1 | import re |
---|
2 | from xml.dom import minidom |
---|
3 | from types import ListType |
---|
4 | from types import TupleType |
---|
5 | |
---|
6 | from zope.interface import implements, classProvides |
---|
7 | from zope.app.annotation.interfaces import IAnnotations |
---|
8 | |
---|
9 | from Products.CMFPlone.Portal import PloneSite |
---|
10 | from Products.CMFCore import utils |
---|
11 | |
---|
12 | from collective.transmogrifier.interfaces import ISection, ISectionBlueprint |
---|
13 | from collective.transmogrifier.utils import defaultMatcher |
---|
14 | |
---|
15 | from quintagroup.transmogrifier.interfaces import IExportDataCorrector, IImportDataCorrector |
---|
16 | from quintagroup.transmogrifier.adapters.exporting import ReferenceExporter |
---|
17 | from quintagroup.transmogrifier.manifest import ManifestExporterSection |
---|
18 | from quintagroup.transmogrifier.logger import VALIDATIONKEY |
---|
19 | |
---|
20 | from quintagroup.transmogrifier.simpleblog2quills.interfaces import IExportItemManipulator, IBlog |
---|
21 | |
---|
22 | # URL of the site, where blog is located (this is needed to fix links in entries) |
---|
23 | SITE_URLS = [] |
---|
24 | IMAGE_FOLDER = 'images' |
---|
25 | # this registries are needed to avoid loosing images with equal ids |
---|
26 | IMAGE_IDS = [] |
---|
27 | IMAGE_PATHS = {} |
---|
28 | |
---|
29 | class BlogManifest(object): |
---|
30 | implements(IExportDataCorrector) |
---|
31 | |
---|
32 | def __init__(self, context): |
---|
33 | self.context = context |
---|
34 | |
---|
35 | def __call__(self, data): |
---|
36 | doc = minidom.parseString(data['data']) |
---|
37 | root = doc.documentElement |
---|
38 | for child in root.getElementsByTagName('record'): |
---|
39 | if child.getAttribute('type') not in ('BlogEntry', 'BlogFolder'): |
---|
40 | root.removeChild(child) |
---|
41 | folder = doc.createElement('record') |
---|
42 | folder.setAttribute('type', 'Large Plone Folder') |
---|
43 | folder.appendChild(doc.createTextNode(IMAGE_FOLDER)) |
---|
44 | root.appendChild(folder) |
---|
45 | data['data'] = doc.toxml('utf-8') |
---|
46 | return data |
---|
47 | |
---|
48 | class BlogFolderManifest(object): |
---|
49 | implements(IExportDataCorrector) |
---|
50 | |
---|
51 | def __init__(self, context): |
---|
52 | self.context = context |
---|
53 | |
---|
54 | def __call__(self, data): |
---|
55 | doc = minidom.parseString(data['data']) |
---|
56 | root = doc.documentElement |
---|
57 | for child in root.getElementsByTagName('record'): |
---|
58 | if child.getAttribute('type') not in ('BlogEntry', 'BlogFolder'): |
---|
59 | root.removeChild(child) |
---|
60 | data['data'] = doc.toxml('utf-8') |
---|
61 | return data |
---|
62 | |
---|
63 | class BlogEntryManifest(object): |
---|
64 | implements(IExportItemManipulator) |
---|
65 | |
---|
66 | def __init__(self, context): |
---|
67 | self.context = context |
---|
68 | |
---|
69 | def __call__(self, item, **kw): |
---|
70 | # remove manifest data from item - content contained in BlogEntry isn't exported |
---|
71 | if '_files' in item and 'manifest' in item['_files']: |
---|
72 | del item['_files']['manifest'] |
---|
73 | return item |
---|
74 | |
---|
75 | def recurseToInterface(item, ifaces): |
---|
76 | """Recurse up the aq_chain until an object providing `iface' is found, |
---|
77 | and return that. |
---|
78 | """ |
---|
79 | if not isinstance(ifaces, (ListType, TupleType)): |
---|
80 | ifaces = [ifaces] |
---|
81 | parent = item.aq_parent |
---|
82 | for iface in ifaces: |
---|
83 | if iface.providedBy(item): |
---|
84 | return item |
---|
85 | for iface in ifaces: |
---|
86 | if iface.providedBy(parent): |
---|
87 | return parent |
---|
88 | if isinstance(parent, PloneSite): |
