Module:Road data/strings/CAN/doc

This module is used to store type data for use with several road templates, including jct and routelist row. Background information and a summary of the syntax are listed below.

Hierarchy and fields
At its most basic level, this module is a nested table of strings. At the top is the root table, named for the abbreviation of a country, state, or province. This table stores the type data for a particular place, which is named in the comment in the first line, and is returned at the end of the module. The table is composed of further tables, one per type. The basic syntax for a type table is:

The three main fields in a type table are,  , and. Generally,  determines the shield that is displayed, if any;   specifies the target of the link generated by a including module, if any; and   determines the displayed abbreviation. Currently, these are the three types used by all countries. By convention, they are always specified, using an empty string if there is no value.

Two additional fields are used in the United States. stores the name of the banner file, while  stores a code representing the width of the shield. Unlike the three main fields, these are generally omitted entirely when unused.

Once a type is defined, it can be referred to later in the code. A common idiom used for bannered types is defining the shield to be the shield field of the base type (from Module:Road data/strings/USA/MO):

Another common idiom is defining a type alias by setting an entry in the root table equal to a previously defined type (from Module:Road data/strings/HKG):

Parser arguments
When the parser function of Module:Road data/parser is called, it is passed up to three parameters. The second one is the field to parse, and the last one is a rarely-used option designed for multiple-shield types. The first and most important parameter is a table of arguments collected by the calling module, which generally includes the state, country, or both; the type and number of the route; and a few miscellaneous arguments. This table of arguments forms the basis of the parser's format string syntax.

The table accessible by the strings includes the following entries by default: The above entries are primarily used to find the string module itself, so they should not be a concern for module writers.
 * : The state or province the route is located in.
 * : The country the route is located in. If the country is not passed by the calling module, the parser will attempt to include it.


 * : The type of the route. This determines the entry of the root table that is used by the parser.
 * : The route "number". This is easily the most important argument for module writers.

The following entries are used less often:
 * : The county the route is located in. This is usually used for county routes in the United States.
 * : This entry is similar in function and utility to the  entry.
 * : A tag used to disambiguate the link target. This is mostly used for bannered routes in the United States.
 * : This rare entry is used exclusively for West Virginia county routes.

Parser hooks, which will be described later, can add entries to this table that may be used by strings.

Basic string syntax
The most basic value that can be used for most type table fields is a specially formatted string, which will be referred to in this documentation as a format string. This is the string that will ultimately be parsed and returned by the parser. A format string is an ordinary string object. The power of these strings comes in the form of two special instructions that are recognized by the parser.

The first is anything in  form. The parser will replace such a statement with the value of the  entry in the arguments table described earlier. This is what allows the route number to be spliced into a shield or link name.

The second special string is in the form of. This functions as a rudimentary if-then-else statement. The parser tests the value of  to see if it is equal to the value specified in. may be empty, in which case the parser tests the existence of the  argument. If the result of the test is true, the statement is replaced with the value of the  block. Otherwise, it is replaced with the value of the  block.

The two statements may be combined. The parser will parse the if-then-else statement first, and then perform the argument inclusion. This combination is commonly used with bannered routes in the United States, where the  argument is tested and the link disambiguation is adjusted accordingly, as follows (from Module:Road data/strings/USA/AL):

When parsing the  field, the parser first checks to see if the   argument was provided. If so, it replaces the statement with. If not, the statement is replaced with the empty string placed in the  block. Then, the parser replaces  with the route number and, if the   argument was provided,   with the value of that argument.

Switching
Some logic is too complicated to represent with only format strings. This framework provides several methods to express complex data. All of these involve storing a nested table as the value of a field.

The most straightforward functionality provided by nested tables is switching. In its most basic form, the table consists of a series of key-value pairs, with the keys being route numbers and the values being the format strings used by those routes. Usually, the format string returned does not need parsing, but the option is there. A  entry should be provided to handle any route numbers not explicitly stated. The following is a representative example of route-based switching (from Module:Road data/strings/USA/AR):

In this example, Highways 917 and 980 have non-standard shield names, which are explicitly provided. Other route numbers use the default format.

Switching on other arguments is also allowed. The name of the argument to be used for switching is stated in the  field of the table. Nesting switches on different arguments is also allowed. A good example that uses both forms of switching can be found in Ontario:

In this example, which is a shield table that is reused by several types in Ontario, the  argument is used for the primary switch. If the route is in Simcoe County, a second switch is performed, this time on the route number.

Existence testing
Another use for tables is existence testing. If a table has the  field set to , the parser will perform existence testing on the result of parsing the   field. If the test fails, the result of parsing the  field is returned. Existence testing may be chained by using a second ifexists table as the value of the first table's  field, and so on. Here's an example of nested existence testing (from Module:Road data/strings/GBR):

Hooks
Due to technical limitations, these string modules cannot contain functions. Rather than force functionality into the string framework, the parser can call functions in a separate hooks module. The functions in this module, Module:Road data/parser/hooks, are more-or-less fully functional functions. The exact functionalities of these hooks are beyond the scope of this documentation. Descriptions of these hooks may be found on their documentation page.

Generally speaking, a hook is called by setting the  field in a table as equal to the name of a hook. Hooks receive two arguments, both tables:, which is the table in the definition; and  , which is simply the table of arguments normally passed to the parser. The hook returns a string, which is then parsed as usual. A powerful feature of hooks is that they can add arbitrary values to the argument table, which may be referenced in the string returned by the hook. Generally, the format string returned by the hook is specified in some form by the  field of the table, though there are exceptions. Here is an example of a hook (from Module:Road data/strings/MEX):

In this example, the parser will process the link by calling the  hook. In short, this hook takes the argument referenced in, passes it through the mask module specified in  , and stores it in the field in the arguments noted in. The hook returns the string given in, which has access to the   argument added by the hook.

Other functionality
Functionality exists to display multiple shields for one route, which is used to display tolled and free shields for routes where they differ. This is done by supplying a table with two values, which are listed without indices. The parser is called twice by the calling module, and it returns one shield per call. An example may be found in Texas:

Structure
Each country has its own module. In the United States and Canada, each state/territory/province also has its own module. Each module begins with a comment stating the name of the country or state, followed by the root table declaration, as follows (from Module:Road data/strings/USA/AS):

The root table is named based on the established abbreviation for the country or state, which is the same as the abbreviation used in the module title. This table stores the various types used in that particular place. Most of the remaining code in the module defines these various types. The module ends by returning the root table:

Aliasing
There are two ways to define a type as an alias. If the type is defined within the module, simply set the new type as equal to the type being aliased, as shown above (from Module:Road data/strings/HKG):

If the type is defined in a separate module, such as a state highway type being used in another state's module, a special syntax may be used to refer to that module (from Module:Road data/strings/USA/NJ):

This code sets the  type as a link to the   type in Module:Road data/strings/USA/NY. The parser will import that module and process the type as if the original module had declared it itself. The alias declaration may not add or override any data in the type table it points to.

Style
There are a few style guidelines that should be followed:
 * 1) Align table fields using spaces. All tables should be aligned so that fields line up with each other, as shown in the above examples.
 * 2) Do not place a table's closing brace on a separate line.
 * 3) Each table field should be on its own line.
 * 4) Add spaces to either side of an assignment operator (equals sign).
 * 5) Leave a blank line between types. Type aliases should be set off from their base type by a blank line, but no blank lines should be placed between the aliases themselves.