Escaping special characters. Coping with non-UTF-8 characters from Microsoft Stream.

python/transcripts.py

@@ -28,11 +28,43 @@ def write_title_tex(title: str, title_file: str) -> None:
     s += "\\date{}" + "\n"
     s += "\\maketitle" + "\n"
     s += "\\setstretch{1.2}" + "\n"
-    output_file = open(title_file, "w")
+    output_file = open(title_file, "w", encoding="UTF-8")
     output_file.write(s)
     output_file.close()
 
 
+def txt_to_tex(text: str) -> str:
+    """
+    A function to escape special text characters, such that 
+    LaTeX renders them correctly.
+    """
+
+    lookup_dict = {
+        "%": "\\%",
+        "$": "\\$",
+        "{": "\\{",
+        "}": "\\}",
+        "_": "\\_",
+        "#": "\\#",
+        "&": "\\&",
+        ">": "\\textgreater",
+        "<": "\\textless"
+    }
+
+    tex = ""
+    for c in text:
+        # Replace the existing backslashes first.
+        if c == "\\":
+            tex += "\\textbackslash"
+            continue
+        if c in lookup_dict.keys():
+            tex += lookup_dict[c]
+            continue
+        tex += c
+    
+    return tex
+
+
 def srt_to_txt(srt_file: str) -> list:
     """
     A function to remove the numbers and times from an srt captions file.
@@ -46,7 +78,11 @@ def srt_to_txt(srt_file: str) -> list:
         if len(line) > 0:
             if line[0].isdigit():
                 continue
-        lines.append(line)
+
+        # Replace special characters.
+        tex = txt_to_tex(line)
+
+        lines.append(tex)
     return lines
 
 
@@ -57,7 +93,13 @@ def vtt_to_txt(vtt_file: str) -> list:
     vtt = webvtt.read(vtt_file)
     lines = []
     for line in vtt:
-        lines.append(line.text + "\n\n")
+        # Microsoft Stream vtt text sometimes includes non-UTF-8 characters.
+        text_line = str(line.text)
+
+        # Replace special characters.
+        tex = txt_to_tex(text_line)
+
+        lines.append(tex + "\n\n")
     return lines
 
 
@@ -115,7 +157,7 @@ def process_files(doc_dir: str, csv_file_path: str) -> None:
         title_file = "title.tex"
         body_file = "body.tex"
         write_title_tex(title, title_file)
-        output_file = open(body_file, "w")
+        output_file = open(body_file, "w", encoding="UTF-8")
         output_file.writelines(lines)
         output_file.close()
 
@@ -158,7 +200,8 @@ if __name__ == "__main__":
     A program to convert .srt and .vtt caption files to
     PDF transcript files.
     """
-    parent_dir = os.path.dirname(os.getcwd())
+    python_dir = os.path.dirname(__file__)
+    parent_dir = os.path.dirname(python_dir)
     data_dir = os.path.join(parent_dir, "data")
     doc_dir = os.path.join(parent_dir, "doc")
     transcript_csv = get_input_file_path(data_dir)