Support mode modifiers

[Aloso]

Section about mode modifiers in popular regex engines

Here are the best supported mode modifiers:

• (?i): case insensitive
• (?m): multi-line mode (^ and $ match begin/end of line)
• (?s): single-line mode (allow . to match \n)
• (?x): free-spacing mode
• (?xx): free-spacing mode even in character classes
• (?U): make greedy quantifiers lazy and vice versa (a*? <--> a*)
• (?n): make all capturing groups without a name non-capturing ((a) --> (?:a))
• (?J): allow duplicate group names
• (?d): opt-out of Unicode line break support, so ., ^ and $ only treat ASCII \n as line break
• (?X): makes invalid escape sequences an error

Some of these aren't needed because rulex has sane defaults:

• s isn't needed because you can use [cp] or Grapheme in rulex
• x and xx isn't needed because rulex is already free-spacing
• n isn't needed because groups are non-capturing by default
• X isn't needed because rulex doesn't have escape sequences, and invalid identifiers are an error

End of section

---

The mode modifiers we want to support are i (ignore case), m (multiline), J (duplicate group names), d (ASCII-only \n line breaks) and U (lazy quantifiers by default). The will be available with the following names:

• [ ] ignore_case
• [ ] multiline
• [ ] reuse_groups
• [ ] ascii_line_breaks
• [x] lazy

They can be enabled and disabled with the enable mode;/disable mode; syntax. For example:

enable lazy;
'hello ' (disable ignore_case; 'world')*

Note that JavaScript doesn't allow mode modifiers within a regex, but the i and m modifiers are instead available as flags for the entire regex. So we can support enable ignore_case; 'hello world', which compiles to /hello world/i. This means that we need to support outputting flags.