[Feature] Replace pdfplumber with pymupdf

.gitignore

@@ -0,0 +1,5 @@
+.DS_Store
+
+out/
+__pycache__/
+results/

twix-ui/backend/app.py

@@ -49,7 +49,6 @@ def process_phrase():
         phrases, cost = twix.extract_phrase(pdf_paths, result_folder)
         print(cost)
 
-
         # Define paths for the bounding box files
         bb_path = os.path.join(result_folder, 'merged_phrases_bounding_box_page_number.json')
         bb_txt_path = os.path.join(os.path.dirname(result_folder), 'merged_raw_phrases_bounding_box_page_number.txt')
@@ -67,7 +66,6 @@ def process_phrase():
                 # Create both txt and json versions in each location
                 txt_dest = os.path.join(dest_path, 'merged_raw_phrases_bounding_box_page_number.txt')
                 json_dest = os.path.join(dest_path, 'merged_raw_phrases_bounding_box_page_number.json')
-
                 if os.path.exists(source_path):
                     # Copy TXT file as-is
                     shutil.copy(source_path, txt_dest)

twix-ui/src/services/api.js

@@ -135,8 +135,8 @@ async function predictFields(files) {
   }
 
   const data = await response.json();
-  // Create and trigger download of content as text file
-  const blob = new Blob([data.content], { type: 'text/plain' });
+  // Create and trigger download of fields as text file
+  const blob = new Blob([data.fields], { type: 'text/plain' });
   const url = window.URL.createObjectURL(blob);
   const a = document.createElement('a');
   a.href = url;
@@ -595,11 +595,14 @@ export function stringifyOrderedJSON(obj) {
   // Build object string with keys in original order
   let result = '{';
   result += allKeys.map(key => {
+    // Stringify the key
+    const keyStr = JSON.stringify(key);
+    // Get the value
     const value = obj[key];
     // Recursively stringify the value
     const valueStr = stringifyOrderedJSON(value);
     // Return the key-value pair
-    return `"${key}":${valueStr}`;
+    return `${keyStr}:${valueStr}`;
   }).join(',');
   result += '}';