Can't scrape dinamic site. A lot of pages with diferent URL

Hi everybody
I need help. The site I want to scrape is a Real estate one. I need to scrape only the bold numbers per page. They are part of text. THESE words /1-СТАЕН; СВЕТА ТРОИЦА; ПАНЕЛ/ are variable. I want to scrape in separate columns number of offers/15/, average price/47969/, and average price per sq. m./1022/. In normal search the steps are:1.Fill filter – district /Света троица/; 1 bedroom /1-стаен/; feature/панел/. 2. Search – result is down-I need the bold numbers. 3. Change feature in filter /brick/.4. Search – result is similar to down-I need the bold numbers.5. Change feature in filter/Epk/.6. Search – result is similar to down-I need the bold numbers. The result might be null. Then repeat all steps for 2 bedroom apartments in the same district. Then for 3 bedrooms apartments in the same district. I want to apply this to 1-2-3 bedrooms apartments in all districts of the city.
I tried many times to scrape. Usually take only the second number/47969/. And after export, the number is not in a separate column. It is part of text. Also problem is that every new search is a new URL. Please help.
site is www.imot.bg
This is needed part of the searched result:
1 - 15 от общо 15 обяви - Продава
Резултат от Вашето търсене на:
Вид имот: 1-СТАЕН,
Местоположение: град София
Под райони: СВЕТА ТРОИЦА,
Особености: ПАНЕЛ
Да не се показват обяви, докладвани за проблем,
Средна цена на имотите в извадката: 47969 euro
Средна цена на кв. м: 1022 euro.
Запази търсенетоOn-line известяване
Страница на резултата от търсене: