20086551557

Committed 10 Dec 2025 03:43AM UTC coverage: 90.303% (+0.3%) from 90.041%

Build # 20086551557

Build Type

push

github

Committed by

bramp

Commit Message

Refactor arrow shaft detection: unified method, stroked line support, multi-head grouping

- Merge _find_simple_shaft, _find_stroked_line_shaft, and _find_cornered_shaft
  into a single unified _find_shaft method that handles all shaft types by
  extracting points and finding closest/furthest from the arrowhead tip
- Add support for stroked line shafts (stroke_color instead of fill_color)
- Add tail_grouping_tolerance config for grouping arrowheads with nearby tails
- Group arrowheads that share the same shaft_block (L-shaped arrows with
  multiple heads at different ends)
- Use union-find algorithm to group arrowheads by shared shaft or tail proximity
- Extract colors_match to shared utils module
- Add comprehensive tests for stroked line shafts, tail correctness, and
  multi-head arrow grouping
- Update golden files for pages 011, 013, 015, 017 with corrected arrow detection

Coverage Stats

204 of 206 new or added lines in 5 files covered. (99.03%)

252 existing lines in 14 files now uncovered.

11855 of 13128 relevant lines covered (90.3%)

0.9 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

99.11

/src/build_a_long/pdf_extract/classifier/classifier.py

"""
Rule-based classifier for labeling page elements.

Pipeline order and dependencies
--------------------------------
The classification pipeline operates in two main phases:

1. **Bottom-up Scoring**: All classifiers run independently to identify potential
   candidates (e.g. page numbers, part counts, step numbers) and score them based
   on heuristics. No construction of final elements happens here.

2. **Top-down Construction**: The root `PageClassifier` is invoked to construct
   the final `Page` object. It recursively requests the construction of its
   dependencies (e.g. "Give me the best PageNumber"), which in turn construct
   their own dependencies. This ensures a consistent and validated object tree.

"""

from __future__ import annotations

import logging

from build_a_long.pdf_extract.classifier.bags import (
    BagNumberClassifier,
    NewBagClassifier,
)
from build_a_long.pdf_extract.classifier.batch_classification_result import (
    BatchClassificationResult,
)
from build_a_long.pdf_extract.classifier.block_filter import (
    filter_duplicate_blocks,
    filter_overlapping_text_blocks,
)
from build_a_long.pdf_extract.classifier.classification_result import (
    ClassificationResult,
)
from build_a_long.pdf_extract.classifier.classifier_config import ClassifierConfig
from build_a_long.pdf_extract.classifier.pages import (
    PageHintCollection,
)
from build_a_long.pdf_extract.classifier.pages.background_classifier import (
    BackgroundClassifier,
)
from build_a_long.pdf_extract.classifier.pages.divider_classifier import (
    DividerClassifier,
)
from build_a_long.pdf_extract.classifier.pages.page_classifier import PageClassifier
from build_a_long.pdf_extract.classifier.pages.page_number_classifier import (
    PageNumberClassifier,
)
from build_a_long.pdf_extract.classifier.pages.preview_classifier import (
    PreviewClassifier,
)
from build_a_long.pdf_extract.classifier.pages.progress_bar_classifier import (
    ProgressBarClassifier,
)
from build_a_long.pdf_extract.classifier.pages.progress_bar_indicator_classifier import (
    ProgressBarIndicatorClassifier,
)
from build_a_long.pdf_extract.classifier.pages.trivia_text_classifier import (
    TriviaTextClassifier,
)
from build_a_long.pdf_extract.classifier.parts import (
    PartCountClassifier,
    PartNumberClassifier,
    PartsClassifier,
    PartsImageClassifier,
    PartsListClassifier,
    PieceLengthClassifier,
    ShineClassifier,
)
from build_a_long.pdf_extract.classifier.removal_reason import RemovalReason
from build_a_long.pdf_extract.classifier.steps import (
    ArrowClassifier,
    DiagramClassifier,
    RotationSymbolClassifier,
    StepClassifier,
    StepCountClassifier,
    StepNumberClassifier,
    SubAssemblyClassifier,
)
from build_a_long.pdf_extract.classifier.text import FontSizeHints, TextHistogram
from build_a_long.pdf_extract.classifier.topological_sort import topological_sort
from build_a_long.pdf_extract.extractor import PageData
from build_a_long.pdf_extract.extractor.bbox import filter_contained
from build_a_long.pdf_extract.extractor.lego_page_elements import (
    PageNumber,
    PartCount,
    PartsList,
    StepNumber,
)
from build_a_long.pdf_extract.extractor.page_blocks import Blocks

