pyproject.toml

[tool.poetry]
name = "html2tei"
version = "1.2.3"
description = "Map the HTML schema of portals to valid TEI XML with the tags and structures used in them using small manual portal-specific configurations."
readme = "README.md"
repository = "https://github.com/ELTE-DH/HTML2TEI"
authors = ["dlazesz", "sarkozizsofia"]
license = "LGPLv3"
# License and required Python version is automatically added
classifiers = [
    "Topic :: Software Development :: Libraries :: Python Modules",
    "Operating System :: OS Independent",
    "Development Status :: 5 - Production/Stable",
]

[tool.poetry.scripts]
html2tei = "html2tei.__main__:main"

[tool.poetry.dependencies]
python = "^3.8"
beautifulsoup4 = "^4.9.0"
lxml = "^4.5.0"
pyyaml = "^6.0.0"
warcio = "^1.7.0"
certifi = "^2021.10.8"
webarticlecurator = "^1.10.2"
mthasher = "^2.0.4"
mplogger = "^1.0.0"
# A list of all of the optional dependencies, some of which are included in the
# below `extras`. They can be opted into by apps.
justext = { version = "^3.0.0", optional = true }
newspaper3k = { version = "^0.2.8", optional = true }
trafilatura = { version = "^1.0.0", optional = true }
langdetect = "^1.0.9"

[tool.poetry.extras]
justext = ["justext"]
newspaper3k = ["newspaper3k"]
trafilatura = ["trafilatura"]
full = ["justext", "newspaper3k", "trafilatura"]

[tool.poetry.dev-dependencies]
pytest = "^7"

[build-system]
requires = ["poetry-core>=1.0.0"]
build-backend = "poetry.core.masonry.api"