Miscellaneous Terminology

This page documents miscellaneous terminology used throughout bpt and CRS.

application

An application is a program that is intended to be run and distributed to users to perform some set of designated tasks.

In bpt, an application executable is created for each source file with an appropriate file stem.

See also

Building Applications with bpt

CMake

CMake is a popular cross-platform build system and project configuration tool for C and C++ projects.

See also

common dependencies

The common dependencies of a project are the dependencies that appear at the top-level of the project’s bpt.yaml file. (Refer: Project.dependencies)

These dependencies are added as direct dependencies of every library in the project, whether that is the default library or each library in the Project.libraries array.

The common test-dependencies are the similar but apply only to the libraries’ test dependencies. (Refer: Project.test-dependencies)

compiler
compile

Compiling is the process of transforming human-readable source code and emits a lower-level code intended to be executed. The compiler is a program that performs the compilation. GCC, Visual C++, and Clang are examples of compilers.

default library

The default library of a project is the library that bpt generates if the libraries (Project.libraries) property is omitted in bpt.yaml. It will have the same name as the project, and its library root will be the same as the project root.

Note

If the libraries property is specified then bpt will not generate a default library.

See also

dependency specifier

bpt allows a few syntaxes to specify a dependency. A specifier provides a name, a version range, and some set of library names to use from the external package.

See also

Dependency Specifiers

environment variables

Every operating system process has a set of environment variables, which is an array of key-value pairs that map a text string key to some text string value. These are commonly used to control the behavior of commands and subprocesses.

For example, the “PATH” environment variable controls how command names are mapped to executable files.

bpt uses some environment variables to control some behavior, such as BPT_LOG_LEVEL and BPT_NO_DEFAULT_REPO.

header-only library

A header-only library is a library that contains no exported compilable source files, and only contains header files.

JSON

JSON is the JavaScript Object Notation, a plaintext format for representing semi-structured data.

JSON values can be strings, numbers, boolean values, a null value, arrays of values, and “objects” which map a string to another JSON value.

Example

example.json
{
  "foo": 123.0,
  "bar": [true, false, null],
  "baz": "quux",
  "another": {
    "nested-object": [],
  }
}

JSON data cannot contain comments. The order of keys within a JSON object is not significant.

JSON5

A superset of JSON that permits comments, single-quote strings, multi-line strings, trailing commas, and bare identifier keys.

library

Within the context of software development, a library is a set of code that is designed to be used by other code to build applications or additional higher-level libraries.

A library contains a set of definitions of entities that can be used by other code. This facilitates code reuse.

A library is the smallest consumable unit of code. That is: You cannot “use” only a subset of a library when building a library or application.

See also

library root

The directory in which the source files of a single library reside. Contains the src/ and/or include/ directories for that library, each of which is a source root.

This path is specified using the libs[].path key of the library in bpt.yaml.

See also

linker
linking

Linking is the process of combining separate translation units (i.e. compiled source files code) into a program. A linker is a program that performs linking.

During linking, references to names across translation units are resolved. If a name is referenced but its definition is not found, the linker will most often fail to perform the linking.

package ID

A package ID is a string that identifies a package. It is composed of the package’s name, version, and the package revision number:

<name>@<version>~<revision>

package revision number

CRS allows packages within a repository to be updated without changing the version of the package itself. This is reserved for changes that only update the package metadata or fix critical issues that render a prior revision to be unusable. The revision number is always a single positive integer and begins at 1.

The package revision number is visible on the package ID as the number following the tilde ~ suffix.

public headers
private headers

Within a library, the public headers and private headers of are the header files that live in the public source root or the private source root, respectively.

The public headers are visible to the library’s users, but the private headers are only available to the library while compiling the library itself.

See also

Source Root Kinds

source root

A directory within a library that defines the structure of a source tree and and acts as a header search path. This directory contains source files.

Within a library root, the src/ and include/ directories are source roots.

See also

test

In bpt, “test” refers to a library-provided execuatble program that can be used to verify that the library implements the correct behavior.

Tests are compiled and linked automatically as part of any bpt build invocation.

See also

Writing Tests with bpt

test dependency

A test dependency is a dependency within a library that is only used for compiling and linking the library’s tests.

Unlike regular dependencies, test dependencies are not transitive.

Test dependencies are declared using the test-dependencies property in bpt.yaml.

tweak-headers

Special header files that are used to inject configuration options into dependency libraries. The library to be configured must be written to support tweak-headers.

Tweak headers are placed in the “tweaks directory”, which is controlled with the bpt build --tweaks-dir option given to bpt build commands.

See also

For more information, refer to this article.

URL

A Uniform Resource Locator is a string that specifies how to find a resource, either on the network/internet or on the local filesystem.

YAML

YAML is a data representation format intended to be written by humans and contain arbitrarily nested data structures.

YAML is a superset of JSON, so every valid JSON document is also a valid equivalent YAML document.

You can learn more about YAML at https://yaml.org.

The example from the JSON definition could also be written in YAML as:

example.yaml
foo: 123.0
bar:
  - true
  - false
  - null
# We can use comments in YAML!
baz: quux
another:
  nested-object: []

YAML can also be written as a “better JSON” by using the data block delimiters:

example.yaml
{
  foo: 123.0,
  bar: [true, false, null],
  # We can still use comments!
  baz: "quux",
  another: {nested-object: []},  # Trailing commas are allowed!
}

Note

Any JSON document is also a valid equivalent YAML document.

Note

bpt parses according to YAML 1.2, so alternate boolean plain scalars (e.g. “no” and “yes”) and sexagesimal datetimes are not supported.