logger = logging.getLogger(__name__)

# Pages with more blocks than this threshold will be skipped during classification.
# This avoids O(n²) algorithms (like duplicate detection) that become prohibitively
# slow on pages with thousands of vector drawings. Such pages are typically info
# pages where each character is a separate vector graphic.
# TODO: Add spatial indexing to handle high-block pages efficiently.
MAX_BLOCKS_PER_PAGE = 1000


# TODO require config, so we don't accidentally use default empty config
def classify_elements(
    page: PageData, config: ClassifierConfig | None = None
) -> ClassificationResult:
    """Classify and label elements on a single page using rule-based heuristics.

    Args:
        page: A single PageData object to classify.
        config: Optional classifier configuration with font/page hints.
            If None, uses default empty configuration (no hints).
            For better classification accuracy, pass a config with
            FontSizeHints computed from multiple pages of the same PDF.

    Returns:
        A ClassificationResult object containing the classification results.
    """
    if config is None:
        config = ClassifierConfig()
    classifier = Classifier(config)

    return classifier.classify(page)


def classify_pages(
    pages: list[PageData], pages_for_hints: list[PageData] | None = None
) -> BatchClassificationResult:
    """Classify and label elements across multiple pages using rule-based heuristics.

    This function performs a three-phase process:
    1. Filtering phase: Mark duplicate/similar blocks as removed on each page
    2. Analysis phase: Build font size hints from text properties (excluding
       removed blocks)
    3. Classification phase: Use hints to guide element classification

    Args:
        pages: A list of PageData objects to classify.
        pages_for_hints: Optional list of pages to use for generating font/page hints.
            If None, uses `pages`. This allows generating hints from all pages
            while only classifying a subset (e.g., when using --pages filter).

    Returns:
        BatchClassificationResult containing per-page results and global histogram
    """

    # TODO There is a bunch of duplication in here between hints and non-hints. Refactor

    # Use all pages for hint generation if provided, otherwise use selected pages
    hint_pages = pages_for_hints if pages_for_hints is not None else pages

    # Phase 1: Filter duplicate blocks on each page and track removals
    # Skip pages with too many blocks to avoid O(n²) performance issues
    removed_blocks_per_page: list[dict[Blocks, RemovalReason]] = []
    skipped_pages: set[int] = set()  # Track page numbers that are skipped

    for page_data in pages:
        # Skip pages with too many blocks - these are likely info/inventory pages
        # with vectorized text that cause O(n²) algorithms to be very slow
        if len(page_data.blocks) > MAX_BLOCKS_PER_PAGE:
            logger.debug(
                f"Page {page_data.page_number}: skipping classification "
                f"({len(page_data.blocks)} blocks exceeds threshold of "
                f"{MAX_BLOCKS_PER_PAGE})"
            )
            skipped_pages.add(page_data.page_number)
            removed_blocks_per_page.append({})
            continue

        kept_blocks = page_data.blocks

        # Filter overlapping text blocks (e.g., "4" and "43" at same origin)
        kept_blocks, text_removed = filter_overlapping_text_blocks(kept_blocks)

        # Filter duplicate image/drawing blocks based on IOU
        kept_blocks, bbox_removed = filter_duplicate_blocks(kept_blocks)

