Show More
@@ -0,0 +1,19 b'' | |||||
|
1 | from nbconvert import ConverterRST | |||
|
2 | import nose.tools as nt | |||
|
3 | ||||
|
4 | import os | |||
|
5 | import glob | |||
|
6 | ||||
|
7 | def clean_dir(): | |||
|
8 | "Remove .rst files created during conversion" | |||
|
9 | map(os.remove, glob.glob("*.rst")) | |||
|
10 | map(os.remove, glob.glob("*.png")) | |||
|
11 | ||||
|
12 | ||||
|
13 | @nt.with_setup(clean_dir, clean_dir) | |||
|
14 | def test_simple(): | |||
|
15 | fname = 'test.ipynb' | |||
|
16 | c = ConverterRST(fname) | |||
|
17 | f = c.render() | |||
|
18 | nt.assert_true('rst' in f, 'changed file extension to rst') | |||
|
19 |
@@ -1,198 +1,222 b'' | |||||
1 | #!/usr/bin/env python |
|
1 | #!/usr/bin/env python | |
2 | """A really simple notebook to rst/html exporter. |
|
2 | """A really simple notebook to rst/html exporter. | |
3 |
|
3 | |||
4 | Usage |
|
4 | Usage | |
5 |
|
5 | |||
6 | ./nb2html.py file.ipynb |
|
6 | ./nb2html.py file.ipynb | |
7 |
|
7 | |||
8 | Produces 'file.rst' and 'file.html', along with auto-generated figure files |
|
8 | Produces 'file.rst' and 'file.html', along with auto-generated figure files | |
9 | called nb_figure_NN.png. |
|
9 | called nb_figure_NN.png. | |
10 |
|
10 | |||
11 | """ |
|
11 | """ | |
12 |
|
12 | |||
13 | import os |
|
13 | import os | |
14 | import subprocess |
|
14 | import subprocess | |
15 | import sys |
|
15 | import sys | |
16 |
|
16 | |||
17 | from IPython.nbformat import current as nbformat |
|
17 | from IPython.nbformat import current as nbformat | |
18 | from IPython.utils.text import wrap_paragraphs, indent |
|
18 | from IPython.utils.text import wrap_paragraphs, indent | |
19 |
|
19 | |||
20 |
|
20 | |||
21 | # Cell converters |
|
21 | # Cell converters | |
22 |
|
22 | |||
23 | def unknown_cell(cell): |
|
23 | def unknown_cell(cell): | |
24 | """Default converter for cells of unknown type. |
|
24 | """Default converter for cells of unknown type. | |
25 | """ |
|
25 | """ | |
26 |
|
26 | |||
27 | return rst_directive('.. warning:: Unknown cell') + \ |
|
27 | return rst_directive('.. warning:: Unknown cell') + \ | |
28 | [repr(cell)] |
|
28 | [repr(cell)] | |
29 |
|
29 | |||
30 | def heading_cell(cell): |
|
|||
31 | """convert a heading cell to rst |
|
|||
32 |
|
||||
33 | Returns list.""" |
|
|||
34 | heading_level = {1:'=', 2:'-', 3:'`', 4:'\'', 5:'.',6:'~'} |
|
|||
35 | marker = heading_level[cell.level] |
|
|||
36 | return ['{0}\n{1}\n'.format(cell.source, marker*len(cell.source))] |
|
|||
37 |
|
||||
38 | def markdown_cell(cell): |
|
|||
39 | """convert a markdown cell to rst |
|
|||
40 |
|
||||
41 | Returns list.""" |
|
|||
42 | return [cell.source] |
|
|||
43 |
|
||||
44 |
|
30 | |||
45 | def rst_directive(directive, text=''): |
|
31 | def rst_directive(directive, text=''): | |
46 | out = [directive, ''] |
|
32 | out = [directive, ''] | |
47 | if text: |
|
33 | if text: | |
48 | out.extend([indent(text), '']) |
|
34 | out.extend([indent(text), '']) | |
49 | return out |
|
35 | return out | |
50 |
|
36 | |||
51 | def code_cell(cell): |
|
|||
52 | """Convert a code cell to rst |
|
|||
53 |
|
||||
54 | Returns list.""" |
|
|||
55 |
|
||||
56 | if not cell.input: |
|
|||
57 | return [] |
|
|||
58 |
|
||||
59 | lines = ['In[%s]:' % cell.prompt_number, ''] |
|
|||
60 | lines.extend(rst_directive('.. code:: python', cell.input)) |
|
|||
61 |
|
||||
62 | for output in cell.outputs: |
|
|||
63 | conv = converters.get(output.output_type, unknown_cell) |
|
|||
64 | lines.extend(conv(output)) |
|
|||
65 |
|
||||
66 | return lines |
|
|||
67 |
|
||||
68 | # Converters for parts of a cell. |
|
37 | # Converters for parts of a cell. | |
