exstruct 0.2.3__tar.gz → 0.2.21__tar.gz

{exstruct-0.2.3 → exstruct-0.2.21}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: exstruct
-Version: 0.2.3
+Version: 0.2.21
 Summary: Excel to structured JSON (tables, shapes, charts) for LLM/RAG pipelines
 Keywords: excel,structure,data,exstruct
 Author: harumiWeb
@@ -60,12 +60,12 @@ Description-Content-Type: text/markdown
 
 ![ExStruct Image](/docs/assets/icon.webp)
 
-ExStruct reads Excel workbooks and outputs structured data (cells, table candidates, shapes, charts, print areas/views, hyperlinks) as JSON by default, with optional YAML/TOON formats. It targets both COM/Excel environments (rich extraction) and non-COM environments (cells + table candidates + print areas), with tunable detection heuristics and multiple output modes to fit LLM/RAG pipelines.
+ExStruct reads Excel workbooks and outputs structured data (tables, shapes, charts, hyperlinks) as JSON by default, with optional YAML/TOON formats. It targets both COM/Excel environments (rich extraction) and non-COM environments (cells + table candidates), with tunable detection heuristics and multiple output modes to fit LLM/RAG pipelines.
 
 ## Features
 
-- **Excel → Structured JSON**: cells, shapes, charts, table candidates, and print areas/views per sheet.
-- **Output modes**: `light` (cells + table candidates + print areas; no COM, shapes/charts empty), `standard` (texted shapes + arrows, charts, print areas), `verbose` (all shapes with width/height, charts with size, print areas). Verbose also emits cell hyperlinks. Size output is flag-controlled.
+- **Excel → Structured JSON**: cells, shapes, charts, and table candidates per sheet.
+- **Output modes**: `light` (cells + table candidates only), `standard` (texted shapes + arrows, charts), `verbose` (all shapes with width/height). Verbose also emits cell hyperlinks.
 - **Formats**: JSON (compact by default, `--pretty` available), YAML, TOON (optional dependencies).
 - **Table detection tuning**: adjust heuristics at runtime via API.
 - **CLI rendering** (Excel required): optional PDF and per-sheet PNGs.
@@ -85,8 +85,7 @@ Optional extras:
 - All extras at once: `pip install exstruct[yaml,toon,render]`
 
 Platform note:
-
-- Full extraction (shapes/charts) targets Windows + Excel (COM via xlwings). On other platforms, use `mode=light` to get cells + `table_candidates`.
+- Full extraction (shapes/charts) targets Windows + Excel (COM via xlwings). On other platforms, use `mode=light` to get cells + `table_candidates` safely.
 
 ## Quick Start (CLI)
 
@@ -96,7 +95,6 @@ exstruct input.xlsx -o out.json --pretty   # pretty JSON to a file
 exstruct input.xlsx --format yaml          # YAML (needs pyyaml)
 exstruct input.xlsx --format toon          # TOON (needs python-toon)
 exstruct input.xlsx --sheets-dir sheets/   # split per sheet in chosen format
-exstruct input.xlsx --print-areas-dir areas/  # split per print area (if any)
 exstruct input.xlsx --mode light           # cells + table candidates only
 exstruct input.xlsx --pdf --image          # PDF and PNGs (Excel required)
 ```
@@ -114,7 +112,7 @@ set_table_detection_params(table_score_threshold=0.3, density_min=0.04)
 wb = extract("input.xlsx", mode="standard")
 export(wb, Path("out.json"), pretty=False)  # compact JSON
 
-# Model helpers: iterate, index, and serialize directly
+# Model helpers: iterate, index, and serialize directly from the models
 first_sheet = wb["Sheet1"]           # __getitem__ access
 for name, sheet in wb:               # __iter__ yields (name, SheetData)
     print(name, len(sheet.rows))
@@ -122,23 +120,19 @@ wb.save("out.json", pretty=True)     # WorkbookData → file (by extension)
 first_sheet.save("sheet.json")       # SheetData → file (by extension)
 print(first_sheet.to_yaml())         # YAML text (requires pyyaml)
 
-# ExStructEngine: per-instance options for extraction/output
-from exstruct import ExStructEngine, StructOptions, OutputOptions
-
-engine = ExStructEngine(
-    options=StructOptions(mode="verbose"),  # verbose includes hyperlinks by default
-    output=OutputOptions(include_shapes=False, pretty=True),
-)
-wb2 = engine.extract("input.xlsx")
-engine.export(wb2, Path("out_filtered.json"))  # drops shapes via OutputOptions
-
-# Enable hyperlinks in other modes
-engine_links = ExStructEngine(options=StructOptions(mode="standard", include_cell_links=True))
-with_links = engine_links.extract("input.xlsx")
-
-# Export per print area (if print areas exist)
-from exstruct import export_print_areas_as
-export_print_areas_as(wb, "areas", fmt="json", pretty=True)
+# ExStructEngine: per-instance options for extraction/output
+from exstruct import ExStructEngine, StructOptions, OutputOptions
+
+engine = ExStructEngine(
+    options=StructOptions(mode="verbose"),  # verbose includes hyperlinks by default
+    output=OutputOptions(include_shapes=False, pretty=True),
+)
+wb2 = engine.extract("input.xlsx")
+engine.export(wb2, Path("out_filtered.json"))  # drops shapes via OutputOptions
+
+# Enable hyperlinks in other modes
+engine_links = ExStructEngine(options=StructOptions(mode="standard", include_cell_links=True))
+with_links = engine_links.extract("input.xlsx")
 ```
 
