ocr-tech-k2 / tests /test_external_outputs.py
prosperious's picture
Add 195 files (runs/hf-recognizer-mixed-v2-full/finalized/package/artifacts/model/model.safetensors ...)
1ef1de7 verified
Raw History Blame Contribute Delete
1.58 kB
from __future__ import annotations
from pathlib import Path
from ocrtech.external_outputs import bundle_page_paths, document_from_markdown
def test_document_from_markdown_parses_text_and_tables(tmp_path: Path) -> None:
document = document_from_markdown(
tmp_path / "sample.md",
"# नेपाल सरकार\n\n| नाम | उमेर |\n| --- | --- |\n| राम | ३२ |\n| Sita | 29 |\n",
metadata={"engine": "fixture"},
)
assert document.text.startswith("नेपाल सरकार")
assert len(document.tables) == 1
grid = {(cell.row, cell.col): cell.text for cell in document.tables[0].cells}
assert grid[(0, 0)] == "नाम"
assert grid[(1, 0)] == "राम"
assert document.pages[0].blocks[0].block_type == "title"
assert any(block.block_type == "table" for block in document.pages[0].blocks)
def test_document_from_markdown_splits_pagebreaks(tmp_path: Path) -> None:
document = document_from_markdown(
tmp_path / "sample.md",
"Page one\n\n<!-- pagebreak -->\n\nPage two\n",
)
assert len(document.pages) == 2
assert document.pages[0].text_lines[0].text == "Page one"
assert document.pages[1].text_lines[0].text == "Page two"
def test_bundle_page_paths_accepts_single_file_and_sorted_directory(tmp_path: Path) -> None:
first = tmp_path / "b.png"
second = tmp_path / "a.png"
first.write_bytes(b"b")
second.write_bytes(b"a")
assert bundle_page_paths(first) == [first]
assert bundle_page_paths(tmp_path) == [second, first]