Hadiil commited on
Commit
267e7b9
·
verified ·
1 Parent(s): 740bd24

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +26 -11
app.py CHANGED
@@ -1,7 +1,7 @@
1
  import os
2
  from fastapi import FastAPI, UploadFile, File, HTTPException, Form
3
  from fastapi.staticfiles import StaticFiles
4
- from fastapi.responses import RedirectResponse
5
  from transformers import pipeline
6
  import logging
7
  from PIL import Image
@@ -206,24 +206,39 @@ plt.show()
206
 
207
  return {"code": code, "filename": code_filename}
208
  except Exception as e:
209
- logger.error(f"Error during visualization codeS generation: {e}")
210
  raise HTTPException(status_code=500, detail=str(e))
211
 
212
  async def extract_text_from_file(file: UploadFile):
213
  try:
214
  file_content = await file.read()
 
 
 
215
  if file.filename.endswith(".pdf"):
216
- doc = fitz.open(stream=file_content, filetype="pdf")
217
- text = ""
218
- for page in doc:
219
- text += page.get_text()
220
- return text
 
 
 
 
221
  elif file.filename.endswith(".docx"):
222
- doc = Document(io.BytesIO(file_content))
223
- text = "\n".join([para.text for para in doc.paragraphs])
224
- return text
 
 
 
 
225
  elif file.filename.endswith(".txt"):
226
- return file_content.decode("utf-8")
 
 
 
 
227
  else:
228
  raise ValueError("Unsupported file format. Please upload a PDF, DOCX, or TXT file.")
229
  except Exception as e:
 
1
  import os
2
  from fastapi import FastAPI, UploadFile, File, HTTPException, Form
3
  from fastapi.staticfiles import StaticFiles
4
+ from fastapi.responses import RedirectResponse, JSONResponse
5
  from transformers import pipeline
6
  import logging
7
  from PIL import Image
 
206
 
207
  return {"code": code, "filename": code_filename}
208
  except Exception as e:
209
+ logger.error(f"Error during visualization code generation: {e}")
210
  raise HTTPException(status_code=500, detail=str(e))
211
 
212
  async def extract_text_from_file(file: UploadFile):
213
  try:
214
  file_content = await file.read()
215
+ if not file_content:
216
+ raise ValueError("Uploaded file is empty.")
217
+
218
  if file.filename.endswith(".pdf"):
219
+ try:
220
+ doc = fitz.open(stream=file_content, filetype="pdf")
221
+ text = ""
222
+ for page in doc:
223
+ text += page.get_text()
224
+ return text
225
+ except Exception as e:
226
+ logger.error(f"Error reading PDF file: {e}")
227
+ raise ValueError("Failed to read PDF file. It might be corrupted or not a valid PDF.")
228
  elif file.filename.endswith(".docx"):
229
+ try:
230
+ doc = Document(io.BytesIO(file_content))
231
+ text = "\n".join([para.text for para in doc.paragraphs])
232
+ return text
233
+ except Exception as e:
234
+ logger.error(f"Error reading DOCX file: {e}")
235
+ raise ValueError("Failed to read DOCX file. It might be corrupted or not a valid DOCX.")
236
  elif file.filename.endswith(".txt"):
237
+ try:
238
+ return file_content.decode("utf-8")
239
+ except Exception as e:
240
+ logger.error(f"Error reading TXT file: {e}")
241
+ raise ValueError("Failed to read TXT file. It might be corrupted or not a valid TXT.")
242
  else:
243
  raise ValueError("Unsupported file format. Please upload a PDF, DOCX, or TXT file.")
244
  except Exception as e: