Allow validator to be called from reads_json and writes_json

IPython/nbformat/current.py

@@ -1,213 +1,223 b''
 """The official API for working with notebooks in the current format version.
 
 Authors:
 
 * Brian Granger
 * Jonathan Frederic
 """
 
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2008-2011  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
 #-----------------------------------------------------------------------------
 
 from __future__ import print_function
 
 from xml.etree import ElementTree as ET
 import re
 
 from IPython.utils.py3compat import unicode_type
 
 from IPython.nbformat.v3 import (
     NotebookNode,
     new_code_cell, new_text_cell, new_notebook, new_output, new_worksheet,
     parse_filename, new_metadata, new_author, new_heading_cell, nbformat,
-    nbformat_minor, to_notebook_json
+    nbformat_minor, nbformat_schema, to_notebook_json
 )
 from IPython.nbformat import v3 as _v_latest
 
 from .reader import reads as reader_reads
 from .reader import versions
 from .convert import convert
+from .validator import nbvalidate
 
 #-----------------------------------------------------------------------------
 # Code
 #-----------------------------------------------------------------------------
 
 current_nbformat = nbformat
 current_nbformat_minor = nbformat_minor
 current_nbformat_module = _v_latest.__name__
 
+
 def docstring_nbformat_mod(func):
     """Decorator for docstrings referring to classes/functions accessed through
     nbformat.current.
 
     Put {nbformat_mod} in the docstring in place of 'IPython.nbformat.v3'.
     """
     func.__doc__ = func.__doc__.format(nbformat_mod=current_nbformat_module)
     return func
 
 
 class NBFormatError(ValueError):
     pass
 
 
 def parse_py(s, **kwargs):
     """Parse a string into a (nbformat, string) tuple."""
     nbf = current_nbformat
     nbm = current_nbformat_minor
     
     pattern = r'# <nbformat>(?P<nbformat>\d+[\.\d+]*)</nbformat>'
     m = re.search(pattern,s)
     if m is not None:
         digits = m.group('nbformat').split('.')
         nbf = int(digits[0])
         if len(digits) > 1:
             nbm = int(digits[1])
 
     return nbf, nbm, s
 
 
 def reads_json(s, **kwargs):
     """Read a JSON notebook from a string and return the NotebookNode object."""
-    return convert(reader_reads(s), current_nbformat)
+    nbjson = reader_reads(s)
+    num_errors = nbvalidate(nbjson)
+    if num_errors > 0:
+        print("Num errors: %d" % num_errors)
+    return convert(nbjson, current_nbformat)
 
 
 def writes_json(nb, **kwargs):
-    return versions[current_nbformat].writes_json(nb, **kwargs)
+    nbjson = versions[current_nbformat].writes_json(nb, **kwargs)
+    num_errors = nbvalidate(nbjson)
+    if num_errors > 0:
+        print("Num errors: %d" % num_errors)
+    return nbjson
 
 
 def reads_py(s, **kwargs):
     """Read a .py notebook from a string and return the NotebookNode object."""
     nbf, nbm, s = parse_py(s, **kwargs)
     if nbf in (2, 3):
         nb = versions[nbf].to_notebook_py(s, **kwargs)
     else:
         raise NBFormatError('Unsupported PY nbformat version: %i' % nbf)
     return nb
 
 
 def writes_py(nb, **kwargs):
     # nbformat 3 is the latest format that supports py
     return versions[3].writes_py(nb, **kwargs)
 
 
 # High level API
 
 
 def reads(s, format, **kwargs):
     """Read a notebook from a string and return the NotebookNode object.
 
     This function properly handles notebooks of any version. The notebook
     returned will always be in the current version's format.
 
     Parameters
     ----------
     s : unicode
         The raw unicode string to read the notebook from.
     format : (u'json', u'ipynb', u'py')
         The format that the string is in.
 
     Returns
     -------
     nb : NotebookNode
         The notebook that was read.
     """
     format = unicode_type(format)
     if format == u'json' or format == u'ipynb':
         return reads_json(s, **kwargs)
     elif format == u'py':
         return reads_py(s, **kwargs)
     else:
         raise NBFormatError('Unsupported format: %s' % format)
 
 
 def writes(nb, format, **kwargs):
     """Write a notebook to a string in a given format in the current nbformat version.
 
     This function always writes the notebook in the current nbformat version.
 
     Parameters
     ----------
     nb : NotebookNode
         The notebook to write.
     format : (u'json', u'ipynb', u'py')
         The format to write the notebook in.
 
     Returns
     -------
     s : unicode
         The notebook string.
     """
     format = unicode_type(format)
     if format == u'json' or format == u'ipynb':
         return writes_json(nb, **kwargs)
     elif format == u'py':
         return writes_py(nb, **kwargs)
     else:
         raise NBFormatError('Unsupported format: %s' % format)
 
 
 def read(fp, format, **kwargs):
     """Read a notebook from a file and return the NotebookNode object.
 
     This function properly handles notebooks of any version. The notebook
     returned will always be in the current version's format.
 
