excel-schema-engine 0.1.8

Declarative Excel schema builder, reader and validator

Excel Schema Engine

A lightweight Python library for generating, validating, and reading Excel files using a declarative schema.

Define your Excel structure once and use it to:

• generate Excel templates
• validate uploaded files
• read rows as structured data
• highlight errors directly in Excel

Built on top of openpyxl.

---

Installation

pip install excel-schema-engine

or, with Poetry:

poetry add excel-schema-engine

---

Quickstart

For a complete runnable example, see:

• examples/quickstart.py – define a schema, build a workbook, write rows, autosize columns.

---

Core concepts

• ExcelSchema: top‑level object that describes the whole workbook (sheets, styles, author for comments).
• SheetSchema: describes a single sheet (its name and list of columns).
• Column: describes a logical column with:
  • key: dictionary key used when reading/writing rows
  • header: text shown in the Excel header
  • style: name of a style defined in ExcelSchema.styles
  • comment: optional Comment shown in the header cell
  • subcolumns: nested columns for multi‑level headers
• CellStyle: strongly‑typed wrapper around openpyxl styles (Font, PatternFill, Alignment, Border).
• ExcelErrorsSchema: defines styles used to mark and highlight errors and fixed values.

All these classes are simple @dataclass models and can be constructed in plain Python.

---

Building Excel templates

See:

• examples/quickstart.py

• Headers:

  • Single‑row headers are created when a Column has no subcolumns.
  • Multi‑row headers are created when a Column has subcolumns (the parent header is merged across all subcolumns).

• Styles and comments:

  • Header styles are looked up by name in schema.styles.
  • Header comments are created using Comment(comment, author, width, height) where author comes from ExcelSchema.author.

---

Writing data rows

Use ExcelBuilder.write_rows to write dictionaries into the sheet based on the schema.

See:

• examples/quickstart.py

• Missing sheet: if the given sheet_name is not present in the schema, a localized error message is printed and nothing is written.

• Column mapping: keys from each row dict must match Column.key values (including nested subcolumns).

---

Validating uploaded files

ExcelValidator checks that the headers in a user‑submitted file match the schema (order and names).

See:

• examples/validation_and_reading.py

validate_headers returns a list of human‑readable error messages (already localized).

The validator checks:

• Number of columns (expected vs found)
• Missing columns
• Wrong header text (position, expected text, actual text)

---

Reading rows as structured data

ExcelReader converts worksheet rows into dictionaries using the same SheetSchema.

See:

• examples/validation_and_reading.py

Each row is a dict where keys are Column.key values and values are cell contents.

---

Highlighting and marking errors in Excel

Use ExcelErrorsSchema and ExcelErrors to visually mark problematic cells and rows.

See:

• examples/mark_errors.py

• mark_error: colors the cell with error_schema.error_fill (or a custom CellStyle) and optionally adds a localized comment.

• highlight_row: colors all cells in a row with error_schema.highlight_fill (or custom style).

• mark_fixed: re‑styles previously marked cells using error_schema.fixed_fill (or custom style) and optionally replaces the comment text.

---

Localization

The library supports multiple languages for validator and builder messages via the Language enum and ValidatorErrComment:

• Supported out of the box:
  • Language.EN – English
  • Language.RU – Russian
  • Language.PL – Polish
  • Language.BOBR – a playful demo language

You can extend or override messages at runtime:

from excel_schema_engine.global_vars import Language, ValidatorErrComment

ValidatorErrComment.messages[Language.FR] = {
    "missing_column": "Colonne manquante : {column}",
    "wrong_header": "Colonne {index} : attendu '{expected}', trouvé '{found}'",
    "columns_count": "Colonnes attendues : {expected}, trouvées : {found}",
    "miss_sheet": "La feuille nommée {sheet_name} est introuvable.",
    "error_prefix": "Erreur : ",
}

Messages are formatted using str.format, so you can use any placeholders expected by the library ({column}, {index}, {expected}, {found}, {sheet_name}, etc.).

---

API overview

Everything you typically need is re‑exported from the package root:

• Schema and styles
  • ExcelSchema
  • SheetSchema
  • Column
  • Comment
  • CellStyle
  • ExcelErrorsSchema
• Builders and readers
  • ExcelBuilder
  • ExcelReader
  • ExcelValidator
  • ExcelErrors
• Utilities
  • autosize_columns

For more advanced usage, you can also import:

• Language and ValidatorErrComment from excel_schema_engine.global_vars

This should be enough to build templates, validate uploaded files, and work with structured Excel data in your applications.