fixed the mypy

Signed-off-by: Peter Staar <taa@zurich.ibm.com>
This commit is contained in:
Peter Staar 2024-10-19 06:23:35 +02:00
parent 70b2ae3fab
commit e60c52586b

View File

@ -15,15 +15,14 @@ from docling_core.types.doc import (
from docling.backend.abstract_backend import DeclarativeDocumentBackend from docling.backend.abstract_backend import DeclarativeDocumentBackend
from docling.datamodel.base_models import InputFormat from docling.datamodel.base_models import InputFormat
from docling.datamodel.document import InputDocument
# from docling.datamodel.document import InputDocument
_log = logging.getLogger(__name__) _log = logging.getLogger(__name__)
class AsciidocBackend(DeclarativeDocumentBackend): class AsciidocBackend(DeclarativeDocumentBackend):
def __init__(self, in_doc: "InputDocument", path_or_stream: Union[BytesIO, Path]): def __init__(self, in_doc: InputDocument, path_or_stream: Union[BytesIO, Path]):
super().__init__(in_doc, path_or_stream) super().__init__(in_doc, path_or_stream)
self.path_or_stream = path_or_stream self.path_or_stream = path_or_stream
@ -76,7 +75,8 @@ class AsciidocBackend(DeclarativeDocumentBackend):
""" """
content = "" content = ""
with open(self.path_or_stream, "r") as fr: if isinstance(self.path_or_stream, Path):
with open(self.path_or_stream.name, "r") as fr:
self.lines = fr.readlines() self.lines = fr.readlines()
# self.lines = file_content.splitlines() # self.lines = file_content.splitlines()
@ -91,7 +91,7 @@ class AsciidocBackend(DeclarativeDocumentBackend):
# Title # Title
if self.is_title(line): if self.is_title(line):
item = self.parse_title(line) item = self.parse_title(line)
doc.add_text(text=item["text"], label="title") doc.add_text(text=item["text"], label=DocItemLabel.TITLE)
# Section headers # Section headers
elif self.is_section_header(line): elif self.is_section_header(line):
@ -125,7 +125,7 @@ class AsciidocBackend(DeclarativeDocumentBackend):
# Plain text # Plain text
elif line: elif line:
item = self.parse_text(line) item = self.parse_text(line)
doc.add_text(text=item["text"], label="text") doc.add_text(text=item["text"], label=DocItemLabel.TEXT)
if in_table and len(table_data) > 0: if in_table and len(table_data) > 0:
data = self.populate_table_as_grid(table_data) data = self.populate_table_as_grid(table_data)