Show More
@@ -1,249 +1,251 | |||||
1 | """Code for converting notebooks to and from v3.""" |
|
1 | """Code for converting notebooks to and from v3.""" | |
2 |
|
2 | |||
3 | # Copyright (c) IPython Development Team. |
|
3 | # Copyright (c) IPython Development Team. | |
4 | # Distributed under the terms of the Modified BSD License. |
|
4 | # Distributed under the terms of the Modified BSD License. | |
5 |
|
5 | |||
6 | import json |
|
6 | import json | |
7 | import re |
|
7 | import re | |
8 |
|
8 | |||
9 | from .nbbase import ( |
|
9 | from .nbbase import ( | |
10 | nbformat, nbformat_minor, |
|
10 | nbformat, nbformat_minor, | |
11 | NotebookNode, |
|
11 | NotebookNode, | |
12 | ) |
|
12 | ) | |
13 |
|
13 | |||
14 | from IPython.nbformat import v3 |
|
14 | from IPython.nbformat import v3 | |
15 | from IPython.utils.log import get_logger |
|
15 | from IPython.utils.log import get_logger | |
16 |
|
16 | |||
17 | def _warn_if_invalid(nb, version): |
|
17 | def _warn_if_invalid(nb, version): | |
18 | """Log validation errors, if there are any.""" |
|
18 | """Log validation errors, if there are any.""" | |
19 | from IPython.nbformat import validate, ValidationError |
|
19 | from IPython.nbformat import validate, ValidationError | |
20 | try: |
|
20 | try: | |
21 | validate(nb, version=version) |
|
21 | validate(nb, version=version) | |
22 | except ValidationError as e: |
|
22 | except ValidationError as e: | |
23 | get_logger().error("Notebook JSON is not valid v%i: %s", version, e) |
|
23 | get_logger().error("Notebook JSON is not valid v%i: %s", version, e) | |
24 |
|
24 | |||
25 | def upgrade(nb, from_version=3, from_minor=0): |
|
25 | def upgrade(nb, from_version=3, from_minor=0): | |
26 | """Convert a notebook to v4. |
|
26 | """Convert a notebook to v4. | |
27 |
|
27 | |||
28 | Parameters |
|
28 | Parameters | |
29 | ---------- |
|
29 | ---------- | |
30 | nb : NotebookNode |
|
30 | nb : NotebookNode | |
31 | The Python representation of the notebook to convert. |
|
31 | The Python representation of the notebook to convert. | |
32 | from_version : int |
|
32 | from_version : int | |
33 | The original version of the notebook to convert. |
|
33 | The original version of the notebook to convert. | |
34 | from_minor : int |
|
34 | from_minor : int | |
35 | The original minor version of the notebook to convert (only relevant for v >= 3). |
|
35 | The original minor version of the notebook to convert (only relevant for v >= 3). | |
36 | """ |
|
36 | """ | |
37 | if from_version == 3: |
|
37 | if from_version == 3: | |
38 | # Validate the notebook before conversion |
|
38 | # Validate the notebook before conversion | |
39 | _warn_if_invalid(nb, from_version) |
|
39 | _warn_if_invalid(nb, from_version) | |
40 |
|
40 | |||
41 | # Mark the original nbformat so consumers know it has been converted |
|
41 | # Mark the original nbformat so consumers know it has been converted | |
42 | orig_nbformat = nb.pop('orig_nbformat', None) |
|
42 | orig_nbformat = nb.pop('orig_nbformat', None) | |
43 | nb.metadata.orig_nbformat = orig_nbformat or 3 |
|
43 | nb.metadata.orig_nbformat = orig_nbformat or 3 | |
44 |
|
44 | |||
45 | # Mark the new format |
|
45 | # Mark the new format | |
46 | nb.nbformat = nbformat |
|
46 | nb.nbformat = nbformat | |
47 | nb.nbformat_minor = nbformat_minor |
|
47 | nb.nbformat_minor = nbformat_minor | |
48 |
|
48 | |||
49 | # remove worksheet(s) |
|
49 | # remove worksheet(s) | |
50 | nb['cells'] = cells = [] |
|
50 | nb['cells'] = cells = [] | |
51 | # In the unlikely event of multiple worksheets, |
|
51 | # In the unlikely event of multiple worksheets, | |
52 | # they will be flattened |
|
52 | # they will be flattened | |
53 | for ws in nb.pop('worksheets', []): |
|
53 | for ws in nb.pop('worksheets', []): | |
