Show More
@@ -1,270 +1,270 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 | import argparse |
|
16 | from IPython.external import argparse | |
17 | from IPython.nbformat import current as nbformat |
|
17 | from IPython.nbformat import current as nbformat | |
18 | from IPython.utils.text import indent |
|
18 | from IPython.utils.text import 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 |
|
30 | |||
31 | def rst_directive(directive, text=''): |
|
31 | def rst_directive(directive, text=''): | |
32 | out = [directive, ''] |
|
32 | out = [directive, ''] | |
33 | if text: |
|
33 | if text: | |
34 | out.extend([indent(text), '']) |
|
34 | out.extend([indent(text), '']) | |
35 | return out |
|
35 | return out | |
36 |
|
36 | |||
37 | # Converters for parts of a cell. |
|
37 | # Converters for parts of a cell. | |
38 |
|
38 | |||
39 |
|
39 | |||
40 | class ConversionException(Exception): |
|
40 | class ConversionException(Exception): | |
41 | pass |
|
41 | pass | |
42 |
|
42 | |||
43 |
|
43 | |||
44 | class Converter(object): |
|
44 | class Converter(object): | |
45 | default_encoding = 'utf-8' |
|
45 | default_encoding = 'utf-8' | |
46 |
|
46 | |||
47 | def __init__(self, infile): |
|
47 | def __init__(self, infile): | |
48 | self.infile = infile |
|
48 | self.infile = infile | |
49 | self.dirpath = os.path.dirname(infile) |
|
49 | self.dirpath = os.path.dirname(infile) | |
50 |
|
50 | |||
51 | @property |
|
51 | @property | |
52 | def extension(self): |
|
52 | def extension(self): | |
53 | raise ConversionException("""extension must be defined in Converter |
|
53 | raise ConversionException("""extension must be defined in Converter | |
54 | subclass""") |
|
54 | subclass""") | |
55 |
|
55 | |||
56 | def dispatch(self, cell_type): |
|
56 | def dispatch(self, cell_type): | |
57 | """return cell_type dependent render method, for example render_code |
|
57 | """return cell_type dependent render method, for example render_code | |
58 | """ |
|
58 | """ | |
59 | return getattr(self, 'render_' + cell_type, unknown_cell) |
|
59 | return getattr(self, 'render_' + cell_type, unknown_cell) | |
60 |
|
60 | |||
61 | def convert(self): |
|
61 | def convert(self): | |
62 | lines = [] |
|
62 | lines = [] | |
63 | for cell in self.nb.worksheets[0].cells: |
|
63 | for cell in self.nb.worksheets[0].cells: | |
64 | conv_fn = self.dispatch(cell.cell_type) |
|
64 | conv_fn = self.dispatch(cell.cell_type) | |
65 | lines.extend(conv_fn(cell)) |
|
65 | lines.extend(conv_fn(cell)) | |
66 | lines.append('') |
|
66 | lines.append('') | |
67 | return '\n'.join(lines) |
|
67 | return '\n'.join(lines) | |
68 |
|
68 | |||
69 | def render(self): |
|
69 | def render(self): | |
70 | "read, convert, and save self.infile" |
|
70 | "read, convert, and save self.infile" | |
71 | self.read() |
|
71 | self.read() | |
72 | self.output = self.convert() |
|
72 | self.output = self.convert() | |
73 | return self.save() |
|
73 | return self.save() | |
74 |
|
74 | |||
75 | def read(self): |
|
75 | def read(self): | |
76 | "read and parse notebook into NotebookNode called self.nb" |
|
76 | "read and parse notebook into NotebookNode called self.nb" | |
77 | with open(self.infile) as f: |
|
77 | with open(self.infile) as f: | |
78 | self.nb = nbformat.read(f, 'json') |
|
78 | self.nb = nbformat.read(f, 'json') | |
79 |
|
79 | |||
80 | def save(self, infile=None, encoding=None): |
|
80 | def save(self, infile=None, encoding=None): | |
81 | "read and parse notebook into self.nb" |
|
81 | "read and parse notebook into self.nb" | |
82 | if infile is None: |
|
82 | if infile is None: | |
83 | infile = os.path.splitext(self.infile)[0] + '.' + self.extension |
|
83 | infile = os.path.splitext(self.infile)[0] + '.' + self.extension | |
84 | if encoding is None: |
|
84 | if encoding is None: | |
85 | encoding = self.default_encoding |
|
85 | encoding = self.default_encoding | |
86 | with open(infile, 'w') as f: |
|
86 | with open(infile, 'w') as f: | |
87 | f.write(self.output.encode(encoding)) |
|
87 | f.write(self.output.encode(encoding)) | |
88 | return infile |
|
88 | return infile | |
89 |
|
89 | |||
90 | def render_heading(self, cell): |
|
90 | def render_heading(self, cell): | |
