Compilation identifier

See also: CompilationIdentifier JavaDoc

A compilation identifier consists of dash ('-') separated name compontents. Its purpose is to uniquely identify a given compilation pass or operation in the current build execution. The compilation identifiers are human-readable.

When a compilation task is run, it is run with an assigned compilation identifier. This identifier will uniquely identify the output location of the compilation, and is associated with a build task in the build system. This ensures that only a single compilation task will produce its outputs for the given location, and the tasks won't overwrite each others outputs.

Let's look at an example, with Java compilation:

saker.java.compile(
	SourceDirectories: src/main,
	Identifier: main
)
saker.java.compile(
	SourceDirectories: src/lib,
	Identifier: lib
)

In the above, we assigned different compilation identifiers for the different compilation tasks. They will run concurrently, and produce their outputs for the associated location. (See Compilation output.)

The compilation identifiers also ensure that no differently configured compilation tasks will run with the same identifier:

saker.java.compile(
	SourceDirectories: src/main,
	Identifier: main
)
saker.java.compile(
	SourceDirectories: src/lib,
	Identifier: main
)

The above will cause a build error thrown by the saker.build system, as different compilation tasks are started for the same compilation identifier.

Note that the above behaviours may be dependent on the actual tasks that use the compilation identifiers.

Format

Compilation identifiers consist of name parts separated by dashes. The names may consist of characters a-z, A-Z, 0-9, _, ., (), [], @. When a compilation identifier is constructed, it is normalized to lowercase letters. The name parts are deduplicated, and not ordered. Empty name components, preceeding and trailing dashes are omitted.

Some compilation identifier examples:

main
test
lib-debug
lib-release
lib-debug-x64
lib-debug-x86

Note that all of the following compilation identifiers are considered to be the same:

lib-debug-x64
x64-lib-debug
-x64---lib--debug--
LIB-X64-dEbUG

Operations

Compilation identifiers have the following operations defined on them. These operations are usually performed by the build tasks, and not in the build script.

Concatenation/union.

The name components of the operand compilation identifiers are merged, and deduplicated. A single compilation identifier is the result that contains all of the operant name parts.

E.g. lib-x64 concatenated with lib-debug results in lib-x64-debug.

Subset

Checks whether all of the name components of an identifier are present in another. This operation is often performed then compiler option merging is being done.

Some examples:

• main is a subset of main
• main is a subset of main-debug
• lib-x64 is a subset of lib-x64-debug (and other variants like lib-debug-x64, order doesn't matter)
• lib-release is not a subset of lib-x64-debug