CTAN update: xint
List selectors [L][n], [L][a:b], in expressions are faster. And items are counted starting at zero, not one ([L][a:b] already followed exactly Python slicing rules). The iter keyword's meaning has changed, new one associated with rseq, former one which is associated to rrseq renamed to iterr. The float operations with large precision handle small inputs faster. The documentation has an example of using iter to implement expandably the Brent-Salamin algorithm for digits of Pi. See CHANGES.{pdf,html} for more.
This package is located at http://mirror.ctan.org/macros/generic/xint More information is at http://www.ctan.org/pkg/xint We are supported by the TeX User Groups. Please join a users group; see http://www.tug.org/usergroups.html .
Thanks for the upload. For the CTAN Team Ina Dau
Xint – Expandable arbitrary precision floating point and integer operations
The calx package and its superset the xintexpr package provide:
- \calxeval: numerical expression parser for floating-point computations.
- \calxinteval: numerical expression parser for computations with big integers.
Floating-point computations are supported up to 100 significant digits of precision. A full set of trigonometric, logarithmic, and exponential functions is implemented.
The two parsers are also available under their legacy names \xintfloateval and \xintiieval, respectively.
Loading xintexpr automatically loads calx and provides a third parser, \xinteval, which handles fractions. It also loads the legacy macro interface packages xintcore, xint, and xintfrac.
All three parsers produce their results by expansion alone. They accept comma-separated inputs and nested structures using square brackets. Users can define custom variables and functions.
| Package | Xint |
| Version | 2.0a1 2026-11-01 |
| Copyright | 2013–2022, 2025, 2026 Jean-François Burnol |
| Maintainer | Jean-François Burnol |