"""Tests for pdfplumber structure methods."""
from __future__ import annotations
from pathlib import Path
from unittest.mock import MagicMock, patch
import pytest
from zotero_chunk_rag.feature_extraction.methods.pdfplumber_structure import (
PdfplumberLines,
PdfplumberText,
_detect_with_strategy,
)
from zotero_chunk_rag.feature_extraction.models import TableContext
from zotero_chunk_rag.feature_extraction.protocols import StructureMethod
def _make_ctx(
page_num: int = 0,
bbox: tuple[float, float, float, float] = (72.0, 200.0, 540.0, 600.0),
) -> TableContext:
"""Create a TableContext with a mock pymupdf page."""
mock_page = MagicMock()
mock_page.get_text.return_value = []
mock_page.get_drawings.return_value = []
mock_page.rect = MagicMock()
mock_page.rect.height = 792.0
mock_page.rect.width = 612.0
return TableContext(
page=mock_page,
page_num=page_num,
bbox=bbox,
pdf_path=Path("test.pdf"),
)
# ---------------------------------------------------------------------------
# PdfplumberLines
# ---------------------------------------------------------------------------
class TestPdfplumberLines:
def test_protocol_conformance(self) -> None:
assert isinstance(PdfplumberLines(), StructureMethod)
def test_name(self) -> None:
assert PdfplumberLines().name == "pdfplumber_lines"
def test_returns_none_no_tables(self) -> None:
"""When pdfplumber finds no tables, detect() returns None."""
mock_cropped = MagicMock()
mock_cropped.find_tables.return_value = []
mock_plumber_page = MagicMock()
mock_plumber_page.within_bbox.return_value = mock_cropped
mock_pdf = MagicMock()
mock_pdf.pages = [mock_plumber_page]
mock_pdf.__enter__ = lambda self: self
mock_pdf.__exit__ = MagicMock(return_value=False)
ctx = _make_ctx()
with patch(
"zotero_chunk_rag.feature_extraction.methods.pdfplumber_structure.pdfplumber"
) as mock_pdfplumber:
mock_pdfplumber.open.return_value = mock_pdf
method = PdfplumberLines()
result = method.detect(ctx)
assert result is None
# ---------------------------------------------------------------------------
# PdfplumberText
# ---------------------------------------------------------------------------
class TestPdfplumberText:
def test_protocol_conformance(self) -> None:
assert isinstance(PdfplumberText(), StructureMethod)
def test_name(self) -> None:
assert PdfplumberText().name == "pdfplumber_text"
# ---------------------------------------------------------------------------
# No coordinate conversion
# ---------------------------------------------------------------------------
class TestPdfplumber:
def test_no_coordinate_conversion(self) -> None:
"""Verify that pdfplumber cell bboxes are used directly without conversion.
Create mock pdfplumber table with known cell bboxes and verify the
resulting boundary positions match the raw values.
"""
mock_table = MagicMock()
# 2x2 grid: cells at known positions
mock_table.cells = [
(72.0, 200.0, 306.0, 400.0),
(306.0, 200.0, 540.0, 400.0),
(72.0, 400.0, 306.0, 600.0),
(306.0, 400.0, 540.0, 600.0),
]
mock_cropped = MagicMock()
mock_cropped.find_tables.return_value = [mock_table]
mock_plumber_page = MagicMock()
mock_plumber_page.within_bbox.return_value = mock_cropped
mock_pdf = MagicMock()
mock_pdf.pages = [mock_plumber_page]
ctx = _make_ctx()
with patch(
"zotero_chunk_rag.feature_extraction.methods.pdfplumber_structure.pdfplumber"
) as mock_pdfplumber:
mock_pdfplumber.open.return_value = mock_pdf
method = PdfplumberLines()
result = method.detect(ctx)
assert result is not None
# Column boundary: 306.0 (the only internal x0)
assert len(result.col_boundaries) == 1
assert result.col_boundaries[0].min_pos == 306.0
# Row boundary: 400.0 (the only internal y0)
assert len(result.row_boundaries) == 1
assert result.row_boundaries[0].min_pos == 400.0
