Skip to main content

A Python-based build system with a Makefile-like experience

Project description

HXMK (Hexcell Make)

PyPI GitHub PyPI - Python Version

HXMK is a python-based build system for the Hexcell Projects. It brings a Makefile like experience to a familiar and simple environment.

Best of all: it has nice colors

colors2 colors

Installation

pip install HXMK

Dependencies:

Manual (latest version)
$ git clone https://github.com/Hexcell/HXMK.git
$ cd HXMK

$ pip install .

Usage

If the project directory contains a hxmk.py file, you can do

hxmk [args|rules]

Arguments are specified like name=value and rules like name.

hxmk abc=123 somerule

In case no rule was given, HXMK will look for the rule @everything and execute it (if it exists). No arguments or rules have to be specified be default.

hxmk # <- this is completely valid

HXMK can also be used to clean directories.

hxmk clean [args|rules]

This will look for a .clean file that contains a glob pattern on each line.

Example of a .clean file:

bin
obj
*.o
__pycache__

clean counts as a rule (although it's not an actual rule), so the rule @everything will not be run if not explicitly stated like this:

hxmk clean everything # will clean and THEN run @everything

hxmk.py

Rules

The simplest rule would look something like this:

@rule()
def everything(c):
    pass

c is an instance of the class Commander, it is used to execute commands. To do so, it overloads the lshift operator.

@rule()
def everything(c):
    c << "echo Commands are executed like this."

Rules can have dependencies. They are given through the return annotation. A rules dependencies are executed before it.

@rule()
def everything(c) -> "other":
    c << "echo What a lovely day it is today."

def other(c):
    c << "echo @everything depends on me so I go first!"

Multiple dependencies are given in a list or tuple, whichever you prefer.

@rule()
def everything(c) -> ("other", "something")

Rules have triggers, which are state if and when a rule shall be executed. If no trigger is specified, the trigger always will be set to True.

For example, to execute a rule whenever at least one of the folders bin and obj are missing, you could do the following:

@rule(not_found=["bin", "obj"])
def dirs(c):
    c << "mkdir -p bin"
    c << "mkdir -p obj"

The following triggers are implemented so far:

  • always, always execute the rule. It is a bool.
  • dependencies, execute it if one or more dependencies were executed. It is a bool.
  • not_found, execute when a specified path is not found (file or folder). It can be a str, list or a tuple.
  • changed, cache a file or list of files. Execute when any of the specified files are not found in the cache or have been changed. It can be a str, list, or a tuple.

If not_found is given, the rule will assume that you are going to create the specified path. If that path is not found after the rule was executed, a warning will be shown.

An example of a caching rule:

@rule(changed=["a.cpp", "b.cpp"], not_found="program")

This above rule will be executed if a.cpp or b.cpp have been changed, or when program was not found. Though for this particular case, pattern rules would be recommended.

Pattern Rules

Pattern Rules are rules that are executed multiple times for multiple files. They look like this:

@pattern("src/*.c -> obj/*.o")
def somerule(c, src, dest):
    pass

(The syntax for the patterns is src -> dest.) This basically means, every .c file in src will be turned into an .o file in obj. The Pattern Rule will be executed for every .c file. This could be used to compile every .c file in a directory.

@pattern("src/*.c -> obj/*.o")
def somerule(c, src, dest):
    c << f"gcc -c {src} -o {dest}"

Pattern Rules can have multiple sources and destinations. In that case, the parameters src and dest are lists.

@pattern("src/*.c include/*.h -> obj/*.o")
def somerule(c, src, dest):
    c << f"gcc -c {src[0]} -o {dest}"

Pattern rules are cached. Before executing, the Pattern Rule checks whether the source files have been modified since the last build and if the destination file exists already. If the destination file does not exist, the rule will be executed, else it will only be executed if the source file was modified or not found in the cache.

Builtins

All builtins are immediately available without having to import anything.

make
make(self, path, args=[], isolate=False)

make will start HXMK in a different folder. Optionally args (in the sense of CLI args) can be passed in to args.

