Select adequate highlighter for cell magic languages

Detect cells that use language extension magics such as %%R or %%bash
and call the appropriate pygmentize lexer.

Closes  #4184

IPython/nbconvert/filters/highlight.py

@@ -1,88 +1,131 b''
 """
 Module containing filter functions that allow code to be highlighted
 from within Jinja templates.
 """
 #-----------------------------------------------------------------------------
 # 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
 #-----------------------------------------------------------------------------
 
-from  pygments import highlight as pygements_highlight
+import re
+
+from pygments import highlight as pygements_highlight
 from pygments.lexers import get_lexer_by_name
 from pygments.formatters import HtmlFormatter
 from pygments.formatters import LatexFormatter
 
 # Our own imports
 from IPython.nbconvert.utils.lexers import IPythonLexer
 
 #-----------------------------------------------------------------------------
 # Globals and constants
 #-----------------------------------------------------------------------------
 
 MULTILINE_OUTPUTS = ['text', 'html', 'svg', 'latex', 'javascript', 'json']
 
+# list of magic language extensions and their associated pygment lexers
+magic_languages = {'%%R': 'r',
+                   '%%bash': 'bash',
+                   '%%octave': 'octave',
+                   '%%perl': 'perl',
+                   '%%ruby': 'ruby'
+                   }
+
+# build a RE to catch language extensions and choose an adequate
+# pygments lexer
+re_languages = "|".join(magic_languages.keys())
+re_magic_language = re.compile(r'^\s*({})\s+'.format(re_languages),
+                               re.MULTILINE)
+
 #-----------------------------------------------------------------------------
 # Utility functions
 #-----------------------------------------------------------------------------
 
 __all__ = [
     'highlight2html',
     'highlight2latex'
 ]
 
 
 def highlight2html(source, language='ipython'):
     """
     Return a syntax-highlighted version of the input source as html output.
-    
+
     Parameters
     ----------
     source : str
         Source code to highlight the syntax of.
     language : str
         Language to highlight the syntax of.
     """
-    
+
     return _pygment_highlight(source, HtmlFormatter(), language)
 
 
 def highlight2latex(source, language='ipython'):
     """
     Return a syntax-highlighted version of the input source as latex output.
-    
+
     Parameters
     ----------
     source : str
         Source code to highlight the syntax of.
     language : str
         Language to highlight the syntax of.
     """
     return _pygment_highlight(source, LatexFormatter(), language)
 
 
+def which_magic_language(source):
+    """
+    When a cell uses another language through a magic extension,
+    the other language is returned.
+    If no language magic is detected, this function returns None.
+
+    Parameters
+    ----------
+    source: str
+        Source code of the cell to highlight
+    """
+
+    m = re_magic_language.search(source)
+
+    if m:
+        # By construction of the re, the matched language must be in the
+        # language dictionnary
+        assert(m.group(1) in magic_languages)
+        return magic_languages[m.group(1)]
+    else:
+        return None
+
+
 def _pygment_highlight(source, output_formatter, language='ipython'):
     """
     Return a syntax-highlighted version of the input source
-    
+
     Parameters
     ----------
     source : str
         Source code to highlight the syntax of.
     output_formatter : Pygments formatter
     language : str
         Language to highlight the syntax of.
     """
-    
+
+    magic_language = which_magic_language(source)
+    if magic_language:
+        language = magic_language
+
     if language == 'ipython':
         lexer = IPythonLexer()
     else:
         lexer = get_lexer_by_name(language, stripall=True)
 
-    return pygements_highlight(source, lexer, output_formatter) 
+    return pygements_highlight(source, lexer, output_formatter)

IPython/nbconvert/filters/tests/test_highlight.py

@@ -1,65 +1,97 b''
 """
 Module with tests for Highlight
 """
 
 #-----------------------------------------------------------------------------
 # 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
 #-----------------------------------------------------------------------------
 
 from ...tests.base import TestsBase
-from ..highlight import highlight2html, highlight2latex
+from ..highlight import highlight2html, highlight2latex, which_magic_language
 
 
 #-----------------------------------------------------------------------------
 # Class
 #-----------------------------------------------------------------------------
 
 class TestHighlight(TestsBase):
     """Contains test functions for highlight.py"""
 
     #Hello world test, magics test, blank string test
     tests = [
         """
         #Hello World Example
 
         def say(text):
             print(text)
 
         say('Hello World!')
         """,
         """
         %%pylab
         plot(x,y, 'r')
         """
-        ]   
+        ]
 
     tokens = [
         ['Hello World Example', 'say', 'text', 'print', 'def'],
         ['pylab', 'plot']]
 
-
     def test_highlight2html(self):
         """highlight2html test"""
         for index, test in enumerate(self.tests):
             self._try_highlight(highlight2html, test, self.tokens[index])
 
-
     def test_highlight2latex(self):
         """highlight2latex test"""
         for index, test in enumerate(self.tests):
             self._try_highlight(highlight2latex, test, self.tokens[index])
 
-
     def _try_highlight(self, method, test, tokens):
         """Try highlighting source, look for key tokens"""
         results = method(test)
         for token in tokens:
             assert token in results
+
+    magic_tests = {
+            'r':
+                """%%R -i x,y -o XYcoef
+                lm.fit <- lm(y~x)
+                par(mfrow=c(2,2))
+                print(summary(lm.fit))
+                plot(lm.fit)
+                XYcoef <- coef(lm.fit)
+                """,
+            'bash':
+                """# the following code is in bash
+                %%bash
+                echo "test" > out
+                """,
+            'ipython':
+                """# this should not be detected
+                print("%%R")
+                """
+            }
+
+    def test_highlight_rmagic(self):
+        """Test %%R magic highlighting"""
+        language = which_magic_language(self.magic_tests['r'])
+        assert language == 'r'
+
+    def test_highlight_bashmagic(self):
+        """Test %%bash magic highlighting"""
+        language = which_magic_language(self.magic_tests['bash'])
+        assert language == 'bash'
+
+    def test_highlight_interferemagic(self):
+        """Test that magic highlighting does not interfere with ipython code"""
+        language = which_magic_language(self.magic_tests['ipython'])
+        assert language == None
  No newline at end of file