Excel Data Types & Formulas
Excel's type system is notoriously complex. Dates are stored as serialized floats (e.g., 44197.5) offset from the 1900 or 1904 epoch. Formulas are stored as raw text strings in an <f> tag, while the cached calculated result sits in a sibling <v> tag. Furthermore, precision loss is common due to IEEE 754 float limits.
These files isolate extreme boundary conditions in the type system. They include long text, zero-padded strings ('00123), deep formula references, and standard Excel error states like #DIV/0! and #REF!.
dates and times
A file focusing purely on serial dates, times, and timestamps across the epoch boundaries.
formulas only
A file exclusively containing standard formulas (SUM, VLOOKUP, IF) and their cached results.
error values
A file containing forced calculation errors (#DIV/0!, #N/A, #REF!, #VALUE!) to test exception handling.
data types
A comprehensive file containing Text, Integers, Floats, Booleans, extremely long strings, and zero-padded text strings.
Frequently Asked Questions
Use Cases
- Testing parser logic for Excel serial dates (converting float values back into ISO-8601 timestamps).
- Validating that parsers extract the calculated `<v>` value of a formula rather than raw `<f>` string.
- Ensuring leading zeros on text-formatted numbers are not incorrectly stripped by aggressive type-casting.
- Testing error-state ingestion (handling #N/A or #VALUE! strings in data pipelines).