69 | figures_counter = 1 |
|
38 | figures_counter = 1 | |
70 |
|
39 | |||
71 | def out_display(output): |
|
40 | class ConversionException(Exception): | |
72 | """convert display data from the output of a code cell to rst. |
|
41 | pass | |
|
42 | ||||
|
43 | class Converter(object): | |||
|
44 | default_encoding = 'utf-8' | |||
|
45 | def __init__(self, fname): | |||
|
46 | self.fname = fname | |||
|
47 | ||||
|
48 | @property | |||
|
49 | def extension(self): | |||
|
50 | raise ConversionException("""extension must be defined in Converter | |||
|
51 | subclass""") | |||
|
52 | ||||
|
53 | def dispatch(self,cell_type): | |||
|
54 | """return cell_type dependent render method, for example render_code | |||
|
55 | """ | |||
|
56 | return getattr(self, 'render_'+cell_type, unknown_cell) | |||
|
57 | ||||
|
58 | def convert(self): | |||
|
59 | lines = [] | |||
|
60 | for cell in self.nb.worksheets[0].cells: | |||
|
61 | conv_fn = self.dispatch(cell.cell_type) | |||
|
62 | lines.extend(conv_fn(cell)) | |||
|
63 | lines.append('') | |||
|
64 | return '\n'.join(lines) | |||
|
65 | ||||
|
66 | def render(self): | |||
|
67 | "read, convert, and save self.fname" | |||
|
68 | self.read() | |||
|
69 | self.output = self.convert() | |||
|
70 | return self.save() | |||
|
71 | ||||
|
72 | def read(self): | |||
|
73 | "read and parse notebook into NotebookNode called self.nb" | |||
|
74 | with open(self.fname) as f: | |||
|
75 | self.nb = nbformat.read(f, 'json') | |||
|
76 | ||||
|
77 | def save(self,fname=None, encoding=None): | |||
|
78 | "read and parse notebook into self.nb" | |||
|
79 | if fname is None: | |||
|
80 | fname = os.path.splitext(self.fname)[0] + '.' + self.extension | |||
|
81 | if encoding is None: | |||
|
82 | encoding = self.default_encoding | |||
|
83 | with open(fname, 'w') as f: | |||
|
84 | f.write(self.output.encode(encoding)) | |||
|
85 | return fname | |||
73 |
|
86 | |||
74 | Returns list. |
|
87 | def render_heading(self,cell): | |
75 | """ |
|
88 | raise NotImplementedError | |
76 | global figures_counter |
|
89 | def render_code(self,cell): | |
|
90 | raise NotImplementedError | |||
|
91 | def render_markdown(self,cell): | |||
|
92 | raise NotImplementedError | |||
|
93 | def render_pyout(self,cell): | |||
|
94 | raise NotImplementedError | |||
|
95 | def render_display_data(self,cell): | |||
|
96 | raise NotImplementedError | |||
77 |
|
97 | |||
78 | lines = [] |
|
98 | class ConverterRST(Converter): | |
|
99 | extension = 'rst' | |||
|
100 | def render_heading(self,cell): | |||
|
101 | """convert a heading cell to rst | |||
79 |
|
|
102 | ||
80 | if 'png' in output: |
|
103 | Returns list.""" | |
81 | fname = 'nb_figure_%s.png' % figures_counter |
|
104 | heading_level = {1:'=', 2:'-', 3:'`', 4:'\'', 5:'.',6:'~'} | |
82 | with open(fname, 'w') as f: |
|
105 | marker = heading_level[cell.level] | |
83 | f.write(output.png.decode('base64')) |
|
106 | return ['{0}\n{1}\n'.format(cell.source, marker*len(cell.source))] | |
84 |
|
107 | |||
85 | figures_counter += 1 |
|
108 | def render_code(self,cell): | |
86 | lines.append('.. image:: %s' % fname) |
|
109 | """Convert a code cell to rst | |
87 | lines.append('') |
|
|||
88 |
|
||||
89 | return lines |
|
|||
90 |
|
|
110 | ||
91 |
|
111 | Returns list.""" | ||
92 | def out_pyout(output): |
|
|||
93 | """convert pyout part of a code cell to rst |
|
|||
94 |
|
112 | |||
95 | Returns list.""" |
|
113 | if not cell.input: | |
|
114 | return [] | |||
96 |
|
115 | |||
97 |
lines = [' |
|
116 | lines = ['In[%s]:' % cell.prompt_number, ''] | |
98 |
|
117 | lines.extend(rst_directive('.. code:: python', cell.input)) | ||