        # Combine all removal mappings into a single dict for this page
        combined_removed_mapping = {
            **text_removed,
            **bbox_removed,
        }

        logger.debug(
            f"Page {page_data.page_number}: "
            f"filtered {len(text_removed)} overlapping text, "
            f"{len(bbox_removed)} duplicate bbox blocks"
        )

        removed_blocks_per_page.append(combined_removed_mapping)

    # Phase 2: Extract font size hints from hint pages (excluding removed blocks)
    # Build pages with non-removed blocks for hint extraction and histogram

    # Filter duplicates from hint pages (may be different from pages to classify)
    hint_pages_without_duplicates = []
    for page_data in hint_pages:
        # Skip high-block pages for hints too (same threshold)
        if len(page_data.blocks) > MAX_BLOCKS_PER_PAGE:
            continue

        # TODO We are re-filtering duplicates here; optimize by changing the API
        # to accept one list of PageData, and seperate by page_numbers.
        kept_blocks = page_data.blocks
        kept_blocks, _ = filter_overlapping_text_blocks(kept_blocks)
        kept_blocks, _ = filter_duplicate_blocks(kept_blocks)

        hint_pages_without_duplicates.append(
            PageData(
                page_number=page_data.page_number,
                bbox=page_data.bbox,
                blocks=kept_blocks,
            )
        )

    # Build pages without duplicates for classification
    pages_without_duplicates = []
    for page_data, removed_mapping in zip(pages, removed_blocks_per_page, strict=True):
        # We need to filter blocks that were removed by ANY filter
        non_removed_blocks = [
            block for block in page_data.blocks if block not in removed_mapping
        ]
        pages_without_duplicates.append(
            PageData(
                page_number=page_data.page_number,
                bbox=page_data.bbox,
                blocks=non_removed_blocks,
            )
        )

    # Generate hints from hint pages, histogram from pages to classify
    font_size_hints = FontSizeHints.from_pages(hint_pages_without_duplicates)
    page_hints = PageHintCollection.from_pages(hint_pages_without_duplicates)
    histogram = TextHistogram.from_pages(pages_without_duplicates)

    # Phase 3: Classify using the hints (on pages without duplicates)
    config = ClassifierConfig(font_size_hints=font_size_hints, page_hints=page_hints)
    classifier = Classifier(config)

    results = []
    for page_data, page_without_duplicates, removed_mapping in zip(
        pages, pages_without_duplicates, removed_blocks_per_page, strict=True
    ):
        # Handle skipped pages
        if page_data.page_number in skipped_pages:
            result = ClassificationResult(
                page_data=page_data,
                skipped_reason=(
                    f"Page has {len(page_data.blocks)} blocks, which exceeds "
                    f"the threshold of {MAX_BLOCKS_PER_PAGE}. This is likely an "
                    f"info/inventory page with vectorized text."
                ),
            )
            results.append(result)
            continue

        # Classify using only non-removed blocks
        result = classifier.classify(page_without_duplicates)

        # Update result to use original page_data (with all blocks)
        result.page_data = page_data

        # Mark removed blocks
        for removed_block, removal_reason in removed_mapping.items():
            result.mark_removed(removed_block, removal_reason)

        results.append(result)

    return BatchClassificationResult(results=results, histogram=histogram)


type Classifiers = (
    PageNumberClassifier
    | ProgressBarClassifier
    | ProgressBarIndicatorClassifier
    | PreviewClassifier
    | BackgroundClassifier
    | DividerClassifier
    | BagNumberClassifier
    | PartCountClassifier
    | PartNumberClassifier
    | StepNumberClassifier
    | StepCountClassifier
    | PieceLengthClassifier
    | PartsClassifier
    | PartsListClassifier
    | PartsImageClassifier
    | ShineClassifier
    | NewBagClassifier
    | DiagramClassifier
    | ArrowClassifier
    | SubAssemblyClassifier
    | StepClassifier
    | TriviaTextClassifier
    | PageClassifier
)


class Classifier:
    """
    Performs a single run of classification based on rules, configuration, and hints.
    This class should be stateless.
    """

