Import and export

Command	Reads	Writes	Notes
:csv save/load	CSV	CSV	Plain text, fast
:xlsx save/load	.xlsx formulas + values	EXCEL mode: formulas + cached values; other modes: values only	:xlsx load switches to EXCEL
:pd save/load	CSV/TSV/JSON	same	Needs gridcalc[extras]; row 1 as headers; JSON keeps value types

:xlsx load translates Excel formulas into gridcalc's EXCEL grammar and reads every worksheet, empty ones included. It replaces the whole workbook, including the code block and saved models. On import:

• Values keep their xlsx type. Text such as 00123 stays a label, booleans become =TRUE/=FALSE, and known error codes become formulas such as =#N/A.

• Shared and array formulas import as their cached values.

• Cells beyond 256 columns or 1024 rows are dropped.

• Chartsheets are skipped.

• Dates in a workbook using the 1904 date system are converted to gridcalc's 1900-based serials.

The load warns with a count of fallback formulas and dropped cells. The headless CLI prints these warnings to stderr; the terminal and desktop apps do not show them.

:xlsx save refuses an empty workbook and any sheet name Excel rejects: empty, longer than 31 characters, containing : \ / ? * [ ], starting or ending with ', named History, or equal to another name ignoring case.

INDIRECT and 3D ranges (Sheet1:Sheet3!A1:B2) are deliberately unsupported, because they would defeat the static dependency graph. Functions outside the auto-loaded library produce #NAME?. The full list of what does not cross the boundary is in Limitations.

The xlsx path goes through a C++ extension wrapping OpenXLSX rather than a pure-Python reader, which is why the core install needs no third-party runtime dependency to read a spreadsheet.