91 | raise NotImplementedError |
|
91 | raise NotImplementedError | |
92 |
|
92 | |||
93 | def render_code(self, cell): |
|
93 | def render_code(self, cell): | |
94 | raise NotImplementedError |
|
94 | raise NotImplementedError | |
95 |
|
95 | |||
96 | def render_markdown(self, cell): |
|
96 | def render_markdown(self, cell): | |
97 | raise NotImplementedError |
|
97 | raise NotImplementedError | |
98 |
|
98 | |||
99 | def render_pyout(self, cell): |
|
99 | def render_pyout(self, cell): | |
100 | raise NotImplementedError |
|
100 | raise NotImplementedError | |
101 |
|
101 | |||
102 | def render_display_data(self, cell): |
|
102 | def render_display_data(self, cell): | |
103 | raise NotImplementedError |
|
103 | raise NotImplementedError | |
104 |
|
104 | |||
105 | def render_stream(self, cell): |
|
105 | def render_stream(self, cell): | |
106 | raise NotImplementedError |
|
106 | raise NotImplementedError | |
107 |
|
107 | |||
108 |
|
108 | |||
109 | class ConverterRST(Converter): |
|
109 | class ConverterRST(Converter): | |
110 | extension = 'rst' |
|
110 | extension = 'rst' | |
111 | figures_counter = 0 |
|
111 | figures_counter = 0 | |
112 |
|
112 | |||
113 | def render_heading(self, cell): |
|
113 | def render_heading(self, cell): | |
114 | """convert a heading cell to rst |
|
114 | """convert a heading cell to rst | |
115 |
|
115 | |||
116 | Returns list.""" |
|
116 | Returns list.""" | |
117 | heading_level = {1: '=', 2: '-', 3: '`', 4: '\'', 5: '.', 6: '~'} |
|
117 | heading_level = {1: '=', 2: '-', 3: '`', 4: '\'', 5: '.', 6: '~'} | |
118 | marker = heading_level[cell.level] |
|
118 | marker = heading_level[cell.level] | |
119 | return ['{0}\n{1}\n'.format(cell.source, marker * len(cell.source))] |
|
119 | return ['{0}\n{1}\n'.format(cell.source, marker * len(cell.source))] | |
120 |
|
120 | |||
121 | def render_code(self, cell): |
|
121 | def render_code(self, cell): | |
122 | """Convert a code cell to rst |
|
122 | """Convert a code cell to rst | |
123 |
|
123 | |||
124 | Returns list.""" |
|
124 | Returns list.""" | |
125 |
|
125 | |||
126 | if not cell.input: |
|
126 | if not cell.input: | |
127 | return [] |
|
127 | return [] | |
128 |
|
128 | |||
129 | lines = ['In[%s]:' % cell.prompt_number, ''] |
|
129 | lines = ['In[%s]:' % cell.prompt_number, ''] | |
130 | lines.extend(rst_directive('.. code:: python', cell.input)) |
|
130 | lines.extend(rst_directive('.. code:: python', cell.input)) | |
131 |
|
131 | |||
132 | for output in cell.outputs: |
|
132 | for output in cell.outputs: | |
133 | conv_fn = self.dispatch(output.output_type) |
|
133 | conv_fn = self.dispatch(output.output_type) | |
134 | lines.extend(conv_fn(output)) |
|
134 | lines.extend(conv_fn(output)) | |
135 |
|
135 | |||
136 | return lines |
|
136 | return lines | |
137 |
|
137 | |||
138 | def render_markdown(self, cell): |
|
138 | def render_markdown(self, cell): | |
139 | """convert a markdown cell to rst |
|
139 | """convert a markdown cell to rst | |
140 |
|
140 | |||
141 | Returns list.""" |
|
141 | Returns list.""" | |
142 | return [cell.source] |
|
142 | return [cell.source] | |
143 |
|
143 | |||
144 | def render_plaintext(self, cell): |
|
144 | def render_plaintext(self, cell): | |
145 | """convert plain text to rst |
|
145 | """convert plain text to rst | |
146 |
|
146 | |||
147 | Returns list.""" |
|
147 | Returns list.""" | |
148 | return [cell.source] |
|
148 | return [cell.source] | |
149 |
|
149 | |||
150 | def render_pyout(self, output): |
|
150 | def render_pyout(self, output): | |
151 | """convert pyout part of a code cell to rst |
|
151 | """convert pyout part of a code cell to rst | |
152 |
|
152 | |||
153 | Returns list.""" |
|
153 | Returns list.""" | |
154 |
|
154 | |||
155 | lines = ['Out[%s]:' % output.prompt_number, ''] |
|
155 | lines = ['Out[%s]:' % output.prompt_number, ''] | |
156 |
|
156 | |||
157 | # output is a dictionary like object with type as a key |
|
157 | # output is a dictionary like object with type as a key | |
158 | if 'latex' in output: |
|
158 | if 'latex' in output: | |
159 | lines.extend(rst_directive('.. math::', output.latex)) |
|
159 | lines.extend(rst_directive('.. math::', output.latex)) | |
160 |
|
160 | |||
