CatalogueOfLife / general

The Catalogue of Life
49 stars 5 forks source link

allow for phonetic match #75

Closed mfrasca closed 4 years ago

mfrasca commented 4 years ago

(only searched through the issues, maybe you already have this?)

in the GPL software I maintain, I assume users will make typing mistakes, like typing Calanchoe when they mean Kalanchoe, or Rhododendron without the 'h'. I have seen such mistakes at several botanical gardens, not just made by amateurs. maybe the Latin root is the source of trouble, but whatever the cause, I have added a 'phonetic' column to my taxonomy table, where I store a simplified version of the epithet, something like this:

3005|Ancalanthus|5|1|3319|Balf. f.|ankalantus
3011|Aphelandrella|5|1|3465|Mildbr.|afilandrila
3012|Lagochilium|5|1|3465|Nees|lakukilum
3119|Cyrtanthera|5|1|3390|Nees|sirtantira
3110|Rhacodiscus|5|1|3390|Lindau|rakudiskus

it would be nice if your web-service would implement a similar phonetic match, too.

mdoering commented 4 years ago

Our search does that with various options. We have a very restricted normalisation for common mispellings like the silent h or double letters. That gets applied everywhere AFAIR. Then we also provide a generic fuzzy matching search which is an option that users can control through a http parameter.