Modules

Modules and the Import System

KCL code is organized in modules. For code in one module to access the code defined in another module, a process called importing must be used.

Importing is undertaken at compile-time in KCL. The advantage is to have static checking enabled.

A regular KCL module is a file on the file system. It is required to have a .k suffix.

Packages

To help manage modules and provide a naming hierarchy, KCL has the concept of packages. In KCL, a package maps to exactly a file system directory, and a regular module maps to a file.

Files directly under a package are considered parts of the package, instead of individual regular modules.

Packages can have sub-packages.

Packages are special modules:

• All packages in KCL are modules.
• A single-file module can never be a package.

All modules have a name.

Sub package names are separated from their parent package name by dots.

To summary, a regular KCL module is a .k file, and a package is a directory on the file system. All .k files directly under the directory are included in the package, other files are ignored. If the directory has subdirectories, they become sub-packages as long as there are .k files underneath.

Intra-Package Name Space Sharing

Inside a package, all .k files are considered parts of the package, instead of regular modules. Code in these files share a single name space and can access names defined in other files, without explicitly granted.

Package Initialization

A package can have the initialization code. The code must exist in only one of the .k files under this package. The interpreter guarantees that the initialization code is executed after all definitions.

Searching

The searching begins when an import statement is used to import a module.

Module Cache

In KCL, only standard system modules are cached. When a cached module is imported, the cached version is used. In other words, KCL runtime would not create another copy of the standard system module in memory.

However, other modules are uncached. Importing a module multiple time would create multiple instances of the module.

Module Names

An import statement specifies the name of the module to import. The syntax is:

import <module_name>

The rule to search with the module name is very simple:

• Step 1: Searches the module name from the standard system modules, then plugins modules.
  • See standard system modules and plugins modules for more details. If matched, the module is imported. Otherwise, continue to Step 2.
• Step 2. Whether a module name starts with a . is checked. If yes, the name is a so-called relative pathname, and we go to Step 5. Otherwise, continue to Step 3.
• Step 3: If the module name does not start with any ., then the compiler searches the nearest root path directory from this directory to the parent, and find the module according to the name just from the root path. If no root path is found, find the module according to the name from the folder the .k file including this import statement exists.
  • root path: the directory contains a kcl.mod file. If matched, the module is imported. Otherwise, continue to Step 4.
• Step 4: Then the compiler checks if the name is the name of any library module that requires explicit loading. If matched, the library module is imported. Otherwise, continue to Step 6.
• Step 5. For relative importing, find the module according to the name from the folder the .k file including this import statement exists. Interpret leading dots using the following rule:
• One dot: Ignore.
• Tow or more dots: Suppose there are n leading dots, then the searching starts at n - 1 levels above this folder. If matched, the module is imported. Otherwise, continue to Step 6.
• Step 6. Module not found, report an error.

Do case-sensitive search when the operating system allows. If case-sensitive search is not allowed, search directories before regular files.

In KCL, the from <> import <> is unsupported, and relative import is performed with the import <> syntax.

Uniqueness of Module

Each module has a unique location path in its scope, so that a module or package could be located with a unique location path, such as a.b.c.

Searching by location path should be supported by the kcl compiler, which needs to provide corresponding searching features through the command line and api form.

Standard System Packages

KCL supports a few standard system modules. The following is the full list of these standard system modules:

Package	Member
datetime	today
	now
	ticks
	date
math	ceil
	exp
	expm1
	factorial
	floor
	gcd
	isfinite
	isinf
	isnan
	log
	log1p
	log2
	log10
	modf
	pow
	sqrt
regex	replace
	match
	compile
	findall
	search
	split
units	n
	u
	m
	k
	K
	M
	G
	T
	P
	Ki
	Mi
	Gi
	Ti
	Pi
	to_n
	to_u
	to_m
	to_K
	to_M
	to_G
	to_T
	to_P
	to_Ki
	to_Mi
	to_Gi
	to_Ti
	to_Pi
json	encode
	decode
	dump_to_file
yaml	encode
	decode
	dump_to_file
net	split_host_port
	join_host_port
	fqdn
	parse_IP
	to_IP4
	to_IP16
	IP_string
	is_IPv4
	is_IP
	is_loopback_IP
	is_multicast_IP
	is_interface_local_multicast_IP
	is_link_local_multicast_IP
	is_link_local_unicast_IP
	is_global_unicast_IP
	is_unspecified_IP
base64	encode
	decode
crypto	md5
	sha1
	sha224
	sha256
	sha384
	sha512

• datetime
  • ticks() -> float Return the current time in seconds since the Epoch. Fractions of a second may be present if the system clock provides them.
  • date() -> str return the %Y-%m-%d %H:%M:%S format date.
  • now() -> str return the local time. e.g. 'Sat Jun 06 16:26:11 1998'
  • today() -> str return the %Y-%m-%d %H:%M:%S.%{ticks} format date.
• math
  • ceil(x) -> int Return the ceiling of x as an Integral. This is the smallest integer >= x.
  • factorial(x) -> int Return x!. Raise a error if x is negative or non-integral.
  • floor(x) -> int Return the floor of x as an Integral. This is the largest integer <= x.
  • gcd(a: int, b: int) -> int Return the greatest common divisor of x and y
  • isfinite(x) -> bool Return True if x is neither an infinity nor a NaN, and False otherwise.
  • isinf(x) -> bool Return True if x is a positive or negative infinity, and False otherwise.
  • isnan(x) -> bool Return True if x is a NaN (not a number), and False otherwise.
  • modf(x) -> Listfloat, float] Return the fractional and integer parts of x. Both results carry the sign of x and are floats.
  • exp(x) -> float Return e raised to the power of x.
  • expm1(x) -> float Return exp(x)-1. This function avoids the loss of precision involved in the direct evaluation of exp(x)-1 for small x.
  • log(x, base=2.71828182845904523536028747135266250) -> float Return the logarithm of x to the base e.
  • log1p(x) -> float Return the natural logarithm of 1+x (base e). The result is computed in a way which is accurate for x near zero.
  • log2(x) -> float Return the base 2 logarithm of x.
  • log10(x) -> float Return the base 10 logarithm of x.
  • pow(x, y) -> float Return x**y (x to the power of y).
  • sqrt(x) -> float Return the square root of x.
• regex
  • replace(string: str, pattern: str, replace: str, count=0) -> str Return the string obtained by replacing the leftmost non-overlapping occurrences of the pattern in string by the replacement.
  • match(string: str, pattern: str) -> bool Try to apply the pattern at the start of the string, returning a bool value True if any match was found, or False if no match was found.
  • compile(pattern: str) -> bool Compile a regular expression pattern, returning a bool value denoting whether the pattern is valid.
  • findall(string: str, pattern: str) -> List[str] Return a list of all non-overlapping matches in the string.
  • search(string: str, pattern: str) -> bool Scan through string looking for a match to the pattern, returning a bool value True if any match was found, or False if no match was found.
  • split(string: str, pattern: str, maxsplit=0) -> List[str] Scan through string looking for a match to the pattern, returning a Match object, or None if no match was found.
• units
  • Unit constants
    • Fixed point: n, u, m, k, K, G, T and P.
    • Power of 2: Ki, Mi, Gi, Ti and Pi.
  • Functions
    • to_n(num: int) -> str Int literal to string with n suffix
    • to_u(num: int) -> str Int literal to string with u suffix
    • to_m(num: int) -> str Int literal to string with m suffix
    • to_K(num: int) -> str Int literal to string with K suffix
    • to_M(num: int) -> str Int literal to string with M suffix
    • to_G(num: int) -> str Int literal to string with G suffix
    • to_T(num: int) -> str Int literal to string with T suffix
    • to_P(num: int) -> str Int literal to string with P suffix
    • to_Ki(num: int) -> str Int literal to string with Ki suffix
    • to_Mi(num: int) -> str Int literal to string with Mi suffix
    • to_Gi(num: int) -> str Int literal to string with Gi suffix
    • to_Ti(num: int) -> str Int literal to string with Ti suffix
    • to_Pi(num: int) -> str Int literal to string with Pi suffix
• json
  • encode(data: any, sort_keys: bool = False, indent: int = None, ignore_private: bool = False, ignore_none: bool = False) -> str Serialize a KCL object data to a JSON formatted str.
  • decode(value: str) -> any Deserialize value (a string instance containing a JSON document) to a KCL object.
  • dump_to_file(data: any, filename: str, ignore_private: bool = False, ignore_none: bool = False) -> None Serialize a KCL object data to a JSON formatted str and write it into the file filename.
• yaml
  • encode(data: any, sort_keys: bool = False, ignore_private: bool = False, ignore_none: bool = False) -> str Serialize a KCL object data to a YAML formatted str.
  • decode(value: str) -> any Deserialize value (a string instance containing a YAML document) to a KCL object.
  • dump_to_file(data: any, filename: str, ignore_private: bool = False, ignore_none: bool = False) -> None Serialize a KCL object data to a YAML formatted str and write it into the file filename.
• net
  • split_host_port(ip_end_point: str) -> List[str] Split the 'host' and 'port' from the ip end point.
  • join_host_port(host, port) -> str Merge the 'host' and 'port'.
  • fqdn(name: str = '') -> str Return Fully Qualified Domain Name (FQDN).
  • parse_IP(ip) -> str Parse 'ip' to a real IP address
  • to_IP4(ip) -> str Get the IP4 form of 'ip'.
  • to_IP16(ip) -> int Get the IP16 form of 'ip'.
  • IP_string(ip: str | int) -> str Get the IP string.
  • is_IPv4(ip: str) -> bool Whether 'ip' is a IPv4 one.
  • is_IP(ip: str) -> bool Whether ip is a valid ip address.
  • is_loopback_IP(ip: str) -> bool Whether 'ip' is a loopback one.
  • is_multicast_IP(ip: str) -> bool Whether 'ip' is a multicast one.
  • is_interface_local_multicast_IP(ip: str) -> bool Whether 'ip' is a interface, local and multicast one.
  • is_link_local_multicast_IP(ip: str) -> bool Whether 'ip' is a link local and multicast one.
  • is_link_local_unicast_IP(ip: str) -> bool Whether 'ip' is a link local and unicast one.
  • is_global_unicast_IP(ip: str) -> bool Whether 'ip' is a global and unicast one.
  • is_unspecified_IP(ip: str) -> bool Whether 'ip' is a unspecified one.
• base64
  • encode(value: str, encoding: str = "utf-8") -> str Encode the string value using the codec registered for encoding.
  • decode(value: str, encoding: str = "utf-8") -> str Decode the string value using the codec registered for encoding.
• crypto
  • md5(value: str, encoding: str = "utf-8") -> str Encrypt the string value using MD5 and the codec registered for encoding.
  • sha1(value: str, encoding: str = "utf-8") -> str Encrypt the string value using SHA1 and the codec registered for encoding.
  • sha224(value: str, encoding: str = "utf-8") -> str Encrypt the string value using SHA224 and the codec registered for encoding.
  • sha256(value: str, encoding: str = "utf-8") -> str Encrypt the string value using SHA256 and the codec registered for encoding.
  • sha384(value: str, encoding: str = "utf-8") -> str Encrypt the string value using SHA384 and the codec registered for encoding.
  • sha512(value: str, encoding: str = "utf-8") -> str Encrypt the string value using SHA512 and the codec registered for encoding.

The Built-in System Package

KCL provides a list of built-in system modules, which are loaded automatically and can be directly used without providing any module name. For example, print is a widely used built-in module.

The following is the full list of these built-in system modules:

• print()
  • The print function.
• multiplyof(a, b)
  • Check if the modular result of a and b is 0
• isunique(inval)
  • Check if a list has duplicated elements
• len(inval) Return the length of a value
• abs(x) Return the absolute value of the argument.
• all(iterable) Return True if bool(x) is True for all values x in the iterable. If the iterable is empty, return True.
• any(iterable) Return True if bool(x) is True for any x in the iterable. If the iterable is empty, return False.
• bin(number) Return the binary representation of an integer.
• hex(number) Return the hexadecimal representation of an integer.
• oct(number) Return the octal representation of an integer.
• ord(c) -> int Return the Unicode code point for a one-character string.
• sorted(iterable) Return a new list containing all items from the iterable in ascending order. A custom key function can be supplied to customize the sort order, and the reverse flag can be set to request the result in descending order.
• range(start, end, step=1) Return the range of a value with start, end and step parameter.
• min(iterable) With a single iterable argument, return its smallest item. The default keyword-only argument specifies an object to return if the provided iterable is empty. With two or more arguments, return the smallest argument.
• max(iterable) With a single iterable argument, return its biggest item. The default keyword-only argument specifies an object to return if the provided iterable is empty. With two or more arguments, return the largest argument.
• sum(iterable, start) Return the sum of a 'start' value (default: 0) plus an iterable of numbers. When the iterable is empty, return the start value. This function is intended specifically for use with numeric values and may reject non-numeric types.
• pow(x, y, z) Equivalent to x**y (with two arguments) or x**y % z (with three arguments). Some types, such as ints, are able to use a more efficient algorithm when invoked using the three argument form.
• round(number, ndigits) Round a number to a given precision in decimal digits. The return value is an integer if ndigits is omitted or None. Otherwise the return value has the same type as the number. ndigits may be negative.
• typeof(x: any, *, full_name: bool = False) -> str Return the type of the value 'x' at runtime. When the 'full_name' is 'True', return the full package type name such as pkg.schema.

Plugin Modules

KCL compiler needs to provide the ability to dynamically expand and load plugin modules without modifying the compiler itself. KCL compiler needs to support flexible pluggable module extension mechanism, so that KCL users can use more abundant built-in function capabilities to simplify writing.

KCL compiler needs to ensure the stability and safety of the expansion mechanism, without affecting the core of the compiler.

Searching extended plugin module is performed after the standard system module. The standard system module has a higher priority in naming. If it exists a standard or built-in system module with the same name, the extended plugin module will be ignored.