    def __init__(self, config: ClassifierConfig):
        self.config = config
        # Sort classifiers topologically based on their dependencies
        self.classifiers = topological_sort(
            [
                PageNumberClassifier(config=config),
                ProgressBarIndicatorClassifier(config=config),
                ProgressBarClassifier(config=config),
                BackgroundClassifier(config=config),
                DividerClassifier(config=config),
                BagNumberClassifier(config=config),
                PartCountClassifier(config=config),
                PartNumberClassifier(config=config),
                StepNumberClassifier(config=config),
                StepCountClassifier(config=config),
                PieceLengthClassifier(config=config),
                PartsClassifier(config=config),
                PartsListClassifier(config=config),
                DiagramClassifier(config=config),
                RotationSymbolClassifier(config=config),
                ArrowClassifier(config=config),
                PartsImageClassifier(config=config),
                ShineClassifier(config=config),
                NewBagClassifier(config=config),
                PreviewClassifier(config=config),
                SubAssemblyClassifier(config=config),
                StepClassifier(config=config),
                TriviaTextClassifier(config=config),
                PageClassifier(config=config),
            ]
        )

    def classify(self, page_data: PageData) -> ClassificationResult:
        """
        Runs the classification logic and returns a result.
        It does NOT modify page_data directly.

        The classification process runs in three phases:
        1. Score all classifiers (bottom-up) - auto-registers classifiers
        2. Construct final elements (top-down starting from Page)
        """
        result = ClassificationResult(page_data=page_data)

        logger.debug(f"Starting classification for page {page_data.page_number}")

        # 1. Score all classifiers (Bottom-Up)
        # Note: score() automatically registers each classifier for its output labels
        for classifier in self.classifiers:
            classifier.score(result)

        # 2. Construct (Top-Down)
        # Find the PageClassifier to start the construction process
        page_classifier = next(
            c for c in self.classifiers if isinstance(c, PageClassifier)
        )
        page_classifier.build_all(result)

        # TODO Do we actualy ever add warnings?
        warnings = self._log_post_classification_warnings(page_data, result)
        for warning in warnings:
            result.add_warning(warning)

        return result

    def _log_post_classification_warnings(
        self, page_data: PageData, result: ClassificationResult
    ) -> list[str]:
        warnings = []

        # Check if there's a page number
        page_numbers = result.get_winners_by_score("page_number", PageNumber)
        if not page_numbers:
            warnings.append(f"Page {page_data.page_number}: missing page number")

        # Get elements by label
        parts_lists = result.get_winners_by_score("parts_list", PartsList)
        part_counts = result.get_winners_by_score("part_count", PartCount)

        for pl in parts_lists:
            inside_counts = filter_contained(part_counts, pl.bbox)
            if not inside_counts:
                warnings.append(
                    f"Page {page_data.page_number}: parts list at {pl.bbox} "
                    f"contains no part counts"
                )

        steps = result.get_winners_by_score("step_number", StepNumber)
        ABOVE_EPS = 2.0
        for step in steps:
            sb = step.bbox
            above = [pl for pl in parts_lists if pl.bbox.y1 <= sb.y0 + ABOVE_EPS]
            if not above:
                warnings.append(
                    f"Page {page_data.page_number}: step number '{step.value}' "
                    f"at {sb} has no parts list above it"
                )
        return warnings

