1        11 x 11       Numeric only

2       M2        13 x 13       Numeric, uppercase letters, space,
                                and the characters `"$%*+-./:"`

3       M3        15 x 15       Latin-1 and Shift JIS

4       M4        17 x 17       Latin-1 and Shift JIS
------------------------------------------------------------------

Table: {#tbl:micrqr_sizes tag=": Micro QR Code Sizes"}

Version M4 can encode up to 35 digits, 21 alphanumerics, 15 bytes or 9 Kanji
characters.

Except for version M1, which is always ECC level L, the amount of ECC codewords
can be adjusted using the `--secure` option (API `option_1`); however ECC level
H is not available for any version, and ECC level Q is only available for
version M4:

----------------------------------------------------------------------
Input    ECC      Error Correction        Recovery      Available for
         Level    Capacity                Capacity      Versions
-----    -----    --------------------    ----------    --------------
1        L        Approx 20% of symbol    Approx 7%     M1, M2, M3, M4

2        M        Approx 37% of symbol    Approx 15%    M2, M3, M4

3        Q        Approx 55% of symbol    Approx 25%    M4
----------------------------------------------------------------------

Table: {#tbl:micrqr_eccs tag=": Micro QR ECC Levels"}

The defaults for symbol size and ECC level depend on the input and whether
either of them is specified.

For barcode readers that support it, non-ASCII data density may be maximized by
using the `--fullmultibyte` switch (API `option_3 = ZINT_FULL_MULTIBYTE`).

Micro QR Code has four different masks designed to minimize unwanted patterns.
The best mask to use is selected automatically by Zint but may be manually
specified by using the `--mask` switch with values 0-3, or in the API by setting
`option_3 = (N + 1) << 8` where N is 0-3. To use with `ZINT_FULL_MULTIBYTE` set
```c
option_3 = ZINT_FULL_MULTIBYTE | (N + 1) << 8
```

### 6.6.5 Rectangular Micro QR Code (rMQR) (ISO 23941)

![`zint -b RMQR -d "0123456"`](images/rmqr.svg){.i2dbig}

A rectangular version of QR Code, rMQR supports encoding of GS1 data, and either
Latin-1 characters or Shift JIS characters, and other encodings using the ECI
mechanism. As with other symbologies data should be entered as UTF-8 with
conversion being handled by Zint. The amount of ECC codewords can be adjusted
using the `--secure` option (API `option_1`), however only ECC levels M and H
are valid for this type of symbol.

Input    ECC Level    Error Correction Capacity    Recovery Capacity
-----    ---------    -------------------------    -----------------
2        M            Approx 37% of symbol         Approx 15%
4        H            Approx 65% of symbol         Approx 30%

Table: {#tbl:rmqr_eccs tag=": rMQR ECC Levels"}

The preferred symbol sizes can be selected using the `--vers` option (API
`option_2`) as shown in the table below. Input values between 33 and 38 fix the
height of the symbol while allowing Zint to determine the minimum symbol width.

--------------------------------------------------------------------------
Input  Version  Symbol Size (HxW)     Input  Version  Symbol Size (HxW)
-----  -------  -----------------  -  -----  -------  --------------------
1      R7x43    7 x 43                20     R13x77   13 x 77

2      R7x59    7 x 59                21     R13x99   13 x 99

3      R7x77    7 x 77                22     R13x139  13 x 139

4      R7x99    7 x 99                23     R15x43   15 x 43

5      R7x139   7 x 139               24     R15x59   15 x 59

6      R9x43    9 x 43                25     R15x77   15 x 77

7      R9x59    9 x 59                26     R15x99   15 x 99

8      R9x77    9 x 77                27     R15x139  15 x 139

9      R9x99    9 x 99                28     R17x43   17 x 43

10     R9x139   9 x 139               29     R17x59   17 x 59

11     R11x27   11 x 27               30     R17x77   17 x 77

12     R11x43   11 x 43               31     R17x99   17 x 99

13     R11x59   11 x 59               32     R17x139  17 x 139

14     R11x77   11 x 77               33     R7xW     7 x automatic width

15     R11x99   11 x 99               34     R9xW     9 x automatic width

16     R11x139  11 x 139              35     R11xW    11 x automatic width

17     R13x27   13 x 27               36     R13xW    13 x automatic width

18     R13x43   13 x 43               37     R15xW    15 x automatic width

19     R13x59   13 x 59               38     R17xW    17 x automatic width
--------------------------------------------------------------------------

Table: {#tbl:rmqr_sizes tag=": rMQR Sizes"}

The largest version R17x139 (32) can encode up to 361 digits, 219 alphanumerics,
150 bytes, or 92 Kanji characters.

For barcode readers that support it, non-ASCII data density may be maximized by
using the `--fullmultibyte` switch or in the API by setting
`option_3 = ZINT_FULL_MULTIBYTE`.

