runbox / runbox7

Runbox 7 web app
Other
127 stars 26 forks source link

Searching for terms having diacritics #1259

Open kotecky opened 2 years ago

kotecky commented 2 years ago

(Following a discussion started here, @gtandersen)

Searching in Runbox does not return results independently on diacritics. Searching for Jérôme and Jerome returns two different sets of results.

The way it should be: Searching should return results independently on diacritics.

Background: most languages contain diacritic signs. In some the frequency is so high (Slavic) that every second word has one or more accents (example: https://cs.wikipedia.org/). This leads to a number of issues when searching. In these languages the email addresses are often different from names. So when looking for an email from Jonáš will not include emails where jonas@abc.org is in To or in Cc.

Another similar issue comes from the way people write: in short or informal emails diacritic signs are often omitted. This leads to the same issue of searches returning only partial results. Very frustrating!

Outlook, Gmail, Firefox and Acrobat (surely among many other software) yield the same results when searching for façon and facon. Runbox should too.

gtandersen commented 2 years ago

Thanks for the suggestion @kotecky -- we will consider this when we make improvements to the search functionality.