Importing and using the extended plugin module should be consistent with the standard or built-in system module.

Replacing Standard System Packages

Replacing standard system modules is not allowed.

Examples

We show more module features through an example.

Suppose we have the following directories and files:

    .
    ├── mod1.k
    ├── mod2.k
    ├── pkg1
    │   ├── def1.k
    │   ├── def2.k
    │   └── def3init.k
    └── pkg2
        ├── file2.k
        └── subpkg3
            └── file3.k

From the structure we can see that pkg1 and pkg2 are two packages, subpkg3 is a subpackage of pkg2, and mod1.k and mod2.k are regular modules.

Importing a Standard System Package

The following statement can import the standard system module math

import math

This is the only way to import a standard system module. After importing a standard system module, functions, variables and schemas defined in it can be used. For example, the following statement uses the log10 function defined in math

a = math.log10(100) # a is 2 after computation.

Importing a Regular Module

In mod1.k, we can import mod2 using one of the following syntaxes.

import mod2

import .mod2

The difference is that in the first syntax, the KCL compiler will first try to check if mod2 matches any of the standard system modules' name. Since it does not match any standard system module's name, the statement will check the directory where mod1.k resists in, like what the second statement does.

Suppose in mod2.k there is a definition of a variable::

a = 100

After importing mod2, we can access a in mod1.k using the following syntax

b = mod2.a

Importing a Package

In mod1.k, we can import pkg1 using one of the following syntaxes.

import pkg1

import .pkg1

The difference is that in the first syntax, the KCL compiler will first try to check if pkg1 matches any of the standard system modules' name. Since it does not match any standard system module's name, the statement will check the directory where mod1.k resists in, like what the second statement does.

We can use similar statements to import pkg2. Note that importing pkg2 will not import subpkg3.

The name of the package is the name of the imported module.

Suppose in file2.k that is inside pkg2 there is a definition to variable foo

foo = 100

This variable can be used in mod1.k after importing pkg2 like the following

bar = pkg2.foo

Importing a Subpackage

To import subpkg3 from mod1.k, one of the following statements can be used.

import pkg2.subpkg3

import .pkg2.subpkg3

The behaviors of these statements are identical.

The name of the subpackage is the name of the imported module.

Suppose in file3.k that is inside subpkg3 there is a definition to variable foo

foo = 100

This variable can be used in mod1.k after importing subpkg3 like the following

bar = subpkg3.foo

Relative Importing

Relative importing is useful when there is code trying to import modules that does not exist recursively inside the current directory.

For example, the following statements, if written in file3.k, can be used to import pkg2, pkg1 and mod2 respectively.

import ...pkg2 # Go two levels up then import pkg2
import ...pkg1 # Go two levels up then import pkg1
import ...mod2 # Go two levels up then import mod2

Importing from a Root Path

Suppose we have a kcl.mod file in the directory to mark it as a root path, then we have the following files:

    .
    |── kcl.mod
    ├── mod1.k
    ├── mod2.k
    ├── pkg1
    │   ├── def1.k
    │   ├── def2.k
    │   └── def3init.k
    └── pkg2
        ├── file2.k
        └── subpkg3
            └── file3.k

In pkg1 def1.k, we can import pkg2.subpkg3 file3 using the following syntaxes.

import pkg2.subpkg3.file3

Importing from the root path is very convenient when the code is trying to import modules from a directory needs to look up multiple directories above this directory. At also, it is helpful to organize a large number of files in a root directory.

Importing a Module Inside a Package

Note that subpkg3 is only implemented with one file file3.k. The file can be regarded as a regular module and imported directly.

In mod1.k, the importing statement would be::

import pkg2.subpkg3.file3

Different from importing subpkg3, now the name of the module is file3. We can access the variable foo defined in this module with the following statement

bar = file3.foo

Precedence of Importing

When an import statement specifies a package to import, the virtual machine first looks for a directory named according to the import statement in the file system.

If such a directory is not found, the virtual machine looks for a single file module.

For example, when the statement import a.b.c appears, the virtual machine first looks for the directory a/b/c from the directory of the current file. If a/b/c is not found, the virtual machine looks for a file named a/b/c.k. If the file is also absent, an error is reported.

Package Implemented with Multiple Files

Package pkg1 is implemented with multiple KCL files.

Multiple files can be used to define variables, schemas and functions, and they can access names defined in other files of this package.

For example, suppose def1.k defines a variable foo, def2.k defines bar, and def3init.k defines a variable baz, when pkg1 is imported by mod1.k, all these variable can be used

import pkg1
a = pkg1.foo + pkg1.bar + pkg1.baz

Inside a module, names defined in a file can be accessed in another file without further importing. For example, suppose bar in def2.k would invoke foo defined in def1.k, it can directly use foo like the following

bar = foo + 1