Restructure: Move 47 cv_* files into ocr/ package

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

.claude/rules/loc-exceptions.txt

@@ -27,6 +27,7 @@
 
 # Algorithmic monolith — detect_column_geometry() allein 411 LOC, nicht weiter teilbar
 **/cv_layout_columns.py | owner=klausur | reason=detect_column_geometry ist eine einzelne 411-LOC Funktion (Whitespace-Gap-Analyse) | review=2026-10-01
+**/ocr/layout/columns.py | owner=klausur | reason=Same file moved to ocr/ package | review=2026-10-01
 
 # Two indivisible route handlers (~230 LOC each) that cannot be split further
 **/vocab_worksheet_compare_api.py | owner=klausur | reason=compare_ocr_methods (234 LOC) + analyze_grid (255 LOC), each a single cohesive handler | review=2026-10-01

klausur-service/backend/cv_box_detect.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/detect\/box_detect.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.detect.box_detect")

klausur-service/backend/cv_box_layout.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/detect\/box_layout.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.detect.box_layout")

klausur-service/backend/cv_cell_grid.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/cell_grid\/cell_grid.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.cell_grid.cell_grid")

klausur-service/backend/cv_cell_grid_build.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/cell_grid\/build.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.cell_grid.build")

klausur-service/backend/cv_cell_grid_helpers.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/cell_grid\/helpers.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.cell_grid.helpers")

klausur-service/backend/cv_cell_grid_legacy.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/cell_grid\/legacy.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.cell_grid.legacy")

klausur-service/backend/cv_cell_grid_merge.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/cell_grid\/merge.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.cell_grid.merge")

klausur-service/backend/cv_cell_grid_streaming.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/cell_grid\/streaming.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.cell_grid.streaming")

klausur-service/backend/cv_cell_grid_vocab.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/cell_grid\/vocab.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.cell_grid.vocab")

klausur-service/backend/cv_color_detect.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/detect\/color_detect.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.detect.color_detect")

klausur-service/backend/cv_doclayout_detect.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/detect\/doclayout_detect.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.detect.doclayout_detect")

klausur-service/backend/cv_graphic_detect.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/detect\/graphic_detect.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.detect.graphic_detect")

klausur-service/backend/cv_gutter_repair.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/gutter\/repair.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.gutter.repair")

klausur-service/backend/cv_gutter_repair_core.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/gutter\/core.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.gutter.core")

klausur-service/backend/cv_gutter_repair_grid.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/gutter\/grid.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.gutter.grid")

klausur-service/backend/cv_ipa_german.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/ipa_german.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.ipa_german")

klausur-service/backend/cv_layout.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/layout\/layout.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.layout.layout")

klausur-service/backend/cv_layout_analyze.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/layout\/analyze.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.layout.analyze")

klausur-service/backend/cv_layout_classify.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/layout\/classify.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.layout.classify")

klausur-service/backend/cv_layout_classify_position.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/layout\/classify_position.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.layout.classify_position")

klausur-service/backend/cv_layout_column_refine.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/layout\/column_refine.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.layout.column_refine")

klausur-service/backend/cv_layout_columns.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/layout\/columns.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.layout.columns")

klausur-service/backend/cv_layout_detection.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/layout\/detection.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.layout.detection")

klausur-service/backend/cv_layout_row_regularize.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/layout\/row_regularize.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.layout.row_regularize")

klausur-service/backend/cv_layout_rows.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/layout\/rows.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.layout.rows")

klausur-service/backend/cv_layout_scoring.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/layout\/scoring.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.layout.scoring")

klausur-service/backend/cv_ocr_cell_filter.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/engines\/cell_filter.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.engines.cell_filter")

klausur-service/backend/cv_ocr_cell_phonetics.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/engines\/cell_phonetics.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.engines.cell_phonetics")

klausur-service/backend/cv_ocr_engines.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/engines\/engines.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.engines.engines")

klausur-service/backend/cv_ocr_ipa_lookup.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/engines\/ipa_lookup.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.engines.ipa_lookup")

klausur-service/backend/cv_ocr_ipa_repair.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/engines\/ipa_repair.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.engines.ipa_repair")

klausur-service/backend/cv_ocr_vocab_postprocess.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/engines\/vocab_postprocess.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.engines.vocab_postprocess")

klausur-service/backend/cv_ocr_word_assembly.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/engines\/word_assembly.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.engines.word_assembly")

klausur-service/backend/cv_preprocessing.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/preprocessing\/preprocessing.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.preprocessing.preprocessing")

