How to search for items with an embedded hyphen?

[cmcknight]

I have a use case for item titles that have embedded hyphens, i.e.:

C-8
AR-7
PGMP-15

I seem to get back a lot of completely unrelated results that do not have the search terms in them. I do find the item I'm looking for about halfway or two-thirds down the list.

This is using the generic search (no ands, etc.). I feel like I'm missing something.

[JustAnotherArchivist]

I suspect that this is what you're looking for: https://github.com/lucaong/minisearch#tokenization

Specifically, the default tokenize function uses this regular expression for splitting, which includes the hyphen (as part of the ,-/ range):

minisearch/src/MiniSearch.ts

Line 1239 in 8c66bd2

const SPACE_OR_PUNCTUATION = /[\n\r -#%-*,-/:;?@[-\]_{}\u00A0\u00A1\u00A7\u00AB\u00B6\u00B7\u00BB\u00BF\u037E\u0387\u055A-\u055F\u0589\u058A\u05BE\u05C0\u05C3\u05C6\u05F3\u05F4\u0609\u060A\u060C\u060D\u061B\u061E\u061F\u066A-\u066D\u06D4\u0700-\u070D\u07F7-\u07F9\u0830-\u083E\u085E\u0964\u0965\u0970\u09FD\u0A76\u0AF0\u0C77\u0C84\u0DF4\u0E4F\u0E5A\u0E5B\u0F04-\u0F12\u0F14\u0F3A-\u0F3D\u0F85\u0FD0-\u0FD4\u0FD9\u0FDA\u104A-\u104F\u10FB\u1360-\u1368\u1400\u166E\u1680\u169B\u169C\u16EB-\u16ED\u1735\u1736\u17D4-\u17D6\u17D8-\u17DA\u1800-\u180A\u1944\u1945\u1A1E\u1A1F\u1AA0-\u1AA6\u1AA8-\u1AAD\u1B5A-\u1B60\u1BFC-\u1BFF\u1C3B-\u1C3F\u1C7E\u1C7F\u1CC0-\u1CC7\u1CD3\u2000-\u200A\u2010-\u2029\u202F-\u2043\u2045-\u2051\u2053-\u205F\u207D\u207E\u208D\u208E\u2308-\u230B\u2329\u232A\u2768-\u2775\u27C5\u27C6\u27E6-\u27EF\u2983-\u2998\u29D8-\u29DB\u29FC\u29FD\u2CF9-\u2CFC\u2CFE\u2CFF\u2D70\u2E00-\u2E2E\u2E30-\u2E4F\u3000-\u3003\u3008-\u3011\u3014-\u301F\u3030\u303D\u30A0\u30FB\uA4FE\uA4FF\uA60D-\uA60F\uA673\uA67E\uA6F2-\uA6F7\uA874-\uA877\uA8CE\uA8CF\uA8F8-\uA8FA\uA8FC\uA92E\uA92F\uA95F\uA9C1-\uA9CD\uA9DE\uA9DF\uAA5C-\uAA5F\uAADE\uAADF\uAAF0\uAAF1\uABEB\uFD3E\uFD3F\uFE10-\uFE19\uFE30-\uFE52\uFE54-\uFE61\uFE63\uFE68\uFE6A\uFE6B\uFF01-\uFF03\uFF05-\uFF0A\uFF0C-\uFF0F\uFF1A\uFF1B\uFF1F\uFF20\uFF3B-\uFF3D\uFF3F\uFF5B\uFF5D\uFF5F-\uFF65]+/u

Both on indexing and on searching, that therefore gets split into C and 8 for example and is functionally equivalent to entering C 8.

[lucaong]

That’s correct @JustAnotherArchivist , the default tokenizer splits by space or punctuation, and - is considered punctuation. One solution is to specify a custom tokenize function that doesn’t split hyphens.

Unfortunately the modified regular expression is quite cumbersome if one needs to take care of non-English characters, diacritics, etc., but this should work:

// This is almost the same regexp as
// the default one, but excludes the
// hyphen by having ,./ instead of ,-/
const CUSTOM_SPACE_OR_PUNCT = /[\n\r -#%-*,./:;?@[-\]_{}\u00A0\u00A1\u00A7\u00AB\u00B6\u00B7\u00BB\u00BF\u037E\u0387\u055A-\u055F\u0589\u058A\u05BE\u05C0\u05C3\u05C6\u05F3\u05F4\u0609\u060A\u060C\u060D\u061B\u061E\u061F\u066A-\u066D\u06D4\u0700-\u070D\u07F7-\u07F9\u0830-\u083E\u085E\u0964\u0965\u0970\u09FD\u0A76\u0AF0\u0C77\u0C84\u0DF4\u0E4F\u0E5A\u0E5B\u0F04-\u0F12\u0F14\u0F3A-\u0F3D\u0F85\u0FD0-\u0FD4\u0FD9\u0FDA\u104A-\u104F\u10FB\u1360-\u1368\u1400\u166E\u1680\u169B\u169C\u16EB-\u16ED\u1735\u1736\u17D4-\u17D6\u17D8-\u17DA\u1800-\u180A\u1944\u1945\u1A1E\u1A1F\u1AA0-\u1AA6\u1AA8-\u1AAD\u1B5A-\u1B60\u1BFC-\u1BFF\u1C3B-\u1C3F\u1C7E\u1C7F\u1CC0-\u1CC7\u1CD3\u2000-\u200A\u2010-\u2029\u202F-\u2043\u2045-\u2051\u2053-\u205F\u207D\u207E\u208D\u208E\u2308-\u230B\u2329\u232A\u2768-\u2775\u27C5\u27C6\u27E6-\u27EF\u2983-\u2998\u29D8-\u29DB\u29FC\u29FD\u2CF9-\u2CFC\u2CFE\u2CFF\u2D70\u2E00-\u2E2E\u2E30-\u2E4F\u3000-\u3003\u3008-\u3011\u3014-\u301F\u3030\u303D\u30A0\u30FB\uA4FE\uA4FF\uA60D-\uA60F\uA673\uA67E\uA6F2-\uA6F7\uA874-\uA877\uA8CE\uA8CF\uA8F8-\uA8FA\uA8FC\uA92E\uA92F\uA95F\uA9C1-\uA9CD\uA9DE\uA9DF\uAA5C-\uAA5F\uAADE\uAADF\uAAF0\uAAF1\uABEB\uFD3E\uFD3F\uFE10-\uFE19\uFE30-\uFE52\uFE54-\uFE61\uFE63\uFE68\uFE6A\uFE6B\uFF01-\uFF03\uFF05-\uFF0A\uFF0C-\uFF0F\uFF1A\uFF1B\uFF1F\uFF20\uFF3B-\uFF3D\uFF3F\uFF5B\uFF5D\uFF5F-\uFF65]+/u

const miniSearch = new MiniSearch({
  // …
  tokenize: (text) => text.split(CUSTOM_SPACE_OR_PUNCT)
})

[lucaong]

Closing this issue for now, as the above solution should work. @cmcknight feel free to comment further if necessary, and I will reopen the issue in case.

[cmcknight]

Hi All,

Sorry for not replying sooner. I've been sidelined due to an illness.

This makes sense and I agree that it can be closed.