1	"""
2	Rule-based classifier for labeling page elements.
3
4	Pipeline order and dependencies
5	--------------------------------
6	The classification pipeline operates in two main phases:
7
8	1. Bottom-up Scoring: All classifiers run independently to identify potential
9	candidates (e.g. page numbers, part counts, step numbers) and score them based
10	on heuristics. No construction of final elements happens here.
11
12	2. Top-down Construction: The root `PageClassifier` is invoked to construct
13	the final `Page` object. It recursively requests the construction of its
14	dependencies (e.g. "Give me the best PageNumber"), which in turn construct
15	their own dependencies. This ensures a consistent and validated object tree.
16
17	"""
18
19	from __future__ import annotations	1✔
20
21	import logging	1✔
22
23	from build_a_long.pdf_extract.classifier.bags import (	1✔
24	BagNumberClassifier,
25	NewBagClassifier,
26	)
27	from build_a_long.pdf_extract.classifier.batch_classification_result import (	1✔
28	BatchClassificationResult,
29	)
30	from build_a_long.pdf_extract.classifier.block_filter import (	1✔
31	filter_duplicate_blocks,
32	filter_overlapping_text_blocks,
33	)
34	from build_a_long.pdf_extract.classifier.classification_result import (	1✔
35	ClassificationResult,
36	)
37	from build_a_long.pdf_extract.classifier.classifier_config import ClassifierConfig	1✔
38	from build_a_long.pdf_extract.classifier.pages import (	1✔
39	PageHintCollection,
40	)
41	from build_a_long.pdf_extract.classifier.pages.background_classifier import (	1✔
42	BackgroundClassifier,
43	)
44	from build_a_long.pdf_extract.classifier.pages.divider_classifier import (	1✔
45	DividerClassifier,
46	)
47	from build_a_long.pdf_extract.classifier.pages.page_classifier import PageClassifier	1✔
48	from build_a_long.pdf_extract.classifier.pages.page_number_classifier import (	1✔
49	PageNumberClassifier,
50	)
51	from build_a_long.pdf_extract.classifier.pages.preview_classifier import (	1✔
52	PreviewClassifier,
53	)
54	from build_a_long.pdf_extract.classifier.pages.progress_bar_classifier import (	1✔
55	ProgressBarClassifier,
56	)
57	from build_a_long.pdf_extract.classifier.pages.progress_bar_indicator_classifier import (	1✔
58	ProgressBarIndicatorClassifier,
59	)
60	from build_a_long.pdf_extract.classifier.pages.trivia_text_classifier import (	1✔
61	TriviaTextClassifier,
62	)
63	from build_a_long.pdf_extract.classifier.parts import (	1✔
64	PartCountClassifier,
65	PartNumberClassifier,
66	PartsClassifier,
67	PartsImageClassifier,
68	PartsListClassifier,
69	PieceLengthClassifier,
70	ShineClassifier,
71	)
72	from build_a_long.pdf_extract.classifier.removal_reason import RemovalReason	1✔
73	from build_a_long.pdf_extract.classifier.steps import (	1✔
74	ArrowClassifier,
75	DiagramClassifier,
76	RotationSymbolClassifier,
77	StepClassifier,
78	StepCountClassifier,
79	StepNumberClassifier,
80	SubAssemblyClassifier,
81	)
82	from build_a_long.pdf_extract.classifier.text import FontSizeHints, TextHistogram	1✔
83	from build_a_long.pdf_extract.classifier.topological_sort import topological_sort	1✔
84	from build_a_long.pdf_extract.extractor import PageData	1✔
85	from build_a_long.pdf_extract.extractor.bbox import filter_contained	1✔
86	from build_a_long.pdf_extract.extractor.lego_page_elements import (	1✔
87	PageNumber,
88	PartCount,
89	PartsList,
90	StepNumber,
91	)
92	from build_a_long.pdf_extract.extractor.page_blocks import Blocks	1✔
93
94	logger = logging.getLogger(__name__)	1✔
95
96	# Pages with more blocks than this threshold will be skipped during classification.
97	# This avoids O(n²) algorithms (like duplicate detection) that become prohibitively
98	# slow on pages with thousands of vector drawings. Such pages are typically info
