Merge pull request #15 from mwouts/v0.3.0

V0.3.0

.travis.yml

@@ -9,7 +9,7 @@ python:
 # command to install dependencies
 install:
   - pip install codecov
-  - pip install pytest pytest-cov
+  - pip install pytest pytest-cov testfixtures
   - pip install notebook
   - pip install -r requirements.txt
   - pip install .

HISTORY.rst

@@ -6,6 +6,14 @@ Release History
 dev
 +++
 
+
+0.3.0 (2018-07-17)
++++++++++++++++++++
+
+**Improvements**
+
+- Introducing support for notebooks as python `.py` or R scripts `.R`
+
 0.2.6 (2018-07-13)
 +++++++++++++++++++
 

README.md

@@ -23,11 +23,12 @@ The format actually supports [many languages](https://yihui
 .name/knitr/demo/engines/).
 
 R markdown is almost like plain markdown. There are only two differences:
-- R markdown has a specific syntax for active code cells, that start with
-```
-```{python}
-```
-These active cells may optionally contain cell options.
+- R markdown has a specific syntax for active code cells: language, and
+optional cell options are enclosed into a pair of curly brackets:
+
+    ```
+    ```{python}
+
 - a YAML header, that describes the notebook title, author, and desired
 output (HTML, slides, PDF...).
 

nbrmd/__init__.py

@@ -9,14 +9,14 @@
 R Markdown notebooks.
 """
 
-from .nbrmd import read, reads, readf, write, writes, writef
+from .nbrmd import readf, writef, writes, reads, notebook_extensions, readme
 
 try:
     from .rmarkdownexporter import RMarkdownExporter
 except ImportError as e:
     RMarkdownExporter = str(e)
 
 try:
-    from .cm import RmdFileContentsManager
+    from .contentsmanager import RmdFileContentsManager
 except ImportError as e:
     RmdFileContentsManager = str(e)

nbrmd/chunk_options.py → nbrmd/cell_metadata.py

@@ -9,6 +9,7 @@
 """
 
 import ast
+import json
 
 _boolean_options_dictionary = [('hide_input', 'echo', True),
                                ('hide_output', 'include', True)]
@@ -36,7 +37,7 @@ def _py_logical_values(rbool):
     raise RLogicalValueError
 
 
-def to_chunk_options(language, metadata):
+def metadata_to_rmd_options(language, metadata):
     options = language.lower()
     if 'name' in metadata:
         options += ' ' + metadata['name'] + ','
@@ -60,7 +61,7 @@ def to_chunk_options(language, metadata):
                     ', '.join(['"{}"'.format(str(v)) for v in co_value])))
         else:
             options += ' {}={},'.format(co_name, str(co_value))
-    return options.strip(',')
+    return options.strip(',').strip()
 
 
 def update_metadata_using_dictionary(name, value, metadata):
@@ -141,7 +142,7 @@ def parse_rmd_options(line):
                     if len(result) and name is '':
                         raise RMarkdownOptionParsingError(
                             'Option line "{}" has no name for '
-                            'option value {}' .format(line, value))
+                            'option value {}'.format(line, value))
                     result.append((name.strip(), value.strip()))
                     name = ''
                     value = ''
@@ -166,7 +167,7 @@ def parse_rmd_options(line):
     return result
 
 
-def to_metadata(options):
+def rmd_options_to_metadata(options):
     options = options.split(' ', 1)
     if len(options) == 1:
         language = options[0]
@@ -207,3 +208,15 @@ def to_metadata(options):
             continue
 
     return language, metadata
+
+
+def json_options_to_metadata(options):
+    try:
+        return json.loads(options)
+    except ValueError:
+        return {}
+
+
+def metadata_to_json_options(metadata):
+    return json.dumps({k: metadata[k] for k in metadata
+                       if k not in _ignore_metadata})

