]> git.openstreetmap.org Git - nominatim.git/blobdiff - settings/icu_tokenizer.yaml
reintroduce brand and remove etymology
[nominatim.git] / settings / icu_tokenizer.yaml
index f682bbcdf8ad4c38a4e891e55f5360e70a7a28d0..530df1a63397e423355ec7a75dc36178e2c2418e 100644 (file)
@@ -1,3 +1,5 @@
+query-preprocessing:
+    - step: normalize
 normalization:
     - ":: lower ()"
     - ":: Hans-Hant"
@@ -24,6 +26,7 @@ transliteration:
     - ":: lower ()"
     - "[^a-z0-9[:Space:]] >"
     - ":: NFC ()"
+    - "[:Space:]+ > ' '"
 sanitizers:
     - step: clean-housenumbers
       filter-kind:
@@ -34,18 +37,23 @@ sanitizers:
         - (\A|.*,)[^\d,]{3,}(,.*|\Z)
     - step: clean-postcodes
       convert-to-address: yes
-      default-pattern: [A-Z0-9- ]{3,12}
+      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.*"]
       whitelist: [bg,ca,cs,da,de,el,en,es,et,eu,fi,fr,gl,hu,it,ja,mg,ms,nl,no,pl,pt,ro,ru,sk,sl,sv,tr,uk,vi]
       use-defaults: all
       mode: append
+    - step: tag-japanese
 token-analysis:
     - analyzer: generic
     - id: "@housenumber"
       analyzer: housenumbers
+    - id: "@postcode"
+      analyzer: postcodes
     - id: bg
       analyzer: generic
       mode: variant-only