hare-module(5) File Formats Manual hare-module(5) NAME hare-module - Hare module layout DESCRIPTION Hare modules are represented as directories in the filesystem. A directory is a valid Hare module if it contains at least one Hare source file (with the file extension .ha), or if it contains a file named README. All files which end in .ha, .s, and .o are treated as inputs to the module, and are respectively treated as Hare sources, assembly sources, and objects to be linked into the final binary. The list of files considered eligible may be filtered by build tags (see BUILD TAGS below). The format for the filename is name[tagset].ext, where the name is user-defined, the ext is either ha, s, or o, and a tagset (see BUILD TAGS) is optionally provided after the name. A file will only be in- cluded if all tags included (with +) in the tagset are present, and no tags excluded (with -) in the tagset are present. Only one file for a given combination of name and ext will be selected for the build; the file selected is the one with the most tag specifiers. If there are two or more such files, the build driver will error out. For example, if the following files are present in a directory: * foo.ha * bar.ha * bar+linux.ha * bar+plan9.ha * baz+x86_64.s * bat-x86_64.ha If the build tags are +linux+x86_64, then the files which are included in the module are foo.ha, bar+linux.ha, and baz+x86_64.s. If the following files are added: * meep+linux-libc.ha * meep+linux+x86_64.ha then the build driver will error out, unless +libc is added to the build tags. Additionally, subdirectories in a module will be considered part of that module if their name consists only of a tagset, e.g. "+linux" or "-x86_64". A directory with only a name (but without a tagset, e.g. "example") is con- sidered a submodule, and as such must be imported separately. For example, "foo::bar" refers to foo/bar/. If a directory name contains both a name and a tagset, the build driver will error out. DEPENDENCY RESOLUTION The "use" directives in each Hare source file used as an input to hare- build(1), hare-run(1), or hare-test(1) are scanned and used to determine the dependencies for a program. This process is repeated for each depen- dency to obtain a complete dependency tree. Dependencies are searched for by examining first the current working direc- tory, then each component of the HAREPATH, which is a list of paths sepa- rated by colons. HAREPATH is obtained from the environment variable of the same name. If the environment variable is unset, a default value is used, which can be viewed with the hare version -v command. Typically, it is set to include the path to the standard library installed on the system, as well as a system-provided location for third-party modules installed via the system package manager. hare-deps(1) may be used to print the dependency tree of a Hare module. BUILD TAGS Build tags allow you to add constraints on what features or platforms are enabled for your build. A tag is a name, consisting of characters which aren't any of '+', '-', or '.', and a '+' or '-' prefix to signal inclusiv- ity or exclusivity. To add or remove build tags, use the -T flag. For example, -T +foo-bar will add the 'foo' tag and remove the 'bar' tag. Some tags are enabled by default, enabling features for the host platform. You can view the default tagset by running hare version -v. To remove all default tags, use -T^. SEE ALSO hare(1), hare-build(1), hare-run(1), hare-test(1) 2026-08-27 hare-module(5)
NAME | DESCRIPTION | DEPENDENCY RESOLUTION | BUILD TAGS | SEE ALSO
Want to link to this manual page? Use this URL:
<https://man.freebsd.org/cgi/man.cgi?query=hare-module&sektion=5&manpath=FreeBSD+Ports+15.1.quarterly>