solved pdf postprocess issue for nb with accented names

IPython/nbconvert/postprocessors/pdf.py

@@ -1,149 +1,160 b''
 """
 Contains writer for writing nbconvert output to PDF.
 """
 #-----------------------------------------------------------------------------
 #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 subprocess
 import os
 import sys
 
 from IPython.utils.traitlets import Integer, List, Bool
+from IPython.utils.encoding import DEFAULT_ENCODING
 
 from .base import PostProcessorBase
 
+
 #-----------------------------------------------------------------------------
 # Classes
 #-----------------------------------------------------------------------------
 class PDFPostProcessor(PostProcessorBase):
     """Writer designed to write to PDF files"""
 
     latex_count = Integer(3, config=True, help="""
         How many times pdflatex will be called.
         """)
 
     latex_command = List(["pdflatex", "{filename}"], config=True, help="""
         Shell command used to compile PDF.""")
 
     bib_command = List(["bibtex", "{filename}"], config=True, help="""
         Shell command used to run bibtex.""")
 
     verbose = Bool(False, config=True, help="""
         Whether or not to display the output of the compile call.
         """)
 
     temp_file_exts = List(['.aux', '.bbl', '.blg', '.idx', '.log', '.out'], 
         config=True, help="""
         Filename extensions of temp files to remove after running.
         """)
     pdf_open = Bool(False, config=True, help="""
         Whether or not to open the pdf after the compile call.
         """)
 
     def run_command(self, command_list, filename, count, log_function):
         """Run command_list count times.
         
         Parameters
         ----------
         command_list : list
             A list of args to provide to Popen. Each element of this
             list will be interpolated with the filename to convert.
         filename : unicode
             The name of the file to convert.
         count : int
             How many times to run the command.
         
         Returns
         -------
         continue : bool
             A boolean indicating if the command was successful (True)
             or failed (False).
         """
+        #HACK: Encode file name with the correct encoding
+        # For Windows must be cp1252 (win application) and we cannot use
+        #    encoding.DEFAULT_ENCODING or sys.stdin.encoding because input
+        #    encoding could be different (cp437 in case of dos console)
+        if sys.platform == 'win32':
+            filename = filename.encode('cp1252')
+        else:
+            filename = filename.encode('utf-8')
+        #END_HACK
         command = [c.format(filename=filename) for c in command_list]
         times = 'time' if count == 1 else 'times'
         self.log.info("Running %s %i %s: %s", command_list[0], count, times, command)
         with open(os.devnull, 'rb') as null:
             stdout = subprocess.PIPE if not self.verbose else None
             for index in range(count):
-                p = subprocess.Popen(command, stdout=stdout, stdin=null)
+                p = subprocess.Popen(command, stdout=stdout, stdin=null, shell=True)
                 out, err = p.communicate()
                 if p.returncode:
                     if self.verbose:
                         # verbose means I didn't capture stdout with PIPE,
                         # so it's already been displayed and `out` is None.
                         out = u''
                     else:
                         out = out.decode('utf-8', 'replace')
                     log_function(command, out)
                     return False # failure
         return True # success
 
     def run_latex(self, filename):
         """Run pdflatex self.latex_count times."""
 
         def log_error(command, out):
             self.log.critical(u"%s failed: %s\n%s", command[0], command, out)
 
         return self.run_command(self.latex_command, filename, 
             self.latex_count, log_error)
 
     def run_bib(self, filename):
         """Run bibtex self.latex_count times."""
         filename = os.path.splitext(filename)[0]
 
         def log_error(command, out):
             self.log.warn('%s had problems, most likely because there were no citations',
                 command[0])
             self.log.debug(u"%s output: %s\n%s", command[0], command, out)
 
         return self.run_command(self.bib_command, filename, 1, log_error)
 
     def clean_temp_files(self, filename):
         """Remove temporary files created by pdflatex/bibtext."""
         self.log.info("Removing temporary LaTeX files")
         filename = os.path.splitext(filename)[0]
         for ext in self.temp_file_exts:
             try:
                 os.remove(filename+ext)
             except OSError:
                 pass
 
     def open_pdf(self, filename):
         """Open the pdf in the default viewer."""
         if sys.platform.startswith('darwin'):
             subprocess.call(('open', filename))
         elif os.name == 'nt':
             os.startfile(filename)
         elif os.name == 'posix':
             subprocess.call(('xdg-open', filename))
         return
 
     def postprocess(self, filename):
         """Build a PDF by running pdflatex and bibtex"""
         self.log.info("Building PDF")
         cont = self.run_latex(filename)
         if cont:
             cont = self.run_bib(filename)
         else:
             self.clean_temp_files(filename)
             return
         if cont:
             cont = self.run_latex(filename)
         self.clean_temp_files(filename)
         filename = os.path.splitext(filename)[0]
         if os.path.isfile(filename+'.pdf'):
             self.log.info('PDF successfully created')
             if self.pdf_open: 
                 self.log.info('Viewer called')
                 self.open_pdf(filename+'.pdf')
         return
  

