19751674072

Committed 28 Nov 2025 01:40AM UTC coverage: 89.023% (-0.8%) from 89.847%

Build # 19751674072

Build Type

push

github

Committed by

bramp

Commit Message

refactor(classifier): reorganize text modules and break circular dependencies

Major Changes:
- Created text/ subdirectory for text-related classifier modules
- Moved text_histogram, text_extractors, font_size_hints to text/
- Created constants.py to resolve circular dependency issue

Module Organization:
- classifier/text/__init__.py: Package exports for text modules
- classifier/text/text_histogram.py: TextHistogram class
- classifier/text/text_extractors.py: Text extraction functions
- classifier/text/font_size_hints.py: FontSizeHints class

Circular Dependency Resolution:
- Created classifier/constants.py with CATALOG_ELEMENT_ID_THRESHOLD
- Removed ClassVar from PageHintCollection
- Updated font_size_hints.py and page_hint_collection.py to import from constants
- Fixed package-level circular import by importing TextHistogram directly from module
- Added TODO to consider moving constant to ClassifierConfig

Bug Fixes:
- Fixed DrawableItem frozen model issue in drawing.py
- Create new instances with depth instead of mutating frozen objects

Import Updates:
- Updated all imports across ~15 files to use new module paths
- Updated classifier/__init__.py to re-export text module classes

Tests:
- All tests passing (42/42 test files)
- Type checking passes
- Code formatted with ruff

Run Details

32 of 32 new or added lines in 23 files covered. (100.0%)

180 existing lines in 19 files now uncovered.

7429 of 8345 relevant lines covered (89.02%)

0.89 hits per line

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

98.85

/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
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.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.progress_bar_classifier import (
    ProgressBarClassifier,
)
from build_a_long.pdf_extract.classifier.parts import (
    PartCountClassifier,
    PartNumberClassifier,
    PartsClassifier,
    PartsImageClassifier,
    PartsListClassifier,
    PieceLengthClassifier,
)
from build_a_long.pdf_extract.classifier.removal_reason import RemovalReason
from build_a_long.pdf_extract.classifier.steps import (
    DiagramClassifier,
    StepClassifier,
    StepNumberClassifier,
)
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.lego_page_elements import (
    PageNumber,
    PartCount,
    PartsList,
    StepNumber,
)
from build_a_long.pdf_extract.extractor.page_blocks import Blocks

logger = logging.getLogger(__name__)


def classify_elements(page: PageData) -> ClassificationResult:
    """Classify and label elements on a single page using rule-based heuristics.

    Args:
        page: A single PageData object to classify.

    Returns:
        A ClassificationResult object containing the classification results.
    """
    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
    """
    # 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
    duplicate_removals: list[dict[Blocks, Blocks]] = []
    for page_data in pages:
        # Get blocks to keep and mapping of removed blocks
        kept_blocks, removed_mapping = filter_duplicate_blocks(page_data.blocks)

        logger.debug(
            f"Page {page_data.page_number}: "
            f"filtered {len(removed_mapping)} duplicate blocks"
        )

        duplicate_removals.append(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:
        # 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, _ = filter_duplicate_blocks(page_data.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, duplicate_removals, strict=True):
        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, duplicate_removals, strict=True
    ):
        # 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 duplicate blocks as removed
        for removed_block, kept_block in removed_mapping.items():
            result.mark_removed(
                removed_block,
                RemovalReason(reason_type="duplicate_bbox", target_block=kept_block),
            )

        results.append(result)

    return BatchClassificationResult(results=results, histogram=histogram)


