docvortex 0.2.1__py3-none-any.whl

docvortex/analyzers/native/pdf/titles.py

@@ -0,0 +1,75 @@
+"""保留既有入口的 Flash PDF 门面，内部实现按职责显式组织。"""
+
+from .title_analysis.body_profile import (
+    _infer_document_body_profile as _infer_document_body_profile,
+    _document_font_is_regular as _document_font_is_regular,
+    _infer_lane_body_profile as _infer_lane_body_profile,
+    _line_uses_document_regular_font as _line_uses_document_regular_font,
+)
+from .title_analysis.common import (
+    _NUMBERED_SECTION_TITLE_RE as _NUMBERED_SECTION_TITLE_RE,
+    _SECTION_NUMBER_ONLY_RE as _SECTION_NUMBER_ONLY_RE,
+    _SECTION_TITLE_TERMINAL_RE as _SECTION_TITLE_TERMINAL_RE,
+    _UNNUMBERED_SECTION_HEADING_RE as _UNNUMBERED_SECTION_HEADING_RE,
+    _build_physical_title_gap_map as _build_physical_title_gap_map,
+    _line_inside_visual_container as _line_inside_visual_container,
+    _line_near_visual_container as _line_near_visual_container,
+)
+from .title_analysis.document_profile import _infer_document_title_profile as _infer_document_title_profile
+from .title_analysis.lane_titles import (
+    _classify_paragraph_titles_in_lane as _classify_paragraph_titles_in_lane,
+    _protect_front_matter_title_types as _protect_front_matter_title_types,
+    _visual_row_has_body_style_sibling as _visual_row_has_body_style_sibling,
+    _is_near_full_mixed_inline_row as _is_near_full_mixed_inline_row,
+    _continues_local_body_row as _continues_local_body_row,
+    _is_continuous_field_row as _is_continuous_field_row,
+    _is_full_width_inline_heading as _is_full_width_inline_heading,
+    _has_following_body_row as _has_following_body_row,
+    _has_following_compact_text_section as _has_following_compact_text_section,
+    _unify_visual_row_title_types as _unify_visual_row_title_types,
+    _infer_front_matter_boundary as _infer_front_matter_boundary,
+    _normalized_title_gap as _normalized_title_gap,
+    _expand_paragraph_title_neighbors as _expand_paragraph_title_neighbors,
+)
+from .title_analysis.page_titles import (
+    _classify_page_titles as _classify_page_titles,
+    _demote_non_structural_anomaly_titles as _demote_non_structural_anomaly_titles,
+    _find_repeated_grid_title_suppressions as _find_repeated_grid_title_suppressions,
+    _find_container_visual_row_title_suppressions as _find_container_visual_row_title_suppressions,
+    _classify_document_title as _classify_document_title,
+    _expand_document_title_across_lanes as _expand_document_title_across_lanes,
+    _classify_additional_document_title_bands as _classify_additional_document_title_bands,
+    _classify_cross_lane_centered_section_titles as _classify_cross_lane_centered_section_titles,
+    _demote_cross_lane_body_continuation_titles as _demote_cross_lane_body_continuation_titles,
+    _classify_cross_lane_emphasized_section_titles as _classify_cross_lane_emphasized_section_titles,
+    _is_wide_leading_title_continuation as _is_wide_leading_title_continuation,
+    _expand_cross_lane_paragraph_title_neighbors as _expand_cross_lane_paragraph_title_neighbors,
+    _demote_hanging_multiline_text_titles as _demote_hanging_multiline_text_titles,
+    _demote_visual_container_caption_titles as _demote_visual_container_caption_titles,
+    _demote_sentence_tail_titles as _demote_sentence_tail_titles,
+    _document_title_fonts_compatible as _document_title_fonts_compatible,
+    _document_title_uses_page_fallback as _document_title_uses_page_fallback,
+)
+from .title_analysis.prototype import (
+    _title_profile_seed_matches_cluster as _title_profile_seed_matches_cluster,
+    _title_profile_alignment as _title_profile_alignment,
+    _matching_document_title_prototype as _matching_document_title_prototype,
+    _line_conflicts_document_title_profile as _line_conflicts_document_title_profile,
+    _title_font_families_compatible as _title_font_families_compatible,
+)
+from .title_analysis.structural import (
+    _normalized_section_title_text as _normalized_section_title_text,
+    _is_plausible_section_number as _is_plausible_section_number,
+    _section_title_has_body_followers as _section_title_has_body_followers,
+    _classify_explicit_section_titles as _classify_explicit_section_titles,
+    _classify_document_structural_titles as _classify_document_structural_titles,
+    _promote_noninitial_document_title_band as _promote_noninitial_document_title_band,
+    _canonical_title_style_key as _canonical_title_style_key,
+    _classify_document_structural_title_candidates as _classify_document_structural_title_candidates,
+    _collect_legacy_paragraph_title_sources as _collect_legacy_paragraph_title_sources,
+    _classify_inline_typography_reset_titles as _classify_inline_typography_reset_titles,
+    _classify_body_height_section_titles as _classify_body_height_section_titles,
+    _body_height_section_followers as _body_height_section_followers,
+)
+
+__all__ = []

docvortex/analyzers/native/pdf/typography.py

@@ -0,0 +1,19 @@
+"""提供原生文本与布局共享的字体族归一化。"""
+
+from __future__ import annotations
+
+import re
+
+
+def _normalized_font_family(
+    signature: tuple[str, int] | None,
+) -> str | None:
+    """移除 PDF 字体子集前缀并归一化字体族名称，供几何续行作软兼容判断。"""
+
+    if signature is None:
+        return None
+    name = re.sub(r"^[A-Z]{6}\+", "", signature[0])
+    return re.sub(r"[\s_-]+", "", name).casefold() or None
+
+
+__all__ = ["_normalized_font_family"]