### 6.6.6 UPNQR (Univerzalnega Plačilnega Naloga QR)

![`zint -b UPNQR -i upn_utf8.txt --quietzones`](images/upnqr.svg){.i2d}

A variation of QR Code used by Združenje Bank Slovenije (Bank Association of
Slovenia). The size, error correction level and ECI are set by Zint and do not
need to be specified. UPNQR is unusual in that it uses Latin-2 (ISO/IEC 8859-2
plus ASCII) formatted data. Zint will accept UTF-8 data and convert it to
Latin-2, or if your data is already Latin-2 formatted use the `--binary` switch
(API `input_mode = DATA_MODE`).

The following example creates a symbol from data saved as a Latin-2 file:

```bash
zint -o upnqr.png -b 143 --scale=3 --binary -i upn.txt
```

A mask may be manually specified or the `--fast` option used as with QRCODE.

### 6.6.7 MaxiCode (ISO 16023)

![`zint -b MAXICODE -d "1Z00004951\GUPSN\G06X610\G159\G1234567\G1/1\G\GY\G1 MAIN
ST\GNY\GNY\R\E" --esc --primary="152382802000000"
--scmvv=96`](images/maxicode.svg){.i2d}

Developed by UPS the MaxiCode symbology employs a grid of hexagons surrounding a
bullseye finder pattern. This symbology is designed for the identification of
parcels. MaxiCode symbols can be encoded in one of five modes. In modes 2 and 3
MaxiCode symbols are composed of two parts named the primary and secondary
messages. The primary message consists of a Structured Carrier Message which
includes various data about the package being sent and the secondary message
usually consists of address data in a data structure. The format of the primary
message required by Zint is given in the following table.

Characters   Meaning
----------   ---------------------------------------------------------------
1 - 9        Postcode data which can consist of up to 9 digits (for mode 2)
             or up to 6 alphanumeric characters (for mode 3). Remaining
             unused characters for mode 3 can be filled with the SPACE
             character (ASCII 32) or omitted.
             (adjust the following character positions according to postcode
             length)
10 - 12      Three-digit country code according to ISO 3166-1.
13 - 15      Three-digit service code. This depends on your parcel courier.

Table: {#tbl:maxicode_scm tag=": MaxiCode Structured Carrier Message Format"}

The primary message can be set at the command prompt using the `--primary`
switch (API `primary`). The secondary message uses the normal data entry method.
For example:

```bash
zint -o test.eps -b 57 --primary="999999999840012" \
    -d "Secondary Message Here"
```

When using the API the primary message must be placed in the `primary` string.
The secondary is entered in the same way as described in [5.2 Encoding and
Saving to File]. When either of these modes is selected Zint will analyse the
primary message and select either mode 2 or mode 3 as appropriate.

As a convenience the secondary message for modes 2 and 3 can be set to be
prefixed by the ISO/IEC 15434 Format `"01"` (transportation) sequence
`"[)>\R01\Gvv"`, where `vv` is a 2-digit version, by using the `--scmvv` switch
(API `option_2 = vv + 1`). For example to use the common version `"96"` (ASC
MH10/SC 8):

```bash
zint -b 57 --primary="152382802840001" --scmvv=96 --esc -d \
  "1Z00004951\GUPSN\G06X610\G159\G1234567\G1/1\G\GY\G1 MAIN ST\GNY\GNY\R\E"
```

will prefix `"[)>\R01\G96"` to the secondary message. (`\R`, `\G` and `\E` are
the escape sequences for Record Separator, Group Separator and End of
Transmission respectively - see Table {@tbl:escape_sequences}.)

Modes 4 to 6 can be accessed using the `--mode` switch (API `option_1`). Modes 4
to 6 do not have a primary message. For example:

```bash
zint -o test.eps -b 57 --mode=4 -d "A MaxiCode Message in Mode 4"
```

Mode 6 is reserved for the maintenance of scanner hardware and should not be
used to encode user data.

This symbology uses Latin-1 character encoding by default but also supports the
ECI encoding mechanism. The maximum length of text which can be placed in a
MaxiCode symbol depends on the type of characters used in the text.

Example maximum data lengths are given in the table below:

-----------------------------------------------------------------------
Mode   Maximum Data Length   Maximum Data Length   Number of Error
       for Capital Letters   for Numeric Digits    Correction Codewords
----   -------------------   -------------------   --------------------
2`*`   84                    126                   50

3`*`   84                    126                   50

4      93                    138                   50

5      77                    113                   66

6      93                    138                   50
-----------------------------------------------------------------------

Table: {#tbl:maxicode_data_length_maxima tag=": MaxiCode Data Length Maxima"}

`*` - secondary only

MaxiCode supports Structured Append of up to 8 symbols, which can be set by
using the `--structapp` option (see [4.17 Structured Append]) (API `structapp`).
It does not support specifying an ID.