99	# pages where each character is a separate vector graphic.
100	# TODO: Add spatial indexing to handle high-block pages efficiently.
101	MAX_BLOCKS_PER_PAGE = 1000	1✔
102
103
104	# TODO require config, so we don't accidentally use default empty config
105	def classify_elements(	1✔
106	page: PageData, config: ClassifierConfig \| None = None
107	) -> ClassificationResult:
108	"""Classify and label elements on a single page using rule-based heuristics.
109
110	Args:
111	page: A single PageData object to classify.
112	config: Optional classifier configuration with font/page hints.
113	If None, uses default empty configuration (no hints).
114	For better classification accuracy, pass a config with
115	FontSizeHints computed from multiple pages of the same PDF.
116
117	Returns:
118	A ClassificationResult object containing the classification results.
119	"""
120	if config is None:	1✔
121	config = ClassifierConfig()	1✔
122	classifier = Classifier(config)	1✔
123
124	return classifier.classify(page)	1✔
125
126
127	def classify_pages(	1✔
128	pages: list[PageData], pages_for_hints: list[PageData] \| None = None
129	) -> BatchClassificationResult:
130	"""Classify and label elements across multiple pages using rule-based heuristics.
131
132	This function performs a three-phase process:
133	1. Filtering phase: Mark duplicate/similar blocks as removed on each page
134	2. Analysis phase: Build font size hints from text properties (excluding
135	removed blocks)
136	3. Classification phase: Use hints to guide element classification
137
138	Args:
139	pages: A list of PageData objects to classify.
140	pages_for_hints: Optional list of pages to use for generating font/page hints.
141	If None, uses `pages`. This allows generating hints from all pages
142	while only classifying a subset (e.g., when using --pages filter).
143
144	Returns:
145	BatchClassificationResult containing per-page results and global histogram
146	"""
147
148	# TODO There is a bunch of duplication in here between hints and non-hints. Refactor
149
150	# Use all pages for hint generation if provided, otherwise use selected pages
151	hint_pages = pages_for_hints if pages_for_hints is not None else pages	1✔
152
153	# Phase 1: Filter duplicate blocks on each page and track removals
154	# Skip pages with too many blocks to avoid O(n²) performance issues
155	removed_blocks_per_page: list[dict[Blocks, RemovalReason]] = []	1✔
156	skipped_pages: set[int] = set() # Track page numbers that are skipped	1✔
157
158	for page_data in pages:	1✔
159	# Skip pages with too many blocks - these are likely info/inventory pages
160	# with vectorized text that cause O(n²) algorithms to be very slow
161	if len(page_data.blocks) > MAX_BLOCKS_PER_PAGE:	1✔
162	logger.debug(	1✔
163	f"Page {page_data.page_number}: skipping classification "
164	f"({len(page_data.blocks)} blocks exceeds threshold of "
165	f"{MAX_BLOCKS_PER_PAGE})"
166	)
167	skipped_pages.add(page_data.page_number)	1✔
168	removed_blocks_per_page.append({})	1✔
169	continue	1✔
170
171	kept_blocks = page_data.blocks	1✔
172
173	# Filter overlapping text blocks (e.g., "4" and "43" at same origin)
174	kept_blocks, text_removed = filter_overlapping_text_blocks(kept_blocks)	1✔
175
176	# Filter duplicate image/drawing blocks based on IOU
177	kept_blocks, bbox_removed = filter_duplicate_blocks(kept_blocks)	1✔
178
179	# Combine all removal mappings into a single dict for this page
180	combined_removed_mapping = {	1✔
181	**text_removed,
182	**bbox_removed,
183	}
184
185	logger.debug(	1✔
186	f"Page {page_data.page_number}: "
187	f"filtered {len(text_removed)} overlapping text, "
188	f"{len(bbox_removed)} duplicate bbox blocks"
189	)
190
191	removed_blocks_per_page.append(combined_removed_mapping)	1✔
192