99 | if 'latex' in output: |
|
|||
100 | lines.extend(rst_directive('.. math::', output.latex)) |
|
|||
101 |
|
118 | |||
102 | if 'text' in output: |
|
119 | for output in cell.outputs: | |
103 | lines.extend(rst_directive('.. parsed-literal::', output.text)) |
|
120 | conv_fn = self.dispatch(output.output_type) | |
|
121 | lines.extend(conv_fn(output)) | |||
104 |
|
122 | |||
105 | return lines |
|
123 | return lines | |
106 |
|
124 | |||
|
125 | def render_markdown(self,cell): | |||
|
126 | """convert a markdown cell to rst | |||
107 |
|
|
127 | ||
108 | converters = dict(heading = heading_cell, |
|
128 | Returns list.""" | |
109 | code = code_cell, |
|
129 | return [cell.source] | |
110 | markdown = markdown_cell, |
|
|||
111 | pyout = out_pyout, |
|
|||
112 | display_data = out_display, |
|
|||
113 | ) |
|
|||
114 |
|
130 | |||
|
131 | def render_pyout(self,output): | |||
|
132 | """convert pyout part of a code cell to rst | |||
115 |
|
|
133 | ||
|
134 | Returns list.""" | |||
116 |
|
135 | |||
117 | def convert_notebook(nb): |
|
136 | lines = ['Out[%s]:' % output.prompt_number, ''] | |
118 | lines = [] |
|
137 | ||
119 | for cell in nb.worksheets[0].cells: |
|
138 | if 'latex' in output: | |
120 | conv = converters.get(cell.cell_type, unknown_cell) |
|
139 | lines.extend(rst_directive('.. math::', output.latex)) | |
121 | lines.extend(conv(cell)) |
|
|||
122 | lines.append('') |
|
|||
123 |
|
||||
124 | return '\n'.join(lines) |
|
|||
125 |
|
140 | |||
|
141 | if 'text' in output: | |||
|
142 | lines.extend(rst_directive('.. parsed-literal::', output.text)) | |||
126 |
|
143 | |||
127 | def nb2rst(fname): |
|
144 | return lines | |
128 | "Convert notebook to rst" |
|
|||
129 |
|
||||
130 | with open(fname) as f: |
|
|||
131 | nb = nbformat.read(f, 'json') |
|
|||
132 |
|
145 | |||
133 | rst = convert_notebook(nb) |
|
146 | def render_display_data(self,output): | |
|
147 | """convert display data from the output of a code cell to rst. | |||
134 |
|
|
148 | ||
135 | newfname = os.path.splitext(fname)[0] + '.rst' |
|
149 | Returns list. | |
136 | with open(newfname, 'w') as f: |
|
150 | """ | |
137 | f.write(rst.encode('utf8')) |
|
151 | global figures_counter | |
138 |
|
152 | |||
139 | return newfname |
|
153 | lines = [] | |
|
154 | ||||
|
155 | if 'png' in output: | |||
|
156 | fname = 'nb_figure_%s.png' % figures_counter | |||
|
157 | with open(fname, 'w') as f: | |||
|
158 | f.write(output.png.decode('base64')) | |||
140 |
|
159 | |||
|
160 | figures_counter += 1 | |||
|
161 | lines.append('.. image:: %s' % fname) | |||
|
162 | lines.append('') | |||
|
163 | ||||
|
164 | return lines | |||
141 |
|
165 | |||
142 | def rst2simplehtml(fname): |
|
166 | def rst2simplehtml(fname): | |
143 | """Convert a rst file to simplified html suitable for blogger. |
|
167 | """Convert a rst file to simplified html suitable for blogger. | |
144 |
|
168 | |||
145 | This just runs rst2html with certain parameters to produce really simple |
|
169 | This just runs rst2html with certain parameters to produce really simple | |
146 | html and strips the document header, so the resulting file can be easily |
|
170 | html and strips the document header, so the resulting file can be easily | |
147 | pasted into a blogger edit window. |
|
171 | pasted into a blogger edit window. | |
148 | """ |
|
172 | """ | |
149 |
|
173 | |||
150 | # This is the template for the rst2html call that produces the cleanest, |
|
174 | # This is the template for the rst2html call that produces the cleanest, | |
151 | # simplest html I could find. This should help in making it easier to |
|
175 | # simplest html I could find. This should help in making it easier to | |