By default all variables from the root module are readable in every module called by make. This behavior can be stopped by setting isolate to True.

# hxmk.py
somevar = "hello"
@rule()
def everything(c):
    make("other")
# other/hxmk.py
@rule()
def everything(c):
    print(somevar)
>>> @all
>>> Entering other
>>> other/@all
>>> "hello"
default
default(arg, val)

default is used to optain a value from the CLI or default to another value in case it's not found. The first argument is the name and the second one is the default value. An example would be:

$ hxmk config=debug
config = default("config", "release")
assert config in ["debug", "release"]

# ...now config can be used
as_args
as_args(l, prefix="", suffix="")

as_args is used to use a list as CLI arguments. The first argument is a list, tuple or dict, the keyword arguments are prefix and suffix.

An example use case would be this:

...
    c << "ld ... %s" % as_args(["a.o", "b.o", "c.o"])

>>> ld ... "a.o" "b.o" "c.o"

Passing in a dict would return "key=value" for each item.

...
    c << "ld ... %s" % as_args({"a": "b", "c": "d"})

>>> ld ... "a=b" "c=d"

When passing in a dict, the prefix and suffix parameters can not be used.

glob
glob(pathname, *, recursive=False)

glob is just the glob function from pythons standard library. For more information, look at Pythons documentation on it.

It could be used in combination with as_args to link all object files.

...
    c << "ld ... %s" % as_args(glob("obj/*.o"))

>>> ld ... "main.o" "utils.o"

Goals

The main goal is to create a simple build too with the same functionality as make (+ more) while maintaining readability and sanity.

Currently, HXMK runs just fine, but there have been and will most likely be many API changes. Bugs are to be expected, if you find one, do not hesitate to create an issue.

TODO

  • rules
  • pattern eules
  • argument parsing
  • build subdirectories
  • cleaning
  • add docstrings and comments for every function
  • full api documentation
  • better cross-platform support / more builtins (mkdir, ...)
  • in depth compatibility testing with various python versions

Contributing

Any form of contribution is welcome ^^

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

HXMK-0.0.9.tar.gz (9.7 kB view details)

Uploaded Source

Built Distribution

HXMK-0.0.9-py3-none-any.whl (11.2 kB view details)

Uploaded Python 3

File details

Details for the file HXMK-0.0.9.tar.gz.

File metadata

  • Download URL: HXMK-0.0.9.tar.gz
  • Upload date:
  • Size: 9.7 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/1.13.0 pkginfo/1.5.0.1 requests/2.21.0 setuptools/40.6.3 requests-toolbelt/0.9.1 tqdm/4.31.1 CPython/3.7.3

File hashes

Hashes for HXMK-0.0.9.tar.gz
Algorithm Hash digest
SHA256 8169967caffe50bda0ecedcbc7ea665aac234e7d0afc637f4466fa3cb79cc8c9
MD5 176cff145b732ff7ef063a7a9df662a1
BLAKE2b-256 42b3492761d67751058cb9eb629d7eea7aa3acce51bfda19560d2d024490b8b1

See more details on using hashes here.

File details

Details for the file HXMK-0.0.9-py3-none-any.whl.

File metadata

  • Download URL: HXMK-0.0.9-py3-none-any.whl
  • Upload date:
  • Size: 11.2 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/1.13.0 pkginfo/1.5.0.1 requests/2.21.0 setuptools/40.6.3 requests-toolbelt/0.9.1 tqdm/4.31.1 CPython/3.7.3

File hashes

Hashes for HXMK-0.0.9-py3-none-any.whl
Algorithm Hash digest
SHA256 147e850a2e8d1b6157d590553a4ec1d8146fb486c77a0046a31209db6464e35b
MD5 de966369851ce5b4d30fd5eb22e4140e
BLAKE2b-256 051f076fa6f63a042bb8f5c24d88c485d0a4dc030d7ba9b7029ea1b20c2de2aa

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page