fixed some formatting errors, still haven't fixed line spacing

main.py

@@ -1,4 +1,5 @@
 import os
+import string
 
 from docx import Document
 import nltk
@@ -223,6 +224,9 @@ def group_by_style(values, detokenizer):
             x['bold'], x['italic'], x['underline'], x['font_name'], x['font_size'], x['font_color'])):
         text = detokenizer.detokenize([item['text'] for item in group])
 
+        if groups and not text.startswith((",", ";", ":", ".", ")")):
+            text = " " + text
+
         groups.append({"text": text,
                        "bold": key[0],
                        "italic": key[1],
@@ -287,7 +291,9 @@ if __name__ == "__main__":
 
     out_doc = Document()
 
-    for original_runs_in_paragraph, translated_paragraph in zip(paragraphs_with_runs, translated_paragraphs):
+    for original_runs_in_paragraph, translated_paragraph, original_paragraph in zip(paragraphs_with_runs,
+                                                                                    translated_paragraphs,
+                                                                                    doc.paragraphs):
         # sometimes we get empty paragraphs for some reason, I think it's just docx shenanigans
         if not original_runs_in_paragraph:
             continue
@@ -297,7 +303,7 @@ if __name__ == "__main__":
         paragraph_with_style = generate_alignments(original_runs_in_paragraph, translated_paragraph, aligner,
                                                    temp_folder, detokenizer)
 
-        para = out_doc.add_paragraph()
+        para = out_doc.add_paragraph(style=original_paragraph.style)
 
         # flatten the paragraph, we don't need it to split into sentences anymore
         paragraph_with_style = [item for sublist in paragraph_with_style for item in sublist]
@@ -306,7 +312,7 @@ if __name__ == "__main__":
         paragraph_with_runs = group_by_style(paragraph_with_style, detokenizer)
 
         for item in paragraph_with_runs:
-            run = para.add_run(item["text"] + " ")
+            run = para.add_run(item["text"])
             # Preserve original run formatting
             run.bold = item['bold']
             run.italic = item['italic']