klausur-service/backend/cv_preprocessing_deskew.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/preprocessing\/deskew.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.preprocessing.deskew")

klausur-service/backend/cv_preprocessing_dewarp.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/preprocessing\/dewarp.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.preprocessing.dewarp")

klausur-service/backend/cv_review.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/review\/review.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.review.review")

klausur-service/backend/cv_review_llm.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/review\/llm.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.review.llm")

klausur-service/backend/cv_review_pipeline.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/review\/pipeline.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.review.pipeline")

klausur-service/backend/cv_review_spell.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/review\/spell.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.review.spell")

klausur-service/backend/cv_syllable_core.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/detect\/syllable\/core.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.detect.syllable.core")

klausur-service/backend/cv_syllable_detect.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/detect\/syllable\/detect.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.detect.syllable.detect")

klausur-service/backend/cv_syllable_merge.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/detect\/syllable\/merge.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.detect.syllable.merge")

klausur-service/backend/cv_vocab_pipeline.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/pipeline.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.pipeline")

klausur-service/backend/cv_vocab_types.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/types.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.types")

klausur-service/backend/cv_words_first.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/words_first.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.words_first")

klausur-service/backend/ocr/__init__.py

@@ -0,0 +1,9 @@
+"""
+OCR package — restructured from cv_* flat modules.
+
+Backward-compatible re-exports: consumers can still use
+``from cv_layout import ...`` etc. via the shim files in backend/.
+"""
+
+from .types import *  # noqa: F401,F403
+from .pipeline import *  # noqa: F401,F403

klausur-service/backend/ocr/cell_grid/__init__.py

@@ -0,0 +1,2 @@
+"""Cell-grid construction sub-package."""
+from .cell_grid import *  # noqa: F401,F403

klausur-service/backend/ocr/cell_grid/build.py

@@ -10,8 +10,8 @@ from typing import Any, Dict, List, Optional, Tuple
 
 import numpy as np
 
-from cv_vocab_types import PageRegion, RowGeometry
+from ..types import PageRegion, RowGeometry
-from cv_ocr_engines import (
+from ..engines.engines import (
     RAPIDOCR_AVAILABLE,
     _assign_row_words_to_columns,
     _clean_cell_text,
@@ -22,7 +22,7 @@ from cv_ocr_engines import (
     ocr_region_rapid,
     ocr_region_trocr,
 )
-from cv_cell_grid_helpers import (
+from .helpers import (
     _MIN_WORD_CONF,
     _ensure_minimum_crop_size,
     _heal_row_gaps,

klausur-service/backend/ocr/cell_grid/cell_grid.py

@@ -16,7 +16,7 @@ DATENSCHUTZ: Alle Verarbeitung erfolgt lokal.
 """
 
 # --- Helpers ---
-from cv_cell_grid_helpers import (  # noqa: F401
+from .helpers import (  # noqa: F401
     _MIN_WORD_CONF,
     _compute_cell_padding,
     _ensure_minimum_crop_size,
@@ -26,26 +26,26 @@ from cv_cell_grid_helpers import (  # noqa: F401
 )
 
 # --- v2 build (current default) ---
-from cv_cell_grid_build import (  # noqa: F401
+from .build import (  # noqa: F401
     _NARROW_COL_THRESHOLD_PCT,
     _ocr_cell_crop,
     build_cell_grid_v2,
 )
 
 # --- Legacy build (DEPRECATED) ---
-from cv_cell_grid_legacy import (  # noqa: F401
+from .legacy import (  # noqa: F401
     _ocr_single_cell,
     build_cell_grid,
 )
 
 # --- Streaming variants ---
-from cv_cell_grid_streaming import (  # noqa: F401
+from .streaming import (  # noqa: F401
     build_cell_grid_streaming,
     build_cell_grid_v2_streaming,
 )
 
 # --- Row merging ---
-from cv_cell_grid_merge import (  # noqa: F401
+from .merge import (  # noqa: F401
     _PHONETIC_ONLY_RE,
     _is_phonetic_only_text,
     _merge_continuation_rows,
@@ -54,7 +54,7 @@ from cv_cell_grid_merge import (  # noqa: F401
 )
 