---|
89 | # Stop when we get to the portal root. |
---|
90 | return None |
---|
91 | return recurseToInterface(parent, ifaces) |
---|
92 | |
---|
93 | def getUniqueId(image_id): |
---|
94 | """ Generate id that is unique in IMAGE_IDS registry. |
---|
95 | """ |
---|
96 | if '.' in image_id: |
---|
97 | name, ext = image_id.rsplit('.', 1) |
---|
98 | ext = '.' + ext |
---|
99 | else: |
---|
100 | name, ext = image_id, '' |
---|
101 | if image_id in IMAGE_IDS: |
---|
102 | c = 1 |
---|
103 | new_id = name + str(c) + ext |
---|
104 | while new_id in IMAGE_IDS: |
---|
105 | c += 1 |
---|
106 | new_id = name + str(c) + ext |
---|
107 | image_id = new_id |
---|
108 | |
---|
109 | return image_id |
---|
110 | |
---|
111 | class BlogEntryExporter(ReferenceExporter): |
---|
112 | implements(IExportDataCorrector) |
---|
113 | |
---|
114 | SRC = re.compile(r'src="([^"]+)"') |
---|
115 | |
---|
116 | def __init__(self, context): |
---|
117 | self.context = context |
---|
118 | self.portal_url = utils.getToolByName(self.context, 'portal_url') |
---|
119 | self.portal = self.portal_url.getPortalObject() |
---|
120 | |
---|
121 | def __call__(self, data): |
---|
122 | data = super(BlogEntryExporter, self).__call__(data) |
---|
123 | doc = minidom.parseString(data['data']) |
---|
124 | try: |
---|
125 | elem = [i for i in doc.getElementsByTagName('field') if i.getAttribute('name') == 'body'][0] |
---|
126 | except IndexError: |
---|
127 | return data |
---|
128 | |
---|
129 | text = elem.firstChild.nodeValue |
---|
130 | urls = self.SRC.findall(text) |
---|
131 | blog = recurseToInterface(self.context, IBlog) |
---|
132 | blog_url = blog.absolute_url() |
---|
133 | blog_path = blog.getPhysicalPath() |
---|
134 | for url in urls: |
---|
135 | url = str(url) |
---|
136 | image_id = url.rsplit('/', 1)[-1] |
---|
137 | # bad link |
---|
138 | if '://' in url and not url.startswith('http://'): |
---|
139 | continue |
---|
140 | if url.startswith('http://'): |
---|
141 | for site in SITE_URLS: |
---|
142 | if url.startswith(site): |
---|
143 | # check whether image is stored in blog |
---|
144 | relative_url = url[len(site):] |
---|
145 | relative_url = relative_url.strip('/') |
---|
146 | # if link is broken we'll get an AttributeError |
---|
147 | try: |
---|
148 | image = self.portal.unrestrictedTraverse(relative_url) |
---|
149 | except AttributeError: |
---|
150 | break |
---|
151 | in_blog = recurseToInterface(image, IBlog) is not None and True or False |
---|
152 | if in_blog: |
---|
153 | image_id = self.fixImageId(image, image_id, blog_path) |
---|
154 | new_url = '/'.join((blog_url, IMAGE_FOLDER, image_id)) |
---|
155 | text = text.replace(url, new_url, 1) |
---|
156 | break |
---|
157 | else: |
---|
158 | if url.startswith('/'): |
---|
159 | # if link is broken we'll get an AttributeError |
---|
160 | try: |
---|
161 | image = self.portal.unrestrictedTraverse(url.strip('/')) |
---|
162 | except AttributeError: |
---|
163 | continue |
---|
164 | else: |
---|
165 | # if link is broken we'll get an AttributeError |
---|
166 | try: |
---|
167 | image = self.context.unrestrictedTraverse(url) |
---|
168 | except AttributeError: |
---|
169 | continue |
---|
170 | in_blog = recurseToInterface(image, IBlog) is not None and True or False |
---|
171 | if in_blog: |
---|
172 | image_id = self.fixImageId(image, image_id, blog_path) |
---|
173 | path = self.context.getPhysicalPath() |
---|
174 | level = len(path) - len(blog_path) - 1 |
---|
175 | new_url = '/'.join(['..' for i in range(level)]) |
---|
176 | new_url = '/'.join([new_url, IMAGE_FOLDER, image_id]) |
---|
177 | text = text.replace(url, new_url, 1) |
---|
178 | elif url.startswith('../'): |
---|
