Spaces:
Build error
Build error
Update app.py
Browse files
app.py
CHANGED
|
@@ -8,8 +8,17 @@ import pinecone
|
|
| 8 |
import tempfile
|
| 9 |
import shutil
|
| 10 |
|
|
|
|
|
|
|
|
|
|
| 11 |
# Dynamically add Poppler to PATH if necessary
|
| 12 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
|
| 14 |
# Initialize Streamlit app
|
| 15 |
st.title("PDF Image to Text/Word Converter with Search Capability")
|
|
@@ -30,7 +39,11 @@ if uploaded_file:
|
|
| 30 |
|
| 31 |
# Convert PDF pages to images
|
| 32 |
st.write("Converting PDF to images...")
|
| 33 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 34 |
|
| 35 |
# Extract text from images
|
| 36 |
st.write("Extracting text from images...")
|
|
|
|
| 8 |
import tempfile
|
| 9 |
import shutil
|
| 10 |
|
| 11 |
+
# Debug: Display current PATH
|
| 12 |
+
st.write("System PATH:", os.environ["PATH"])
|
| 13 |
+
|
| 14 |
# Dynamically add Poppler to PATH if necessary
|
| 15 |
+
poppler_path = "C:\\Program Files\\poppler-24.08.0\\bin"
|
| 16 |
+
if poppler_path not in os.environ["PATH"]:
|
| 17 |
+
os.environ["PATH"] += os.pathsep + poppler_path
|
| 18 |
+
|
| 19 |
+
# Debug: Confirm Poppler is accessible
|
| 20 |
+
if shutil.which("pdfinfo") is None:
|
| 21 |
+
st.error("Poppler's pdfinfo executable not found in PATH!")
|
| 22 |
|
| 23 |
# Initialize Streamlit app
|
| 24 |
st.title("PDF Image to Text/Word Converter with Search Capability")
|
|
|
|
| 39 |
|
| 40 |
# Convert PDF pages to images
|
| 41 |
st.write("Converting PDF to images...")
|
| 42 |
+
try:
|
| 43 |
+
images = convert_from_path(pdf_path, output_folder=temp_dir)
|
| 44 |
+
except Exception as e:
|
| 45 |
+
st.error(f"Error during PDF to image conversion: {e}")
|
| 46 |
+
st.stop()
|
| 47 |
|
| 48 |
# Extract text from images
|
| 49 |
st.write("Extracting text from images...")
|