MaxiCode uses a different scaling than other symbols for raster output, see
[4.9.3 MaxiCode Raster Scaling], and also for EMF vector output, when the scale
is multiplied by 20 instead of 2.

### 6.6.8 Aztec Code (ISO 24778)

![`zint -b AZTEC -d "123456789012"`](images/aztec.svg){.i2d}

Invented by Andrew Longacre at Welch Allyn Inc in 1995 the Aztec Code symbol is
a matrix symbol with a distinctive bullseye finder pattern. Zint can generate
Compact Aztec Code (sometimes called Small Aztec Code) as well as 'full-range'
Aztec Code symbols and by default will automatically select symbol type and size
dependent on the length of the data to be encoded. Error correction codewords
will normally be generated to fill at least 23% of the symbol. Two options are
available to change this behaviour:

The size of the symbol can be specified using the `--vers` option (API
`option_2`) to a value between 1 and 36 according to the following table. The
symbols marked with an asterisk (`*`) in the table below are 'compact' symbols,
meaning they have a smaller bullseye pattern at the centre of the symbol.

Input  Symbol Size      Input  Symbol Size      Input  Symbol Size
-----  -----------  --  -----  -----------  --  -----  -----------
1      15 x 15`*`       13     53 x 53          25     105 x 105
2      19 x 19`*`       14     57 x 57          26     109 x 109
3      23 x 23`*`       15     61 x 61          27     113 x 113
4      27 x 27`*`       16     67 x 67          28     117 x 117
5      19 x 19          17     71 x 71          29     121 x 121
6      23 x 23          18     75 x 75          30     125 x 125
7      27 x 27          19     79 x 79          31     131 x 131
8      31 x 31          20     83 x 83          32     135 x 135
9      37 x 37          21     87 x 87          33     139 x 139
10     41 x 41          22     91 x 91          34     143 x 143
11     45 x 45          23     95 x 95          35     147 x 147
12     49 x 49          24     101 x 101        36     151 x 151

Table: {#tbl:aztec_sizes tag=": Aztec Code Sizes"}

Note that in symbols which have a specified size the amount of error correction
is dependent on the length of the data input and Zint will allow error
correction capacities as low as 3 codewords.

Alternatively the amount of error correction data can be specified by setting
the `--secure` option (API `option_1`) to a value from the following table.

Mode   Error Correction Capacity
----   -------------------------
1      >10% + 3 codewords
2      >23% + 3 codewords
3      >36% + 3 codewords
4      >50% + 3 codewords

Table: {#tbl:aztec_eccs tag=": Aztec Code Error Correction Modes"}

It is not possible to select both symbol size and error correction capacity for
the same symbol. If both options are selected then the error correction capacity
selection will be ignored.

Aztec Code supports ECI encoding and can encode up to a maximum length of
approximately 3823 numeric or 3067 alphabetic characters or 1914 bytes of data.
A separate symbology ID (`BARCODE_HIBC_AZTEC`) can be used to encode Health
Industry Barcode (HIBC) data.

Aztec Code supports Structured Append of up to 26 symbols and an optional
alphanumeric ID of up to 32 characters, which can be set by using the
`--structapp` option (see [4.17 Structured Append]) (API `structapp`). The ID
cannot contain spaces. If an ID is not given, no ID is encoded.

### 6.6.9 Aztec Runes (ISO 24778)

![`zint -b AZRUNE -d "125"`](images/azrune.svg){.i2d}

A truncated version of compact Aztec Code for encoding whole integers between 0
and 255, as defined in ISO/IEC 24778 Annex A. Includes Reed-Solomon error
correction. It does not support Structured Append.

### 6.6.10 Code One

![`zint -b CODEONE -d "1234567890123456789012"`](images/codeone.svg){.i2d}

A matrix symbology developed by Ted Williams in 1992 which encodes data in a way
similar to Data Matrix, Code One is able to encode the Latin-1 character set or
GS1 data, and also supports the ECI mechanism. There are two types of Code One
symbol - fixed-ratio symbols which are roughly square (versions A through to H)
and variable-width versions (versions S and T). These can be selected by using
`--vers` (API `option_2`) as shown in the table below:

------------------------------------------------------------
Input   Version   Size         Numeric         Alphanumeric
                  (W x H)      Data Capacity   Data Capacity
-----   -------   ----------   -------------   -------------
1       A         16 x 18      22              13

2       B         22 x 22      44              27

3       C         28 x 28      104             64

4       D         40 x 42      217             135

5       E         52 x 54      435             271

6       F         70 x 76      886             553

7       G         104 x 98     1755            1096

8       H         148 x 134    3550            2218

9       S         width x 8    18              N/A

10      T         width x 16   90              55
------------------------------------------------------------

Table: {#tbl:codeone_sizes tag=": Code One Sizes"}

Version S symbols can only encode numeric data. The width of version S and
version T symbols is determined by the length of the input data.

