13 lines
563 B
Python
13 lines
563 B
Python
from kbqa.indexing.loaders import LoadedPage
|
|
from kbqa.indexing.splitter import split_pages
|
|
|
|
|
|
def test_split_pages_is_deterministic() -> None:
|
|
document_id = "12345678-1234-5678-1234-567812345678"
|
|
pages = [LoadedPage(content="知识库" * 300, page_number=1)]
|
|
first = split_pages(document_id, pages, 100, 10)
|
|
second = split_pages(document_id, pages, 100, 10)
|
|
assert [chunk.id for chunk in first] == [chunk.id for chunk in second]
|
|
assert first[0].page_number == 1
|
|
assert [chunk.order_index for chunk in first] == list(range(len(first)))
|