IPython/nbconvert/tests/test_nbconvertapp.py

@@ -1,218 +1,217 b''
 # -*- coding: utf-8 -*- 
 """Test NbConvertApp"""
 
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2013 The IPython Development Team
 #
 #  Distributed under the terms of the BSD License.  The full license is in
 #  the file COPYING, distributed as part of this software.
 #-----------------------------------------------------------------------------
 
 #-----------------------------------------------------------------------------
 # Imports
 #-----------------------------------------------------------------------------
 
 import os
 import glob
 import sys
 
 from .base import TestsBase
 
 import IPython.testing.tools as tt
 from IPython.testing import decorators as dec
 
 
 #-----------------------------------------------------------------------------
 # Constants
 #-----------------------------------------------------------------------------
 
 
 #-----------------------------------------------------------------------------
 # Classes and functions
 #-----------------------------------------------------------------------------
 
 class TestNbConvertApp(TestsBase):
     """Collection of NbConvertApp tests"""
 
 
     def test_notebook_help(self):
         """Will help show if no notebooks are specified?"""
         with self.create_temp_cwd():
             out, err = self.call('nbconvert --log-level 0', ignore_return_code=True)
             self.assertIn("see '--help-all'", out)
     
     def test_help_output(self):
         """ipython nbconvert --help-all works"""
         tt.help_all_output_test('nbconvert')
 
     def test_glob(self):
         """
         Do search patterns work for notebook names?
         """
         with self.create_temp_cwd(['notebook*.ipynb']):
             self.call('nbconvert --to python *.ipynb --log-level 0')
             assert os.path.isfile('notebook1.py')
             assert os.path.isfile('notebook2.py')
 
 
     def test_glob_subdir(self):
         """
         Do search patterns work for subdirectory notebook names?
         """
         with self.create_temp_cwd():
             self.copy_files_to(['notebook*.ipynb'], 'subdir/')
             self.call('nbconvert --to python --log-level 0 ' + 
                       os.path.join('subdir', '*.ipynb'))
             assert os.path.isfile('notebook1.py')
             assert os.path.isfile('notebook2.py')
 
 
     def test_explicit(self):
         """
         Do explicit notebook names work?
         """
         with self.create_temp_cwd(['notebook*.ipynb']):
             self.call('nbconvert --log-level 0 --to python notebook2')
             assert not os.path.isfile('notebook1.py')
             assert os.path.isfile('notebook2.py')
 
 
     @dec.onlyif_cmds_exist('pdflatex')
     @dec.onlyif_cmds_exist('pandoc')
     def test_filename_spaces(self):
         """
         Generate PDFs with graphics if notebooks have spaces in the name?
         """
         with self.create_temp_cwd(['notebook2.ipynb']):
             os.rename('notebook2.ipynb', 'notebook with spaces.ipynb')
             o,e = self.call('nbconvert --log-level 0 --to latex '
                             '"notebook with spaces" --post PDF '
                             '--PDFPostProcessor.verbose=True')
             assert os.path.isfile('notebook with spaces.tex')
             assert os.path.isdir('notebook with spaces_files')
             assert os.path.isfile('notebook with spaces.pdf')
 
     @dec.onlyif_cmds_exist('pdflatex')
     @dec.onlyif_cmds_exist('pandoc')
     def test_post_processor(self):
         """
         Do post processors work?
         """
         with self.create_temp_cwd(['notebook1.ipynb']):
             self.call('nbconvert --log-level 0 --to latex notebook1 '
                       '--post PDF --PDFPostProcessor.verbose=True')
             assert os.path.isfile('notebook1.tex')
             assert os.path.isfile('notebook1.pdf')
 