Code One supports Structured Append of up to 128 symbols, which can be set by
using the `--structapp` option (see [4.17 Structured Append]) (API `structapp`).
It does not support specifying an ID. Structured Append is not supported with
GS1 data nor for Version S symbols.

### 6.6.11 Grid Matrix

![`zint -b GRIDMATRIX --eci=29 -d "AAT2556 电池充电器+降压转换器
 200mA至2A tel:86 019 82512738"`](images/gridmatrix.svg){.i2d}

Grid Matrix groups modules in a chequerboard pattern, and by default supports
the GB 2312 standard set, which includes Hanzi, ASCII and a small number of
ISO/IEC 8859-1 characters. Input should be entered as UTF-8 with conversion to
GB 2312 being carried out automatically by Zint. Up to around 1529 alphanumeric
characters or 2751 digits may be encoded. The symbology also supports the ECI
mechanism. Support for GS1 data has not yet been implemented.

The size of the symbol and the error correction capacity can be specified. If
you specify both of these values then Zint will make a 'best-fit' attempt to
satisfy both conditions. The symbol size can be specified using the `--vers`
option (API `option_2`), and the error correction capacity can be specified by
using the `--secure` option (API `option_1`), according to the following tables.

Input   Symbol Size       Input   Symbol Size
-----   -----------   -   -----   -----------
1       18 x 18           8       102 x 102
2       30 x 30           9       114 x 114
3       42 x 42           10      126 x 126
4       54 x 54           11      138 x 138
5       66 x 66           12      150 x 150
6       78 x 78           13      162 x 162
7       90 x 90

Table: {#tbl:gridmatrix_sizes tag=": Grid Matrix Sizes"}

Mode    Error Correction Capacity
----    -------------------------
1       Approximately 10%
2       Approximately 20%
3       Approximately 30%
4       Approximately 40%
5       Approximately 50%

Table: {#tbl:gridmatrix_eccs tag=": Grid Matrix Error Correction Modes"}

Non-ASCII data density may be maximized by using the `--fullmultibyte` switch
(API `option_3 = ZINT_FULL_MULTIBYTE`), but check that your barcode reader
supports this before using.

Grid Matrix supports Structured Append of up to 16 symbols and a numeric ID
(file signature), which can be set by using the `--structapp` option (see [4.17
Structured Append]) (API `structapp`). The ID ranges from 0 (default) to 255.

### 6.6.12 DotCode

![`zint -b DOTCODE -d "[01]00012345678905[17]201231[10]ABC123456"
--gs1`](images/dotcode.svg){.i2d}

DotCode uses a grid of dots in a rectangular formation to encode characters up
to a maximum of approximately 450 characters (or 900 numeric digits). The
symbology supports ECI encoding and GS1 data encoding. By default Zint will
produce a symbol which is approximately square, however the width of the symbol
can be adjusted by using the `--cols` option (API `option_2`) (maximum 200).
Outputting DotCode to raster images (BMP, GIF, PCX, PNG, TIF) will require
setting the scale of the image to a larger value than the default (e.g.
approximately 10) for the dots to be plotted correctly. Approximately 33% of the
resulting symbol is comprised of error correction codewords.

DotCode has two sets of 4 masks, designated 0-3 and 0'-3', the second `"prime"`
set being the same as the first with corners lit. The best mask to use is
selected automatically by Zint but may be manually specified by using the
`--mask` switch with values 0-7, where 4-7 denote 0'-3', or in the API by
setting `option_3 = (N + 1) << 8` where N is 0-7.

DotCode supports Structured Append of up to 35 symbols, which can be set by
using the `--structapp` option (see [4.17 Structured Append]) (API `structapp`).
It does not support specifying an ID.

### 6.6.13 Han Xin Code (ISO 20830)

![`zint -b HANXIN -d "Hanxin Code symbol"`](images/hanxin.svg){.i2d}

Also known as Chinese Sensible Code, Han Xin is capable of encoding characters
in either the Latin-1 character set or the GB 18030 character set (which is a
UTF, i.e. includes all Unicode characters, optimized for Chinese characters) and
is also able to support the ECI mechanism. Support for the encoding of GS1 data
has not yet been implemented.

The size of the symbol can be specified using the `--vers` option (API
`option_2`) to a value between 1 and 84 according to the following table.

