Specification: Data Sheet Standard (DSS) v1.0

Status: Proposal / Draft v1.0
Author: Vincenzo Manto @ datastripes.com / ilovecsv.com / ihatecsv.com
Extension: .DSS
MIME Type: text/DSS
Encoding: UTF-8

---

1. Abstract

The Data Sheet Standard (DSS) is a text-based, human-readable data format designed to represent multi-sheet spreadsheet data. Unlike CSV, it supports multiple tabs and sparse data placement via an anchor-based system. Unlike XLSX, it is non-binary, non-XML, and fully compatible with version control systems (Git-friendly).

2. Design Principles

1. Human Readable: A user should be able to understand the data by opening the file in a simple text editor.
2. Sparse Data Support: Only populated cells are stored. No "padding" with empty commas is required to reach a specific coordinate.
3. Multi-Sheet: A single file can contain multiple named sheets.
4. Git-Friendly: Changes to a single cell result in a predictable, single-line diff.
5. Simplicity: Parsers should be implementable in under 100 lines of code.

---

3. File Structure

A DSS file consists of three optional/mandatory layers:

1. Global Metadata (Optional)
2. Sheet Declarations (Mandatory for multi-sheet)
3. Data Anchors (Mandatory)

3.1 Header (Global Metadata)

The file begins with an optional metadata block encapsulated by triple dashes ---.

---
format: DSS 1.0
created_by: Datastripes
date: 2023-10-27
---

3.2 Sheet Declaration

Sheets are defined by a name enclosed in square brackets.

• Syntax: [Sheet Name]
• Rules: Sheet names must be unique within the file. Characters [ and ] are reserved.

3.3 Data Anchors (The Coordinate System)

Anchors define where a block of data starts on the 2D grid using standard A1 notation.

• Syntax: @ Coordinate (e.g., @ A1, @ M20)
• Behavior: All subsequent lines of data are relative to this anchor until a new anchor or sheet is declared.

---

4. Data Syntax

Inside an anchored block, data follows a strict CSV-inspired structure (RFC 4180 compliant).

4.1 Delimiters

• The default delimiter is the comma ,.
• Rows are separated by standard line breaks (LF or CRLF).

4.2 Data Types

DSS is "type-aware" through notation:

• String: Wrapped in double quotes (e.g., "Sales Report"). Mandatory if the string contains commas or line breaks.
• Numeric: Unquoted digits (e.g., 123, 45.67).
• Boolean: true or false (case-insensitive).
• Null/Empty: Represented by an empty value between commas or a explicit null.
• Formula: (Optional) Prefixed with =, stored as a string (e.g., "=SUM(B2:B10)").

---

5. Formal Grammar (EBNF-like)

<file>        ::= [<metadata>] <sheet_list>
<metadata>    ::= "---" <newline> { <key_value> <newline> } "---" <newline>
<sheet_list>  ::= { <sheet_block> }
<sheet_block> ::= "[" <name> "]" <newline> { <anchor_block> }
<anchor_block>::= "@" <coordinate> <newline> <csv_data>
<coordinate>  ::= [A-Z]+[0-9]+
<csv_data>    ::= { <row> <newline> }
<row>         ::= <value> { "," <value> }

---

6. Implementation Example

Below is an example of a .DSS file representing a complex, sparse spreadsheet.

---
project: Financial Forecast
version: 2.1
---

[Quarterly Report]
@ A1
"Department", "Budget", "Actual"
"Marketing", 50000, 48500
"R&D", 120000, 131000

@ G1
"Status: Over Budget"
"Risk Level: Low"

@ A10
"Notes:"
"The R&D department exceeded budget due to hardware acquisition."

[Settings]
@ B2
"Tax Rate", 0.22
"Currency", "EUR"

---

7. Parsing Logic (Reference for Developers)

To parse a DSS file:

1. Initialize a dictionary of sheets.
2. Iterate through lines:
   • If line starts with [, create a new sheet object/key.
   • If line starts with @, parse the A1 coordinate into row/column indices (e.g., B2 -> row 1, col 1). This is the Active Cursor.
   • If line contains data, split by comma (respecting quotes). Map each value to Active Cursor + current_offset. Increment Active Cursor row for the next line.
3. Ignore lines starting with # (comments).

---

8. Why DSS vs Others

Feature	CSV	XLSX (XML)	DSS
Multi-Sheet	No	Yes	Yes
Sparse Data	No (requires , padding)	Yes	Yes
Git-Friendly	Yes	No (Binary/Compressed)	Yes
Human Readable	Yes	No	Yes
Parsing Overhead	Extremely Low	High	Low

---

9. Future Extensions

• Cell Styling: Optional CSS-like block for formatting (e.g., $ A1:B10 { font-weight: bold; }).
• Encryption: Support for encrypted data blocks within specific sheets.

---

Standardization Note:
This format is released under the CC BY 4.0 (Creative Commons Attribution 4.0 International License). It is free for use, modification, and distribution in any software, specifically intended for the Datastripes ecosystem and the ilovecsv.com toolkit.