     @dec.onlyif_cmds_exist('pandoc')
     def test_spurious_cr(self):
         """Check for extra CR characters"""
         with self.create_temp_cwd(['notebook2.ipynb']):
             self.call('nbconvert --log-level 0 --to latex notebook2')
             assert os.path.isfile('notebook2.tex')
             with open('notebook2.tex') as f:
                 tex = f.read()
             self.call('nbconvert --log-level 0 --to html notebook2')
             assert os.path.isfile('notebook2.html')
             with open('notebook2.html') as f:
                 html = f.read()
         self.assertEqual(tex.count('\r'), tex.count('\r\n'))
         self.assertEqual(html.count('\r'), html.count('\r\n'))
 
     @dec.onlyif_cmds_exist('pandoc')
     def test_png_base64_html_ok(self):
         """Is embedded png data well formed in HTML?"""
         with self.create_temp_cwd(['notebook2.ipynb']):
             self.call('nbconvert --log-level 0 --to HTML '
                       'notebook2.ipynb --template full')
             assert os.path.isfile('notebook2.html')
             with open('notebook2.html') as f:
                 assert "data:image/png;base64,b'" not in f.read()
 
     @dec.onlyif_cmds_exist('pandoc')
     def test_template(self):
         """
         Do export templates work?
         """
         with self.create_temp_cwd(['notebook2.ipynb']):
             self.call('nbconvert --log-level 0 --to slides '  
                       'notebook2.ipynb --template reveal')
             assert os.path.isfile('notebook2.slides.html')
             with open('notebook2.slides.html') as f:
                 assert '/reveal.css' in f.read()
 
 
     def test_glob_explicit(self):
         """
         Can a search pattern be used along with matching explicit notebook names?
         """
         with self.create_temp_cwd(['notebook*.ipynb']):
             self.call('nbconvert --log-level 0 --to python '
                       '*.ipynb notebook1.ipynb notebook2.ipynb')
             assert os.path.isfile('notebook1.py')
             assert os.path.isfile('notebook2.py')
 
 
     def test_explicit_glob(self):
         """
         Can explicit notebook names be used and then a matching search pattern?
         """
         with self.create_temp_cwd(['notebook*.ipynb']):
             self.call('nbconvert --log-level 0 --to=python '
                       'notebook1.ipynb notebook2.ipynb *.ipynb')
             assert os.path.isfile('notebook1.py')
             assert os.path.isfile('notebook2.py')
 
 
     def test_default_config(self):
         """
         Does the default config work?
         """
         with self.create_temp_cwd(['notebook*.ipynb', 'ipython_nbconvert_config.py']):
             self.call('nbconvert --log-level 0')
             assert os.path.isfile('notebook1.py')
             assert not os.path.isfile('notebook2.py')
 
 
     def test_override_config(self):
         """
         Can the default config be overriden?
         """
         with self.create_temp_cwd(['notebook*.ipynb',
                                    'ipython_nbconvert_config.py',
                                    'override.py']):
             self.call('nbconvert --log-level 0 --config="override.py"')
             assert not os.path.isfile('notebook1.py')
             assert os.path.isfile('notebook2.py')
 
     def test_accents_in_filename(self):
         """
         Can notebook names include accents?
         """
         with self.create_temp_cwd(['nb*.ipynb']):
             self.call('nbconvert --log-level 0 --to python nb1_*')
             assert os.path.isfile(u'nb1_análisis.py')
     
     @dec.onlyif_cmds_exist('pandoc')        
     def test_accents_in_command_line(self):
         """
         Are accents allowed in arguments of command line?
         """
         with self.create_temp_cwd(['nb*.ipynb']):
             self.call('nbconvert --to latex nb1_* ' 
                       '--SphinxTransform.author="análisis"')
             assert os.path.isfile(u'nb1_análisis.tex')
 
     @dec.onlyif_cmds_exist('pdflatex')
     @dec.onlyif_cmds_exist('pandoc')
     def test_filename_spaces(self):
         """
         Generate PDFs if notebooks have an accent in their name?
         """
         with self.create_temp_cwd(['nb*.ipynb']):
             o,e = self.call('nbconvert --log-level 0 --to latex '
                             '"nb1_*" --post PDF '
                             '--PDFPostProcessor.verbose=True')
-            assert os.path.isfile('nb1_análisis.tex')
-            assert os.path.isdir('nb1_análisis_files')
-            assert os.path.isfile('nb1_análisis.pdf')
+            assert os.path.isfile(u'nb1_análisis.tex')
+            assert os.path.isfile(u'nb1_análisis.pdf')