Skip to main content

Topic: search engine contest (Read 524 times)

  • Quetzal
  • [*][*][*]
search engine contest
hi all
strange title, i do admit it easily

then, i've got a nice ine-line dictionnaie Glosbe.com wich provide plenty of dictionnaire in many many language
i'm curious and like to have à look for spell in other language

i tried to built one serarchEngine in Otterbro... but their whatever their address is just "simple" ...glosbe/fr/en/{SearchTerms}
impossible to modify the letter of second dictionnaire by this way [adress bar] glo en/inaction => ...glosbe/fr/{ en/inaction }

the slash of en/  that should glue fr/ to give fr/en/{} is ever miserabaliy tranform in a en%2f{searchTerms} in the adress bar (normal)

do there is a way to type correctly this, or i have to make a new serachengine for any language of this website ???

thanks a lot for readding ;)

  • beastie
  • [*][*][*]
Re: search engine contest
Reply #1
or i have to make a new serachengine for any language of this website
That's what I do. How can Otter know the language of the word you're searching for?

Automatic language detection may happen on the side of the target website, as is the case for Google Translate. But not all websites do that.

Theoretically Otter may detect the HTML element's language through the lang attribute, but some pages don't use the attribute properly, or at all. Otter may also detect the word's language (using Google's Cloud API, CFFI-CLD2, NLTK, etc.) and pass it to the search engine as a variable similar to {searchTerms}, say {searchLanguage}.

IMHO these methods are all prone to errors and quite complicated to be implemented in a clean way, so personally I'll just keep creating multiple search engines.

  • ersi
  • [*][*][*][*][*]
Re: search engine contest
Reply #2
Yes, I also create multiple search engines when I need multiple languages either on translation websites or on Wikipedia. Just make sensible keywords (aliases) for them and you will be able to find them no matter how many there are.

Sensible such as
enfr for English to French
defr for German to French
rufr for Russian to French
wifr for Wikipedia in French
wien for Wikipedia in English
etc. in the same pattern. Of course you may find a better pattern suitable for yourself.

{searchTerms} is good only for actual search terms.
  • Last Edit: 2019-08-11, 09:26:13 by ersi

  • Quetzal
  • [*][*][*]
Re: search engine contest
Reply #3
well, i was thinking, that it was may be possible to "emulate" the slash "/" that ends the adress of the wesite between /fr/en {searchterms}

but if it's not, i will do many search engines...

and here a new problem... i already got a lot of search engine... and make a long list... but it's another problem (of the  display in Otter this list)

thanks a lot for the try... if a find a way, i will come back here ;)



  • Frenzie
  • [*][*][*][*][*]
  • Administrator
Re: search engine contest
Reply #4
The DuckDuckGo bangs could also be of interest (if you happen to use that search engine).

  • ersi
  • [*][*][*][*][*]
Re: search engine contest
Reply #5
The DuckDuckGo bangs could also be of interest (if you happen to use that search engine).
Ah yes, such as !m for google maps and !imdb for internet movie database.
Here's the growing list https://duckduckgo.com/bang
Translation engines are still more convenient by configuring the browser's search engines, or by bookmarking.