Advertisement
Guest User

Untitled

a guest
Jun 27th, 2019
65
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 2.39 KB | None | 0 0
  1. import pandas as pd
  2. import os
  3. import re
  4.  
  5. root = r"my_dir"
  6.  
  7. agg_df = pd.DataFrame()
  8.  
  9. for directory, subdirectory, files in os.walk(root):
  10. for file in files:
  11. if file.endswith('.xlsm'):
  12. filepath = os.path.join(directory, file)
  13. xls = pd.ExcelFile(filepath)
  14. for i in xls.sheet_names:
  15. if re.search(r'Apples', i):
  16. df_temp = pd.read_excel(filepath, sheet_names=i)
  17. df_temp['Filepath'] = filepath
  18. df_temp['Sheet_Name'] = i
  19. agg_df = agg_df.append(df_temp)
  20. elif re.search(r'Oranges', i):
  21. df_temp = pd.read_excel(filepath, sheet_names=i)
  22. df_temp['Filepath'] = filepath
  23. df_temp['Sheet_Name'] = i
  24. agg_df = agg_df.append(df_temp)
  25. elif re.search('Grapes', i):
  26. df_temp = pd.read_excel(filepath, sheet_names=i)
  27. df_temp['Filepath'] = filepath
  28. df_temp['Sheet_Name'] = i
  29. agg_df = agg_df.append(df_temp)
  30. elif re.search(r'Tomatoes', i):
  31. df_temp = pd.read_excel(filepath, sheet_names=i)
  32. df_temp['Filepath'] = filepath
  33. df_temp['Sheet_Name'] = i
  34. agg_df = agg_df.append(df_temp)
  35. elif re.search(r'Peaches', i):
  36. df_temp = pd.read_excel(filepath, sheet_names=i)
  37. df_temp['Filepath'] = filepath
  38. df_temp['Sheet_Name'] = i
  39. agg_df = agg_df.append(df_temp)
  40. elif re.search(r'Pears', i):
  41. df_temp = pd.read_excel(filepath, sheet_names=i)
  42. df_temp['Filepath'] = filepath
  43. df_temp['Sheet_Name'] = i
  44. agg_df = agg_df.append(df_temp)
  45. elif re.search(r'Bananas', i):
  46. df_temp = pd.read_excel(filepath, sheet_names=i)
  47. df_temp['Filepath'] = filepath
  48. df_temp['Sheet_Name'] = i
  49. agg_df = agg_df.append(df_temp)
  50. elif re.search(r'Mangos', i):
  51. df_temp = pd.read_excel(filepath, sheet_names=i)
  52. df_temp['Filepath'] = filepath
  53. df_temp['Sheet_Name'] = i
  54. agg_df = agg_df.append(df_temp)
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement