AkylaiBva commited on
Commit
f8aff73
·
verified ·
1 Parent(s): 842055d

Update tools.py

Browse files
Files changed (1) hide show
  1. tools.py +13 -18
tools.py CHANGED
@@ -40,24 +40,13 @@ import openpyxl
40
  import re
41
  from PIL import Image
42
  import pytesseract
43
- try:
44
- from duckduckgo_search import DDGS
45
- except ImportError:
46
- from ddgs import DDGS
47
 
48
 
49
  def web_search(query, max_results=3):
50
- with DDGS() as ddgs:
51
- results = ddgs.text(query, max_results=max_results)
52
- if not results:
53
- return "No results found."
54
- output = []
55
- for r in results:
56
- title = r.get('title')
57
- url = r.get('href')
58
- snippet = r.get('body')
59
- output.append(f"{title}\n{url}\n{snippet}\n")
60
- return "\n".join(output)
61
 
62
  def wiki_search(query):
63
  wikipedia.set_lang("en")
@@ -77,7 +66,7 @@ def get_youtube_transcript(url):
77
  from youtube_transcript_api import YouTubeTranscriptApi
78
  video_id = url.split("v=")[-1]
79
  transcript = YouTubeTranscriptApi.get_transcript(video_id)
80
- return " ".join([t["text"] for t in transcript])
81
 
82
  def speech_recognition(path):
83
  print(path)
@@ -92,8 +81,14 @@ def reverse_string(s):
92
 
93
  def query_image(path):
94
  print(path)
95
- img = Image.open(path)
96
- text = pytesseract.image_to_string(img)
 
 
 
 
 
 
97
  return text
98
 
99
  def query_video(path):
 
40
  import re
41
  from PIL import Image
42
  import pytesseract
43
+ from smolagents import DuckDuckGoSearchTool
44
+
 
 
45
 
46
 
47
  def web_search(query, max_results=3):
48
+ result = DuckDuckGoSearchTool(query)
49
+ return result
 
 
 
 
 
 
 
 
 
50
 
51
  def wiki_search(query):
52
  wikipedia.set_lang("en")
 
66
  from youtube_transcript_api import YouTubeTranscriptApi
67
  video_id = url.split("v=")[-1]
68
  transcript = YouTubeTranscriptApi.get_transcript(video_id)
69
+ return transcript
70
 
71
  def speech_recognition(path):
72
  print(path)
 
81
 
82
  def query_image(path):
83
  print(path)
84
+ import easyocr
85
+
86
+ reader = easyocr.Reader(['en']) # Add 'ru', 'tr', etc. if needed
87
+ result = reader.readtext('image.jpg')
88
+
89
+ text = ""
90
+ for bbox, text, confidence in result:
91
+ text += f"{text} (confidence: {confidence})"
92
  return text
93
 
94
  def query_video(path):