Input Symbol Size    Input Symbol Size    Input Symbol Size
----- ----------- -- ----- ----------- -- ----- -----------
1     23 x 23        29    79 x 79        57    135 x 135
2     25 x 25        30    81 x 81        58    137 x 137
3     27 x 27        31    83 x 83        59    139 x 139
4     29 x 29        32    85 x 85        60    141 x 141
5     31 x 31        33    87 x 87        61    143 x 143
6     33 x 33        34    89 x 89        62    145 x 145
7     35 x 35        35    91 x 91        63    147 x 147
8     37 x 37        36    93 x 93        64    149 x 149
9     39 x 39        37    95 x 95        65    151 x 151
10    41 x 41        38    97 x 97        66    153 x 153
11    43 x 43        39    99 x 99        67    155 x 155
12    45 x 45        40    101 x 101      68    157 x 157
13    47 x 47        41    103 x 103      69    159 x 159
14    49 x 49        42    105 x 105      70    161 x 161
15    51 x 51        43    107 x 107      71    163 x 163
16    53 x 53        44    109 x 109      72    165 x 165
17    55 x 55        45    111 x 111      73    167 x 167
18    57 x 57        46    113 x 113      74    169 x 169
19    59 x 59        47    115 x 115      75    171 x 171
20    61 x 61        48    117 x 117      76    173 x 173
21    63 x 63        49    119 x 119      77    175 x 175
22    65 x 65        50    121 x 121      78    177 x 177
23    67 x 67        51    123 x 123      79    179 x 179
24    69 x 69        52    125 x 125      80    181 x 181
25    71 x 71        53    127 x 127      81    183 x 183
26    73 x 73        54    129 x 129      82    185 x 185
27    75 x 75        55    131 x 131      83    187 x 187
28    77 x 77        56    133 x 133      84    189 x 189

Table: {#tbl:hanxin_sizes tag=": Han Xin Sizes"}

The largest version (84) can encode 7827 digits, 4350 ASCII characters, up to
2175 Chinese characters, or 3261 bytes, making it the most capacious of all the
barcodes supported by Zint.

There are four levels of error correction capacity available for Han Xin Code
which can be set by using the `--secure` option (API `option_1`) to a value from
the following table.

Mode   Recovery Capacity
----   -----------------
1      Approx 8%
2      Approx 15%
3      Approx 23%
4      Approx 30%

Table: {#tbl:hanxin_eccs tag=": Han Xin Error Correction Modes"}

Non-ASCII data density may be maximized by using the `--fullmultibyte` switch
(API `option_3 = ZINT_FULL_MULTIBYTE`), but check that your barcode reader
supports this before using.

Han Xin has four different masks designed to minimize unwanted patterns. The
best mask to use is selected automatically by Zint but may be manually specified
by using the `--mask` switch with values 0-3, or in the API by setting
`option_3 = (N + 1) << 8` where N is 0-3. To use with `ZINT_FULL_MULTIBYTE` set
```c
option_3 = ZINT_FULL_MULTIBYTE | (N + 1) << 8
```

### 6.6.14 Ultracode

![`zint -b ULTRA -d
"HEIMASÍÐA KENNARAHÁSKÓLA ÍSLANDS"`](images/ultra.svg){.ultra}

This symbology uses a grid of coloured elements to encode data. ECI and GS1
modes are supported. The amount of error correction can be set using the
`--secure` option (API `option_1`) to a value as shown in the following table.

Value  EC Level  Amount of symbol holding error correction data
-----  --------  ----------------------------------------------
1      EC0       0% - Error detection only
2      EC1       Approx 5%
3      EC2       Approx 9% - Default value
4      EC3       Approx 17%
5      EC4       Approx 25%
6      EC5       Approx 33%

Table: {#tbl:ultra_eccs tag=": Ultracode Error Correction Values"}

Zint does not currently implement data compression by default, but this can be
initiated through the API by setting

```c
symbol->option_3 = ULTRA_COMPRESSION;
```

With compression, up to 504 digits, 375 alphanumerics or 252 bytes can be
encoded.

Revision 2 of Ultracode (2023) may be specified using `--vers=2` (API
`option_2 = 2`).

* * *
WARNING: Revision 2 of Ultracode was only finalized December 2023 and Zint has
not yet been updated to support it. Do not use.

* * *

Ultracode supports Structured Append of up to 8 symbols and an optional numeric
ID (File Number), which can be set by using the `--structapp` option (see [4.17
Structured Append]) (API `structapp`). The ID ranges from 1 to 80088. If an ID
is not given, no ID is encoded.

\clearpage

## 6.7 Other Barcode-Like Markings

### 6.7.1 Facing Identification Mark (FIM)

![`zint -b FIM --compliantheight -d "C"`](images/fim.svg){.trk}

Used by the United States Postal Service (USPS), the FIM symbology is used to
assist automated mail processing. There are only 5 valid symbols which can be
generated using the characters A-E as shown in the table below.

Code Letter   Usage
-----------   --------------------------------------------------------------
A             Used for courtesy reply mail and metered reply mail with a
              pre-printed POSTNET symbol.
B             Used for business reply mail without a pre-printed zip code.
C             Used for business reply mail with a pre-printed zip code.
D             Used for Information Based Indicia (IBI) postage.
E             Used for customized mail with a USPS Intelligent Mail barcode.

