fix n^2 performance issue in coalesce_streams preprocessor

for n consecutive stream outputs, `\r` fix would be compiled n times,
and applied to each output (n-i) times.

- move pattern to module level
- apply replacement after coalescing outputs

IPython/nbconvert/preprocessors/coalescestreams.py

@@ -1,83 +1,85 b''
 """Module that allows latex output notebooks to be conditioned before
 they are converted.  Exposes a decorator (@cell_preprocessor) in
 addition to the coalesce_streams pre-proccessor.
 """
 #-----------------------------------------------------------------------------
 # Copyright (c) 2013, the IPython Development Team.
 #
 # Distributed under the terms of the Modified BSD License.
 #
 # The full license is in the file COPYING.txt, distributed with this software.
 #-----------------------------------------------------------------------------
 
 #-----------------------------------------------------------------------------
 # Imports
 #-----------------------------------------------------------------------------
 import re
 
 #-----------------------------------------------------------------------------
 # Functions
 #-----------------------------------------------------------------------------
 def cell_preprocessor(function):
     """
     Wrap a function to be executed on all cells of a notebook
     
     The wrapped function should have these parameters:
     
     cell : NotebookNode cell
         Notebook cell being processed
     resources : dictionary
         Additional resources used in the conversion process.  Allows
         preprocessors to pass variables into the Jinja engine.
     index : int
         Index of the cell being processed
     """
     
     def wrappedfunc(nb, resources):
         for worksheet in nb.worksheets :
             for index, cell in enumerate(worksheet.cells):
                 worksheet.cells[index], resources = function(cell, resources, index)
         return nb, resources
     return wrappedfunc
 
+cr_pat = re.compile(r'.*\r(?=[^\n])')
 
 @cell_preprocessor
 def coalesce_streams(cell, resources, index):
     """
     Merge consecutive sequences of stream output into single stream
     to prevent extra newlines inserted at flush calls
     
     Parameters
     ----------
     cell : NotebookNode cell
         Notebook cell being processed
     resources : dictionary
         Additional resources used in the conversion process.  Allows
         transformers to pass variables into the Jinja engine.
     index : int
         Index of the cell being processed
     """
     
     outputs = cell.get('outputs', [])
     if not outputs:
         return cell, resources
     
     last = outputs[0]
     new_outputs = [last]
-    
     for output in outputs[1:]:
         if (output.output_type == 'stream' and
             last.output_type == 'stream' and
             last.stream == output.stream
         ):
             last.text += output.text
 
-            # Respect \r characters.
-            cr_pat = re.compile(r'.*\r(?=[^\n])')
-            last.text = cr_pat.sub('', last.text)
         else:
             new_outputs.append(output)
             last = output
+    
+    # process \r characters
+    for output in new_outputs:
+        if output.output_type == 'stream':
+            output.text = cr_pat.sub('', output.text)
 
     cell.outputs = new_outputs
     return cell, resources