docs(search): cover ?/bracket wildcards and the no-stemmer fallback

This commit is contained in:
Trenton Holmes authored and GitHub committed 2026-09-14 21:05:58 -07:00
1 parent 6c603e9855
commit add910c0a7
1 file changed
+31 -6
+31 -6
View File
@@ -988,14 +988,39 @@ Matching inexact words:
``` ```
invoice* invoice*
title:Invoice* title:Invoice*
20[12]?
20[!0]?
``` ```
Wildcards are matched against the _stemmed_ terms stored in the index, not `*` matches any run of characters, `?` matches exactly one character, and
against the words as they appear in the document. Each literal part of a `[...]` matches any one character from the set or range inside the brackets
pattern is tried both as you typed it and in its stemmed form, so a trailing (`[!...]` negates it) — the same glob syntax as a Unix shell. `20[12]?`
`*` matches a word and its inflections (`invoice*` finds "invoice", "invoices" matches years from 2010-2029, and `20[!0]?` excludes the 2000s. Digits are a
and "invoiced") as well as longer words whose stored term still begins with reliable place to see this literally, since they pass through lowercasing,
what you typed (`copy*` finds "copyright" alongside "copy" and "copies"). ASCII-folding and stemming unchanged; matching `?` or a bracket class against
ordinary stemmed text is subject to the same stemming quirks described below
for `*`, and just as easy to get a surprising zero-result from.
A `[...]` class is only recognized as pattern syntax when the value also
contains a `*` or `?` elsewhere; on its own it is ordinary text. A bracket
range that would otherwise be ambiguous with a single character, such as
`title:200[1-9]`, is rejected with an error rather than silently searched as
a literal string that will not match anything real.
Wildcards on the text fields are matched against the _stemmed_ terms stored
in the index, not against the words as they appear in the document, but only
when a [supported stemmer language](#PAPERLESS_SEARCH_LANGUAGE) is
configured. Each literal part of a pattern is tried both as you typed it and
in its stemmed form, so a trailing `*` matches a word and its inflections
(`invoice*` finds "invoice", "invoices" and "invoiced") as well as longer
words whose stored term still begins with what you typed (`copy*` finds
"copyright" alongside "copy" and "copies"). If the configured language has no
stemmer, or none is configured and none can be inferred from the OCR
language, the index keeps unstemmed (lowercased, ASCII-folded) terms instead,
and only the literal typed prefix is tried -- `copy*` then still finds "copy"
and "copyright" (both literally start with "copy"), but no longer finds
"copies" (stored as-typed, its surface spelling "copies" does not start with
"copy").
It is still not a plain prefix search over the original text. A trailing `*` It is still not a plain prefix search over the original text. A trailing `*`
matches a stored term when either the run you typed or its stemmed form is a matches a stored term when either the run you typed or its stemmed form is a