fix(compile): use rune slice to truncate history summary — avoids UTF-8 boundary corruption

[mpeter]

Root cause

buildCompiledArticle in tools/compile.go truncates learning content for the history table using a byte slice:

summary = summary[:80] + "..."

In Go, string[i:j] slices by bytes, not characters. When a multibyte UTF-8 character (e.g. em-dash U+2014, encoded as 0xE2 0x80 0x94) spans the byte-80 boundary, the slice cuts the character mid-encoding. Appending "..." then places 0x2E (period) as the third byte of the incomplete sequence, producing 0xE2 0x80 0x2E — invalid UTF-8.

Fix

Convert to []rune before slicing, then back to string:

if runes := []rune(summary); len(runes) > 80 {
    summary = string(runes[:80]) + "..."
}

Reproducer

Any learning whose information field contains a multibyte character (em-dash, curly quote, arrows, CJK, etc.) where the UTF-8 encoding spans byte offset 80. Observed in production on the process-git compiled article — every compile produced 0xE2 0x80 0x2E causing UnicodeDecodeError in Python consumers of the article.

Testing

go test ./tools/...  # 438 tests pass