Table: {#tbl:fim_characters tag=": Valid FIM Characters"}

### 6.7.2 Flattermarken

![`zint -b FLAT -d "1304056"`](images/flat.svg){.lin}

Used for the recognition of page sequences in print-shops, the Flattermarken is
not a true barcode symbol and requires precise knowledge of the position of the
mark on the page. The Flattermarken system can encode numeric data up to a
maximum of 128 digits and does not include a check digit.


# 7. Legal and Version Information

## 7.1 License

Zint, libzint and Zint Barcode Studio are Copyright © 2024 Robin Stuart. All
historical versions are distributed under the GNU General Public License version
3 or later. Versions 2.5 and later are released under a dual license: the
encoding library is released under the BSD (3 clause) license whereas the GUI,
Zint Barcode Studio, and the CLI are released under the GNU General Public
License version 3 or later.

Telepen is a trademark of SB Electronic Systems Ltd.

QR Code is a registered trademark of Denso Wave Incorporated.

Mailmark is a registered trademark of Royal Mail Group Ltd.

Microsoft, Windows and the Windows logo are either registered trademarks or
trademarks of Microsoft Corporation in the United States and/or other countries.

Linux is the registered trademark of Linus Torvalds in the U.S. and other
countries.

Mac and macOS are trademarks of Apple Inc., registered in the U.S. and other
countries.

The Zint logo is derived from "SF Planetary Orbiter" font by ShyFoundary.

Zint.org.uk website design and hosting provided by Robert Elliott.

## 7.2 Patent Issues

All of the code in Zint is developed using information in the public domain,
usually freely available on the Internet. Some of the techniques used may be
subject to patents and other intellectual property legislation. It is my belief
that any patents involved in the technology underlying symbologies utilised by
Zint are 'unadopted', that is the holder does not object to their methods being
used.

Any methods patented or owned by third parties or trademarks or registered
trademarks used within Zint or in this document are and remain the property of
their respective owners and do not indicate endorsement or affiliation with
those owners, companies or organisations.

## 7.3 Version Information

The current stable version of Zint is 2.13.0, released on 18th December 2023.

See `"ChangeLog"` in the project root directory for information on all releases.

## 7.4 Sources of Information

Below is a list of some of the sources used in rough chronological order:

- Nick Johnson's Barcode Specifications
- Bar Code 1 Specification Source Page
- SB Electronic Systems Telepen website
- Pharmacode specifications from Laetus
- Morovia RM4SCC specification
- Australia Post's 'A Guide to Printing the 4-State Barcode' and bcsample source
   code
- Plessey algorithm from GNU-Barcode v0.98 by Leonid A. Broukhis
- GS1 General Specifications v 8.0 Issue 2
- PNG: The Definitive Guide and wpng source code by Greg Reolofs
- PDF417 specification and pdf417 source code by Grand Zebu
- Barcode Reference, TBarCode/X User Documentation and TBarCode/X demonstration
   program from Tec-It
- IEC16022 source code by Stefan Schmidt et al
- United States Postal Service Specification USPS-B-3200
- Adobe Systems Incorporated Encapsulated PostScript File Format Specification
- BSI Online Library
- Libdmtx Data Matrix ECC200 decoding library

## 7.5 Standards Compliance

Zint was developed to provide compliance with the following British and
international standards:

