rich-struct

rich's print, with 25 themes (23 popular editor themes and two originals) and helpers for Markdown, JSON, YAML, XML, Python objects and tables (lists, numpy arrays, pandas DataFrames and Series).

pip install rich-struct          # add [yaml] to dump objects with print_yaml

from rich.markdown import Markdown
from rich.theme import Theme
from rich_struct import ThemeName, print, print_md, print_json, print_yaml, print_xml, print_obj, print_syntax, print_table

print("[bold]hello[/bold]", {"a": 1})                 # drop-in rich.print, Midnight Slate by default
print(Markdown("## Printing structs"), theme=ThemeName.DRACULA)

print_md("# Title\n\n```python\nprint('hi')\n```")
print_json({"a": [1, 2, None]}, theme="Nord")         # enum value as a string works too
print_yaml("a: 1\nb: [x, y]")
print_xml("<a><b c='d'/></a>")
print_obj(my_dataclass, theme=ThemeName.GITHUB_LIGHT)
print_syntax("SELECT 1", "sql")                       # any pygments lexer
print_table(df, title="Sales")                        # DataFrame, Series, numpy array, list...
print(df)                                             # arrays and frames become tables in print too

print_md("# Mine", theme=Theme({"markdown.h1": "bold magenta", "syntax.keyword": "red"}))

Every function takes theme=: a ThemeName, its string value, or a rich Theme. To change the theme for all of them at once, call set_default_theme("Dracula") (same accepted values; no argument resets to Midnight Slate). Passing None to any of them prints None (JSON: null) in the theme's colour for null values instead of failing. ThemeName.X.theme gives you the underlying rich Theme to use with your own Console.

Screenshots

One per kind of struct, each in a different theme. The demo notebook shows every kind in all 25 themes; open it in Jupyter or VS Code to see the colours.

Markdown — print_md, Midnight Slate (the default)

JSON — print_json, Nord

YAML — print_yaml, One Dark Pro

XML — print_xml, GitHub Dark

Python objects — print_obj, Catppuccin Mocha

Code — print_syntax, One Light

Lists — print_table, Catppuccin Latte

NumPy arrays — print_table, Solarized Dark

pandas DataFrames — print_table, GitHub Light

pandas Series — print_table, Night Owl

Tables

print_table(data, title=None, max_rows=20, max_cols=10) accepts:

• pandas DataFrame and Series, numpy arrays of any dimension (more than 2: one table per slice)
• lists of dicts (records), lists of lists (a grid), and any other sequence such as array.array or range (one column)

Numbers are right-aligned, floats in a column share their decimals, and missing values are dimmed. Long data shows its head and tail around a … gap, and columns that don't fit the terminal width go into the gap as well. to_table(data) returns the renderable instead of printing it. numpy and pandas are not dependencies.

Themes

Themes are built from each theme's colour palette, not ported token-by-token from the editor.

Output is painted on the theme's background across the full terminal width. Pass background=False to keep the terminal's own background; nothing is painted or padded when the output has no colour (e.g. piped to a file). A custom Theme gets a background by defining a background style, e.g. "background": "#c0caf5 on #1a1b26".

Code colours come from syntax.* styles (text, comment, keyword, string, number, function, type, tag, attribute, operator); a custom Theme may override any of them, and may set syntax.block (e.g. "on #111c2e") to shade code blocks inside Markdown, and repr.background to give print_obj output its own background.