161 | if 'text' in output: |
|
161 | if 'text' in output: | |
162 | lines.extend(rst_directive('.. parsed-literal::', output.text)) |
|
162 | lines.extend(rst_directive('.. parsed-literal::', output.text)) | |
163 |
|
163 | |||
164 | return lines |
|
164 | return lines | |
165 |
|
165 | |||
166 | def render_display_data(self, output): |
|
166 | def render_display_data(self, output): | |
167 | """convert display data from the output of a code cell to rst. |
|
167 | """convert display data from the output of a code cell to rst. | |
168 |
|
168 | |||
169 | Returns list. |
|
169 | Returns list. | |
170 | """ |
|
170 | """ | |
171 | lines = [] |
|
171 | lines = [] | |
172 |
|
172 | |||
173 | if 'png' in output: |
|
173 | if 'png' in output: | |
174 | infile = 'nb_figure_%s.png' % self.figures_counter |
|
174 | infile = 'nb_figure_%s.png' % self.figures_counter | |
175 | fullname = os.path.join(self.dirpath, infile) |
|
175 | fullname = os.path.join(self.dirpath, infile) | |
176 | with open(fullname, 'w') as f: |
|
176 | with open(fullname, 'w') as f: | |
177 | f.write(output.png.decode('base64')) |
|
177 | f.write(output.png.decode('base64')) | |
178 |
|
178 | |||
179 | self.figures_counter += 1 |
|
179 | self.figures_counter += 1 | |
180 | lines.append('.. image:: %s' % infile) |
|
180 | lines.append('.. image:: %s' % infile) | |
181 | lines.append('') |
|
181 | lines.append('') | |
182 |
|
182 | |||
183 | return lines |
|
183 | return lines | |
184 |
|
184 | |||
185 | def render_stream(self, output): |
|
185 | def render_stream(self, output): | |
186 | """convert stream part of a code cell to rst |
|
186 | """convert stream part of a code cell to rst | |
187 |
|
187 | |||
188 | Returns list.""" |
|
188 | Returns list.""" | |
189 |
|
189 | |||
190 | lines = [] |
|
190 | lines = [] | |
191 |
|
191 | |||
192 | if 'text' in output: |
|
192 | if 'text' in output: | |
193 | lines.extend(rst_directive('.. parsed-literal::', output.text)) |
|
193 | lines.extend(rst_directive('.. parsed-literal::', output.text)) | |
194 |
|
194 | |||
195 | return lines |
|
195 | return lines | |
196 |
|
196 | |||
197 |
|
197 | |||
198 | def rst2simplehtml(infile): |
|
198 | def rst2simplehtml(infile): | |
199 | """Convert a rst file to simplified html suitable for blogger. |
|
199 | """Convert a rst file to simplified html suitable for blogger. | |
200 |
|
200 | |||
201 | This just runs rst2html with certain parameters to produce really simple |
|
201 | This just runs rst2html with certain parameters to produce really simple | |
202 | html and strips the document header, so the resulting file can be easily |
|
202 | html and strips the document header, so the resulting file can be easily | |
203 | pasted into a blogger edit window. |
|
203 | pasted into a blogger edit window. | |
204 | """ |
|
204 | """ | |
205 |
|
205 | |||
206 | # This is the template for the rst2html call that produces the cleanest, |
|
206 | # This is the template for the rst2html call that produces the cleanest, | |
207 | # simplest html I could find. This should help in making it easier to |
|
207 | # simplest html I could find. This should help in making it easier to | |
208 | # paste into the blogspot html window, though I'm still having problems |
|
208 | # paste into the blogspot html window, though I'm still having problems | |
209 | # with linebreaks there... |
|
209 | # with linebreaks there... | |
210 | cmd_template = ("rst2html --link-stylesheet --no-xml-declaration " |
|
210 | cmd_template = ("rst2html --link-stylesheet --no-xml-declaration " | |
211 | "--no-generator --no-datestamp --no-source-link " |
|
211 | "--no-generator --no-datestamp --no-source-link " | |
212 | "--no-toc-backlinks --no-section-numbering " |
|
212 | "--no-toc-backlinks --no-section-numbering " | |
213 | "--strip-comments ") |
|
213 | "--strip-comments ") | |
214 |
|
214 | |||
215 | cmd = "%s %s" % (cmd_template, infile) |
|
215 | cmd = "%s %s" % (cmd_template, infile) | |
216 | proc = subprocess.Popen(cmd, |
|
216 | proc = subprocess.Popen(cmd, | |
217 | stdout=subprocess.PIPE, |
|
217 | stdout=subprocess.PIPE, | |
218 | stderr=subprocess.PIPE, |
|
218 | stderr=subprocess.PIPE, | |
219 | shell=True) |
|
219 | shell=True) | |