54 | # upgrade each cell |
|
54 | # upgrade each cell | |
55 | for cell in ws['cells']: |
|
55 | for cell in ws['cells']: | |
56 | cells.append(upgrade_cell(cell)) |
|
56 | cells.append(upgrade_cell(cell)) | |
57 | # upgrade metadata |
|
57 | # upgrade metadata | |
58 | nb.metadata.pop('name', '') |
|
58 | nb.metadata.pop('name', '') | |
59 | # Validate the converted notebook before returning it |
|
59 | # Validate the converted notebook before returning it | |
60 | _warn_if_invalid(nb, nbformat) |
|
60 | _warn_if_invalid(nb, nbformat) | |
61 | return nb |
|
61 | return nb | |
62 | elif from_version == 4: |
|
62 | elif from_version == 4: | |
63 | # nothing to do |
|
63 | # nothing to do | |
64 | if from_minor != nbformat_minor: |
|
64 | if from_minor != nbformat_minor: | |
65 | nb.metadata.orig_nbformat_minor = from_minor |
|
65 | nb.metadata.orig_nbformat_minor = from_minor | |
66 | nb.nbformat_minor = nbformat_minor |
|
66 | nb.nbformat_minor = nbformat_minor | |
67 |
|
67 | |||
68 | return nb |
|
68 | return nb | |
69 | else: |
|
69 | else: | |
70 | raise ValueError('Cannot convert a notebook directly from v%s to v4. ' \ |
|
70 | raise ValueError('Cannot convert a notebook directly from v%s to v4. ' \ | |
71 | 'Try using the IPython.nbformat.convert module.' % from_version) |
|
71 | 'Try using the IPython.nbformat.convert module.' % from_version) | |
72 |
|
72 | |||
73 | def upgrade_cell(cell): |
|
73 | def upgrade_cell(cell): | |
74 | """upgrade a cell from v3 to v4 |
|
74 | """upgrade a cell from v3 to v4 | |
75 |
|
75 | |||
76 | heading cell -> markdown heading |
|
76 | heading cell -> markdown heading | |
77 | code cell: |
|
77 | code cell: | |
78 | - remove language metadata |
|
78 | - remove language metadata | |
79 | - cell.input -> cell.source |
|
79 | - cell.input -> cell.source | |
80 | - cell.prompt_number -> cell.execution_count |
|
80 | - cell.prompt_number -> cell.execution_count | |
81 | - update outputs |
|
81 | - update outputs | |
82 | """ |
|
82 | """ | |
83 | cell.setdefault('metadata', NotebookNode()) |
|
83 | cell.setdefault('metadata', NotebookNode()) | |
84 | if cell.cell_type == 'code': |
|
84 | if cell.cell_type == 'code': | |
85 | cell.pop('language', '') |
|
85 | cell.pop('language', '') | |
86 | if 'collapsed' in cell: |
|
86 | if 'collapsed' in cell: | |
87 | cell.metadata['collapsed'] = cell.pop('collapsed') |
|
87 | cell.metadata['collapsed'] = cell.pop('collapsed') | |
88 | cell.source = cell.pop('input', '') |
|
88 | cell.source = cell.pop('input', '') | |
89 | cell.execution_count = cell.pop('prompt_number', None) |
|
89 | cell.execution_count = cell.pop('prompt_number', None) | |
90 | cell.outputs = upgrade_outputs(cell.outputs) |
|
90 | cell.outputs = upgrade_outputs(cell.outputs) | |
91 | elif cell.cell_type == 'heading': |
|
91 | elif cell.cell_type == 'heading': | |
92 | cell.cell_type = 'markdown' |
|
92 | cell.cell_type = 'markdown' | |
93 | level = cell.pop('level', 1) |
|
93 | level = cell.pop('level', 1) | |
94 | cell.source = '{hashes} {single_line}'.format( |
|
94 | cell.source = '{hashes} {single_line}'.format( | |
95 | hashes='#' * level, |
|
95 | hashes='#' * level, | |
96 | single_line = ' '.join(cell.get('source', '').splitlines()), |
|
96 | single_line = ' '.join(cell.get('source', '').splitlines()), | |
97 | ) |
|
97 | ) | |
98 | elif cell.cell_type == 'html': |
|
98 | elif cell.cell_type == 'html': | |
99 | # Technically, this exists. It will never happen in practice. |
|
99 | # Technically, this exists. It will never happen in practice. | |
100 | cell.cell_type = 'markdown' |
|
100 | cell.cell_type = 'markdown' | |
101 | return cell |
|
101 | return cell | |
102 |
|
102 | |||
103 | def downgrade_cell(cell): |
|
103 | def downgrade_cell(cell): | |