179 | # remove '../' from the start of sting |
---|
180 | new_url = url[3:] |
---|
181 | text = text.replace(url, new_url, 1) |
---|
182 | elif url.startswith('/'): |
---|
183 | # these links didn't work so rewrite them with '..' |
---|
184 | # find how many level self.context is under portal root |
---|
185 | level = len(self.context.getPhysicalPath()) - 3 |
---|
186 | new_url = '/'.join(['..' for i in range(level)]) |
---|
187 | new_url = new_url + url |
---|
188 | text = text.replace(url, new_url, 1) |
---|
189 | |
---|
190 | elem.firstChild.nodeValue = text |
---|
191 | data['data'] = doc.toxml('utf-8') |
---|
192 | return data |
---|
193 | |
---|
194 | def fixImageId(self, image, image_id, blog_path): |
---|
195 | """ Check whether image is good or generate new if it's bad. |
---|
196 | """ |
---|
197 | image_path = '/'.join(image.getPhysicalPath()) |
---|
198 | if image_id in IMAGE_IDS and image_path not in IMAGE_PATHS: |
---|
199 | image_id = getUniqueId(image_id) |
---|
200 | if image_id not in IMAGE_IDS: |
---|
201 | IMAGE_IDS.append(image_id) |
---|
202 | IMAGE_PATHS[image_path] = '/'.join(blog_path[2:] + (IMAGE_FOLDER, image_id)) |
---|
203 | |
---|
204 | return image_id |
---|
205 | |
---|
206 | class PathRewriter(object): |
---|
207 | implements(IExportItemManipulator) |
---|
208 | |
---|
209 | def __init__(self, context): |
---|
210 | self.context = context |
---|
211 | |
---|
212 | def __call__(self, item, **kw): |
---|
213 | pathkey = kw.get('path') |
---|
214 | if pathkey is None: |
---|
215 | return item |
---|
216 | |
---|
217 | path = item[pathkey] |
---|
218 | blog = recurseToInterface(self.context, IBlog) |
---|
219 | if blog is None: |
---|
220 | return item |
---|
221 | |
---|
222 | blog_path = blog.getPhysicalPath() |
---|
223 | full_path = '/'.join(self.context.getPhysicalPath()) |
---|
224 | image_id = path.rsplit('/', 1)[-1] |
---|
225 | modified = False |
---|
226 | |
---|
227 | if full_path in IMAGE_PATHS: |
---|
228 | new_path = IMAGE_PATHS[full_path] |
---|
229 | else: |
---|
230 | unique_id = getUniqueId(image_id) |
---|
231 | modified = image_id != unique_id |
---|
232 | new_path = '/'.join(blog_path[2:] + (IMAGE_FOLDER, unique_id)) |
---|
233 | |
---|
234 | IMAGE_IDS.append(image_id) |
---|
235 | IMAGE_PATHS[full_path] = new_path |
---|
236 | |
---|
237 | # change item's path |
---|
238 | item[pathkey] = new_path |
---|
239 | item['_oldpath'] = path |
---|
240 | |
---|
241 | # now we need to fix object id in .marshall.xml |
---|
242 | if modified: |
---|
243 | if '_files' in item and 'marshall' in item['_files']: |
---|
244 | doc = minidom.parseString(item['_files']['marshall']['data']) |
---|
245 | elem = [i for i in doc.getElementsByTagName('field') if i.getAttribute('name') == 'id'][0] |
---|
246 | elem.firstChild.nodeValue = '\n\t\t%s\n\t' % unique_id |
---|
247 | item['_files']['marshall']['data'] = doc.toxml('utf-8') |
---|
248 | |
---|
249 | return item |
---|
250 | |
---|
251 | class ImageFolderSection(object): |
---|
252 | """ This section will generate manifest files for image folders in blog. |
---|
253 | """ |
---|
254 | classProvides(ISectionBlueprint) |
---|
255 | implements(ISection) |
---|
256 | |
---|
257 | def __init__(self, transmogrifier, name, options, previous): |
---|
258 | self.previous = previous |
---|
259 | self.transmogrifier = transmogrifier |
---|
260 | |
---|
261 | self.flagkey = defaultMatcher(options, 'old-path-key', name, 'oldpath') |
---|
262 | self.typekey = defaultMatcher(options, 'type-key', name, 'type') |
---|
263 | self.pathkey = defaultMatcher(options, 'path-key', name, 'path') |
---|
264 | |
---|
265 | |
---|
266 | site_urls = options.get('site-urls', '') |
---|
267 | site_urls = filter(None, [i.strip() for i in site_urls.splitlines()]) |