152 | # paste into the blogspot html window, though I'm still having problems |
|
176 | # paste into the blogspot html window, though I'm still having problems | |
153 | # with linebreaks there... |
|
177 | # with linebreaks there... | |
154 | cmd_template = ("rst2html --link-stylesheet --no-xml-declaration " |
|
178 | cmd_template = ("rst2html --link-stylesheet --no-xml-declaration " | |
155 | "--no-generator --no-datestamp --no-source-link " |
|
179 | "--no-generator --no-datestamp --no-source-link " | |
156 | "--no-toc-backlinks --no-section-numbering " |
|
180 | "--no-toc-backlinks --no-section-numbering " | |
157 | "--strip-comments ") |
|
181 | "--strip-comments ") | |
158 |
|
182 | |||
159 | cmd = "%s %s" % (cmd_template, fname) |
|
183 | cmd = "%s %s" % (cmd_template, fname) | |
160 | proc = subprocess.Popen(cmd, |
|
184 | proc = subprocess.Popen(cmd, | |
161 | stdout=subprocess.PIPE, |
|
185 | stdout=subprocess.PIPE, | |
162 | stderr=subprocess.PIPE, |
|
186 | stderr=subprocess.PIPE, | |
163 | shell=True) |
|
187 | shell=True) | |
164 | html, stderr = proc.communicate() |
|
188 | html, stderr = proc.communicate() | |
165 | if stderr: |
|
189 | if stderr: | |
166 | raise IOError(stderr) |
|
190 | raise IOError(stderr) | |
167 |
|
191 | |||
168 | # Make an iterator so breaking out holds state. Our implementation of |
|
192 | # Make an iterator so breaking out holds state. Our implementation of | |
169 | # searching for the html body below is basically a trivial little state |
|
193 | # searching for the html body below is basically a trivial little state | |
170 | # machine, so we need this. |
|
194 | # machine, so we need this. | |
171 | walker = iter(html.splitlines()) |
|
195 | walker = iter(html.splitlines()) | |
172 |
|
196 | |||
173 | # Find start of main text, break out to then print until we find end /div. |
|
197 | # Find start of main text, break out to then print until we find end /div. | |
174 | # This may only work if there's a real title defined so we get a 'div class' |
|
198 | # This may only work if there's a real title defined so we get a 'div class' | |
175 | # tag, I haven't really tried. |
|
199 | # tag, I haven't really tried. | |
176 | for line in walker: |
|
200 | for line in walker: | |
177 | if line.startswith('<body>'): |
|
201 | if line.startswith('<body>'): | |
178 | break |
|
202 | break | |
179 |
|
203 | |||
180 | newfname = os.path.splitext(fname)[0] + '.html' |
|
204 | newfname = os.path.splitext(fname)[0] + '.html' | |
181 | with open(newfname, 'w') as f: |
|
205 | with open(newfname, 'w') as f: | |
182 | for line in walker: |
|
206 | for line in walker: | |
183 | if line.startswith('</body>'): |
|
207 | if line.startswith('</body>'): | |
184 | break |
|
208 | break | |
185 | f.write(line) |
|
209 | f.write(line) | |
186 | f.write('\n') |
|
210 | f.write('\n') | |
187 |
|
211 | |||
188 | return newfname |
|
212 | return newfname | |
189 |
|
213 | |||
190 |
|
214 | |||
191 | def main(fname): |
|
215 | def main(fname): | |
192 | """Convert a notebook to html in one step""" |
|
216 | """Convert a notebook to html in one step""" | |
193 | newfname = nb2rst(fname) |
|
217 | newfname = nb2rst(fname) | |
194 | rst2simplehtml(newfname) |
|
218 | #rst2simplehtml(newfname) | |
195 |
|
219 | |||
196 |
|
220 | |||
197 | if __name__ == '__main__': |
|
221 | if __name__ == '__main__': | |
198 | main(sys.argv[1]) |
|
222 | main(sys.argv[1]) |
1 | NO CONTENT: file renamed from test.ipynb to tests/test.ipynb |
|
NO CONTENT: file renamed from test.ipynb to tests/test.ipynb |
General Comments 0
You need to be logged in to leave comments.
Login now