     Parameters
     ----------
     fp : file
         Any file-like object with a read method.
     format : (u'json', u'ipynb', u'py')
         The format that the string is in.
 
     Returns
     -------
     nb : NotebookNode
         The notebook that was read.
     """
     return reads(fp.read(), format, **kwargs)
 
 
 def write(nb, fp, format, **kwargs):
     """Write a notebook to a file in a given format in the current nbformat version.
 
     This function always writes the notebook in the current nbformat version.
 
     Parameters
     ----------
     nb : NotebookNode
         The notebook to write.
     fp : file
         Any file-like object with a write method.
     format : (u'json', u'ipynb', u'py')
         The format to write the notebook in.
 
     Returns
     -------
     s : unicode
         The notebook string.
     """
     return fp.write(writes(nb, format, **kwargs))
 
 def _convert_to_metadata():
     """Convert to a notebook having notebook metadata."""
     import glob
     for fname in glob.glob('*.ipynb'):
         print('Converting file:',fname)
         with open(fname,'r') as f:
             nb = read(f,u'json')
         md = new_metadata()
         if u'name' in nb:
             md.name = nb.name
             del nb[u'name']            
         nb.metadata = md
         with open(fname,'w') as f:
             write(nb, f, u'json')
 

IPython/nbformat/v3/__init__.py

@@ -1,74 +1,75 b''
 """The main API for the v3 notebook format.
 
 Authors:
 
 * Brian Granger
 """
 
 #-----------------------------------------------------------------------------
 #  Copyright (C) 2008-2011  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
 #-----------------------------------------------------------------------------
 
 from .nbbase import (
     NotebookNode,
     new_code_cell, new_text_cell, new_notebook, new_output, new_worksheet,
-    new_metadata, new_author, new_heading_cell, nbformat, nbformat_minor
+    new_metadata, new_author, new_heading_cell, nbformat, nbformat_minor,
+    nbformat_schema
 )
 
 from .nbjson import reads as reads_json, writes as writes_json
 from .nbjson import reads as read_json, writes as write_json
 from .nbjson import to_notebook as to_notebook_json
 
 from .nbpy import reads as reads_py, writes as writes_py
 from .nbpy import reads as read_py, writes as write_py
 from .nbpy import to_notebook as to_notebook_py
 
 from .convert import downgrade, upgrade
 
 #-----------------------------------------------------------------------------
 # Code
 #-----------------------------------------------------------------------------
 
 def parse_filename(fname):
     """Parse a notebook filename.
 
     This function takes a notebook filename and returns the notebook
     format (json/py) and the notebook name. This logic can be
     summarized as follows:
 
     * notebook.ipynb -> (notebook.ipynb, notebook, json)
     * notebook.json  -> (notebook.json, notebook, json)
     * notebook.py    -> (notebook.py, notebook, py)
     * notebook       -> (notebook.ipynb, notebook, json)
 
     Parameters
     ----------
     fname : unicode
         The notebook filename. The filename can use a specific filename
         extention (.ipynb, .json, .py) or none, in which case .ipynb will
         be assumed.
 
     Returns
     -------
     (fname, name, format) : (unicode, unicode, unicode)
         The filename, notebook name and format.
     """
     if fname.endswith(u'.ipynb'):
         format = u'json'
     elif fname.endswith(u'.json'):
         format = u'json'
     elif fname.endswith(u'.py'):
         format = u'py'
     else:
         fname = fname + u'.ipynb'
         format = u'json'
     name = fname.split('.')[0]
     return fname, name, format
 

IPython/nbformat/v3/nbbase.py

@@ -1,204 +1,206 b''
 """The basic dict based notebook format.
 
 The Python representation of a notebook is a nested structure of
 dictionary subclasses that support attribute access
 (IPython.utils.ipstruct.Struct). The functions in this module are merely
 helpers to build the structs in the right form.
 """
 
 # Copyright (c) IPython Development Team.
 # Distributed under the terms of the Modified BSD License.
 
 import pprint
 import uuid
+import os
 
 from IPython.utils.ipstruct import Struct
 from IPython.utils.py3compat import cast_unicode, unicode_type
 
 #-----------------------------------------------------------------------------
 # Code
 #-----------------------------------------------------------------------------
 
 # Change this when incrementing the nbformat version
 nbformat = 3
 nbformat_minor = 0
+nbformat_schema = 'v3.withref.json'
 
 class NotebookNode(Struct):
     pass
 
 
 def from_dict(d):
     if isinstance(d, dict):
         newd = NotebookNode()
         for k,v in d.items():
             newd[k] = from_dict(v)
         return newd
     elif isinstance(d, (tuple, list)):
         return [from_dict(i) for i in d]
     else:
         return d
 
 
 def new_output(output_type, output_text=None, output_png=None,
     output_html=None, output_svg=None, output_latex=None, output_json=None,
     output_javascript=None, output_jpeg=None, prompt_number=None,
     ename=None, evalue=None, traceback=None, stream=None, metadata=None):
     """Create a new output, to go in the ``cell.outputs`` list of a code cell.
     """
     output = NotebookNode()
     output.output_type = unicode_type(output_type)
 