---|
268 | for i in site_urls: |
---|
269 | SITE_URLS.append(i) |
---|
270 | |
---|
271 | self.anno = IAnnotations(transmogrifier) |
---|
272 | |
---|
273 | def __iter__(self): |
---|
274 | folders = {} |
---|
275 | |
---|
276 | # safely get logging storage |
---|
277 | if VALIDATIONKEY in self.anno: |
---|
278 | log_storage = self.anno[VALIDATIONKEY] |
---|
279 | else: |
---|
280 | log_storage = None |
---|
281 | |
---|
282 | for item in self.previous: |
---|
283 | item_keys = item.keys() |
---|
284 | pathkey = self.pathkey(*item_keys)[0] |
---|
285 | typekey = self.typekey(*item_keys)[0] |
---|
286 | oldpathkey = self.flagkey(*item_keys)[0] |
---|
287 | |
---|
288 | # collect data about images moved to folders |
---|
289 | if pathkey and typekey and oldpathkey: |
---|
290 | path = item[pathkey] |
---|
291 | old_path = item[oldpathkey] |
---|
292 | type_ = item[typekey] |
---|
293 | folder_path, image_id = path.rsplit('/', 1) |
---|
294 | folders.setdefault(folder_path, []).append((image_id, type_)) |
---|
295 | |
---|
296 | # update logging data (path) for this item |
---|
297 | if log_storage and log_storage[-1] == old_path: |
---|
298 | log_storage.pop() |
---|
299 | log_storage.append(path) |
---|
300 | |
---|
301 | yield item |
---|
302 | |
---|
303 | # generate manifests for those image folders |
---|
304 | items = [] |
---|
305 | for folder, entries in folders.items(): |
---|
306 | items.append({'_entries': entries, pathkey: folder}) |
---|
307 | exporter = ManifestExporterSection(self.transmogrifier, 'manifest', {'blueprint': 'manifest'}, iter(items)) |
---|
308 | for item in exporter: |
---|
309 | yield item |
---|
310 | |
---|
311 | # clean registries |
---|
312 | while IMAGE_IDS: IMAGE_IDS.pop() |
---|
313 | while SITE_URLS: SITE_URLS.pop() |
---|
314 | IMAGE_PATHS.clear() |
---|
315 | |
---|
316 | class WorkflowImporter(object): |
---|
317 | """ This adapter tries to convert all possible workflow histories to |
---|
318 | simple_publication_workflow history. |
---|
319 | """ |
---|
320 | implements(IImportDataCorrector) |
---|
321 | |
---|
322 | def __init__(self, context): |
---|
323 | self.context = context |
---|
324 | |
---|
325 | def __call__(self, data): |
---|
326 | doc = minidom.parseString(data['data']) |
---|
327 | wh = [i for i in doc.getElementsByTagName('cmf:workflow')] |
---|
328 | if not wh: |
---|
329 | # we don't have such workflow history |
---|
330 | return data |
---|
331 | |
---|
332 | wh = wh[0] |
---|
333 | workflow_id = wh.getAttribute('id') |
---|
334 | if workflow_id == 'simple_publication_workflow': |
---|
335 | return data |
---|
336 | wh.setAttribute('id', 'simple_publication_workflow') |
---|
337 | if workflow_id == 'simpleblog_workflow': |
---|
338 | self.fixSimpleBlogWorkflow(wh) |
---|
339 | else: |
---|
340 | self.fixWorkflow(wh) |
---|
341 | |
---|
342 | data['data'] = doc.toxml('utf-8') |
---|
343 | return data |
---|
344 | |
---|
345 | def fixSimpleBlogWorkflow(self, wh): |
---|
346 | for history in wh.getElementsByTagName('cmf:history'): |
---|
347 | for var in history.getElementsByTagName('cmf:var'): |
---|
348 | id_ = var.getAttribute('id') |
---|
349 | value = var.getAttribute('value') |
---|
350 | if id_ == 'review_state' and value == 'draft': |
---|
351 | var.setAttribute('value', 'private') |
---|
352 | |
---|
353 | def fixWorkflow(self, wh): |
---|
354 | for history in wh.getElementsByTagName('cmf:history'): |
---|
355 | for var in history.getElementsByTagName('cmf:var'): |
---|
356 | id_ = var.getAttribute('id') |
---|
357 | value = var.getAttribute('value') |
---|
358 | if id_ == 'review_state' and value == 'visible': |
---|
359 | var.setAttribute('value', 'published') |
---|