Add more functionality to TranscriptFormatter base class

Due to the behavior of the CLI and API, needed more flexibility for combining 1 or many transcripts for a given formatter.

- Now can specify a DELIMITER to separate multiple transcripts on.
- Can also specify how those items are combine overriding the combine class method.

Remove unused imports
Fix adjust some lines to meet PEP

youtube_transcript_api/formatters.py

@@ -1,13 +1,9 @@
-from abc import ABCMeta
+from abc import ABC
 from abc import abstractclassmethod
 from collections import defaultdict
 import json
 import re
 
-from xml.etree import ElementTree
-
-from ._html_unescaping import unescape
-
 
 def parse_timecode(time):
     """Converts a `time` into a formatted transcript timecode.
@@ -31,15 +27,29 @@ def parse_timecode(time):
     return f"{hours}:{mins}:{secs},{ms}"
 
 
-class TranscriptFormatter(metaclass=ABCMeta):
+class TranscriptFormatter(ABC):
-    """
+    """Abstract Base TranscriptFormatter class
-    Abstract Base TranscriptFormatter class
 
     This class should be inherited from to create additional
      custom transcript formatters.
-    
     """
     HTML_TAG_REGEX = re.compile(r'<[^>]*>', re.IGNORECASE)
+    DELIMITER = ''
+
+    @classmethod
+    def combine(cls, transcripts):
+        """Subclass may override this class method.
+
+        Default behavior of this method will ''.join() the str() 
+         of each transcript in transcripts.
+
+        :param transcripts: a list of many transcripts
+        :type transcript_data: list[<formatted transcript>, ...]
+        :return: A string joined on the `cls.DELIMITER` to combine transcripts
+        :rtype: str
+        """
+        return cls.DELIMITER.join(
+                str(transcript) for transcript in transcripts)
 
     @abstractclassmethod
     def format(cls, transcript_data):
@@ -56,9 +66,15 @@ class TranscriptFormatter(metaclass=ABCMeta):
 
 class JSONTranscriptFormatter(TranscriptFormatter):
     """Formatter for outputting JSON data"""
+    DELIMITER = ','
+
+    @classmethod
+    def combine(cls, transcripts):
+        return json.dumps(transcripts)
+
     @classmethod
     def format(cls, transcript_data):
-        return [json.dumps(transcript_data)] if transcript_data else []
+        return transcript_data
 
 
 class TextTranscriptFormatter(TranscriptFormatter):
@@ -66,39 +82,40 @@ class TextTranscriptFormatter(TranscriptFormatter):
 
     Converts the fetched transcript data into separated lines of
      plain text separated by newline breaks (\n) with no timecodes.
-    
     """
+    DELIMITER = '\n\n'
+
     @classmethod
     def format(cls, transcript_data):
-        return ['\n'.join(line['text'] for transcript in transcript_data
+        return '{}\n'.format('\n'.join(
-                            for line in transcript)]
+                    line['text']for line in transcript_data))
 
 
 class SRTTranscriptFormatter(TranscriptFormatter):
     """Formatter for outputting the SRT Format
 
     Converts the fetched transcript data into a simple .srt file format.
-
     """
+    DELIMITER = '\n\n'
+
     @classmethod
     def format(cls, transcript_data):
-        contents = []
+        output = []
-        for transcript in transcript_data:
+        for frame, item in enumerate(transcript_data, start=1):
-            content = []
-            for frame, item in enumerate(transcript, start=1):
             start_time = float(item.get('start'))
             duration = float(item.get('dur', '0.0'))
 
             end_time = parse_timecode(start_time + duration)
             start_time = parse_timecode(start_time)
 
-                content.append("{frame}\n".format(frame=frame))
+            output.append("{frame}\n".format(frame=frame))
-                content.append("{start_time} --> {end_time}\n".format(
+            output.append("{start_time} --> {end_time}\n".format(
                 start_time=start_time, end_time=end_time))
-                content.append("{text}\n\n".format(text=item.get('text')))
+            output.append("{text}".format(text=item.get('text')))
+            if frame < len(transcript_data):
+                output.append('\n\n')
 
-            contents.append(''.join(content))
+        return '{}\n'.format(''.join(output))
-        return ['\n\n'.join(contents)]
 
 
 class TranscriptFormatterFactory:
@@ -124,7 +141,7 @@ class TranscriptFormatterFactory:
         if not issubclass(formatter_class, TranscriptFormatter):
             raise TypeError(
                 f'{formatter_class} must be a subclass of TranscriptFormatter')
-        self._formatters.update({name:formatter_class})
+        self._formatters.update({name: formatter_class})
 
     def add_formatters(self, formatters_dict):
         """Allow creation of multiple transcript formatters at a time.