 **Note (non-COM environments):** If Excel COM is unavailable, extraction still runs and returns cells + `table_candidates`; `shapes`/`charts` will be empty.
@@ -158,11 +152,11 @@ set_table_detection_params(
 
 Use higher thresholds to reduce false positives; lower them if true tables are missed.
 
-## Output Modes
-
-- **light**: cells + table candidates (no COM needed).
-- **standard**: texted shapes + arrows, charts (COM if available), table candidates. Hyperlinks are off unless `include_cell_links=True`.
-- **verbose**: all shapes (with width/height), charts, table candidates, and cell hyperlinks.
+## Output Modes
+
+- **light**: cells + table candidates (no COM needed).
+- **standard**: texted shapes + arrows, charts (COM if available), table candidates. Hyperlinks are off unless `include_cell_links=True`.
+- **verbose**: all shapes (with width/height), charts, table candidates, and cell hyperlinks.
 
 ## Error Handling / Fallbacks
 
@@ -191,7 +185,6 @@ To show how well exstruct can structure Excel, we parse a workbook that combines
 (Screenshot below is the actual sample Excel sheet)
 ![Sample Excel](/docs/assets/demo_sheet.png)
 Sample workbook: `sample/sample.xlsx`
-Sample workbook: `sample/sample.xlsx`
 
 ### 1. Input: Excel Sheet Overview
 
@@ -378,12 +371,6 @@ In short, **exstruct = “an engine that converts Excel into a format AI can und
 - Default JSON is compact to reduce tokens; use `--pretty` or `pretty=True` when readability matters.
 - Field `table_candidates` replaces `tables`; adjust downstream consumers accordingly.
 
-## Print Areas (PrintArea / PrintAreaView)
-
-- `SheetData.print_areas` holds print areas (cell coordinates) in light/standard/verbose.
-- Use `export_print_areas_as(...)` or CLI `--print-areas-dir` to write one file per print area (nothing is written if none exist).
-- `PrintAreaView` includes rows and table candidates inside the area, plus shapes/charts that overlap the area (size-less shapes are treated as points). `normalize=True` rebases row/col indices to the area origin.
-
 ## License
 
 BSD-3-Clause. See `LICENSE` for details.
@@ -391,3 +378,18 @@ BSD-3-Clause. See `LICENSE` for details.
 ## Documentation
 
