X-Git-Url: https://git.openstreetmap.org./nominatim.git/blobdiff_plain/fb54bd3fcff5a8a44ae59f0d552ed316b08966b3..55b7edbab545346f7caa0cf9c708f18f74e16013:/settings/icu_tokenizer.yaml diff --git a/settings/icu_tokenizer.yaml b/settings/icu_tokenizer.yaml index a3c62e67..bebd49e9 100644 --- a/settings/icu_tokenizer.yaml +++ b/settings/icu_tokenizer.yaml @@ -25,6 +25,13 @@ transliteration: - "[^a-z0-9[:Space:]] >" - ":: NFC ()" sanitizers: + - step: clean-housenumbers + filter-kind: + - housenumber + - conscriptionnumber + - streetnumber + convert-to-name: + - (\A|.*,)[^\d,]{3,}(,.*|\Z) - step: split-name-list - step: strip-brace-terms - step: tag-analyzer-by-language @@ -34,6 +41,8 @@ sanitizers: mode: append token-analysis: - analyzer: generic + - id: "@housenumber" + analyzer: housenumbers - id: bg analyzer: generic mode: variant-only @@ -59,6 +68,13 @@ token-analysis: mode: variant-only variants: - !include icu-rules/variants-de.yaml + mutations: + - pattern: ä + replacements: ["ä", "ae"] + - pattern: ö + replacements: ["ö", "oe"] + - pattern: ü + replacements: ["ü", "ue"] - id: el analyzer: generic mode: variant-only