104 | """downgrade a cell from v4 to v3 |
|
104 | """downgrade a cell from v4 to v3 | |
105 |
|
105 | |||
106 | code cell: |
|
106 | code cell: | |
107 | - set cell.language |
|
107 | - set cell.language | |
108 | - cell.input <- cell.source |
|
108 | - cell.input <- cell.source | |
109 | - cell.prompt_number <- cell.execution_count |
|
109 | - cell.prompt_number <- cell.execution_count | |
110 | - update outputs |
|
110 | - update outputs | |
111 | markdown cell: |
|
111 | markdown cell: | |
112 | - single-line heading -> heading cell |
|
112 | - single-line heading -> heading cell | |
113 | """ |
|
113 | """ | |
114 | if cell.cell_type == 'code': |
|
114 | if cell.cell_type == 'code': | |
115 | cell.language = 'python' |
|
115 | cell.language = 'python' | |
116 | cell.input = cell.pop('source', '') |
|
116 | cell.input = cell.pop('source', '') | |
117 | cell.prompt_number = cell.pop('execution_count', None) |
|
117 | cell.prompt_number = cell.pop('execution_count', None) | |
118 | cell.collapsed = cell.metadata.pop('collapsed', False) |
|
118 | cell.collapsed = cell.metadata.pop('collapsed', False) | |
119 | cell.outputs = downgrade_outputs(cell.outputs) |
|
119 | cell.outputs = downgrade_outputs(cell.outputs) | |
120 | elif cell.cell_type == 'markdown': |
|
120 | elif cell.cell_type == 'markdown': | |
121 | source = cell.get('source', '') |
|
121 | source = cell.get('source', '') | |
122 | if '\n' not in source and source.startswith('#'): |
|
122 | if '\n' not in source and source.startswith('#'): | |
123 | prefix, text = re.match(r'(#+)\s*(.*)', source).groups() |
|
123 | prefix, text = re.match(r'(#+)\s*(.*)', source).groups() | |
124 | cell.cell_type = 'heading' |
|
124 | cell.cell_type = 'heading' | |
125 | cell.source = text |
|
125 | cell.source = text | |
126 | cell.level = len(prefix) |
|
126 | cell.level = len(prefix) | |
127 | return cell |
|
127 | return cell | |
128 |
|
128 | |||
129 | _mime_map = { |
|
129 | _mime_map = { | |
130 | "text" : "text/plain", |
|
130 | "text" : "text/plain", | |
131 | "html" : "text/html", |
|
131 | "html" : "text/html", | |
132 | "svg" : "image/svg+xml", |
|
132 | "svg" : "image/svg+xml", | |
133 | "png" : "image/png", |
|
133 | "png" : "image/png", | |
134 | "jpeg" : "image/jpeg", |
|
134 | "jpeg" : "image/jpeg", | |
135 | "latex" : "text/latex", |
|
135 | "latex" : "text/latex", | |
136 | "json" : "application/json", |
|
136 | "json" : "application/json", | |
137 | "javascript" : "application/javascript", |
|
137 | "javascript" : "application/javascript", | |
138 | }; |
|
138 | }; | |
139 |
|
139 | |||
140 | def to_mime_key(d): |
|
140 | def to_mime_key(d): | |
141 | """convert dict with v3 aliases to plain mime-type keys""" |
|
141 | """convert dict with v3 aliases to plain mime-type keys""" | |
142 | for alias, mime in _mime_map.items(): |
|
142 | for alias, mime in _mime_map.items(): | |
143 | if alias in d: |
|
143 | if alias in d: | |
144 | d[mime] = d.pop(alias) |
|
144 | d[mime] = d.pop(alias) | |
145 | return d |
|
145 | return d | |
146 |
|
146 | |||
147 | def from_mime_key(d): |
|
147 | def from_mime_key(d): | |
148 | """convert dict with mime-type keys to v3 aliases""" |
|
148 | """convert dict with mime-type keys to v3 aliases""" | |
149 | for alias, mime in _mime_map.items(): |
|
149 | for alias, mime in _mime_map.items(): | |
150 | if mime in d: |
|
150 | if mime in d: | |
151 | d[alias] = d.pop(mime) |
|
151 | d[alias] = d.pop(mime) | |
152 | return d |
|
152 | return d | |
153 |
|
153 | |||
154 | def upgrade_output(output): |
|
154 | def upgrade_output(output): | |
155 | """upgrade a single code cell output from v3 to v4 |
|
155 | """upgrade a single code cell output from v3 to v4 | |
156 |
|
156 | |||
157 | - pyout -> execute_result |
|