nbrmd/cells.py

@@ -0,0 +1,198 @@
+from .languages import cell_language
+from .cell_metadata import metadata_to_rmd_options, rmd_options_to_metadata, \
+    json_options_to_metadata, metadata_to_json_options
+from nbformat.v4.nbbase import new_code_cell, new_markdown_cell
+import re
+
+
+def code_to_rmd(source, metadata, default_language):
+    lines = []
+    language = cell_language(source) or default_language
+    options = metadata_to_rmd_options(language, metadata)
+    lines.append(u'```{{{}}}'.format(options))
+    lines.extend(source)
+    lines.append(u'```')
+    return lines
+
+
+def cell_to_text(self,
+                 cell,
+                 next_cell=None,
+                 default_language='python'):
+    source = cell.get('source').splitlines()
+    metadata = cell.get('metadata', {})
+    skipline = True
+    if 'noskipline' in metadata:
+        skipline = not metadata['noskipline']
+        del metadata['noskipline']
+
+    lines = []
+    if cell.cell_type == 'code':
+        if self.ext == '.Rmd':
+            lines.extend(code_to_rmd(source, metadata, default_language))
+        else:
+            language = cell_language(source) or default_language
+            if language == default_language:
+                if self.ext == '.R':
+                    options = metadata_to_rmd_options(language, metadata)[2:]
+                    if options != '':
+                        lines.append('#+ ' + options)
+                else:
+                    options = metadata_to_json_options(metadata)
+                    if options != '{}':
+                        lines.append('#+ ' + options)
+                lines.extend(source)
+            else:
+                lines.extend(self.markdown_escape(
+                    code_to_rmd(source, metadata, default_language)))
+
+            # Two blank lines before next code cell
+            if next_cell and next_cell.cell_type == 'code':
+                lines.append('')
+    else:
+        if source == []:
+            source = ['']
+        lines.extend(self.markdown_escape(source))
+
+        # Two blank lines between consecutive markdown cells
+        if self.ext == '.Rmd' and next_cell \
+                and next_cell.cell_type == 'markdown':
+            lines.append('')
+
+    if skipline and next_cell:
+        lines.append('')
+
+    return lines
+
+
+_start_code_rmd = re.compile(r"^```\{(.*)\}\s*$")
+_start_code_md = re.compile(r"^```(.*)$")
+_end_code_md = re.compile(r"^```\s*$")
+_option_code_rpy = re.compile(r"^#\+(.*)$")
+_blank = re.compile(r"^\s*$")
+
+
+def start_code_rmd(line):
+    return _start_code_rmd.match(line)
+
+
+def start_code_rpy(line):
+    return _option_code_rpy.match(line)
+
+
+def text_to_cell(self, lines):
+    if self.start_code(lines[0]):
+        return self.code_to_cell(lines, parse_opt=True)
+    elif self.prefix != '' and not lines[0].startswith(self.prefix):
+        return self.code_to_cell(lines, parse_opt=False)
+    else:
+        return self.markdown_to_cell(lines)
+
+
+def parse_code_options(line, ext):
+    if ext == '.Rmd':
+        return rmd_options_to_metadata(_start_code_rmd.findall(line)[0])
+    elif ext == '.R':
+        return rmd_options_to_metadata(_option_code_rpy.findall(line)[0])
+    else:  # .py
+        return 'python', json_options_to_metadata(_option_code_rpy.findall(
+            line)[0])
+
+
+def code_to_cell(self, lines, parse_opt):
+    # Parse options
+    if parse_opt:
+        language, metadata = parse_code_options(lines[0], self.ext)
+        if self.ext == '.Rmd':
+            metadata['language'] = language
+    else:
+        metadata = {}
+
+    # Find end of cell and return
+    if self.ext == '.Rmd':
+        for pos, line in enumerate(lines):
+            if pos > 0 and _end_code_md.match(line):
+                next_line_blank = pos + 1 == len(lines) or \
+                                  _blank.match(lines[pos + 1])
+                if next_line_blank and pos + 2 != len(lines):
+                    return new_code_cell(
+                        source='\n'.join(lines[1:pos]), metadata=metadata), \
+                           pos + 2
+                else:
+                    r = new_code_cell(
+                        source='\n'.join(lines[1:pos]),
+                        metadata=metadata)
+                    r.metadata['noskipline'] = True
+                    return r, pos + 1
+    else:
+        prev_blank = False
+        for pos, line in enumerate(lines):
+            if parse_opt and pos == 0:
+                continue
+
+            if self.prefix != '' and line.startswith(self.prefix):
+                if prev_blank:
+                    return new_code_cell(
+                        source='\n'.join(lines[parse_opt:(pos - 1)]),
+                        metadata=metadata), pos
+                else:
+                    r = new_code_cell(
+                        source='\n'.join(lines[parse_opt:pos]),
+                        metadata=metadata)
+                    r.metadata['noskipline'] = True
+                    return r, pos
+
+            if _blank.match(line):
+                if prev_blank:
+                    # Two blank lines at the end == empty code cell
+                    return new_code_cell(
+                        source='\n'.join(lines[parse_opt:(pos - 1)]),
+                        metadata=metadata), min(pos + 1, len(lines) - 1)
+                prev_blank = True
+            else:
+                prev_blank = False
+
+    # Unterminated cell?
+    return new_code_cell(
+        source='\n'.join(lines[parse_opt:]),
+        metadata=metadata), len(lines)
+
+
+def markdown_to_cell(self, lines):
+    md = []
+    for pos, line in enumerate(lines):
+        # Markdown stops with the end of comments
+        if line.startswith(self.prefix):
+            md.append(self.markdown_unescape(line))
+        elif _blank.match(line):
+            return new_markdown_cell(source='\n'.join(md)), pos + 1
+        else:
+            r = new_markdown_cell(source='\n'.join(md))
+            r.metadata['noskipline'] = True
+            return r, pos
+
+    # still here => unterminated markdown
+    return new_markdown_cell(source='\n'.join(md)), len(lines)
+
+
+def markdown_to_cell_rmd(lines):
+    prev_blank = False
+    for pos, line in enumerate(lines):
+        if start_code_rmd(line):
+            if prev_blank and pos > 1:
+                return new_markdown_cell(
+                    source='\n'.join(lines[:(pos - 1)])), pos
+            else:
+
+                r = new_markdown_cell(
+                    source='\n'.join(lines[:pos]))
+                r.metadata['noskipline'] = True
+                return r, pos
+
+        if _blank.match(line) and prev_blank:
+            return new_markdown_cell(
+                source='\n'.join(lines[:(pos - 1)])), pos + 1
+        prev_blank = _blank.match(line)
+
+    # Unterminated cell?
+    return new_markdown_cell(source='\n'.join(lines)), len(lines)