### 7.5.1 Symbology Standards

- ISO/IEC 24778:2008 Information technology - Automatic identification and data
     capture techniques - Aztec Code bar code symbology specification
- SEMI T1-95 Specification for Back Surface Bar Code Marking of Silicon Wafers
      (BC412) (1996)
- ANSI/AIM BC12-1998 - Uniform Symbology Specification Channel Code
- BS EN 798:1996 Bar coding - Symbology specifications - 'Codabar'
- AIM Europe ISS-X-24 - Uniform Symbology Specification Codablock-F (1995)
- ISO/IEC 15417:2007 Information technology - Automatic identification and data
     capture techniques - Code 128 bar code symbology specification
- BS EN 12323:2005 AIDC technologies - Symbology specifications - Code 16K
- ISO/IEC 16388:2007 Information technology - Automatic identification and data
     capture techniques - Code 39 bar code symbology specification
- ANSI/AIM BC6-2000 - Uniform Symbology Specification Code 49
- ANSI/AIM BC5-1995 - Uniform Symbology Specification Code 93
- AIM Uniform Symbology Specification Code One (1994)
- ISO/IEC 16022:2006 Information technology - Automatic identification and data
     capture techniques - Data Matrix ECC200 bar code symbology specification
- ISO/IEC 21471:2020 Information technology - Automatic identification and data
     capture techniques - Extended rectangular data matrix (DMRE) bar code
     symbology specification
- AIM TSC1705001 (v 4.0 Draft 0.15) - Information technology - Automatic
     identification and data capture techniques - Bar code symbology
     specification - DotCode (Revised 28th May 2019)
- ISO/IEC 15420:2009 Information technology - Automatic identification and data
     capture techniques - EAN/UPC bar code symbology specification
- AIMD014 (v 1.63) - Information technology, Automatic identification and data
     capture techniques - Bar code symbology specification - Grid Matrix
     (Released 9th Dec 2008)
- ISO/IEC 24723:2010 Information technology - Automatic identification and data
     capture techniques - GS1 Composite bar code symbology specification
- ISO/IEC 24724:2011 Information technology - Automatic identification and data
     capture techniques - GS1 DataBar bar code symbology specification
- ISO/IEC 20830:2021 Information technology - Automatic identification and data
     capture techniques - Han Xin Code bar code symbology specification
- ISO/IEC 16390:2007 Information technology - Automatic identification and data
     capture techniques - Interleaved 2 of 5 bar code symbology specification
- ISO/IEC 16023:2000 Information technology - International symbology
     specification - MaxiCode
- ISO/IEC 24728:2006 Information technology - Automatic identification and data
     capture techniques - MicroPDF417 bar code symbology specification
- ISO/IEC 15438:2015 Information technology - Automatic identification and data
     capture techniques - PDF417 bar code symbology specification
- ISO/IEC 18004:2015 Information technology - Automatic identification and data
     capture techniques - QR Code bar code symbology specification
- ISO/IEC 23941:2022 Information technology - Automatic identification and data
     capture techniques - Rectangular Micro QR Code (rMQR) bar code symbology
     specification
- AIMD/TSC15032-43 (v 0.99c) - International Technical Specification - Ultracode
     Symbology (Draft) (Released 4th Nov 2015)

A number of other specification documents have also been referenced, such as
MIL-STD-1189 Rev. B (1989) (LOGMARS), USPS DMM 300 2006 (2011) (POSTNET, PLANET,
FIM) and USPS-B-3200 (2015) (IMAIL). Those not named include postal and delivery
company references in particular.

### 7.5.2 General Standards

- AIM ITS/04-001 International Technical Standard - Extended Channel
     Interpretations Part 1: Identification Schemes and Protocol (Released 24th
     May 2004)
- AIM ITS/04-023 International Technical Standard - Extended Channel
     Interpretations Part 3: Register (Version 2, February 2022)
- GS1 General Specifications Release 24.0 (Jan 2024)
- ANSI/HIBC 2.6-2016 - The Health Industry Bar Code (HIBC) Supplier Labeling
     Standard


# Annex A. Character Encoding

This section is intended as a quick reference to the character sets used by
Zint. All symbologies use standard ASCII input as shown in section A.1, but some
support extended characters as shown in the subsequent section [A.2 Latin
Alphabet No. 1 (ISO/IEC 8859-1)].

## A.1 ASCII Standard

The ubiquitous ASCII standard is well known to most computer users. It's
reproduced here for reference.

Hex   0       1       2       3       4       5       6       7
---   ------  ------  ------  ------  ------  ------  ------  ------
0     `NUL`   `DLE`   `SPACE` `0`     `@`     `P`     `` ` `` `p`
1     `SOH`   `DC1`   `!`     `1`     `A`     `Q`     `a`     `q`
2     `STX`   `DC2`   `"`     `2`     `B`     `R`     `b`     `r`
3     `ETX`   `DC3`   `#`     `3`     `C`     `S`     `c`     `s`
4     `EOT`   `DC4`   `$`     `4`     `D`     `T`     `d`     `t`
5     `ENQ`   `NAK`   `%`     `5`     `E`     `U`     `e`     `u`
6     `ACK`   `SYN`   `&`     `6`     `F`     `V`     `f`     `v`
7     `BEL`   `ETB`   `'`     `7`     `G`     `W`     `g`     `w`
8     `BS`    `CAN`   `(`     `8`     `H`     `X`     `h`     `x`
9     `TAB`   `EM`    `)`     `9`     `I`     `Y`     `i`     `y`
A     `LF`    `SUB`   `*`     `:`     `J`     `Z`     `j`     `z`
B     `VT`    `ESC`   `+`     `;`     `K`     `[`     `k`     `{`
C     `FF`    `FS`    `,`     `<`     `L`     `\`     `l`     `|`
D     `CR`    `GS`    `-`     `=`     `M`     `]`     `m`     `}`
E     `SO`    `RS`    `.`     `>`     `N`     `^`     `n`     `~`
F     `SI`    `US`    `/`     `?`     `O`     `_`     `o`     `DEL`

Table: {#tbl:ascii tag=": ASCII"}

## A.2 Latin Alphabet No. 1 (ISO/IEC 8859-1)

ISO/IEC 8859-1 defines additional characters common in western European
languages like French, German, Italian and Spanish. This extension is the
default encoding of many barcodes (see Table @tbl:default_character_sets) when a
codepoint above hex 9F is encoded. Note that codepoints hex 80 to 9F are not
defined.