157 | - pyout -> execute_result | |
158 | - pyerr -> error |
|
158 | - pyerr -> error | |
159 | - output.type -> output.data.mime/type |
|
159 | - output.type -> output.data.mime/type | |
160 | - mime-type keys |
|
160 | - mime-type keys | |
161 | - stream.stream -> stream.name |
|
161 | - stream.stream -> stream.name | |
162 | """ |
|
162 | """ | |
163 | if output['output_type'] in {'pyout', 'display_data'}: |
|
163 | if output['output_type'] in {'pyout', 'display_data'}: | |
164 | output.setdefault('metadata', NotebookNode()) |
|
164 | output.setdefault('metadata', NotebookNode()) | |
165 | if output['output_type'] == 'pyout': |
|
165 | if output['output_type'] == 'pyout': | |
166 | output['output_type'] = 'execute_result' |
|
166 | output['output_type'] = 'execute_result' | |
167 | output['execution_count'] = output.pop('prompt_number', None) |
|
167 | output['execution_count'] = output.pop('prompt_number', None) | |
168 |
|
168 | |||
169 | # move output data into data sub-dict |
|
169 | # move output data into data sub-dict | |
170 | data = {} |
|
170 | data = {} | |
171 | for key in list(output): |
|
171 | for key in list(output): | |
172 | if key in {'output_type', 'execution_count', 'metadata'}: |
|
172 | if key in {'output_type', 'execution_count', 'metadata'}: | |
173 | continue |
|
173 | continue | |
174 | data[key] = output.pop(key) |
|
174 | data[key] = output.pop(key) | |
175 | to_mime_key(data) |
|
175 | to_mime_key(data) | |
176 | output['data'] = data |
|
176 | output['data'] = data | |
177 | to_mime_key(output.metadata) |
|
177 | to_mime_key(output.metadata) | |
178 | if 'application/json' in data: |
|
178 | if 'application/json' in data: | |
179 | data['application/json'] = json.loads(data['application/json']) |
|
179 | data['application/json'] = json.loads(data['application/json']) | |
180 | # promote ascii bytes (from v2) to unicode |
|
180 | # promote ascii bytes (from v2) to unicode | |
181 | for key in ('image/png', 'image/jpeg'): |
|
181 | for key in ('image/png', 'image/jpeg'): | |
182 | if key in data and isinstance(data[key], bytes): |
|
182 | if key in data and isinstance(data[key], bytes): | |
183 | data[key] = data[key].decode('ascii') |
|
183 | data[key] = data[key].decode('ascii') | |
184 | elif output['output_type'] == 'pyerr': |
|
184 | elif output['output_type'] == 'pyerr': | |
185 | output['output_type'] = 'error' |
|
185 | output['output_type'] = 'error' | |
186 | elif output['output_type'] == 'stream': |
|
186 | elif output['output_type'] == 'stream': | |
187 | output['name'] = output.pop('stream') |
|
187 | output['name'] = output.pop('stream') | |
188 | return output |
|
188 | return output | |
189 |
|
189 | |||
190 | def downgrade_output(output): |
|
190 | def downgrade_output(output): | |
191 | """downgrade a single code cell output to v3 from v4 |
|
191 | """downgrade a single code cell output to v3 from v4 | |
192 |
|
192 | |||
193 | - pyout <- execute_result |
|
193 | - pyout <- execute_result | |
194 | - pyerr <- error |
|
194 | - pyerr <- error | |
195 | - output.data.mime/type -> output.type |
|
195 | - output.data.mime/type -> output.type | |
196 | - un-mime-type keys |
|
196 | - un-mime-type keys | |
197 | - stream.stream <- stream.name |
|
197 | - stream.stream <- stream.name | |
198 | """ |
|
198 | """ | |
199 | if output['output_type'] in {'execute_result', 'display_data'}: |
|
199 | if output['output_type'] in {'execute_result', 'display_data'}: | |
200 | if output['output_type'] == 'execute_result': |
|
200 | if output['output_type'] == 'execute_result': | |
201 | output['output_type'] = 'pyout' |
|
201 | output['output_type'] = 'pyout' | |
202 | output['prompt_number'] = output.pop('execution_count', None) |
|
202 | output['prompt_number'] = output.pop('execution_count', None) | |
