Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
@@ -120,8 +120,8 @@ def read_document(file, clean=True):
|
|
120 |
return content, len(content)
|
121 |
except Exception as e:
|
122 |
return f"Error reading PDF: {e}", 0
|
123 |
-
elif mime
|
124 |
-
# XLSX Handling
|
125 |
try:
|
126 |
wb = load_workbook(io.BytesIO(file_content))
|
127 |
content = ''
|
@@ -134,7 +134,7 @@ def read_document(file, clean=True):
|
|
134 |
content = clean_text(content)
|
135 |
return content, len(content)
|
136 |
except Exception as e:
|
137 |
-
return f"Error reading XLSX: {e}", 0
|
138 |
elif mime == "text/plain":
|
139 |
try:
|
140 |
content = file_content.decode('utf-8')
|
@@ -151,17 +151,18 @@ def read_document(file, clean=True):
|
|
151 |
return content, len(content)
|
152 |
except Exception as e:
|
153 |
return f"Error reading CSV file: {e}", 0
|
154 |
-
elif mime
|
|
|
155 |
try:
|
156 |
return extract_text_from_docx(file_content, clean)
|
157 |
except Exception as e:
|
158 |
-
return f"Error reading DOCX: {e}", 0
|
159 |
-
elif mime
|
|
|
160 |
try:
|
161 |
return extract_text_from_pptx(file_content, clean)
|
162 |
except Exception as e:
|
163 |
-
return f"Error reading PPTX: {e}", 0
|
164 |
-
|
165 |
else:
|
166 |
try:
|
167 |
content = file_content.decode('utf-8')
|
@@ -170,6 +171,7 @@ def read_document(file, clean=True):
|
|
170 |
return content, len(content)
|
171 |
except Exception as e:
|
172 |
return f"Error reading file: {e}", 0
|
|
|
173 |
|
174 |
|
175 |
# --- Chat Functions ---
|
|
|
120 |
return content, len(content)
|
121 |
except Exception as e:
|
122 |
return f"Error reading PDF: {e}", 0
|
123 |
+
elif mime in ["application/vnd.openxmlformats-officedocument.spreadsheetml.sheet", "application/vnd.ms-excel"]:
|
124 |
+
# XLSX and XLS Handling
|
125 |
try:
|
126 |
wb = load_workbook(io.BytesIO(file_content))
|
127 |
content = ''
|
|
|
134 |
content = clean_text(content)
|
135 |
return content, len(content)
|
136 |
except Exception as e:
|
137 |
+
return f"Error reading XLSX/XLS: {e}", 0
|
138 |
elif mime == "text/plain":
|
139 |
try:
|
140 |
content = file_content.decode('utf-8')
|
|
|
151 |
return content, len(content)
|
152 |
except Exception as e:
|
153 |
return f"Error reading CSV file: {e}", 0
|
154 |
+
elif mime in ["application/vnd.openxmlformats-officedocument.wordprocessingml.document", "application/msword", "application/vnd.oasis.opendocument.text"]:
|
155 |
+
# DOCX, DOC, and ODT Handling
|
156 |
try:
|
157 |
return extract_text_from_docx(file_content, clean)
|
158 |
except Exception as e:
|
159 |
+
return f"Error reading DOCX/DOC/ODT: {e}", 0
|
160 |
+
elif mime in ["application/vnd.openxmlformats-officedocument.presentationml.presentation", "application/vnd.ms-powerpoint", "application/vnd.oasis.opendocument.presentation"]:
|
161 |
+
# PPTX, PPT, and ODP Handling
|
162 |
try:
|
163 |
return extract_text_from_pptx(file_content, clean)
|
164 |
except Exception as e:
|
165 |
+
return f"Error reading PPTX/PPT/ODP: {e}", 0
|
|
|
166 |
else:
|
167 |
try:
|
168 |
content = file_content.decode('utf-8')
|
|
|
171 |
return content, len(content)
|
172 |
except Exception as e:
|
173 |
return f"Error reading file: {e}", 0
|
174 |
+
|
175 |
|
176 |
|
177 |
# --- Chat Functions ---
|