 - API Reference (GitHub Pages): https://harumiweb.github.io/exstruct/
+# Engine option cheat sheet
+
+| Option class   | Field               | Meaning |
+| -------------- | ------------------- | ------- |
+| StructOptions  | mode                | "light"/"standard"/"verbose" |
+|                | table_params        | Dict passed to `set_table_detection_params` (table_score_threshold, density_min, coverage_min, min_nonempty_cells) |
+|                | include_cell_links  | Include cell hyperlinks in `rows[*].links` (None -> auto: verbose=True, others=False) |
+| OutputOptions  | fmt                 | Default format ("json"/"yaml"/"yml"/"toon") |
+|                | pretty / indent     | Pretty-print JSON and control indent |
+|                | include_rows        | Include rows (False to drop) |
+|                | include_shapes      | Include shapes |
+|                | include_charts      | Include charts |
+|                | include_tables      | Include table_candidates |
+|                | sheets_dir          | Optional directory for per-sheet exports |
+|                | stream              | Default stream when output_path is None |

{exstruct-0.2.3 → exstruct-0.2.21}/README.md

@@ -4,12 +4,12 @@
 
 ![ExStruct Image](/docs/assets/icon.webp)
 
-ExStruct reads Excel workbooks and outputs structured data (cells, table candidates, shapes, charts, print areas/views, hyperlinks) as JSON by default, with optional YAML/TOON formats. It targets both COM/Excel environments (rich extraction) and non-COM environments (cells + table candidates + print areas), with tunable detection heuristics and multiple output modes to fit LLM/RAG pipelines.
+ExStruct reads Excel workbooks and outputs structured data (tables, shapes, charts, hyperlinks) as JSON by default, with optional YAML/TOON formats. It targets both COM/Excel environments (rich extraction) and non-COM environments (cells + table candidates), with tunable detection heuristics and multiple output modes to fit LLM/RAG pipelines.
 
 ## Features
 
-- **Excel → Structured JSON**: cells, shapes, charts, table candidates, and print areas/views per sheet.
-- **Output modes**: `light` (cells + table candidates + print areas; no COM, shapes/charts empty), `standard` (texted shapes + arrows, charts, print areas), `verbose` (all shapes with width/height, charts with size, print areas). Verbose also emits cell hyperlinks. Size output is flag-controlled.
+- **Excel → Structured JSON**: cells, shapes, charts, and table candidates per sheet.
+- **Output modes**: `light` (cells + table candidates only), `standard` (texted shapes + arrows, charts), `verbose` (all shapes with width/height). Verbose also emits cell hyperlinks.
 - **Formats**: JSON (compact by default, `--pretty` available), YAML, TOON (optional dependencies).
 - **Table detection tuning**: adjust heuristics at runtime via API.
 - **CLI rendering** (Excel required): optional PDF and per-sheet PNGs.
@@ -29,8 +29,7 @@ Optional extras:
 - All extras at once: `pip install exstruct[yaml,toon,render]`
 
 Platform note:
-
-- Full extraction (shapes/charts) targets Windows + Excel (COM via xlwings). On other platforms, use `mode=light` to get cells + `table_candidates`.
+- Full extraction (shapes/charts) targets Windows + Excel (COM via xlwings). On other platforms, use `mode=light` to get cells + `table_candidates` safely.
 
 ## Quick Start (CLI)
 
@@ -40,7 +39,6 @@ exstruct input.xlsx -o out.json --pretty   # pretty JSON to a file
 exstruct input.xlsx --format yaml          # YAML (needs pyyaml)
 exstruct input.xlsx --format toon          # TOON (needs python-toon)
 exstruct input.xlsx --sheets-dir sheets/   # split per sheet in chosen format
-exstruct input.xlsx --print-areas-dir areas/  # split per print area (if any)
 exstruct input.xlsx --mode light           # cells + table candidates only
 exstruct input.xlsx --pdf --image          # PDF and PNGs (Excel required)
 ```
@@ -58,7 +56,7 @@ set_table_detection_params(table_score_threshold=0.3, density_min=0.04)
 wb = extract("input.xlsx", mode="standard")
 export(wb, Path("out.json"), pretty=False)  # compact JSON
 
-# Model helpers: iterate, index, and serialize directly
+# Model helpers: iterate, index, and serialize directly from the models
 first_sheet = wb["Sheet1"]           # __getitem__ access
 for name, sheet in wb:               # __iter__ yields (name, SheetData)
     print(name, len(sheet.rows))
@@ -66,23 +64,19 @@ wb.save("out.json", pretty=True)     # WorkbookData → file (by extension)
 first_sheet.save("sheet.json")       # SheetData → file (by extension)
 print(first_sheet.to_yaml())         # YAML text (requires pyyaml)
 
-# ExStructEngine: per-instance options for extraction/output
-from exstruct import ExStructEngine, StructOptions, OutputOptions
-
-engine = ExStructEngine(
-    options=StructOptions(mode="verbose"),  # verbose includes hyperlinks by default
-    output=OutputOptions(include_shapes=False, pretty=True),
-)
-wb2 = engine.extract("input.xlsx")
-engine.export(wb2, Path("out_filtered.json"))  # drops shapes via OutputOptions
-
-# Enable hyperlinks in other modes
-engine_links = ExStructEngine(options=StructOptions(mode="standard", include_cell_links=True))
-with_links = engine_links.extract("input.xlsx")
-
-# Export per print area (if print areas exist)
-from exstruct import export_print_areas_as
-export_print_areas_as(wb, "areas", fmt="json", pretty=True)
+# ExStructEngine: per-instance options for extraction/output
+from exstruct import ExStructEngine, StructOptions, OutputOptions
+
+engine = ExStructEngine(
+    options=StructOptions(mode="verbose"),  # verbose includes hyperlinks by default
+    output=OutputOptions(include_shapes=False, pretty=True),
+)
+wb2 = engine.extract("input.xlsx")
+engine.export(wb2, Path("out_filtered.json"))  # drops shapes via OutputOptions
+
+# Enable hyperlinks in other modes
+engine_links = ExStructEngine(options=StructOptions(mode="standard", include_cell_links=True))
+with_links = engine_links.extract("input.xlsx")
 ```
 
 **Note (non-COM environments):** If Excel COM is unavailable, extraction still runs and returns cells + `table_candidates`; `shapes`/`charts` will be empty.
@@ -102,11 +96,11 @@ set_table_detection_params(
 
 Use higher thresholds to reduce false positives; lower them if true tables are missed.
 
-## Output Modes
-
-- **light**: cells + table candidates (no COM needed).
-- **standard**: texted shapes + arrows, charts (COM if available), table candidates. Hyperlinks are off unless `include_cell_links=True`.
-- **verbose**: all shapes (with width/height), charts, table candidates, and cell hyperlinks.
+## Output Modes
+
+- **light**: cells + table candidates (no COM needed).
+- **standard**: texted shapes + arrows, charts (COM if available), table candidates. Hyperlinks are off unless `include_cell_links=True`.
+- **verbose**: all shapes (with width/height), charts, table candidates, and cell hyperlinks.
 
 ## Error Handling / Fallbacks
 
@@ -135,7 +129,6 @@ To show how well exstruct can structure Excel, we parse a workbook that combines
 (Screenshot below is the actual sample Excel sheet)
 ![Sample Excel](/docs/assets/demo_sheet.png)
 Sample workbook: `sample/sample.xlsx`
-Sample workbook: `sample/sample.xlsx`
 
 ### 1. Input: Excel Sheet Overview
 
@@ -322,12 +315,6 @@ In short, **exstruct = “an engine that converts Excel into a format AI can und
 - Default JSON is compact to reduce tokens; use `--pretty` or `pretty=True` when readability matters.
 - Field `table_candidates` replaces `tables`; adjust downstream consumers accordingly.
 
-## Print Areas (PrintArea / PrintAreaView)
-
-- `SheetData.print_areas` holds print areas (cell coordinates) in light/standard/verbose.
-- Use `export_print_areas_as(...)` or CLI `--print-areas-dir` to write one file per print area (nothing is written if none exist).
-- `PrintAreaView` includes rows and table candidates inside the area, plus shapes/charts that overlap the area (size-less shapes are treated as points). `normalize=True` rebases row/col indices to the area origin.
-
 ## License
 
 BSD-3-Clause. See `LICENSE` for details.
@@ -335,3 +322,18 @@ BSD-3-Clause. See `LICENSE` for details.
 ## Documentation
 
