rav4s
/
iFixit-Repairability-Dashboard
mirror of https://github.com/Rav4s/iFixit-Repairability-Dashboard
1
1
Fork 0
Code Issues Projects Releases Wiki Activity
Pulls repairability scores for various devices from iFixit, which are then displayed on a website in a few categories.
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
2 Commits
1 Branch
0 Tags
170 KiB
Tree: 8690994b93
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from '8690994b93'
${ noResults }
iFixit-Repairability-Dashboard/scraper.py

53 lines
2.0 KiB
Raw Normal View History

Initail Commit
5 years ago
Added code to scrape iFixit smartphone website and generate lists of each metric
5 years ago
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
  1. import time
  2. from bs4 import BeautifulSoup
  3. import requests
  4. import re
  6. def get_device_manufacturers(device_names):
  7. list_of_manufacturers_without_tags = []
  8. for i in device_names:
  9. i = str(i)
  10. i = i.replace('<div class="cell device-name">\n','')
  11. i = i.lstrip()
  12. i = i.split('<')[0]
  13. i = i.rstrip()
  14. list_of_manufacturers_without_tags.append(i)
  15. return list_of_manufacturers_without_tags
  17. def get_device_models(device_names, soup):
  18. list_of_models_without_tags = []
  19. list_of_models = soup.findAll("span", {"class": "selected"})
  20. for i in list_of_models:
  21. i = str(i)
  22. i = i.replace('<span class="selected">','').replace('</span>','')
  23. list_of_models_without_tags.append(i)
  24. return list_of_models_without_tags
  26. def format_device_scores(device_scores, soup):
  27. list_of_scores_without_tag = []
  28. list_of_scores = soup.findAll("h3")
  29. for i in list_of_scores:
  30. i = str(i)
  31. i = i.replace('<h3>','').replace('</h3>','')
  32. list_of_scores_without_tag.append(i)
  33. return list_of_scores_without_tag
  35. def data_from_each(list_of_manufacturers_without_tags, list_of_models_without_tags, list_of_scores_without_tag):
  36. j = 0
  37. for i in list_of_manufacturers_without_tags:
  38. print(list_of_manufacturers_without_tags[j], list_of_models_without_tags[j], list_of_scores_without_tag[j])
  39. j = j + 1
  41. def main():
  42. link = "https://www.ifixit.com/smartphone-repairability"
  43. page = requests.get(link)
  44. soup = BeautifulSoup(page.content, 'html.parser')
  45. device_names = soup.findAll("div", {"class": "cell device-name"})
  46. device_scores = soup.findAll("div", {"class": "cell device-score"})
  47. list_of_manufacturers_without_tags = get_device_manufacturers(device_names)
  48. list_of_models_without_tags = get_device_models(device_names, soup)
  49. list_of_scores_without_tag = format_device_scores(device_scores, soup)
  50. data_from_each(list_of_manufacturers_without_tags, list_of_models_without_tags, list_of_scores_without_tag)
  52. if __name__ == '__main__':
  53. main()