fix header name for browser languages

[nominatim.git] / settings / icu_tokenizer.yaml

diff --git a/settings/icu_tokenizer.yaml b/settings/icu_tokenizer.yaml
index 16339970c2ae5727cf606fe686a607814cfec2a2..1fa467befebfa4c5977103d6731064dd760791ff 100644 (file)
--- a/settings/icu_tokenizer.yaml
+++ b/settings/icu_tokenizer.yaml
@@ -24,6 +24,7 @@ transliteration:
     - ":: lower ()"
     - "[^a-z0-9[:Space:]] >"
     - ":: NFC ()"
     - ":: lower ()"
     - "[^a-z0-9[:Space:]] >"
     - ":: NFC ()"

+    - "[:Space:]+ > ' '"

 sanitizers:
     - step: clean-housenumbers
       filter-kind:
 sanitizers:
     - step: clean-housenumbers
       filter-kind:


@@ -37,6 +38,7 @@ sanitizers:
       default-pattern: "[A-Z0-9- ]{3,12}"
     - step: clean-tiger-tags
     - step: split-name-list
       default-pattern: "[A-Z0-9- ]{3,12}"
     - step: clean-tiger-tags
     - step: split-name-list

+      delimiters: ;

     - step: strip-brace-terms
     - step: tag-analyzer-by-language
       filter-kind: [".*name.*"]
     - step: strip-brace-terms
     - step: tag-analyzer-by-language
       filter-kind: [".*name.*"]