 - API Reference (GitHub Pages): https://harumiweb.github.io/exstruct/
+# Engine option cheat sheet
+
+| Option class   | Field               | Meaning |
+| -------------- | ------------------- | ------- |
+| StructOptions  | mode                | "light"/"standard"/"verbose" |
+|                | table_params        | Dict passed to `set_table_detection_params` (table_score_threshold, density_min, coverage_min, min_nonempty_cells) |
+|                | include_cell_links  | Include cell hyperlinks in `rows[*].links` (None -> auto: verbose=True, others=False) |
+| OutputOptions  | fmt                 | Default format ("json"/"yaml"/"yml"/"toon") |
+|                | pretty / indent     | Pretty-print JSON and control indent |
+|                | include_rows        | Include rows (False to drop) |
+|                | include_shapes      | Include shapes |
+|                | include_charts      | Include charts |
+|                | include_tables      | Include table_candidates |
+|                | sheets_dir          | Optional directory for per-sheet exports |
+|                | stream              | Default stream when output_path is None |

{exstruct-0.2.3 → exstruct-0.2.21}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "exstruct"
-version = "0.2.3"
+version = "0.2.21"
 description = "Excel to structured JSON (tables, shapes, charts) for LLM/RAG pipelines"
 readme = "README.md"
 license = { file = "LICENSE" }

exstruct-0.2.21/src/exstruct/__init__.py

@@ -0,0 +1,122 @@
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Literal, Optional, TextIO
+
+from .core.integrate import extract_workbook
+from .core.cells import set_table_detection_params
+from .engine import ExStructEngine, OutputOptions, StructOptions
+from .io import save_as_json, save_as_toon, save_as_yaml, save_sheets, serialize_workbook
+from .models import CellRow, Chart, ChartSeries, Shape, SheetData, WorkbookData
+from .render import export_pdf, export_sheet_images
+
+__all__ = [
+    "extract",
+    "export",
+    "export_sheets",
+    "export_pdf",
+    "export_sheet_images",
+    "process_excel",
+    "ExtractionMode",
+    "CellRow",
+    "Shape",
+    "ChartSeries",
+    "Chart",
+    "SheetData",
+    "WorkbookData",
+    "set_table_detection_params",
+    "ExStructEngine",
+    "StructOptions",
+    "OutputOptions",
+]
+
+
+ExtractionMode = Literal["light", "standard", "verbose"]
+
+
+def extract(file_path: str | Path, mode: ExtractionMode = "standard") -> WorkbookData:
+    """Extract workbook semantic structure and return WorkbookData."""
+    include_links = True if mode == "verbose" else False
+    engine = ExStructEngine(options=StructOptions(mode=mode, include_cell_links=include_links))
+    return engine.extract(file_path, mode=mode)
+
+
+def export(
+    data: WorkbookData,
+    path: str | Path,
+    fmt: Optional[Literal["json", "yaml", "yml", "toon"]] = None,
+    *,
+    pretty: bool = False,
+    indent: int | None = None,
+) -> None:
+    """Export WorkbookData to supported file formats (json/yaml/toon)."""
+    dest = Path(path)
+    format_hint = (fmt or dest.suffix.lstrip(".") or "json").lower()
+    match format_hint:
+        case "json":
+            save_as_json(data, dest, pretty=pretty, indent=indent)
+        case "yaml" | "yml":
+            save_as_yaml(data, dest)
+        case "toon":
+            save_as_toon(data, dest)
+        case _:
+            raise ValueError(f"Unsupported export format: {format_hint}")
+
+
+def export_sheets(data: WorkbookData, dir_path: str | Path) -> dict[str, Path]:
+    """
+    Export each sheet as a JSON file (book_name + SheetData) into a directory.
+    Returns a mapping of sheet name to written path.
+    """
+    return save_sheets(data, Path(dir_path), fmt="json")
+
+
+def export_sheets_as(
+    data: WorkbookData,
+    dir_path: str | Path,
+    fmt: Literal["json", "yaml", "yml", "toon"] = "json",
+    *,
+    pretty: bool = False,
+    indent: int | None = None,
+) -> dict[str, Path]:
+    """
+    Export each sheet in the given format (json/yaml/toon), including book_name and SheetData; returns sheet name → path map.
+    """
+    return save_sheets(data, Path(dir_path), fmt=fmt, pretty=pretty, indent=indent)
+
+
+def process_excel(
+    file_path: Path,
+    output_path: Path | None = None,
+    out_fmt: str = "json",
+    image: bool = False,
+    pdf: bool = False,
+    dpi: int = 72,
+    mode: ExtractionMode = "standard",
+    pretty: bool = False,
+    indent: int | None = None,
+    sheets_dir: Path | None = None,
+    stream: TextIO | None = None,
+) -> None:
+    """
+    Convenience wrapper for CLI: export workbook and optionally PDF/PNG images (Excel required for rendering).
+    - If output_path is None, writes the serialized workbook to stdout (or provided stream).
+    - If sheets_dir is given, also writes per-sheet files into that directory.
+    """
+    engine = ExStructEngine(
+        options=StructOptions(mode=mode),
+        output=OutputOptions(fmt=out_fmt, pretty=pretty, indent=indent, sheets_dir=sheets_dir, stream=stream),
+    )
+    engine.process(
+        file_path=file_path,
+        output_path=output_path,
+        out_fmt=out_fmt,
+        image=image,
+        pdf=pdf,
+        dpi=dpi,
+        mode=mode,
+        pretty=pretty,
+        indent=indent,
+        sheets_dir=sheets_dir,
+        stream=stream,
+    )