 # --- Vocab extraction ---
-from cv_cell_grid_vocab import (  # noqa: F401
+from .vocab import (  # noqa: F401
     _cells_to_vocab_entries,
     build_word_grid,
 )

klausur-service/backend/ocr/cell_grid/helpers.py

@@ -13,7 +13,7 @@ from typing import List
 
 import numpy as np
 
-from cv_vocab_types import RowGeometry
+from ..types import RowGeometry
 
 logger = logging.getLogger(__name__)
 

klausur-service/backend/ocr/cell_grid/legacy.py

@@ -12,8 +12,8 @@ from typing import Any, Dict, List, Optional, Tuple
 
 import numpy as np
 
-from cv_vocab_types import PageRegion, RowGeometry
+from ..types import PageRegion, RowGeometry
-from cv_ocr_engines import (
+from ..engines.engines import (
     RAPIDOCR_AVAILABLE,
     _assign_row_words_to_columns,
     _clean_cell_text,
@@ -22,7 +22,7 @@ from cv_ocr_engines import (
     ocr_region_rapid,
     ocr_region_trocr,
 )
-from cv_cell_grid_helpers import (
+from .helpers import (
     _MIN_WORD_CONF,
     _compute_cell_padding,
     _ensure_minimum_crop_size,

klausur-service/backend/ocr/cell_grid/merge.py

@@ -11,7 +11,7 @@ import logging
 import re
 from typing import Any, Dict, List
 
-from cv_ocr_engines import _RE_ALPHA
+from ..engines.engines import _RE_ALPHA
 
 logger = logging.getLogger(__name__)
 

klausur-service/backend/ocr/cell_grid/streaming.py

@@ -13,17 +13,17 @@ from typing import Any, Dict, Generator, List, Optional, Tuple
 
 import numpy as np
 
-from cv_vocab_types import PageRegion, RowGeometry
+from ..types import PageRegion, RowGeometry
-from cv_ocr_engines import (
+from ..engines.engines import (
     RAPIDOCR_AVAILABLE,
     _assign_row_words_to_columns,
 )
-from cv_cell_grid_helpers import (
+from .helpers import (
     _heal_row_gaps,
     _is_artifact_row,
 )
-from cv_cell_grid_build import _ocr_cell_crop
+from .build import _ocr_cell_crop
-from cv_cell_grid_legacy import _ocr_single_cell
+from .legacy import _ocr_single_cell
 
 logger = logging.getLogger(__name__)
 

klausur-service/backend/ocr/cell_grid/vocab.py

@@ -10,13 +10,13 @@ DATENSCHUTZ: Alle Verarbeitung erfolgt lokal.
 import logging
 from typing import Any, Dict, List
 
-from cv_ocr_engines import (
+from ..engines.engines import (
     _attach_example_sentences,
     _fix_phonetic_brackets,
     _split_comma_entries,
 )
-from cv_cell_grid_legacy import build_cell_grid
+from .legacy import build_cell_grid
-from cv_cell_grid_merge import (
+from .merge import (
     _merge_continuation_rows,
     _merge_phonetic_continuation_rows,
     _merge_wrapped_rows,

klausur-service/backend/ocr/detect/__init__.py

@@ -0,0 +1,2 @@
+"""Detection sub-package (boxes, graphics, colors, syllables, doclayout)."""
+from .box_detect import *  # noqa: F401,F403

klausur-service/backend/ocr/detect/box_detect.py

@@ -21,7 +21,7 @@ from typing import List, Optional, Tuple
 import cv2
 import numpy as np
 
-from cv_vocab_types import DetectedBox, PageZone
+from ..types import DetectedBox, PageZone
 
 logger = logging.getLogger(__name__)
 

klausur-service/backend/ocr/detect/graphic_detect.py

@@ -127,7 +127,7 @@ def detect_graphic_elements(
     backend = os.environ.get("GRAPHIC_DETECT_BACKEND", "auto")
     if backend in ("doclayout", "auto"):
         try:
-            from cv_doclayout_detect import detect_layout_regions, is_doclayout_available
+            from .doclayout_detect import detect_layout_regions, is_doclayout_available
             if is_doclayout_available():
                 regions = detect_layout_regions(img_bgr)
                 if regions:

klausur-service/backend/ocr/detect/syllable/__init__.py

@@ -0,0 +1,2 @@
+"""Syllable detection sub-package."""
+from .detect import *  # noqa: F401,F403

klausur-service/backend/ocr/detect/syllable/detect.py

@@ -10,7 +10,7 @@ DATENSCHUTZ: Alle Verarbeitung erfolgt lokal.
 """
 
 # Core: init, validation, autocorrect
-from cv_syllable_core import (  # noqa: F401
+from .core import (  # noqa: F401
     _IPA_RE,
     _STOP_WORDS,
     _get_hyphenators,
@@ -23,7 +23,7 @@ from cv_syllable_core import (  # noqa: F401
 )
 
 # Merge: gap merging, syllabify, insert
-from cv_syllable_merge import (  # noqa: F401
+from .merge import (  # noqa: F401
     _try_merge_pipe_gaps,
     merge_word_gaps_in_zones,
     _try_merge_word_gaps,

klausur-service/backend/ocr/detect/syllable/merge.py

@@ -13,7 +13,7 @@ from typing import Any, Dict, List, Optional
 
 import numpy as np
 
-from cv_syllable_core import (
+from .core import (
     _get_hyphenators,
     _hyphenate_word,
     _IPA_RE,

klausur-service/backend/ocr/engines/__init__.py

@@ -0,0 +1,2 @@
+"""OCR engines sub-package."""
+from .engines import *  # noqa: F401,F403

klausur-service/backend/ocr/engines/cell_filter.py

@@ -16,7 +16,7 @@ from typing import Any, Dict, List, Optional
 
 import numpy as np
 
-from cv_vocab_types import PageRegion, RowGeometry
+from ..types import PageRegion, RowGeometry
 
 logger = logging.getLogger(__name__)
 

klausur-service/backend/ocr/engines/cell_phonetics.py

@@ -11,14 +11,14 @@ import logging
 import re
 from typing import Any, Dict, List
 
-from cv_vocab_types import IPA_AVAILABLE
+from ..types import IPA_AVAILABLE
 
-from cv_ocr_ipa_lookup import (
+from .ipa_lookup import (
     _insert_missing_ipa,
     _replace_phonetics_in_text,
     _text_has_garbled_ipa,
 )
-from cv_ocr_ipa_repair import (
+from .ipa_repair import (
     _has_non_dict_trailing,
     _insert_headword_ipa,
     _strip_post_bracket_garbled,

klausur-service/backend/ocr/engines/engines.py

@@ -24,7 +24,7 @@ from typing import Any, Dict, List, Optional, Tuple
 
 import numpy as np
 
-from cv_vocab_types import (
+from ..types import (
     IPA_AVAILABLE,
     PageRegion,
     RowGeometry,
@@ -47,7 +47,7 @@ except ImportError:
 
 # ── Re-exports from sub-modules (backward compatibility) ──────────────────
 
-from cv_ocr_word_assembly import (  # noqa: F401
+from .word_assembly import (  # noqa: F401
     _group_words_into_lines,
     _words_to_reading_order_lines,
     _rejoin_hyphenated,
@@ -55,7 +55,7 @@ from cv_ocr_word_assembly import (  # noqa: F401
     _words_to_spaced_text,
 )
 
-from cv_ocr_vocab_postprocess import (  # noqa: F401
+from .vocab_postprocess import (  # noqa: F401
     _CHAR_CONFUSION_RULES,
     _DE_INDICATORS_FOR_EN_I,
     _fix_character_confusion,
@@ -66,7 +66,7 @@ from cv_ocr_vocab_postprocess import (  # noqa: F401
     _attach_example_sentences,
 )
 
-from cv_ocr_ipa_lookup import (  # noqa: F401
+from .ipa_lookup import (  # noqa: F401
     _PHONETIC_BRACKET_RE,
     _IPA_CHARS,
     _MIN_WORD_CONF,
@@ -80,20 +80,20 @@ from cv_ocr_ipa_lookup import (  # noqa: F401
     _insert_missing_ipa,
 )
 
-from cv_ocr_ipa_repair import (  # noqa: F401
+from .ipa_repair import (  # noqa: F401
     _has_non_dict_trailing,
     _strip_post_bracket_garbled,
     fix_ipa_continuation_cell,
     _insert_headword_ipa,
 )
 
-from cv_ocr_cell_phonetics import (  # noqa: F401
+from .cell_phonetics import (  # noqa: F401
     fix_cell_phonetics,
     _has_ipa_gap,
     _sync_word_boxes_after_ipa_insert,
 )
 
-from cv_ocr_cell_filter import (  # noqa: F401
+from .cell_filter import (  # noqa: F401
     _RE_REAL_WORD,
     _RE_ALPHA,
     _COMMON_SHORT_WORDS,

klausur-service/backend/ocr/engines/ipa_lookup.py

@@ -23,7 +23,7 @@ import logging
 import re
 from typing import Any, Dict, List, Optional
 
-from cv_vocab_types import (
+from ..types import (
     IPA_AVAILABLE,
     _britfone_dict,
     _ipa_convert_american,

klausur-service/backend/ocr/engines/ipa_repair.py

@@ -16,8 +16,8 @@ import logging
 import re
 from typing import Any, Dict, List, Optional
 
-from cv_vocab_types import IPA_AVAILABLE
+from ..types import IPA_AVAILABLE
-from cv_ocr_ipa_lookup import (
+from .ipa_lookup import (
     _lookup_ipa,
     _GRAMMAR_BRACKET_WORDS,
 )

klausur-service/backend/ocr/gutter/__init__.py

@@ -0,0 +1,2 @@
+"""Gutter repair sub-package."""
+from .repair import *  # noqa: F401,F403

klausur-service/backend/ocr/gutter/grid.py

@@ -11,7 +11,7 @@ import logging
 import time
 from typing import Any, Dict, List, Tuple
 
-from cv_gutter_repair_core import (
+from .core import (
     _init_spellcheckers,
     _is_ipa_text,
     _is_known,

klausur-service/backend/ocr/gutter/repair.py

@@ -10,7 +10,7 @@ DATENSCHUTZ: Alle Verarbeitung erfolgt lokal.
 """
 
 # Core: spellchecker, data types, repair helpers
-from cv_gutter_repair_core import (  # noqa: F401
+from .core import (  # noqa: F401
     _init_spellcheckers,
     _is_known,
     _spell_candidates,
@@ -29,7 +29,7 @@ from cv_gutter_repair_core import (  # noqa: F401
 )
 
 # Grid: analysis and application
-from cv_gutter_repair_grid import (  # noqa: F401
+from .grid import (  # noqa: F401
     analyse_grid_for_gutter_repair,
     apply_gutter_suggestions,
 )

klausur-service/backend/ocr/ipa_german.py

@@ -26,7 +26,7 @@ def _lookup_ipa_de(word: str) -> Optional[str]:
 
     Returns IPA string or None if not found.
     """
-    from cv_vocab_types import _de_ipa_dict, _epitran_de, DE_IPA_AVAILABLE
+    from .types import _de_ipa_dict, _epitran_de, DE_IPA_AVAILABLE
 
     if not DE_IPA_AVAILABLE and _epitran_de is None:
         return None
@@ -109,7 +109,7 @@ def insert_german_ipa(
     Returns:
         Number of cells modified.
     """
-    from cv_vocab_types import DE_IPA_AVAILABLE, _epitran_de
+    from .types import DE_IPA_AVAILABLE, _epitran_de
 
     if not DE_IPA_AVAILABLE and _epitran_de is None:
         logger.warning("German IPA not available — skipping")

klausur-service/backend/ocr/layout/__init__.py

@@ -0,0 +1,2 @@
+"""Layout analysis sub-package."""
+from .layout import *  # noqa: F401,F403

klausur-service/backend/ocr/layout/analyze.py

@@ -13,8 +13,8 @@ from typing import List
 
 import numpy as np
 
-from cv_vocab_types import PageRegion
+from ..types import PageRegion
-from cv_layout_detection import _find_content_bounds
+from .detection import _find_content_bounds
 
 logger = logging.getLogger(__name__)
 
@@ -246,7 +246,7 @@ def analyze_layout(layout_img: np.ndarray, ocr_img: np.ndarray) -> List[PageRegi
 
     # Add header/footer info (gap-based detection with fallback)
     # Lazy import to avoid circular dependency with cv_layout.py
-    from cv_layout_detection import _add_header_footer
+    from .detection import _add_header_footer
     _add_header_footer(regions, top_y, bottom_y, w, h, inv=inv)
 
     top_region = next((r.type for r in regions if r.type in ('header', 'margin_top')), 'none')

klausur-service/backend/ocr/layout/classify.py

@@ -11,16 +11,16 @@ from typing import Dict, List, Optional
 
 import numpy as np
 
-from cv_vocab_types import ColumnGeometry, PageRegion
+from ..types import ColumnGeometry, PageRegion
 
-from cv_layout_scoring import (
+from .scoring import (
     _score_language,
     _score_role,
     _score_dictionary_signals,
     _classify_dictionary_columns,
 )
 
-from cv_layout_classify_position import (
+from .classify_position import (
     _classify_by_position_enhanced,
     _classify_by_position_fallback,
 )
@@ -211,7 +211,7 @@ def classify_column_types(geometries: List[ColumnGeometry],
     # _add_header_footer lives in cv_layout (avoids circular import at module
     # level).  Lazy-import here so the module can be tested independently when
     # cv_layout hasn't been modified yet.
-    from cv_layout_detection import _add_header_footer  # noqa: E402
+    from .detection import _add_header_footer  # noqa: E402
 
     content_h = bottom_y - top_y
 

klausur-service/backend/ocr/layout/classify_position.py

@@ -11,7 +11,7 @@ Extracted from cv_layout_classify.py during file-size split.
 import logging
 from typing import Dict, List, Optional
 
-from cv_vocab_types import ColumnGeometry, PageRegion
+from ..types import ColumnGeometry, PageRegion
 
 logger = logging.getLogger(__name__)
 

klausur-service/backend/ocr/layout/column_refine.py

@@ -16,7 +16,7 @@ from typing import Dict, List, Optional, Tuple
 
 import numpy as np
 
-from cv_vocab_types import ColumnGeometry
+from ..types import ColumnGeometry
 
 logger = logging.getLogger(__name__)
 

klausur-service/backend/ocr/layout/columns.py

@@ -19,8 +19,8 @@ from typing import Dict, List, Optional, Tuple
 
 import numpy as np
 
-from cv_vocab_types import ColumnGeometry
+from ..types import ColumnGeometry
-from cv_layout_detection import _find_content_bounds
+from .detection import _find_content_bounds
 
 logger = logging.getLogger(__name__)
 

klausur-service/backend/ocr/layout/detection.py

@@ -13,7 +13,7 @@ from typing import List, Optional, Tuple
 
 import numpy as np
 
-from cv_vocab_types import (
+from ..types import (
     DocumentTypeResult,
     PageRegion,
 )

klausur-service/backend/ocr/layout/layout.py

@@ -21,14 +21,14 @@ from typing import Any, Dict, List, Optional, Tuple
 
 import numpy as np
 
-from cv_vocab_types import ColumnGeometry, DetectedBox, PageRegion
+from ..types import ColumnGeometry, DetectedBox, PageRegion
 
 logger = logging.getLogger(__name__)
 
 
 # ── Re-exports (backward compatibility) ───────────────────────────────────
 
-from cv_layout_detection import (  # noqa: F401
+from .detection import (  # noqa: F401
     detect_document_type,
     create_ocr_image,
     create_layout_image,
@@ -39,46 +39,46 @@ from cv_layout_detection import (  # noqa: F401
     _add_header_footer,
 )
 
-from cv_layout_analyze import (  # noqa: F401
+from .analyze import (  # noqa: F401
     analyze_layout,
 )
 
-from cv_layout_columns import (  # noqa: F401
+from .columns import (  # noqa: F401
     detect_column_geometry,
     _detect_columns_by_clustering,
     _build_geometries_from_starts,
 )
 
-from cv_layout_column_refine import (  # noqa: F401
+from .column_refine import (  # noqa: F401
     _detect_sub_columns,
     _split_broad_columns,
     expand_narrow_columns,
 )
 
-from cv_layout_rows import (  # noqa: F401
+from .rows import (  # noqa: F401
     detect_row_geometry,
     _build_rows_from_word_grouping,
 )
 
-from cv_layout_row_regularize import (  # noqa: F401
+from .row_regularize import (  # noqa: F401
     _regularize_row_grid,
 )
 
-from cv_layout_scoring import (  # noqa: F401
+from .scoring import (  # noqa: F401
     _score_language,
     _score_role,
     _score_dictionary_signals,
     _classify_dictionary_columns,
 )
 
-from cv_layout_classify import (  # noqa: F401
+from .classify import (  # noqa: F401
     _build_margin_regions,
     positional_column_regions,
     classify_column_types,
     _classify_by_content,
 )
 
-from cv_layout_classify_position import (  # noqa: F401
+from .classify_position import (  # noqa: F401
     _classify_by_position_enhanced,
     _classify_by_position_fallback,
 )
@@ -143,7 +143,7 @@ def detect_column_geometry_zoned(
        per content zone on the corresponding sub-image.
     4. If no boxes: delegates entirely to detect_column_geometry().
     """
-    from cv_box_detect import detect_boxes, split_page_into_zones
+    from ..detect.box_detect import detect_boxes, split_page_into_zones
 
     geo_result = detect_column_geometry(ocr_img, dewarped_bgr)
     if geo_result is None:

klausur-service/backend/ocr/layout/row_regularize.py

@@ -13,7 +13,7 @@ from typing import Dict, List
 
 import numpy as np
 
-from cv_vocab_types import RowGeometry
+from ..types import RowGeometry
 
 logger = logging.getLogger(__name__)
 

klausur-service/backend/ocr/layout/rows.py

@@ -20,9 +20,9 @@ try:
 except ImportError:
     cv2 = None  # type: ignore[assignment]
 
