I could imagine consumers of this library wanting to ignore fixed periods when getting the string for a day period in certain locales. Although I wouldn't necessarily advise this usage, someone might prefix the English language day period with the word "Good" as a greeting to users ("Good morning", "Good afternoon", etc.). Without filtering on fixed periods, this would lead to awkward construction

This should work:

cat test.txt | unic-inspector

https://github.com/unicode-org/icu4x/blob/icu%400.1.0/components/icu/examples/tui.rs#L51

The range U+0000 to U+007F is ASCII, so it doesn't seem right to build latin1_set from that range. If I read the code correctly, it'll erroneously print "latin1 only: false" for any valid Latin 1 codepoint in the U+0080 to U+00FF range.