Wajahat698 commited on
Commit
a2b6da3
·
verified ·
1 Parent(s): 47ae234

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +28 -2
app.py CHANGED
@@ -157,10 +157,36 @@ def merge_markdown_contents(contents):
157
  Merge multiple Markdown contents into a single Markdown string.
158
  """
159
  merged_content = "\n\n---\n\n".join(contents)
160
- return merged_content
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
161
 
162
  def upload_to_firebase(user_id, file):
163
- content = convert_file_to_md(file)
164
  if not content:
165
  return None, "Failed to convert file to content."
166
 
 
157
  Merge multiple Markdown contents into a single Markdown string.
158
  """
159
  merged_content = "\n\n---\n\n".join(contents)
160
+ return
161
+
162
+
163
+
164
+ def extract_text_from_file(file):
165
+ """
166
+ Extract text content from an uploaded file based on its type.
167
+ Args:
168
+ file: The uploaded file object.
169
+ Returns:
170
+ str: Extracted text content.
171
+ """
172
+ try:
173
+ if file.name.endswith(".pdf"):
174
+ pdf_reader = PdfReader(file)
175
+ return "\n".join(page.extract_text() for page in pdf_reader.pages if page.extract_text())
176
+ elif file.name.endswith(".docx"):
177
+ doc = docx.Document(file)
178
+ return "\n".join(paragraph.text for paragraph in doc.paragraphs if paragraph.text)
179
+ elif file.name.endswith(".txt"):
180
+ return file.read().decode("utf-8")
181
+ else:
182
+ return None
183
+ except Exception as e:
184
+ st.error(f"Error extracting text: {e}")
185
+ return None
186
+
187
 
188
  def upload_to_firebase(user_id, file):
189
+ content = extract_text_from_file (file)
190
  if not content:
191
  return None, "Failed to convert file to content."
192