193	# Phase 2: Extract font size hints from hint pages (excluding removed blocks)
194	# Build pages with non-removed blocks for hint extraction and histogram
195
196	# Filter duplicates from hint pages (may be different from pages to classify)
197	hint_pages_without_duplicates = []	1✔
198	for page_data in hint_pages:	1✔
199	# Skip high-block pages for hints too (same threshold)
200	if len(page_data.blocks) > MAX_BLOCKS_PER_PAGE:	1✔
201	continue	1✔
202
203	# TODO We are re-filtering duplicates here; optimize by changing the API
204	# to accept one list of PageData, and seperate by page_numbers.
205	kept_blocks = page_data.blocks	1✔
206	kept_blocks, _ = filter_overlapping_text_blocks(kept_blocks)	1✔
207	kept_blocks, _ = filter_duplicate_blocks(kept_blocks)	1✔
208
209	hint_pages_without_duplicates.append(	1✔
210	PageData(
211	page_number=page_data.page_number,
212	bbox=page_data.bbox,
213	blocks=kept_blocks,
214	)
215	)
216
217	# Build pages without duplicates for classification
218	pages_without_duplicates = []	1✔
219	for page_data, removed_mapping in zip(pages, removed_blocks_per_page, strict=True):	1✔
220	# We need to filter blocks that were removed by ANY filter
221	non_removed_blocks = [	1✔
222	block for block in page_data.blocks if block not in removed_mapping
223	]
224	pages_without_duplicates.append(	1✔
225	PageData(
226	page_number=page_data.page_number,
227	bbox=page_data.bbox,
228	blocks=non_removed_blocks,
229	)
230	)
231
232	# Generate hints from hint pages, histogram from pages to classify
233	font_size_hints = FontSizeHints.from_pages(hint_pages_without_duplicates)	1✔
234	page_hints = PageHintCollection.from_pages(hint_pages_without_duplicates)	1✔
235	histogram = TextHistogram.from_pages(pages_without_duplicates)	1✔
236
237	# Phase 3: Classify using the hints (on pages without duplicates)
238	config = ClassifierConfig(font_size_hints=font_size_hints, page_hints=page_hints)	1✔
239	classifier = Classifier(config)	1✔
240
241	results = []	1✔
242	for page_data, page_without_duplicates, removed_mapping in zip(	1✔
243	pages, pages_without_duplicates, removed_blocks_per_page, strict=True
244	):
245	# Handle skipped pages
246	if page_data.page_number in skipped_pages:	1✔
247	result = ClassificationResult(	1✔
248	page_data=page_data,
249	skipped_reason=(
250	f"Page has {len(page_data.blocks)} blocks, which exceeds "
251	f"the threshold of {MAX_BLOCKS_PER_PAGE}. This is likely an "
252	f"info/inventory page with vectorized text."
253	),
254	)
255	results.append(result)	1✔
256	continue	1✔
257
258	# Classify using only non-removed blocks
259	result = classifier.classify(page_without_duplicates)	1✔
260
261	# Update result to use original page_data (with all blocks)
262	result.page_data = page_data	1✔
263
264	# Mark removed blocks
265	for removed_block, removal_reason in removed_mapping.items():	1✔
266	result.mark_removed(removed_block, removal_reason)	1✔
267
268	results.append(result)	1✔
269
270	return BatchClassificationResult(results=results, histogram=histogram)	1✔
271
272
273	type Classifiers = (	1✔
274	PageNumberClassifier
275	\| ProgressBarClassifier
276	\| ProgressBarIndicatorClassifier
277	\| PreviewClassifier
278	\| BackgroundClassifier
279	\| DividerClassifier
280	\| BagNumberClassifier
281	\| PartCountClassifier
282	\| PartNumberClassifier
283	\| StepNumberClassifier
284	\| StepCountClassifier
285	\| PieceLengthClassifier
286	\| PartsClassifier
287	\| PartsListClassifier
288	\| PartsImageClassifier
289	\| ShineClassifier
290	\| NewBagClassifier
291	\| DiagramClassifier
292	\| ArrowClassifier
293	\| SubAssemblyClassifier
294	\| StepClassifier
295	\| TriviaTextClassifier
296	\| PageClassifier
297	)
298
299
300	class Classifier:	1✔
301	"""
