text-spreadsheet for Python

Let a language model read a whole multi-sheet spreadsheet as text. The model calls one tool with the path of a file on your machine, and Excel, ODS, CSV, SQLite, Parquet and Arrow come back as MTSV: one plain-text file where a tab separates fields, a line break separates records, and a form feed separates sheets.

The version is the version field of pyproject.toml.

• Repository
• Specification

Install

In an environment of its own:

pipx install text-spreadsheet

Then tell your host to launch it. For Claude Code:

claude mcp add text-spreadsheet -- text-spreadsheet

Other hosts take the same command in a configuration file, under mcpServers or servers.

The tool

The model calls read with a path:

read("/home/me/book.xlsx")

and gets a map of the file: its sheets, how many records each holds, the columns of each, where each sheet's lines are, and what the conversion left behind.

<FF>file
source	artifact	converted
/home/me/book.xlsx	/home/me/.cache/text-spreadsheet/home/me/book.mtsv	yes
<FF>sheets
sheet	sheet name	records	first line	last line
1	People	4	1	6
2	Orders	200	7	208
<FF>columns
sheet	position	field name
1	1	Name
1	2	Age
<FF>left behind
what
cell type n

Adding an address returns that part of the file instead of the map:

read("/home/me/book.xlsx", sheet="2")
read("/home/me/book.xlsx", sheet="2", rows="40-120")
read("/home/me/book.xlsx", sheet="2", rows="40-120", fields="2;4")

Each address counts from 1 and is written as 2, 1;3 or 1-3, as RFC 7111 writes a selection of a tabular file. An address the file cannot answer is refused rather than guessed.

What is kept

Every conversion leaves an MTSV copy of the file under your cache directory — ~/.cache/text-spreadsheet on Linux, ~/Library/Caches/text-spreadsheet on macOS — mirroring the path of the file it came from. A later call reads that copy unless the file has changed since, and the map names where it is, so it can also be read directly.

Nothing else is kept, and nothing outside this machine is contacted.

As a library

from text_spreadsheet import read

text = read("/home/me/book.xlsx")

read returns MTSV text, whatever the format it came from. Whatever a spreadsheet holds that MTSV does not — formatting, formulas, types — is left behind, and named in the map.

Layout

Each module hides one decision, named beside it.

src/text_spreadsheet/
  _cache       where the copy lives, and whether it is fresh
  _report      what an integration left behind, as it reported it
  _map         what a file is made of
  _slice       how an address is written, and how a cut is made
  __init__     the operation: convert, keep, and answer
  _server      the tool, its annotations, and the transport
  __main__     the command a host launches
tests/         one test file per module above

Use points one way: __main__ to _server to read, and read to the four below it. Nothing points back up, and nothing but _server knows the protocol.

Test

From the root of the repository:

python3 -m venv .venv
.venv/bin/pip install ./python
.venv/bin/python -m unittest discover -s python/tests

License

MIT