Simple

The type is made of four straightforward variants:

• Empty
• String (built-in string)
• Enriched (enrichment, rich string)
• Join (separator, rich strings)

Composable

Just like the built-in string, rich strings form a monoid.

• Identity: for all rich strings rs, empty ++ rs = rs ++ empty = rs.
• Associativity: for all rich strings rs1, rs2 and rs3, (rs1 ++ rs2) ++ rs3 = rs1 ++ (rs2 ++ rs3).

Optimizing

Rich strings are carefully crafted for common cases. Operations on rich strings avoid unnecessary structural deep-nesting when possible.

• Identity: empty ++ rs = rs ++ empty = rs.
• Inner concatenation: join ~on:sep rss1 ++ join ~on:sep rss2 = join ~on:sep (rss1 @ rss2).
• Inner consing: rs ++ join ~on:empty rss = join ~on:empty (rs :: rss).
• Enrichement grouping: enrich e rs1 ++ enrich e rs2 = enrich e (rs1 ++ rs2).

Extensible

With a mostly transparent interface, rich strings can be extended to fulfill specific needs.

For example:

• Indexing
• Labelling
• ANSI formatting

Tested

As a library, rich-string is tested from the ground-up, with a focus on checking that its advertised properties hold.

Convinced?

Add rich-string to your dune-project, in the depends section:

(package
  (name your-package)
  (depends
    ...
    (rich-string (>= 1.0))))

rich-string is available on the OPAM index.

opam install rich-string

Made with 🩷 by lexa. Support me on Ko-fi. Find me on Bluesky.

rich-string's development does not involve any large language model.