Link resolution

nbconvert/exporters/latex.py

@@ -9,9 +9,9 @@
 from traitlets.config import Config
 
 from nbconvert.filters.highlight import Highlight2Latex
+from nbconvert.filters.filter_links import resolve_references
 from .templateexporter import TemplateExporter
 
-
 class LatexExporter(TemplateExporter):
     """
     Exports to a Latex template.  Inherit from this class if your template is
@@ -44,6 +44,11 @@ def _template_skeleton_path_default(self):
 
     output_mimetype = 'text/latex'
 
+    def default_filters(self):
+        for x in super(LatexExporter, self).default_filters():
+            yield x 
+        yield ('resolve_references', resolve_references)
+
     @property
     def default_config(self):
         c = Config({

nbconvert/filters/filter_links.py

@@ -0,0 +1,38 @@
+#!/usr/bin/env python3
+"""A pandoc filter used in converting notebooks to Latex.
+Converts links between notebooks to Latex cross-references.
+"""
+import re
+
+from pandocfilters import RawInline, applyJSONFilters
+
+def resolve_references(source):
+    """
+    This applies the resolve_one_reference to the text passed in via the source argument.
+
+    This expects content in the form of a string encoded JSON object as represented
+    internally in ``pandoc``.
+    """
+    return applyJSONFilters([resolve_one_reference], source)
+
+def resolve_one_reference(key, val, fmt, meta):
+    """
+    This takes a tuple of arguments that are compatible with ``pandocfilters.walk()`` that
+    allows identifying hyperlinks in the document and transforms them into valid LaTeX 
+    \\ref{} calls so that linking to headers between cells is possible.
+
+    See the documentation in ``pandocfilters.walk()`` for further information on the meaning
+    and specification of ``key``, ``val``, ``fmt``, and ``meta``. 
+    """
+
+    if key == 'Link':
+        target = val[2][0]
+        m = re.match(r'#(.+)$', target)
+        if m:
+            # pandoc automatically makes labels for headings.
+            label = m.group(1).lower()
+            label = re.sub(r'[^\w-]+', '', label) # Strip HTML entities
+            return RawInline('tex', r'Section \ref{%s}' % label)
+
+    # Other elements will be returned unchanged.
+

nbconvert/templates/latex/document_contents.tplx

@@ -64,7 +64,7 @@
 
 % Render markdown
 ((* block markdowncell scoped *))
-    ((( cell.source | citation2latex | strip_files_prefix | convert_pandoc('markdown', 'latex') )))
+    ((( cell.source | citation2latex | strip_files_prefix | convert_pandoc('markdown', 'json',extra_args=[]) | resolve_references | convert_pandoc('json','latex'))))
 ((* endblock markdowncell *))
 
 % Don't display unknown types

setup.py

@@ -193,6 +193,7 @@ def run(self):
     'nbformat',
     'entrypoints',
     'bleach',
+    'pandocfilters>=1.4.1',
 ]
 
 extras_require = setuptools_args['extras_require'] = {