type Classifiers = (
    PageNumberClassifier
    | ProgressBarClassifier
    | BagNumberClassifier
    | PartCountClassifier
    | PartNumberClassifier
    | StepNumberClassifier
    | PieceLengthClassifier
    | PartsClassifier
    | PartsListClassifier
    | PartsImageClassifier
    | NewBagClassifier
    | DiagramClassifier
    | StepClassifier
    | 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),
                ProgressBarClassifier(config),
                BagNumberClassifier(config),
                PartCountClassifier(config),
                PartNumberClassifier(config),
                StepNumberClassifier(config),
                PieceLengthClassifier(config),
                PartsClassifier(config),
                PartsListClassifier(config),
                PartsImageClassifier(config),
                NewBagClassifier(config),
                DiagramClassifier(config),
                StepClassifier(config),
                PageClassifier(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 = [t for t in part_counts if t.bbox.fully_inside(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 filter_duplicate_blocks	1✔
31	from build_a_long.pdf_extract.classifier.classification_result import (	1✔
32	ClassificationResult,
33	)
34	from build_a_long.pdf_extract.classifier.classifier_config import ClassifierConfig	1✔
35	from build_a_long.pdf_extract.classifier.pages import (	1✔
36	PageHintCollection,
37	)
38	from build_a_long.pdf_extract.classifier.pages.page_classifier import PageClassifier	1✔
39	from build_a_long.pdf_extract.classifier.pages.page_number_classifier import (	1✔
40	PageNumberClassifier,
41	)
42	from build_a_long.pdf_extract.classifier.pages.progress_bar_classifier import (	1✔
43	ProgressBarClassifier,
44	)
45	from build_a_long.pdf_extract.classifier.parts import (	1✔
46	PartCountClassifier,
47	PartNumberClassifier,
48	PartsClassifier,
49	PartsImageClassifier,
50	PartsListClassifier,
51	PieceLengthClassifier,
52	)
53	from build_a_long.pdf_extract.classifier.removal_reason import RemovalReason	1✔
54	from build_a_long.pdf_extract.classifier.steps import (	1✔
55	DiagramClassifier,
56	StepClassifier,
57	StepNumberClassifier,
58	)
59	from build_a_long.pdf_extract.classifier.text import FontSizeHints, TextHistogram	1✔
60	from build_a_long.pdf_extract.classifier.topological_sort import topological_sort	1✔
61	from build_a_long.pdf_extract.extractor import PageData	1✔
62	from build_a_long.pdf_extract.extractor.lego_page_elements import (	1✔
63	PageNumber,
64	PartCount,
65	PartsList,
66	StepNumber,
67	)
68	from build_a_long.pdf_extract.extractor.page_blocks import Blocks	1✔
69
70	logger = logging.getLogger(__name__)	1✔
71
72
73	def classify_elements(page: PageData) -> ClassificationResult:	1✔
74	"""Classify and label elements on a single page using rule-based heuristics.
75
76	Args:
77	page: A single PageData object to classify.
78
79	Returns:
80	A ClassificationResult object containing the classification results.
81	"""
82	config = ClassifierConfig()	1✔
83	classifier = Classifier(config)	1✔
84
85	return classifier.classify(page)	1✔
86
87
88	def classify_pages(	1✔
89	pages: list[PageData], pages_for_hints: list[PageData] \| None = None
90	) -> BatchClassificationResult:
91	"""Classify and label elements across multiple pages using rule-based heuristics.
92
93	This function performs a three-phase process:
94	1. Filtering phase: Mark duplicate/similar blocks as removed on each page
95	2. Analysis phase: Build font size hints from text properties (excluding
96	removed blocks)
97	3. Classification phase: Use hints to guide element classification
98
99	Args:
100	pages: A list of PageData objects to classify.
101	pages_for_hints: Optional list of pages to use for generating font/page hints.
102	If None, uses `pages`. This allows generating hints from all pages
103	while only classifying a subset (e.g., when using --pages filter).
104
105	Returns:
106	BatchClassificationResult containing per-page results and global histogram
107	"""
108	# Use all pages for hint generation if provided, otherwise use selected pages
109	hint_pages = pages_for_hints if pages_for_hints is not None else pages	1✔
110
111	# Phase 1: Filter duplicate blocks on each page and track removals
112	duplicate_removals: list[dict[Blocks, Blocks]] = []	1✔
113	for page_data in pages:	1✔
114	# Get blocks to keep and mapping of removed blocks
115	kept_blocks, removed_mapping = filter_duplicate_blocks(page_data.blocks)	1✔
116
117	logger.debug(	1✔
118	f"Page {page_data.page_number}: "
119	f"filtered {len(removed_mapping)} duplicate blocks"
120	)
121
122	duplicate_removals.append(removed_mapping)	1✔
123
124	# Phase 2: Extract font size hints from hint pages (excluding removed blocks)
125	# Build pages with non-removed blocks for hint extraction and histogram
126
127	# Filter duplicates from hint pages (may be different from pages to classify)
128	hint_pages_without_duplicates = []	1✔
129	for page_data in hint_pages:	1✔
130	# TODO We are re-filtering duplicates here; optimize by changing the API
131	# to accept one list of PageData, and seperate by page_numbers.
132	kept_blocks, _ = filter_duplicate_blocks(page_data.blocks)	1✔
133	hint_pages_without_duplicates.append(	1✔
134	PageData(
135	page_number=page_data.page_number,
136	bbox=page_data.bbox,
137	blocks=kept_blocks,
138	)
139	)
140
141	# Build pages without duplicates for classification