302	Performs a single run of classification based on rules, configuration, and hints.
303	This class should be stateless.
304	"""
305
306	def __init__(self, config: ClassifierConfig):	1✔
307	self.config = config	1✔
308	# Sort classifiers topologically based on their dependencies
309	self.classifiers = topological_sort(	1✔
310	[
311	PageNumberClassifier(config=config),
312	ProgressBarIndicatorClassifier(config=config),
313	ProgressBarClassifier(config=config),
314	BackgroundClassifier(config=config),
315	DividerClassifier(config=config),
316	BagNumberClassifier(config=config),
317	PartCountClassifier(config=config),
318	PartNumberClassifier(config=config),
319	StepNumberClassifier(config=config),
320	StepCountClassifier(config=config),
321	PieceLengthClassifier(config=config),
322	PartsClassifier(config=config),
323	PartsListClassifier(config=config),
324	DiagramClassifier(config=config),
325	RotationSymbolClassifier(config=config),
326	ArrowClassifier(config=config),
327	PartsImageClassifier(config=config),
328	ShineClassifier(config=config),
329	NewBagClassifier(config=config),
330	PreviewClassifier(config=config),
331	SubAssemblyClassifier(config=config),
332	StepClassifier(config=config),
333	TriviaTextClassifier(config=config),
334	PageClassifier(config=config),
335	]
336	)
337
338	def classify(self, page_data: PageData) -> ClassificationResult:	1✔
339	"""
340	Runs the classification logic and returns a result.
341	It does NOT modify page_data directly.
342
343	The classification process runs in three phases:
344	1. Score all classifiers (bottom-up) - auto-registers classifiers
345	2. Construct final elements (top-down starting from Page)
346	"""
347	result = ClassificationResult(page_data=page_data)	1✔
348
349	logger.debug(f"Starting classification for page {page_data.page_number}")	1✔
350
351	# 1. Score all classifiers (Bottom-Up)
352	# Note: score() automatically registers each classifier for its output labels
353	for classifier in self.classifiers:	1✔
354	classifier.score(result)	1✔
355
356	# 2. Construct (Top-Down)
357	# Find the PageClassifier to start the construction process
358	page_classifier = next(	1✔
359	c for c in self.classifiers if isinstance(c, PageClassifier)
360	)
361	page_classifier.build_all(result)	1✔
362
363	# TODO Do we actualy ever add warnings?
364	warnings = self._log_post_classification_warnings(page_data, result)	1✔
365	for warning in warnings:	1✔
366	result.add_warning(warning)	1✔
367
368	return result	1✔
369
370	def _log_post_classification_warnings(	1✔
371	self, page_data: PageData, result: ClassificationResult
372	) -> list[str]:
373	warnings = []	1✔
374
375	# Check if there's a page number
376	page_numbers = result.get_winners_by_score("page_number", PageNumber)	1✔
377	if not page_numbers:	1✔
378	warnings.append(f"Page {page_data.page_number}: missing page number")	1✔
379
380	# Get elements by label
381	parts_lists = result.get_winners_by_score("parts_list", PartsList)	1✔
382	part_counts = result.get_winners_by_score("part_count", PartCount)	1✔
383
384	for pl in parts_lists:	1✔
385	inside_counts = filter_contained(part_counts, pl.bbox)	1✔
386	if not inside_counts:	1✔
UNCOV 387	warnings.append(	×
388	f"Page {page_data.page_number}: parts list at {pl.bbox} "
389	f"contains no part counts"
390	)
391
392	steps = result.get_winners_by_score("step_number", StepNumber)	1✔
393	ABOVE_EPS = 2.0	1✔
394	for step in steps:	1✔
395	sb = step.bbox	1✔
396	above = [pl for pl in parts_lists if pl.bbox.y1 <= sb.y0 + ABOVE_EPS]	1✔
397	if not above:	1✔
398	warnings.append(	1✔
399	f"Page {page_data.page_number}: step number '{step.value}' "
400	f"at {sb} has no parts list above it"
401	)
402	return warnings	1✔

bramp / build-along / 20086551557

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous