|
1 #!/usr/bin/env python |
|
2 # |
|
3 # Copyright 2007 Google Inc. |
|
4 # |
|
5 # Licensed under the Apache License, Version 2.0 (the "License"); |
|
6 # you may not use this file except in compliance with the License. |
|
7 # You may obtain a copy of the License at |
|
8 # |
|
9 # http://www.apache.org/licenses/LICENSE-2.0 |
|
10 # |
|
11 # Unless required by applicable law or agreed to in writing, software |
|
12 # distributed under the License is distributed on an "AS IS" BASIS, |
|
13 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
|
14 # See the License for the specific language governing permissions and |
|
15 # limitations under the License. |
|
16 # |
|
17 |
|
18 """A mix-in handler for bulk loading data into an application. |
|
19 |
|
20 For complete documentation, see the Tools and Libraries section of the |
|
21 documentation. |
|
22 |
|
23 To use this in your app, first write a script, e.g. bulkload.py, that |
|
24 instantiates a Loader for each entity kind you want to import and call |
|
25 bulkload.main(instance). For example: |
|
26 |
|
27 person = bulkload.Loader( |
|
28 'Person', |
|
29 [('name', str), |
|
30 ('email', datastore_types.Email), |
|
31 ('birthdate', lambda x: datetime.datetime.fromtimestamp(float(x))), |
|
32 ]) |
|
33 |
|
34 if __name__ == '__main__': |
|
35 bulkload.main(person) |
|
36 |
|
37 See the Loader class for more information. Then, add a handler for it in your |
|
38 app.yaml, e.g.: |
|
39 |
|
40 urlmap: |
|
41 - regex: /load |
|
42 handler: |
|
43 type: 1 |
|
44 path: bulkload.py |
|
45 requires_login: true |
|
46 admin_only: true |
|
47 |
|
48 Finally, deploy your app and run bulkload_client.py. For example, to load the |
|
49 file people.csv into a dev_appserver running on your local machine: |
|
50 |
|
51 ./bulkload_client.py --filename people.csv --kind Person --cookie ... \ |
|
52 --url http://localhost:8080/load |
|
53 |
|
54 The kind parameter is used to look up the Loader instance that will be used. |
|
55 The bulkload handler should usually be admin_only, so that non-admins can't use |
|
56 the shell to modify your app's data. The bulkload client uses the cookie |
|
57 parameter to piggyback its HTTP requests on your login session. A GET request |
|
58 to the URL specified for your bulkload script will give you a cookie parameter |
|
59 you can use (/load in the example above). If your bulkload handler is not |
|
60 admin_only, you may omit the cookie parameter. |
|
61 |
|
62 If you want to do extra processing before the entities are stored, you can |
|
63 subclass Loader and override HandleEntity. HandleEntity is called once with |
|
64 each entity that is imported from the CSV data. You can return one or more |
|
65 entities from HandleEntity to be stored in its place, or None if nothing |
|
66 should be stored. |
|
67 |
|
68 For example, this loads calendar events and stores them as |
|
69 datastore_entities.Event entities. It also populates their author field with a |
|
70 reference to the corresponding datastore_entites.Contact entity. If no Contact |
|
71 entity exists yet for the given author, it creates one and stores it first. |
|
72 |
|
73 class EventLoader(bulkload.Loader): |
|
74 def __init__(self): |
|
75 EventLoader.__init__(self, 'Event', |
|
76 [('title', str), |
|
77 ('creator', str), |
|
78 ('where', str), |
|
79 ('startTime', lambda x: |
|
80 datetime.datetime.fromtimestamp(float(x))), |
|
81 ]) |
|
82 |
|
83 def HandleEntity(self, entity): |
|
84 event = datastore_entities.Event(entity.title) |
|
85 event.update(entity) |
|
86 |
|
87 creator = event['creator'] |
|
88 if creator: |
|
89 contact = datastore.Query('Contact', {'title': creator}).Get(1) |
|
90 if not contact: |
|
91 contact = [datastore_entities.Contact(creator)] |
|
92 datastore.Put(contact[0]) |
|
93 event['author'] = contact[0].key() |
|
94 |
|
95 return event |
|
96 |
|
97 if __name__ == '__main__': |
|
98 bulkload.main(EventLoader()) |
|
99 """ |
|
100 |
|
101 |
|
102 |
|
103 |
|
104 |
|
105 import Cookie |
|
106 import StringIO |
|
107 import csv |
|
108 import httplib |
|
109 import os |
|
110 import sys |
|
111 import traceback |
|
112 import types |
|
113 |
|
114 |
|
115 import google |
|
116 import wsgiref.handlers |
|
117 |
|
118 from google.appengine.api import datastore |
|
119 from google.appengine.api import datastore_types |
|
120 from google.appengine.ext import webapp |
|
121 from google.appengine.ext.bulkload import constants |
|
122 |
|
123 |
|
124 def Validate(value, type): |
|
125 """ Checks that value is non-empty and of the right type. |
|
126 |
|
127 Raises ValueError if value is None or empty, TypeError if it's not the given |
|
128 type. |
|
129 |
|
130 Args: |
|
131 value: any value |
|
132 type: a type or tuple of types |
|
133 """ |
|
134 if not value: |
|
135 raise ValueError('Value should not be empty; received %s.' % value) |
|
136 elif not isinstance(value, type): |
|
137 raise TypeError('Expected a %s, but received %s (a %s).' % |
|
138 (type, value, value.__class__)) |
|
139 |
|
140 |
|
141 class Loader(object): |
|
142 """ A base class for creating datastore entities from CSV input data. |
|
143 |
|
144 To add a handler for bulk loading a new entity kind into your datastore, |
|
145 write a subclass of this class that calls Loader.__init__ from your |
|
146 class's __init__. |
|
147 |
|
148 If you need to run extra code to convert entities from CSV, create new |
|
149 properties, or otherwise modify the entities before they're inserted, |
|
150 override HandleEntity. |
|
151 """ |
|
152 |
|
153 __loaders = {} |
|
154 __kind = None |
|
155 __properties = None |
|
156 |
|
157 def __init__(self, kind, properties): |
|
158 """ Constructor. |
|
159 |
|
160 Populates this Loader's kind and properties map. Also registers it with |
|
161 the bulk loader, so that all you need to do is instantiate your Loader, |
|
162 and the bulkload handler will automatically use it. |
|
163 |
|
164 Args: |
|
165 kind: a string containing the entity kind that this loader handles |
|
166 |
|
167 properties: list of (name, converter) tuples. |
|
168 |
|
169 This is used to automatically convert the CSV columns into properties. |
|
170 The converter should be a function that takes one argument, a string |
|
171 value from the CSV file, and returns a correctly typed property value |
|
172 that should be inserted. The tuples in this list should match the |
|
173 columns in your CSV file, in order. |
|
174 |
|
175 For example: |
|
176 [('name', str), |
|
177 ('id_number', int), |
|
178 ('email', datastore_types.Email), |
|
179 ('user', users.User), |
|
180 ('birthdate', lambda x: datetime.datetime.fromtimestamp(float(x))), |
|
181 ('description', datastore_types.Text), |
|
182 ] |
|
183 """ |
|
184 Validate(kind, basestring) |
|
185 self.__kind = kind |
|
186 |
|
187 Validate(properties, list) |
|
188 for name, fn in properties: |
|
189 Validate(name, basestring) |
|
190 assert callable(fn), ( |
|
191 'Conversion function %s for property %s is not callable.' % (fn, name)) |
|
192 |
|
193 self.__properties = properties |
|
194 |
|
195 Loader.__loaders[kind] = self |
|
196 |
|
197 |
|
198 def kind(self): |
|
199 """ Return the entity kind that this Loader handes. |
|
200 """ |
|
201 return self.__kind |
|
202 |
|
203 |
|
204 def CreateEntity(self, values): |
|
205 """ Creates an entity from a list of property values. |
|
206 |
|
207 Args: |
|
208 values: list of str |
|
209 |
|
210 Returns: |
|
211 list of datastore.Entity |
|
212 |
|
213 The returned entities are populated with the property values from the |
|
214 argument, converted to native types using the properties map given in |
|
215 the constructor, and passed through HandleEntity. They're ready to be |
|
216 inserted. |
|
217 |
|
218 Raises an AssertionError if the number of values doesn't match the number |
|
219 of properties in the properties map. |
|
220 """ |
|
221 Validate(values, list) |
|
222 assert len(values) == len(self.__properties), ( |
|
223 'Expected %d CSV columns, found %d.' % |
|
224 (len(self.__properties), len(values))) |
|
225 |
|
226 entity = datastore.Entity(self.__kind) |
|
227 for (name, converter), val in zip(self.__properties, values): |
|
228 entity[name] = converter(val) |
|
229 |
|
230 entities = self.HandleEntity(entity) |
|
231 |
|
232 if entities is not None: |
|
233 if not isinstance(entities, list): |
|
234 entities = [entities] |
|
235 |
|
236 for entity in entities: |
|
237 if not isinstance(entity, datastore.Entity): |
|
238 raise TypeError('Expected a datastore.Entity, received %s (a %s).' % |
|
239 (entity, entity.__class__)) |
|
240 |
|
241 return entities |
|
242 |
|
243 |
|
244 def HandleEntity(self, entity): |
|
245 """ Subclasses can override this to add custom entity conversion code. |
|
246 |
|
247 This is called for each entity, after its properties are populated from |
|
248 CSV but before it is stored. Subclasses can override this to add custom |
|
249 entity handling code. |
|
250 |
|
251 The entity to be inserted should be returned. If multiple entities should |
|
252 be inserted, return a list of entities. If no entities should be inserted, |
|
253 return None or []. |
|
254 |
|
255 Args: |
|
256 entity: datastore.Entity |
|
257 |
|
258 Returns: |
|
259 datastore.Entity or list of datastore.Entity |
|
260 """ |
|
261 return entity |
|
262 |
|
263 |
|
264 @staticmethod |
|
265 def RegisteredLoaders(): |
|
266 """ Returns a list of the Loader instances that have been created. |
|
267 """ |
|
268 return dict(Loader.__loaders) |
|
269 |
|
270 |
|
271 class BulkLoad(webapp.RequestHandler): |
|
272 """ A handler for bulk load requests. |
|
273 """ |
|
274 |
|
275 def get(self): |
|
276 """ Handle a GET. Just show an info page. |
|
277 """ |
|
278 page = self.InfoPage(self.request.uri) |
|
279 self.response.out.write(page) |
|
280 |
|
281 |
|
282 def post(self): |
|
283 """ Handle a POST. Reads CSV data, converts to entities, and stores them. |
|
284 """ |
|
285 self.response.headers['Content-Type'] = 'text/plain' |
|
286 response, output = self.Load(self.request.get(constants.KIND_PARAM), |
|
287 self.request.get(constants.CSV_PARAM)) |
|
288 self.response.set_status(response) |
|
289 self.response.out.write(output) |
|
290 |
|
291 |
|
292 def InfoPage(self, uri): |
|
293 """ Renders an information page with the POST endpoint and cookie flag. |
|
294 |
|
295 Args: |
|
296 uri: a string containing the request URI |
|
297 Returns: |
|
298 A string with the contents of the info page to be displayed |
|
299 """ |
|
300 page = """ |
|
301 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" |
|
302 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> |
|
303 <html><head> |
|
304 <title>Bulk Loader</title> |
|
305 </head><body>""" |
|
306 |
|
307 page += ('The bulk load endpoint is: <a href="%s">%s</a><br />\n' % |
|
308 (uri, uri)) |
|
309 |
|
310 cookies = os.environ.get('HTTP_COOKIE', None) |
|
311 if cookies: |
|
312 cookie = Cookie.BaseCookie(cookies) |
|
313 for param in ['ACSID', 'dev_appserver_login']: |
|
314 value = cookie.get(param) |
|
315 if value: |
|
316 page += ("Pass this flag to the client: --cookie='%s=%s'\n" % |
|
317 (param, value.value)) |
|
318 break |
|
319 |
|
320 else: |
|
321 page += 'No cookie found!\n' |
|
322 |
|
323 page += '</body></html>' |
|
324 return page |
|
325 |
|
326 |
|
327 def Load(self, kind, data): |
|
328 """ Parses CSV data, uses a Loader to convert to entities, and stores them. |
|
329 |
|
330 On error, fails fast. Returns a "bad request" HTTP response code and |
|
331 includes the traceback in the output. |
|
332 |
|
333 Args: |
|
334 kind: a string containing the entity kind that this loader handles |
|
335 data: a string containing the CSV data to load |
|
336 |
|
337 Returns: |
|
338 tuple (response code, output) where: |
|
339 response code: integer HTTP response code to return |
|
340 output: string containing the HTTP response body |
|
341 """ |
|
342 Validate(kind, basestring) |
|
343 Validate(data, basestring) |
|
344 output = [] |
|
345 |
|
346 try: |
|
347 loader = Loader.RegisteredLoaders()[kind] |
|
348 except KeyError: |
|
349 output.append('Error: no Loader defined for kind %s.' % kind) |
|
350 return (httplib.BAD_REQUEST, ''.join(output)) |
|
351 |
|
352 buffer = StringIO.StringIO(data) |
|
353 reader = csv.reader(buffer, skipinitialspace=True) |
|
354 |
|
355 try: |
|
356 csv.field_size_limit(800000) |
|
357 except AttributeError: |
|
358 pass |
|
359 |
|
360 entities = [] |
|
361 |
|
362 line_num = 1 |
|
363 for columns in reader: |
|
364 if columns: |
|
365 try: |
|
366 output.append('\nLoading from line %d...' % line_num) |
|
367 new_entities = loader.CreateEntity(columns) |
|
368 if new_entities: |
|
369 entities.extend(new_entities) |
|
370 output.append('done.') |
|
371 except: |
|
372 exc_info = sys.exc_info() |
|
373 stacktrace = traceback.format_exception(*exc_info) |
|
374 output.append('error:\n%s' % stacktrace) |
|
375 return (httplib.BAD_REQUEST, ''.join(output)) |
|
376 |
|
377 line_num += 1 |
|
378 |
|
379 for entity in entities: |
|
380 datastore.Put(entity) |
|
381 |
|
382 return (httplib.OK, ''.join(output)) |
|
383 |
|
384 |
|
385 def main(*loaders): |
|
386 """Starts bulk upload. |
|
387 |
|
388 Raises TypeError if not, at least one Loader instance is given. |
|
389 |
|
390 Args: |
|
391 loaders: One or more Loader instance. |
|
392 """ |
|
393 if not loaders: |
|
394 raise TypeError('Expected at least one argument.') |
|
395 |
|
396 for loader in loaders: |
|
397 if not isinstance(loader, Loader): |
|
398 raise TypeError('Expected a Loader instance; received %r' % loader) |
|
399 |
|
400 application = webapp.WSGIApplication([('.*', BulkLoad)]) |
|
401 wsgiref.handlers.CGIHandler().run(application) |
|
402 |
|
403 if __name__ == '__main__': |
|
404 main() |