Chapter 07: Strings, JSON, and Regex

Why this chapter matters

Most useful programs cross text boundaries: command output, config files, logs, HTTP payloads, form fields, templates, and package metadata. This chapter teaches a practical order: use direct string words first, JSON words at structured boundaries, and regex only when pattern matching is truly the right tool.

What you will build

You will build a small log-cleaning workflow that cleans text, moves through JSON, and uses a regex for the one part direct string words do not express.

Concepts in plain English

A string is text. JSON is a text format for structured data. A regex is a compact pattern language for matching text. Regex is powerful, but direct words such as trim, split, contains?, starts_with?, and replace are usually easier to read.

Words introduced

Primary words: trim, trim_start, trim_end, blank?, slice, index_of, last_index_of, repeat, lines, chars, split, replace, contains?, starts_with?, ends_with?, uppercase, lowercase, concat, to_string, json_encode, json_decode, regex, matches?, regex_find, regex_replace, and captures.

Guided example

Run the log cleaner:

bashrco run examples/learn/07-strings-json-and-regex/log-cleaner.rco

The first half uses direct string words:

ricochet"  warn: user=ada  " trim cleaned var
$cleaned uppercase println
$cleaned "user=" contains? println
$cleaned ":" split count println

The JSON section turns a Ricochet map into a JSON string and back:

ricochetpayload map
$payload "ok" true put drop
$payload "name" "Ada" put drop
$payload json_encode json var
$json json_decode value "name" at println

Regex work follows the same result habit:

ricochet"\\d+" regex value digits var
"ticket-42" $digits regex_find "text" at println
$digits "ticket-42" "#" regex_replace println

How to read the code

" warn: user=ada " trim cleaned var starts with a raw string, trims whitespace, and names the cleaned result. The later lines read $cleaned instead of repeating the cleanup pipeline.

json_decode returns a Result. The example unwraps with value because the JSON was produced by json_encode immediately before. When JSON comes from a file, request, or user input, check ok? before unwrapping.

Try it

Try these direct string words before reaching for regex:

ricochet"Ricochet" 0 4 slice println
"Ada\nGrace" lines count println
"abc" chars "," join println
"warning" "warn" starts_with? println

Then intentionally decode bad JSON and inspect the error:

ricochet"not json" json_decode error "kind" at println

Check your understanding

• Use trim to clean edges before comparing text.
• Use split and lines for simple structure.
• Use json_decode at data boundaries and handle its Result.
• Use regex when direct string words cannot express the pattern clearly.

Common mistakes

• Treating JSON parse failure as an exception instead of a result path.
• Using regex where direct string words are simpler.
• Forgetting to escape backslashes in regex strings.
• Assuming concat is the only way to join text. Collections can use join.

What you know now

You can process text, encode and decode structured payloads, and keep regex work explicit and result-checked.