great-tables 1.0.0

v1.0.0

on Python PyPI

Great Tables v1.0.0 is here. This milestone release introduces these powerful new formatting methods: fmt_fraction() for mixed fractions, fmt_chem() for chemical formulas, fmt_index() for alphabetical indices, and fmt_url()/fmt_email() for clickable links and buttons. The new tab_style_body() method lets you style cells based on their data values, and opt_footnote_spec() gives you fine-grained control over footnote presentation. Cell body values are now HTML-escaped by default, a breaking change that strengthens security against XSS. A new load_dataset() function provides a single entry point for all bundled example datasets.

Breaking Changes

• Automatic HTML escaping — Unformatted body cell values are now automatically HTML-escaped before rendering. Cells that previously contained intentional raw HTML will now display the literal markup. Use the html() helper, fmt_passthrough(escape=False), fmt_markdown(), or other formatting methods to opt out of escaping for specific cells. (#881)

New Features

• fmt_fraction() — Format numeric values as mixed fractions (e.g., 2.5 becomes "2 1/2") with control over precision via accuracy= and a choice between "inline" and "diagonal" typesetting layouts. (#886)
• fmt_chem() — Format text as properly typeset chemical formulas and reactions, automatically rendering subscripts, superscripts, isotope notation, and reaction arrows. (#887)
• fmt_index() — Convert numeric values to alphabetical index characters (1 → "A", 2 → "B") with "repeat" and "excel" modes for values beyond the alphabet. (#888)
• fmt_url() and fmt_email() — Transform URLs and email addresses into clickable links with options for custom labels, button styling, and configurable colors. (#891)
• tab_style_body() — Apply styles to body cells based on their data values using exact matches, regex patterns, or predicate functions, with optional expansion to entire rows or columns. (#883)
• Footnote specs and ordering — New opt_footnote_spec() controls how footnote marks are styled using a compact DSL, and opt_footnote_order() controls the arrangement of marked vs. unmarked footnotes in the footer. (#884)
• load_dataset() — Load any of Great Tables' bundled datasets as a Pandas or Polars DataFrame through a single, consistent entry point. (#858)

Bug Fixes

• Interactive table sorting now uses original numeric values rather than formatted string representations, ensuring correct sort order for numeric columns. (#882)
• Booleans, missing values, and zero now render consistently across all backends. (#880)
• Nanoplots now drop positions whose x value is missing instead of breaking. (#889)
• Nanoplots warn when a curved data line is dropped due to insufficient x values. (#867)
• LaTeX output now correctly escapes backslash, tilde, and circumflex characters. (#872)
• LaTeX spanner rows now emit one cell per column. (#870)
• The ISO date style now correctly pads the year. (#869)
• Locale names are now accepted regardless of case. (#874)
• Options matching now uses prefix instead of substring, preventing false matches. (#873)
• HTTP scheme matching is now case-insensitive. (#876)
• Fractional parts are now preserved when scaling padding values. (#879)
• tab_spanner() is now a no-op when no columns or spanners are selected, instead of raising an error. (#865)
• The LSL currency symbol was corrected. (#890)

Enhancements

• cols_*() methods no longer raise an error when a list of column names are not all present in the data. (#860)

Documentation

• Added a glossary reference page. (#862)
• Added documentation for loc.footnotes(). (#885)