From d2b4a528465dc01e8db92b61293c458c7911a333 Mon Sep 17 00:00:00 2001 From: Matthias Baumgartner Date: Mon, 31 Oct 2022 12:21:22 +0100 Subject: essential interfaces (reader, extractor, errors) --- bsie/base/reader.py | 48 ++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 48 insertions(+) create mode 100644 bsie/base/reader.py (limited to 'bsie/base/reader.py') diff --git a/bsie/base/reader.py b/bsie/base/reader.py new file mode 100644 index 0000000..f29e451 --- /dev/null +++ b/bsie/base/reader.py @@ -0,0 +1,48 @@ +"""The Reader classes return high-level content structures from files. + +The Reader fulfills two purposes: + First, it brokers between multiple libraries and file formats. + Second, it separates multiple aspects of a file into distinct content types. + +Part of the bsie module. +A copy of the license is provided with the project. +Author: Matthias Baumgartner, 2022 +""" +# imports +import abc +import typing + +# inner-module imports +from bsie.utils.bsfs import URI, typename + +# exports +__all__: typing.Sequence[str] = ( + 'Aggregator', + 'Reader', + ) + + +## code ## + +class Reader(abc.ABC): + """Read and return some content from a file.""" + + # In what data structure content is returned + CONTENT_TYPE = typing.Union[typing.Any] + # NOTE: Child classes must also assign a typing.Union even if there's + # only one options + + def __str__(self) -> str: + return typename(self) + + def __repr__(self) -> str: + return f'{typename(self)}()' + + # FIXME: How about using contexts instead of calls? + @abc.abstractmethod + def __call__(self, path: URI) -> CONTENT_TYPE: + """Return some content of the file at *path*. + Raises a `ReaderError` if the reader cannot make sense of the file format. + """ + +## EOF ## -- cgit v1.2.3 From b96c6e2096c387b70e2a4c1f0bc53b6044a0dc6f Mon Sep 17 00:00:00 2001 From: Matthias Baumgartner Date: Fri, 25 Nov 2022 14:36:27 +0100 Subject: decouple readers and extractors; use strings for reference and repeated type annotations --- bsie/base/reader.py | 11 ++--------- 1 file changed, 2 insertions(+), 9 deletions(-) (limited to 'bsie/base/reader.py') diff --git a/bsie/base/reader.py b/bsie/base/reader.py index f29e451..e59abef 100644 --- a/bsie/base/reader.py +++ b/bsie/base/reader.py @@ -12,12 +12,11 @@ Author: Matthias Baumgartner, 2022 import abc import typing -# inner-module imports +# bsie imports from bsie.utils.bsfs import URI, typename # exports __all__: typing.Sequence[str] = ( - 'Aggregator', 'Reader', ) @@ -27,20 +26,14 @@ __all__: typing.Sequence[str] = ( class Reader(abc.ABC): """Read and return some content from a file.""" - # In what data structure content is returned - CONTENT_TYPE = typing.Union[typing.Any] - # NOTE: Child classes must also assign a typing.Union even if there's - # only one options - def __str__(self) -> str: return typename(self) def __repr__(self) -> str: return f'{typename(self)}()' - # FIXME: How about using contexts instead of calls? @abc.abstractmethod - def __call__(self, path: URI) -> CONTENT_TYPE: + def __call__(self, path: URI) -> typing.Any: """Return some content of the file at *path*. Raises a `ReaderError` if the reader cannot make sense of the file format. """ -- cgit v1.2.3 From 9ce32829b2bb85907a34a543bfcaa9183d1e362c Mon Sep 17 00:00:00 2001 From: Matthias Baumgartner Date: Fri, 25 Nov 2022 14:39:18 +0100 Subject: string conversion and equality checks --- bsie/base/reader.py | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'bsie/base/reader.py') diff --git a/bsie/base/reader.py b/bsie/base/reader.py index e59abef..b7eabf7 100644 --- a/bsie/base/reader.py +++ b/bsie/base/reader.py @@ -32,6 +32,12 @@ class Reader(abc.ABC): def __repr__(self) -> str: return f'{typename(self)}()' + def __eq__(self, other: typing.Any) -> bool: + return isinstance(other, type(self)) + + def __hash__(self) -> int: + return hash(type(self)) + @abc.abstractmethod def __call__(self, path: URI) -> typing.Any: """Return some content of the file at *path*. -- cgit v1.2.3 From 49cf03fc212c813862453de5352436dc90d1e458 Mon Sep 17 00:00:00 2001 From: Matthias Baumgartner Date: Thu, 15 Dec 2022 16:50:53 +0100 Subject: imports and init files --- bsie/base/reader.py | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'bsie/base/reader.py') diff --git a/bsie/base/reader.py b/bsie/base/reader.py index b7eabf7..cbabd36 100644 --- a/bsie/base/reader.py +++ b/bsie/base/reader.py @@ -13,7 +13,7 @@ import abc import typing # bsie imports -from bsie.utils.bsfs import URI, typename +from bsie.utils import bsfs # exports __all__: typing.Sequence[str] = ( @@ -27,10 +27,10 @@ class Reader(abc.ABC): """Read and return some content from a file.""" def __str__(self) -> str: - return typename(self) + return bsfs.typename(self) def __repr__(self) -> str: - return f'{typename(self)}()' + return f'{bsfs.typename(self)}()' def __eq__(self, other: typing.Any) -> bool: return isinstance(other, type(self)) @@ -39,7 +39,7 @@ class Reader(abc.ABC): return hash(type(self)) @abc.abstractmethod - def __call__(self, path: URI) -> typing.Any: + def __call__(self, path: bsfs.URI) -> typing.Any: """Return some content of the file at *path*. Raises a `ReaderError` if the reader cannot make sense of the file format. """ -- cgit v1.2.3