2017-01-14 23:30:36 +03:00
|
|
|
# coding: utf-8
|
|
|
|
from __future__ import unicode_literals, print_function
|
2015-10-25 15:15:51 +03:00
|
|
|
|
2018-03-27 20:23:02 +03:00
|
|
|
import contextlib
|
|
|
|
from pathlib import Path
|
2019-09-28 19:05:00 +03:00
|
|
|
from fabric.api import local, lcd, env
|
2017-01-14 23:30:36 +03:00
|
|
|
from os import path, environ
|
2018-03-27 20:23:02 +03:00
|
|
|
import shutil
|
2018-05-01 18:51:22 +03:00
|
|
|
import sys
|
2014-09-11 14:28:38 +04:00
|
|
|
|
2015-01-03 13:02:21 +03:00
|
|
|
|
|
|
|
PWD = path.dirname(__file__)
|
2019-09-28 19:05:00 +03:00
|
|
|
ENV = environ["VENV_DIR"] if "VENV_DIR" in environ else ".env"
|
2018-03-27 20:23:02 +03:00
|
|
|
VENV_DIR = Path(PWD) / ENV
|
2015-01-03 13:02:21 +03:00
|
|
|
|
|
|
|
|
2018-03-27 20:23:02 +03:00
|
|
|
@contextlib.contextmanager
|
2019-09-28 19:05:00 +03:00
|
|
|
def virtualenv(name, create=False, python="/usr/bin/python3.6"):
|
2018-03-27 20:23:02 +03:00
|
|
|
python = Path(python).resolve()
|
|
|
|
env_path = VENV_DIR
|
|
|
|
if create:
|
|
|
|
if env_path.exists():
|
|
|
|
shutil.rmtree(str(env_path))
|
2019-09-28 19:05:00 +03:00
|
|
|
local("{python} -m venv {env_path}".format(python=python, env_path=VENV_DIR))
|
|
|
|
|
2018-03-27 20:23:02 +03:00
|
|
|
def wrapped_local(cmd, env_vars=[], capture=False, direct=False):
|
2019-09-28 19:05:00 +03:00
|
|
|
return local(
|
|
|
|
"source {}/bin/activate && {}".format(env_path, cmd),
|
|
|
|
shell="/bin/bash",
|
|
|
|
capture=False,
|
|
|
|
)
|
|
|
|
|
2018-03-27 20:23:02 +03:00
|
|
|
yield wrapped_local
|
|
|
|
|
|
|
|
|
2019-09-28 19:05:00 +03:00
|
|
|
def env(lang="python3.6"):
|
2018-03-27 20:23:02 +03:00
|
|
|
if VENV_DIR.exists():
|
2019-09-28 19:05:00 +03:00
|
|
|
local("rm -rf {env}".format(env=VENV_DIR))
|
|
|
|
if lang.startswith("python3"):
|
|
|
|
local("{lang} -m venv {env}".format(lang=lang, env=VENV_DIR))
|
2018-03-27 20:23:02 +03:00
|
|
|
else:
|
2019-09-28 19:05:00 +03:00
|
|
|
local("{lang} -m pip install virtualenv --no-cache-dir".format(lang=lang))
|
|
|
|
local(
|
|
|
|
"{lang} -m virtualenv {env} --no-cache-dir".format(lang=lang, env=VENV_DIR)
|
|
|
|
)
|
2018-03-27 20:23:02 +03:00
|
|
|
with virtualenv(VENV_DIR) as venv_local:
|
2019-09-28 19:05:00 +03:00
|
|
|
print(venv_local("python --version", capture=True))
|
|
|
|
venv_local("pip install --upgrade setuptools --no-cache-dir")
|
|
|
|
venv_local("pip install pytest --no-cache-dir")
|
|
|
|
venv_local("pip install wheel --no-cache-dir")
|
|
|
|
venv_local("pip install -r requirements.txt --no-cache-dir")
|
|
|
|
venv_local("pip install pex --no-cache-dir")
|
2015-01-03 13:02:21 +03:00
|
|
|
|
|
|
|
|
|
|
|
def install():
|
2018-03-27 20:23:02 +03:00
|
|
|
with virtualenv(VENV_DIR) as venv_local:
|
2019-09-28 19:05:00 +03:00
|
|
|
venv_local("pip install dist/*.tar.gz")
|
2015-01-03 13:02:21 +03:00
|
|
|
|
2014-07-05 22:49:34 +04:00
|
|
|
|
|
|
|
def make():
|
2018-03-27 20:23:02 +03:00
|
|
|
with lcd(path.dirname(__file__)):
|
2019-09-28 19:05:00 +03:00
|
|
|
local(
|
|
|
|
"export PYTHONPATH=`pwd` && source .env/bin/activate && python setup.py build_ext --inplace",
|
|
|
|
shell="/bin/bash",
|
|
|
|
)
|
|
|
|
|
2015-01-03 13:02:21 +03:00
|
|
|
|
2017-11-07 14:11:08 +03:00
|
|
|
def sdist():
|
2018-03-27 20:23:02 +03:00
|
|
|
with virtualenv(VENV_DIR) as venv_local:
|
2017-11-07 14:11:08 +03:00
|
|
|
with lcd(path.dirname(__file__)):
|
2019-09-28 19:05:00 +03:00
|
|
|
local("python -m pip install -U setuptools srsly")
|
|
|
|
local("python setup.py sdist")
|
|
|
|
|
2015-01-03 13:02:21 +03:00
|