203 |
|
203 | |||
204 | # promote data dict to top-level output namespace |
|
204 | # promote data dict to top-level output namespace | |
205 | data = output.pop('data', {}) |
|
205 | data = output.pop('data', {}) | |
206 | if 'application/json' in data: |
|
206 | if 'application/json' in data: | |
207 | data['application/json'] = json.dumps(data['application/json']) |
|
207 | data['application/json'] = json.dumps(data['application/json']) | |
208 | from_mime_key(data) |
|
208 | from_mime_key(data) | |
209 | output.update(data) |
|
209 | output.update(data) | |
210 | from_mime_key(output.get('metadata', {})) |
|
210 | from_mime_key(output.get('metadata', {})) | |
211 | elif output['output_type'] == 'error': |
|
211 | elif output['output_type'] == 'error': | |
212 | output['output_type'] = 'pyerr' |
|
212 | output['output_type'] = 'pyerr' | |
213 | elif output['output_type'] == 'stream': |
|
213 | elif output['output_type'] == 'stream': | |
214 | output['stream'] = output.pop('name') |
|
214 | output['stream'] = output.pop('name') | |
215 | return output |
|
215 | return output | |
216 |
|
216 | |||
217 | def upgrade_outputs(outputs): |
|
217 | def upgrade_outputs(outputs): | |
218 | """upgrade outputs of a code cell from v3 to v4""" |
|
218 | """upgrade outputs of a code cell from v3 to v4""" | |
219 | return [upgrade_output(op) for op in outputs] |
|
219 | return [upgrade_output(op) for op in outputs] | |
220 |
|
220 | |||
221 | def downgrade_outputs(outputs): |
|
221 | def downgrade_outputs(outputs): | |
222 | """downgrade outputs of a code cell to v3 from v4""" |
|
222 | """downgrade outputs of a code cell to v3 from v4""" | |
223 | return [downgrade_output(op) for op in outputs] |
|
223 | return [downgrade_output(op) for op in outputs] | |
224 |
|
224 | |||
225 | def downgrade(nb): |
|
225 | def downgrade(nb): | |
226 | """Convert a v4 notebook to v3. |
|
226 | """Convert a v4 notebook to v3. | |
227 |
|
227 | |||
228 | Parameters |
|
228 | Parameters | |
229 | ---------- |
|
229 | ---------- | |
230 | nb : NotebookNode |
|
230 | nb : NotebookNode | |
231 | The Python representation of the notebook to convert. |
|
231 | The Python representation of the notebook to convert. | |
232 | """ |
|
232 | """ | |
233 | if nb.nbformat != nbformat: |
|
233 | if nb.nbformat != nbformat: | |
234 | return nb |
|
234 | return nb | |
235 |
|
235 | |||
236 | # Validate the notebook before conversion |
|
236 | # Validate the notebook before conversion | |
237 | _warn_if_invalid(nb, nbformat) |
|
237 | _warn_if_invalid(nb, nbformat) | |
238 |
|
238 | |||
239 | nb.nbformat = v3.nbformat |
|
239 | nb.nbformat = v3.nbformat | |
240 | nb.nbformat_minor = v3.nbformat_minor |
|
240 | nb.nbformat_minor = v3.nbformat_minor | |
241 | cells = [ downgrade_cell(cell) for cell in nb.pop('cells') ] |
|
241 | cells = [ downgrade_cell(cell) for cell in nb.pop('cells') ] | |
242 | nb.worksheets = [v3.new_worksheet(cells=cells)] |
|
242 | nb.worksheets = [v3.new_worksheet(cells=cells)] | |
243 | nb.metadata.setdefault('name', '') |
|
243 | nb.metadata.setdefault('name', '') | |
244 | nb.metadata.pop('orig_nbformat', None) |
|
|||
245 | nb.metadata.pop('orig_nbformat_minor', None) |
|
|||
246 |
|
244 | |||
247 | # Validate the converted notebook before returning it |
|
245 | # Validate the converted notebook before returning it | |
248 | _warn_if_invalid(nb, v3.nbformat) |
|
246 | _warn_if_invalid(nb, v3.nbformat) | |
|
247 | ||||
|
248 | nb.orig_nbformat = nb.metadata.pop('orig_nbformat', nbformat) | |||
|
249 | nb.orig_nbformat_minor = nb.metadata.pop('orig_nbformat_minor', nbformat_minor) | |||
|
250 | ||||
249 | return nb |
|
251 | return nb |
General Comments 0
You need to be logged in to leave comments.
Login now