220 | html, stderr = proc.communicate() |
|
220 | html, stderr = proc.communicate() | |
221 | if stderr: |
|
221 | if stderr: | |
222 | raise IOError(stderr) |
|
222 | raise IOError(stderr) | |
223 |
|
223 | |||
224 | # Make an iterator so breaking out holds state. Our implementation of |
|
224 | # Make an iterator so breaking out holds state. Our implementation of | |
225 | # searching for the html body below is basically a trivial little state |
|
225 | # searching for the html body below is basically a trivial little state | |
226 | # machine, so we need this. |
|
226 | # machine, so we need this. | |
227 | walker = iter(html.splitlines()) |
|
227 | walker = iter(html.splitlines()) | |
228 |
|
228 | |||
229 | # Find start of main text, break out to then print until we find end /div. |
|
229 | # Find start of main text, break out to then print until we find end /div. | |
230 | # This may only work if there's a real title defined so we get a 'div class' |
|
230 | # This may only work if there's a real title defined so we get a 'div class' | |
231 | # tag, I haven't really tried. |
|
231 | # tag, I haven't really tried. | |
232 | for line in walker: |
|
232 | for line in walker: | |
233 | if line.startswith('<body>'): |
|
233 | if line.startswith('<body>'): | |
234 | break |
|
234 | break | |
235 |
|
235 | |||
236 | newfname = os.path.splitext(infile)[0] + '.html' |
|
236 | newfname = os.path.splitext(infile)[0] + '.html' | |
237 | with open(newfname, 'w') as f: |
|
237 | with open(newfname, 'w') as f: | |
238 | for line in walker: |
|
238 | for line in walker: | |
239 | if line.startswith('</body>'): |
|
239 | if line.startswith('</body>'): | |
240 | break |
|
240 | break | |
241 | f.write(line) |
|
241 | f.write(line) | |
242 | f.write('\n') |
|
242 | f.write('\n') | |
243 |
|
243 | |||
244 | return newfname |
|
244 | return newfname | |
245 |
|
245 | |||
246 |
|
246 | |||
247 | def main(infile, format='rst'): |
|
247 | def main(infile, format='rst'): | |
248 | """Convert a notebook to html in one step""" |
|
248 | """Convert a notebook to html in one step""" | |
249 | if format == 'rst': |
|
249 | if format == 'rst': | |
250 | converter = ConverterRST(infile) |
|
250 | converter = ConverterRST(infile) | |
251 | converter.render() |
|
251 | converter.render() | |
252 | elif format == 'html': |
|
252 | elif format == 'html': | |
253 | #Currently, conversion to html is a 2 step process, nb->rst->html |
|
253 | #Currently, conversion to html is a 2 step process, nb->rst->html | |
254 | converter = ConverterRST(infile) |
|
254 | converter = ConverterRST(infile) | |
255 | rstfname = converter.render() |
|
255 | rstfname = converter.render() | |
256 | rst2simplehtml(rstfname) |
|
256 | rst2simplehtml(rstfname) | |
257 |
|
257 | |||
258 |
|
258 | |||
259 | if __name__ == '__main__': |
|
259 | if __name__ == '__main__': | |
260 | parser = argparse.ArgumentParser(description='nbconvert: Convert IPython notebooks to other formats') |
|
260 | parser = argparse.ArgumentParser(description='nbconvert: Convert IPython notebooks to other formats') | |
261 |
|
261 | |||
262 | # TODO: consider passing file like object around, rather than filenames |
|
262 | # TODO: consider passing file like object around, rather than filenames | |
263 | # would allow us to process stdin, or even http streams |
|
263 | # would allow us to process stdin, or even http streams | |
264 | #parser.add_argument('infile', nargs='?', type=argparse.FileType('r'), default=sys.stdin) |
|
264 | #parser.add_argument('infile', nargs='?', type=argparse.FileType('r'), default=sys.stdin) | |
265 |
|
265 | |||
266 | #Require a filename as a positional argument |
|
266 | #Require a filename as a positional argument | |
267 | parser.add_argument('infile', nargs=1) |
|
267 | parser.add_argument('infile', nargs=1) | |
268 | parser.add_argument('-f', '--format', default='rst') |
|
268 | parser.add_argument('-f', '--format', default='rst') | |
269 | args = parser.parse_args() |
|
269 | args = parser.parse_args() | |
270 | main(infile=args.infile[0], format=args.format) |
|
270 | main(infile=args.infile[0], format=args.format) |
General Comments 0
You need to be logged in to leave comments.
Login now