Escape sequences are tokenized but never interpreted; backslashes leak into output #21

New issue

Closed

opened 2026-08-07 13:33:07 +02:00 by Hemera · 0 comments

Hemera commented

2026-08-07 13:33:07 +02:00

Owner

String parsing uses take_escaped(none_of(['"', '\\']), '\\', one_of(['"', 'n'])) (src/lexer/mod.rs:445), which accepts \" and \n, but the backslash is captured verbatim and nothing unescapes the value afterwards.

Verified output:

{{= "a\nb" }} → outputs the 4 literal characters a\nb (backslash + n), not a newline.
{{= "say \"hi\"" }} → outputs say \"hi\" (backslashes preserved).

So the escape syntax works (the string doesn't terminate early) but the escapes have no effect on the value. Either interpret \n/\" (and reject unknown escapes), or document that backslashes are kept literally.

String parsing uses `take_escaped(none_of(['"', '\\']), '\\', one_of(['"', 'n']))` (`src/lexer/mod.rs:445`), which accepts `\"` and `\n`, but the backslash is captured verbatim and nothing unescapes the value afterwards. Verified output: - `{{= "a\nb" }}` → outputs the 4 literal characters `a\nb` (backslash + `n`), not a newline. - `{{= "say \"hi\"" }}` → outputs `say \"hi\"` (backslashes preserved). So the escape *syntax* works (the string doesn't terminate early) but the escapes have no effect on the value. Either interpret `\n`/`\"` (and reject unknown escapes), or document that backslashes are kept literally.

Hemera added the

automated-🤖

label

2026-08-07 13:34:42 +02:00

Hemera referenced this issue from a pull request that will close it,

2026-08-09 09:47:24 +02:00

Use a String instead of input to allow for escaping #28

Hemera closed this issue

2026-08-09 09:48:14 +02:00