|
|
|
2018-03-27 20:23:02 +03:00
|
|
|
def wheel():
|
|
|
|
with virtualenv(VENV_DIR) as venv_local:
|
|
|
|
with lcd(path.dirname(__file__)):
|
2019-09-28 19:05:00 +03:00
|
|
|
venv_local("python setup.py bdist_wheel")
|
|
|
|
|
2018-03-27 20:23:02 +03:00
|
|
|
|
|
|
|
def pex():
|
|
|
|
with virtualenv(VENV_DIR) as venv_local:
|
|
|
|
with lcd(path.dirname(__file__)):
|
2019-09-28 19:05:00 +03:00
|
|
|
sha = local("git rev-parse --short HEAD", capture=True)
|
|
|
|
venv_local(
|
|
|
|
"pex dist/*.whl -e spacy -o dist/spacy-%s.pex" % sha, direct=True
|
|
|
|
)
|
2018-03-27 20:23:02 +03:00
|
|
|
|
|
|
|
|
2015-01-03 13:02:21 +03:00
|
|
|
def clean():
|
|
|
|
with lcd(path.dirname(__file__)):
|
2019-09-28 19:05:00 +03:00
|
|
|
local("rm -f dist/*.whl")
|
|
|
|
local("rm -f dist/*.pex")
|
2018-03-27 20:23:02 +03:00
|
|
|
with virtualenv(VENV_DIR) as venv_local:
|
2019-09-28 19:05:00 +03:00
|
|
|
venv_local("python setup.py clean --all")
|
2015-01-03 13:02:21 +03:00
|
|
|
|
2014-07-05 22:49:34 +04:00
|
|
|
|
2015-01-03 13:02:21 +03:00
|
|
|
def test():
|
2018-03-27 20:23:02 +03:00
|
|
|
with virtualenv(VENV_DIR) as venv_local:
|
2015-01-03 13:02:21 +03:00
|
|
|
with lcd(path.dirname(__file__)):
|
2019-09-28 19:05:00 +03:00
|
|
|
venv_local("pytest -x tests")
|
|
|
|
|
2018-03-27 20:23:02 +03:00
|
|
|
|
|
|
|
def train():
|
2019-09-28 19:05:00 +03:00
|
|
|
args = environ.get("SPACY_TRAIN_ARGS", "")
|
2018-03-27 20:23:02 +03:00
|
|
|
with virtualenv(VENV_DIR) as venv_local:
|
2019-09-28 19:05:00 +03:00
|
|
|
venv_local("spacy train {args}".format(args=args))
|
2018-05-01 18:51:22 +03:00
|
|
|
|
|
|
|
|
2019-09-28 19:05:00 +03:00
|
|
|
def conll17(treebank_dir, experiment_dir, vectors_dir, config, corpus=""):
|
|
|
|
is_not_clean = local("git status --porcelain", capture=True)
|
2018-05-01 18:51:22 +03:00
|
|
|
if is_not_clean:
|
|
|
|
print("Repository is not clean")
|
|
|
|
print(is_not_clean)
|
|
|
|
sys.exit(1)
|
2019-09-28 19:05:00 +03:00
|
|
|
git_sha = local("git rev-parse --short HEAD", capture=True)
|
|
|
|
config_checksum = local("sha256sum {config}".format(config=config), capture=True)
|
|
|
|
experiment_dir = Path(experiment_dir) / "{}--{}".format(
|
|
|
|
config_checksum[:6], git_sha
|
|
|
|
)
|
2018-05-01 18:51:22 +03:00
|
|
|
if not experiment_dir.exists():
|
|
|
|
experiment_dir.mkdir()
|
2019-09-28 19:05:00 +03:00
|
|
|
test_data_dir = Path(treebank_dir) / "ud-test-v2.0-conll2017"
|
2018-05-01 18:51:22 +03:00
|
|
|
assert test_data_dir.exists()
|
|
|
|
assert test_data_dir.is_dir()
|
|
|
|
if corpus:
|
|
|
|
corpora = [corpus]
|
|
|
|
else:
|
2019-09-28 19:05:00 +03:00
|
|
|
corpora = ["UD_English", "UD_Chinese", "UD_Japanese", "UD_Vietnamese"]
|
2018-05-01 18:51:22 +03:00
|
|
|
|
2019-09-28 19:05:00 +03:00
|
|
|
local(
|
|
|
|
"cp {config} {experiment_dir}/config.json".format(
|
|
|
|
config=config, experiment_dir=experiment_dir
|
|
|
|
)
|
|
|
|
)
|
2018-05-01 18:51:22 +03:00
|
|
|
with virtualenv(VENV_DIR) as venv_local:
|
|
|
|
for corpus in corpora:
|
2019-09-28 19:05:00 +03:00
|
|
|
venv_local(
|
|
|
|
"spacy ud-train {treebank_dir} {experiment_dir} {config} {corpus} -v {vectors_dir}".format(
|
|
|
|
treebank_dir=treebank_dir,
|
|
|
|
experiment_dir=experiment_dir,
|
|
|
|
config=config,
|
|
|
|
corpus=corpus,
|
|
|
|
vectors_dir=vectors_dir,
|
|
|
|
)
|
|
|
|
)
|
|
|
|
venv_local(
|
|
|
|
"spacy ud-run-test {test_data_dir} {experiment_dir} {corpus}".format(
|
|
|
|
test_data_dir=test_data_dir,
|
|
|
|
experiment_dir=experiment_dir,
|
|
|
|
config=config,
|
|
|
|
corpus=corpus,
|
|
|
|
)
|
|
|
|
)
|