{exstruct-0.2.3 → exstruct-0.2.21}/src/exstruct/cli/main.py

@@ -57,11 +57,6 @@ def build_parser() -> argparse.ArgumentParser:
         type=Path,
         help="Optional directory to write one file per sheet (format follows --format).",
     )
-    parser.add_argument(
-        "--print-areas-dir",
-        type=Path,
-        help="Optional directory to write one file per print area (format follows --format).",
-    )
     return parser
 
 
@@ -85,7 +80,6 @@ def main(argv: list[str] | None = None) -> int:
             mode=args.mode,
             pretty=args.pretty,
             sheets_dir=args.sheets_dir,
-            print_areas_dir=args.print_areas_dir,
         )
         return 0
     except Exception as e:

{exstruct-0.2.3 → exstruct-0.2.21}/src/exstruct/core/charts.py

@@ -1,7 +1,7 @@
 from __future__ import annotations
 
 import logging
-from typing import Dict, List, Optional, Literal
+from typing import Dict, List, Optional
 
 import xlwings as xw
 
@@ -166,7 +166,7 @@ def parse_series_formula(formula: str) -> Optional[Dict[str, Optional[str]]]:
     }
 
 
-def get_charts(sheet: xw.Sheet, mode: Literal["light", "standard", "verbose"] = "standard") -> List[Chart]:
+def get_charts(sheet: xw.Sheet) -> List[Chart]:
     """Parse charts in a sheet into Chart models; failed charts carry an error field."""
     charts: List[Chart] = []
     for ch in sheet.charts:
@@ -182,14 +182,6 @@ def get_charts(sheet: xw.Sheet, mode: Literal["light", "standard", "verbose"] =
             chart_type_label = XL_CHART_TYPE_MAP.get(
                 chart_type_num, f"unknown_{chart_type_num}"
             )
-            chart_width: Optional[int] = None
-            chart_height: Optional[int] = None
-            try:
-                chart_width = int(ch.width)
-                chart_height = int(ch.height)
-            except Exception:
-                chart_width = None
-                chart_height = None
 
             for s in chart_com.SeriesCollection():
                 parsed = parse_series_formula(getattr(s, "Formula", ""))
@@ -228,8 +220,6 @@ def get_charts(sheet: xw.Sheet, mode: Literal["light", "standard", "verbose"] =
                 title=title,
                 y_axis_title=y_axis_title,
                 y_axis_range=y_axis_range,  # type: ignore
-                w=chart_width,
-                h=chart_height,
                 series=series_list,
                 l=int(ch.left),
                 t=int(ch.top),