     if metadata is None:
         metadata = {}
     if not isinstance(metadata, dict):
         raise TypeError("metadata must be dict")
     output.metadata = metadata
 
     if output_type != 'pyerr':
         if output_text is not None:
             output.text = cast_unicode(output_text)
         if output_png is not None:
             output.png = cast_unicode(output_png)
         if output_jpeg is not None:
             output.jpeg = cast_unicode(output_jpeg)
         if output_html is not None:
             output.html = cast_unicode(output_html)
         if output_svg is not None:
             output.svg = cast_unicode(output_svg)
         if output_latex is not None:
             output.latex = cast_unicode(output_latex)
         if output_json is not None:
             output.json = cast_unicode(output_json)
         if output_javascript is not None:
             output.javascript = cast_unicode(output_javascript)
 
     if output_type == u'pyout':
         if prompt_number is not None:
             output.prompt_number = int(prompt_number)
 
     if output_type == u'pyerr':
         if ename is not None:
             output.ename = cast_unicode(ename)
         if evalue is not None:
             output.evalue = cast_unicode(evalue)
         if traceback is not None:
             output.traceback = [cast_unicode(frame) for frame in list(traceback)]
 
     if output_type == u'stream':
         output.stream = 'stdout' if stream is None else cast_unicode(stream)
     
     return output
 
 
 def new_code_cell(input=None, prompt_number=None, outputs=None,
     language=u'python', collapsed=False, metadata=None):
     """Create a new code cell with input and output"""
     cell = NotebookNode()
     cell.cell_type = u'code'
     if language is not None:
         cell.language = cast_unicode(language)
     if input is not None:
         cell.input = cast_unicode(input)
     if prompt_number is not None:
         cell.prompt_number = int(prompt_number)
     if outputs is None:
         cell.outputs = []
     else:
         cell.outputs = outputs
     if collapsed is not None:
         cell.collapsed = bool(collapsed)
     cell.metadata = NotebookNode(metadata or {})
 
     return cell
 
 def new_text_cell(cell_type, source=None, rendered=None, metadata=None):
     """Create a new text cell."""
     cell = NotebookNode()
     # VERSIONHACK: plaintext -> raw
     # handle never-released plaintext name for raw cells
     if cell_type == 'plaintext':
         cell_type = 'raw'
     if source is not None:
         cell.source = cast_unicode(source)
     if rendered is not None:
         cell.rendered = cast_unicode(rendered)
     cell.metadata = NotebookNode(metadata or {})
     cell.cell_type = cell_type
     return cell
 
 
 def new_heading_cell(source=None, rendered=None, level=1, metadata=None):
     """Create a new section cell with a given integer level."""
     cell = NotebookNode()
     cell.cell_type = u'heading'
     if source is not None:
         cell.source = cast_unicode(source)
     if rendered is not None:
         cell.rendered = cast_unicode(rendered)
     cell.level = int(level)
     cell.metadata = NotebookNode(metadata or {})
     return cell
 
 
 def new_worksheet(name=None, cells=None, metadata=None):
     """Create a worksheet by name with with a list of cells."""
     ws = NotebookNode()
     if name is not None:
         ws.name = cast_unicode(name)
     if cells is None:
         ws.cells = []
     else:
         ws.cells = list(cells)
     ws.metadata = NotebookNode(metadata or {})
     return ws
 
 
 def new_notebook(name=None, metadata=None, worksheets=None):
     """Create a notebook by name, id and a list of worksheets."""
     nb = NotebookNode()
     nb.nbformat = nbformat
     nb.nbformat_minor = nbformat_minor
     if worksheets is None:
         nb.worksheets = []
     else:
         nb.worksheets = list(worksheets)
     if metadata is None:
         nb.metadata = new_metadata()
     else:
         nb.metadata = NotebookNode(metadata)
     if name is not None:
         nb.metadata.name = cast_unicode(name)
     return nb
 
 
 def new_metadata(name=None, authors=None, license=None, created=None,
     modified=None, gistid=None):
     """Create a new metadata node."""
     metadata = NotebookNode()
     if name is not None:
         metadata.name = cast_unicode(name)
     if authors is not None:
         metadata.authors = list(authors)
     if created is not None:
         metadata.created = cast_unicode(created)
     if modified is not None:
         metadata.modified = cast_unicode(modified)
     if license is not None:
         metadata.license = cast_unicode(license)
     if gistid is not None:
         metadata.gistid = cast_unicode(gistid)
     return metadata
 
 def new_author(name=None, email=None, affiliation=None, url=None):
     """Create a new author."""
     author = NotebookNode()
     if name is not None:
         author.name = cast_unicode(name)
     if email is not None:
         author.email = cast_unicode(email)
     if affiliation is not None:
         author.affiliation = cast_unicode(affiliation)
     if url is not None:
         author.url = cast_unicode(url)
     return author