142	pages_without_duplicates = []	1✔
143	for page_data, removed_mapping in zip(pages, duplicate_removals, strict=True):	1✔
144	non_removed_blocks = [	1✔
145	block for block in page_data.blocks if block not in removed_mapping
146	]
147	pages_without_duplicates.append(	1✔
148	PageData(
149	page_number=page_data.page_number,
150	bbox=page_data.bbox,
151	blocks=non_removed_blocks,
152	)
153	)
154
155	# Generate hints from hint pages, histogram from pages to classify
156	font_size_hints = FontSizeHints.from_pages(hint_pages_without_duplicates)	1✔
157	page_hints = PageHintCollection.from_pages(hint_pages_without_duplicates)	1✔
158	histogram = TextHistogram.from_pages(pages_without_duplicates)	1✔
159
160	# Phase 3: Classify using the hints (on pages without duplicates)
161	config = ClassifierConfig(font_size_hints=font_size_hints, page_hints=page_hints)	1✔
162	classifier = Classifier(config)	1✔
163
164	results = []	1✔
165	for page_data, page_without_duplicates, removed_mapping in zip(	1✔
166	pages, pages_without_duplicates, duplicate_removals, strict=True
167	):
168	# Classify using only non-removed blocks
169	result = classifier.classify(page_without_duplicates)	1✔
170
171	# Update result to use original page_data (with all blocks)
172	result.page_data = page_data	1✔
173
174	# Mark duplicate blocks as removed
175	for removed_block, kept_block in removed_mapping.items():	1✔
176	result.mark_removed(	1✔
177	removed_block,
178	RemovalReason(reason_type="duplicate_bbox", target_block=kept_block),
179	)
180
181	results.append(result)	1✔
182
183	return BatchClassificationResult(results=results, histogram=histogram)	1✔
184
185
186	type Classifiers = (	1✔
187	PageNumberClassifier
188	\| ProgressBarClassifier
189	\| BagNumberClassifier
190	\| PartCountClassifier
191	\| PartNumberClassifier
192	\| StepNumberClassifier
193	\| PieceLengthClassifier
194	\| PartsClassifier
195	\| PartsListClassifier
196	\| PartsImageClassifier
197	\| NewBagClassifier
198	\| DiagramClassifier
199	\| StepClassifier
200	\| PageClassifier
201	)
202
203
204	class Classifier:	1✔
205	"""
206	Performs a single run of classification based on rules, configuration, and hints.
207	This class should be stateless.
208	"""
209
210	def __init__(self, config: ClassifierConfig):	1✔
211	self.config = config	1✔
212	# Sort classifiers topologically based on their dependencies
213	self.classifiers = topological_sort(	1✔
214	[
215	PageNumberClassifier(config),
216	ProgressBarClassifier(config),
217	BagNumberClassifier(config),
218	PartCountClassifier(config),
219	PartNumberClassifier(config),
220	StepNumberClassifier(config),
221	PieceLengthClassifier(config),
222	PartsClassifier(config),
223	PartsListClassifier(config),
224	PartsImageClassifier(config),
225	NewBagClassifier(config),
226	DiagramClassifier(config),
227	StepClassifier(config),
228	PageClassifier(config),
229	]
230	)
231
232	def classify(self, page_data: PageData) -> ClassificationResult:	1✔
233	"""
234	Runs the classification logic and returns a result.
235	It does NOT modify page_data directly.
236
237	The classification process runs in three phases:
238	1. Score all classifiers (bottom-up) - auto-registers classifiers
239	2. Construct final elements (top-down starting from Page)
240	"""
241	result = ClassificationResult(page_data=page_data)	1✔
242
243	logger.debug(f"Starting classification for page {page_data.page_number}")	1✔
244
245	# 1. Score all classifiers (Bottom-Up)
246	# Note: score() automatically registers each classifier for its output labels
247	for classifier in self.classifiers:	1✔
248	classifier.score(result)	1✔
249
250	# 2. Construct (Top-Down)
251	# Find the PageClassifier to start the construction process
252	page_classifier = next(	1✔
253	c for c in self.classifiers if isinstance(c, PageClassifier)
254	)
255	page_classifier.build_all(result)	1✔
256
257	# TODO Do we actualy ever add warnings?
258	warnings = self._log_post_classification_warnings(page_data, result)	1✔
259	for warning in warnings:	1✔
260	result.add_warning(warning)	1✔
261
262	return result	1✔
263
264	def _log_post_classification_warnings(	1✔
265	self, page_data: PageData, result: ClassificationResult
266	) -> list[str]:
267	warnings = []	1✔
268
269	# Check if there's a page number
270	page_numbers = result.get_winners_by_score("page_number", PageNumber)	1✔
271	if not page_numbers:	1✔
272	warnings.append(f"Page {page_data.page_number}: missing page number")	1✔
273
274	# Get elements by label
275	parts_lists = result.get_winners_by_score("parts_list", PartsList)	1✔
276	part_counts = result.get_winners_by_score("part_count", PartCount)	1✔
277
278	for pl in parts_lists:	1✔
279	inside_counts = [t for t in part_counts if t.bbox.fully_inside(pl.bbox)]	1✔
280	if not inside_counts:	1✔
UNCOV 281	warnings.append(	×
282	f"Page {page_data.page_number}: parts list at {pl.bbox} "
283	f"contains no part counts"
284	)
285
286	steps = result.get_winners_by_score("step_number", StepNumber)	1✔
287	ABOVE_EPS = 2.0	1✔
288	for step in steps:	1✔
289	sb = step.bbox	1✔
290	above = [pl for pl in parts_lists if pl.bbox.y1 <= sb.y0 + ABOVE_EPS]	1✔
291	if not above:	1✔
292	warnings.append(	1✔
293	f"Page {page_data.page_number}: step number '{step.value}' "
294	f"at {sb} has no parts list above it"
295	)
296	return warnings	1✔

bramp / build-along / 19751674072

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