-from cv_vocab_types import RowGeometry
+from ..types import RowGeometry
-from cv_ocr_word_assembly import _group_words_into_lines
+from ..engines.word_assembly import _group_words_into_lines
-from cv_layout_row_regularize import _regularize_row_grid
+from .row_regularize import _regularize_row_grid
 
 logger = logging.getLogger(__name__)
 

klausur-service/backend/ocr/layout/scoring.py

@@ -11,7 +11,7 @@ import logging
 from collections import Counter
 from typing import Any, Dict, List, Optional
 
-from cv_vocab_types import (
+from ..types import (
     ColumnGeometry,
     ENGLISH_FUNCTION_WORDS,
     GERMAN_FUNCTION_WORDS,

klausur-service/backend/ocr/pipeline.py

@@ -14,24 +14,24 @@ Lizenz: Apache 2.0 (kommerziell nutzbar)
 DATENSCHUTZ: Alle Verarbeitung erfolgt lokal.
 """
 
-from cv_vocab_types import *       # noqa: F401,F403
+from .types import *       # noqa: F401,F403
-from cv_preprocessing import *     # noqa: F401,F403
+from .preprocessing.preprocessing import *     # noqa: F401,F403
-from cv_layout import *            # noqa: F401,F403
+from .layout.layout import *            # noqa: F401,F403
-from cv_ocr_engines import *       # noqa: F401,F403
+from .engines.engines import *       # noqa: F401,F403
-from cv_cell_grid import *         # noqa: F401,F403
+from .cell_grid.cell_grid import *         # noqa: F401,F403
-from cv_box_detect import *         # noqa: F401,F403
+from .detect.box_detect import *         # noqa: F401,F403
-from cv_review import *            # noqa: F401,F403
+from .review.review import *            # noqa: F401,F403
 
 # Private names used by consumers — not covered by wildcard re-exports.
-from cv_preprocessing import _apply_shear  # noqa: F401
+from .preprocessing.preprocessing import _apply_shear  # noqa: F401
-from cv_layout import (  # noqa: F401
+from .layout.layout import (  # noqa: F401
     _detect_header_footer_gaps,
     _detect_sub_columns,
     _split_broad_columns,
 )
-from cv_ocr_engines import (  # noqa: F401
+from .engines.engines import (  # noqa: F401
     _fix_character_confusion,
     _fix_phonetic_brackets,
 )
-from cv_cell_grid import _cells_to_vocab_entries  # noqa: F401
+from .cell_grid.cell_grid import _cells_to_vocab_entries  # noqa: F401
-from cv_words_first import build_grid_from_words  # noqa: F401
+from .words_first import build_grid_from_words  # noqa: F401

klausur-service/backend/ocr/preprocessing/__init__.py

@@ -0,0 +1,2 @@
+"""Preprocessing sub-package (deskew, dewarp, image I/O)."""
+from .preprocessing import *  # noqa: F401,F403

klausur-service/backend/ocr/preprocessing/deskew.py

@@ -11,7 +11,7 @@ from typing import Any, Dict, Tuple
 
 import numpy as np
 
-from cv_vocab_types import (
+from ..types import (
     CV2_AVAILABLE,
     TESSERACT_AVAILABLE,
 )

klausur-service/backend/ocr/preprocessing/dewarp.py

@@ -16,7 +16,7 @@ from typing import Any, Dict, List, Tuple
 
 import numpy as np
 
-from cv_vocab_types import (
+from ..types import (
     CV2_AVAILABLE,
     TESSERACT_AVAILABLE,
 )

klausur-service/backend/ocr/preprocessing/preprocessing.py

@@ -17,7 +17,7 @@ from typing import Tuple
 
 import numpy as np
 
-from cv_vocab_types import (
+from ..types import (
     CV2_AVAILABLE,
     TESSERACT_AVAILABLE,
 )
@@ -38,7 +38,7 @@ except ImportError:
     Image = None  # type: ignore[assignment,misc]
 
 # Re-export all deskew functions
-from cv_preprocessing_deskew import (  # noqa: F401
+from .deskew import (  # noqa: F401
     deskew_image,
     deskew_image_by_word_alignment,
     deskew_image_iterative,
@@ -48,7 +48,7 @@ from cv_preprocessing_deskew import (  # noqa: F401
 )
 