Hex   8       9       A       B       C       D       E       F
---   ------  ------  ------  ------  ------  ------  ------  ------
0                     `NBSP`  `°`     `À`     `Ð`     `à`     `ð`
1                     `¡`     `±`     `Á`     `Ñ`     `á`     `ñ`
2                     `¢`     `²`     `Â`     `Ò`     `â`     `ò`
3                     `£`     `³`     `Ã`     `Ó`     `ã`     `ó`
4                     `¤`     `´`     `Ä`     `Ô`     `ä`     `ô`
5                     `¥`     `μ`     `Å`     `Õ`     `å`     `õ`
6                     `¦`     `¶`     `Æ`     `Ö`     `æ`     `ö`
7                     `§`     `·`     `Ç`     `×`     `ç`     `÷`
8                     `¨`     `¸`     `È`     `Ø`     `è`     `ø`
9                     `©`     `¹`     `É`     `Ù`     `é`     `ù`
A                     `ª`     `º`     `Ê`     `Ú`     `ê`     `ú`
B                     `«`     `»`     `Ë`     `Û`     `ë`     `û`
C                     `¬`     `¼`     `Ì`     `Ü`     `ì`     `ü`
D                     `SHY`   `½`     `Í`     `Ý`     `í`     `ý`
E                     `®`     `¾`     `Î`     `Þ`     `î`     `þ`
F                     `¯`     `¿`     `Ï`     `ß`     `ï`     `ÿ`

Table: {#tbl:iso_iec_8869_1 tag=": ISO/IEC 8859-1"}


# Annex B. Qt Backend QZint

Used internally by Zint Barcode Studio to display the preview, the Qt Backend
`QZint` renders a barcode by drawing the vector representation (see [5.5
Buffering Symbols in Memory (vector)]) provided by the Zint library `libzint`.

The main class is `Zint::QZint`, which has getter/setter properties that
correspond to the `zint_symbol` structure (see [5.7 Setting Options]), and a
main method `render()` which takes a Qt `QPainter` to paint with, and a `QRectF`
rectangular area specifying where to paint into:

```c++
/* Encode and display barcode in `paintRect` using `painter`.
   Note: legacy argument `mode` is not used */
void render(QPainter& painter, const QRectF& paintRect,
            AspectRatioMode mode = IgnoreAspectRatio);
```

`render()` will emit one of two Qt signals - `encoded` on successful encoding
and drawing, or `errored` on failure. The client can connect and act
appropriately, for instance:

```c++
connect(qzint, SIGNAL(encoded()), SLOT(on_encoded()));
connect(qzint, SIGNAL(errored()), SLOT(on_errored()));
```

where `qzint` is an instance of `Zint::QZint` and `on_encoded()` and
`on_error()` are Qt slot methods provided by the caller. On error, the error
value and message can be retrieved by the methods `getError()` and `lastError()`
respectively.

The other main method is `save_to_file()`:

```c++
/* Encode and print barcode to file `filename`.
   Only sets `getError()` on error, not on warning */
bool save_to_file(const QString& filename); // `ZBarcode_Print()`
```

which takes a `filename` to output to. It too will emit an `errored` signal on
failure, returning `false` (but nothing on success, which just returns `true`).
Note that rotation is achieved through the setter method `setRotateAngleValue()`
(as opposed to the `rotate_angle` argument used by `ZBarcode_Print()`).

Various other methods are available, for instance methods for testing symbology
capabilities, and utility methods such as `defaultXdim()` and `getAsCLI()`.

For full details, see `"backend_qt/qzint.h"`.


# Annex C. Tcl Backend Binding

A Tcl binding is available in the `"backend_tcl`" sub-directory. To make on
Unix:

```bash
cd backend_tcl
autoconf
./configure
make
sudo make install
```

For Windows, a Visual Studio 6.0 project file is available at
`"backend_tcl\zint_tcl.dsp"`. This can also be opened (and converted) by more
modern Visual Studio versions, though some fixing up of the project
configuration will likely be required.

Once built and installed, invoke the Tcl/Tk CLI `"wish"`:

```bash
wish
```

and ignoring the Tk window click back to the command prompt `"%"` and type:

```bash
require package zint
zint help
```

which will show the usage message, with options very similiar to the Zint CLI.
(One notable difference is that boolean options such as `-bold` take a `1` or
`0` as an argument.)

A demonstration Tcl/Tk program which is also useful in itself is available at
`"backend_tcl/demo/demo.tcl"`. To run type:

```bash
wish demo/demo.tcl
```
which will display the following window.

![Tcl/Tk demonstration program window](images/tcl_demo.png){.pop}

You can select the symbology, enter the data to encode, and set options (which
are the same as those given in the usage message). A raster preview of the
configured barcode is displayed once the `"Generate"` button is pressed.


# Annex D. Man Page ZINT(1)