exstruct-0.2.21/src/exstruct/core/integrate.py

@@ -0,0 +1,140 @@
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Dict, List, Literal
+
+import logging
+import os
+
+import xlwings as xw
+
+from ..models import CellRow, Shape, SheetData, WorkbookData
+from .cells import (
+    detect_tables,
+    detect_tables_openpyxl,
+    extract_sheet_cells,
+    extract_sheet_cells_with_links,
+)
+from .charts import get_charts
+from .shapes import get_shapes_with_position
+
+logger = logging.getLogger(__name__)
+_ALLOWED_MODES: set[str] = {"light", "standard", "verbose"}
+
+
+def _find_open_workbook(file_path: Path) -> xw.Book | None:
+    """Return an existing workbook if already open in Excel; otherwise None."""
+    try:
+        for app in xw.apps:
+            for wb in app.books:
+                try:
+                    if Path(wb.fullname).resolve() == file_path.resolve():
+                        return wb
+                except Exception:
+                    continue
+    except Exception:
+        return None
+    return None
+
+
+def _open_workbook(file_path: Path) -> tuple[xw.Book, bool]:
+    """
+    Open workbook:
+    - If already open, reuse and do not close Excel on exit.
+    - Otherwise create invisible Excel (visible=False) and close when done.
+    Returns (workbook, should_close_app).
+    """
+    existing = _find_open_workbook(file_path)
+    if existing:
+        return existing, False
+    app = xw.App(add_book=False, visible=False)
+    wb = app.books.open(str(file_path))
+    return wb, True
+
+
+def integrate_sheet_content(
+    cell_data: Dict[str, List[CellRow]],
+    shape_data: Dict[str, List[Shape]],
+    workbook: xw.Book,
+    mode: Literal["light", "standard", "verbose"] = "standard",
+) -> Dict[str, SheetData]:
+    """Integrate cells, shapes, charts, and tables into SheetData per sheet."""
+    result: Dict[str, SheetData] = {}
+    for sheet_name, rows in cell_data.items():
+        sheet_shapes = shape_data.get(sheet_name, [])
+        sheet = workbook.sheets[sheet_name]
+
+        sheet_model = SheetData(
+            rows=rows,
+            shapes=sheet_shapes,
+            charts=[] if mode == "light" else get_charts(sheet),
+            table_candidates=detect_tables(sheet),
+        )
+
+        result[sheet_name] = sheet_model
+    return result
+
+
+def extract_workbook(
+    file_path: Path,
+    mode: Literal["light", "standard", "verbose"] = "standard",
+    *,
+    include_cell_links: bool = False,
+) -> WorkbookData:
+    """Extract workbook and return WorkbookData; fallback to cells+tables if Excel COM is unavailable."""
+    if mode not in _ALLOWED_MODES:
+        raise ValueError(f"Unsupported mode: {mode}")
+
+    cell_data = extract_sheet_cells_with_links(file_path) if include_cell_links else extract_sheet_cells(file_path)
+
+    def _cells_and_tables_only(reason: str) -> WorkbookData:
+        sheets: Dict[str, SheetData] = {}
+        for sheet_name, rows in cell_data.items():
+            try:
+                tables = detect_tables_openpyxl(file_path, sheet_name)
+            except Exception:
+                tables = []
+            sheets[sheet_name] = SheetData(
+                rows=rows,
+                shapes=[],
+                charts=[],
+                table_candidates=tables,
+            )
+        logger.warning(
+            "%s Falling back to cells+tables only; shapes and charts will be empty.",
+            reason,
+        )
+        return WorkbookData(book_name=file_path.name, sheets=sheets)
+
+    if mode == "light":
+        return _cells_and_tables_only("Light mode selected.")
+
+    if os.getenv("SKIP_COM_TESTS"):
+        return _cells_and_tables_only("SKIP_COM_TESTS is set; skipping COM/xlwings access.")
+
+    try:
+        wb, close_app = _open_workbook(file_path)
+    except Exception as e:
+        return _cells_and_tables_only(
+            f"xlwings/Excel COM is unavailable. ({e!r})"
+        )
+
+    try:
+        try:
+            shape_data = get_shapes_with_position(wb, mode=mode)
+            merged = integrate_sheet_content(cell_data, shape_data, wb, mode=mode)
+            return WorkbookData(book_name=file_path.name, sheets=merged)
+        except Exception as e:
+            logger.warning("Shape extraction failed; falling back to cells+tables. (%r)", e)
+            return _cells_and_tables_only(
+                f"Shape extraction failed ({e!r})."
+            )
+    finally:
+        # Close only if we created the app to avoid shutting user sessions.
+        try:
+            if close_app:
+                app = wb.app
+                wb.close()
+                app.quit()
+        except Exception:
+            pass