 # Re-export all dewarp functions
-from cv_preprocessing_dewarp import (  # noqa: F401
+from .dewarp import (  # noqa: F401
     _apply_shear,
     _detect_shear_angle,
     _detect_shear_by_hough,

klausur-service/backend/ocr/review/__init__.py

@@ -0,0 +1,2 @@
+"""Review sub-package (spell, LLM, pipeline orchestration)."""
+from .review import *  # noqa: F401,F403

klausur-service/backend/ocr/review/llm.py

@@ -183,7 +183,7 @@ async def llm_review_entries(
     model: str = None,
 ) -> Dict:
     """OCR error correction. Uses spell-checker (REVIEW_ENGINE=spell) or LLM (REVIEW_ENGINE=llm)."""
-    from cv_review_spell import spell_review_entries_sync, _SPELL_AVAILABLE
+    from .spell import spell_review_entries_sync, _SPELL_AVAILABLE
 
     if REVIEW_ENGINE == "spell" and _SPELL_AVAILABLE:
         return spell_review_entries_sync(entries)
@@ -260,8 +260,8 @@ async def llm_review_entries_streaming(
 
     Phase 0 (always): Run _fix_character_confusion and emit any changes.
     """
-    from cv_ocr_engines import _fix_character_confusion
+    from ..engines.engines import _fix_character_confusion
-    from cv_review_spell import spell_review_entries_streaming, _SPELL_AVAILABLE
+    from .spell import spell_review_entries_streaming, _SPELL_AVAILABLE
 
     _CONF_FIELDS = ('english', 'german', 'example')
     originals = [{f: e.get(f, '') for f in _CONF_FIELDS} for e in entries]

klausur-service/backend/ocr/review/pipeline.py

@@ -13,24 +13,24 @@ from typing import Any, Dict, List, Optional
 
 import numpy as np
 
-from cv_vocab_types import (
+from ..types import (
     CV_PIPELINE_AVAILABLE,
     PageRegion,
     PipelineResult,
     VocabRow,
 )
-from cv_preprocessing import (
+from ..preprocessing.preprocessing import (
     deskew_image,
     dewarp_image,
     render_image_high_res,
     render_pdf_high_res,
 )
-from cv_layout import (
+from ..layout.layout import (
     analyze_layout,
     create_layout_image,
     create_ocr_image,
 )
-from cv_ocr_engines import (
+from ..engines.engines import (
     _group_words_into_lines,
 )
 

klausur-service/backend/ocr/review/review.py

@@ -12,7 +12,7 @@ DATENSCHUTZ: Alle Verarbeitung erfolgt lokal.
 """
 
 # Re-export everything for backward compatibility
-from cv_review_pipeline import (  # noqa: F401
+from .pipeline import (  # noqa: F401
     ocr_region,
     run_multi_pass_ocr,
     match_lines_to_vocab,
@@ -20,7 +20,7 @@ from cv_review_pipeline import (  # noqa: F401
     run_cv_pipeline,
 )
 
-from cv_review_spell import (  # noqa: F401
+from .spell import (  # noqa: F401
     _SPELL_AVAILABLE,
     _spell_dict_knows,
     _spell_fix_field,
@@ -31,7 +31,7 @@ from cv_review_spell import (  # noqa: F401
     spell_review_entries_streaming,
 )
 
-from cv_review_llm import (  # noqa: F401
+from .llm import (  # noqa: F401
     OLLAMA_REVIEW_MODEL,
     REVIEW_ENGINE,
     _REVIEW_BATCH_SIZE,

klausur-service/backend/ocr/review/spell.py

@@ -210,7 +210,7 @@ def spell_review_entries_sync(entries: List[Dict]) -> Dict:
     Uses SmartSpellChecker for language-aware corrections with context-based
     disambiguation (a/I), multi-digit substitution, and cross-language guard.
     """
-    from cv_review_llm import _entry_needs_review
+    from .llm import _entry_needs_review
 
     t0 = time.time()
     changes: List[Dict] = []

klausur-service/backend/ocr/words_first.py

@@ -19,7 +19,7 @@ import re
 import statistics
 from typing import Any, Dict, List, Optional, Tuple
 
-from cv_ocr_engines import (
+from .engines.engines import (
     _group_words_into_lines,
     _words_to_reading_order_text,
 )

klausur-service/backend/ocr_image_enhance.py

@@ -0,0 +1,4 @@
+# Backward-compat shim -- module moved to ocr\/image_enhance.py
+import importlib as _importlib
+import sys as _sys
+_sys.modules[__name__] = _importlib.import_module("ocr.image_enhance")