Podwyższenie wymaganej wersji biblioteki librarian do 1.2.5 (wydanej dzisiaj).

[redakcja.git] / apps / api / handlers / library_handlers.py

diff --git a/apps/api/handlers/library_handlers.py b/apps/api/handlers/library_handlers.py
index b30f2e7..4435fe9 100644 (file)
--- a/apps/api/handlers/library_handlers.py
+++ b/apps/api/handlers/library_handlers.py
@@ -1,4 +1,6 @@
 # -*- encoding: utf-8 -*-
 # -*- encoding: utf-8 -*-

+import os.path
+import logging

 
 __author__= "Łukasz Rekucki"
 __date__ = "$2009-09-25 15:49:50$"
 
 __author__= "Łukasz Rekucki"
 __date__ = "$2009-09-25 15:49:50$"


@@ -11,20 +13,28 @@ from datetime import date
 
 from django.core.urlresolvers import reverse
 from django.utils import simplejson as json
 
 from django.core.urlresolvers import reverse
 from django.utils import simplejson as json

+from django.db import IntegrityError

 
 import librarian
 import librarian.html
 
 import librarian
 import librarian.html

-from librarian import dcparser
+from librarian import dcparser, parser

 
 

-from wlrepo import RevisionNotFound, LibraryException, DocumentAlreadyExists
-from explorer.models import PullRequest
+from wlrepo import *
+from explorer.models import PullRequest, GalleryForDocument

 
 # internal imports
 import api.forms as forms
 import api.response as response
 
 # internal imports
 import api.forms as forms
 import api.response as response

-from api.utils import validate_form, hglibrary
+from api.utils import validate_form, hglibrary, natural_order

 from api.models import PartCache
 
 from api.models import PartCache
 

+#
+import settings
+
+
+log = logging.getLogger('platforma.api')
+
+

 #
 # Document List Handlers
 #
 #
 # Document List Handlers
 #


@@ -51,34 +61,36 @@ class LibraryHandler(BaseHandler):
 
         documents = {}
         
 
         documents = {}
         

-        for docid in lib.documents():
+        for docid in lib.documents():            

             documents[docid] = {
                 'url': reverse('document_view', args=[docid]),
                 'name': docid,
                 'parts': []
             }
 
             documents[docid] = {
                 'url': reverse('document_view', args=[docid]),
                 'name': docid,
                 'parts': []
             }
 

-        related = PartCache.objects.defer('part_id')\
+        parts = PartCache.objects.defer('part_id')\

             .values_list('part_id', 'document_id').distinct()
             .values_list('part_id', 'document_id').distinct()

+       
+        document_tree = dict(documents)

 
 

-        for part, docid in related:
+        for part, docid in parts:

             # this way, we won't display broken links
             if not documents.has_key(part):
             # this way, we won't display broken links
             if not documents.has_key(part):

+                log.info("NOT FOUND: %s", part)

                 continue
 
                 continue
 

-            child = documents[part]

             parent = documents[docid]
             parent = documents[docid]

+            child = documents[part]
+
+            # not top-level anymore
+            document_tree.pop(part)
+            parent['parts'].append(child)
+        
+        for doc in documents.itervalues():
+            doc['parts'].sort(key=natural_order(lambda d: d['name']))

             
             

-            if isinstance(parent, dict): # the parent is top-level
-                documents.pop(part)                
-                parent['parts'].append(child)
-                documents[part] = child['parts']
-            else: # not top-level
-                parent.append(child)
-            
-        return {
-            'documents': [d for d in documents.itervalues() if isinstance(d, dict)]
-        }
+        return {'documents': sorted(document_tree.itervalues(),
+            key=natural_order(lambda d: d['name']) ) }

 
     @validate_form(forms.DocumentUploadForm, 'POST')
     @hglibrary
 
     @validate_form(forms.DocumentUploadForm, 'POST')
     @hglibrary


@@ -90,6 +102,9 @@ class LibraryHandler(BaseHandler):
         else:            
             data = request.FILES['ocr_file'].read().decode('utf-8')
 
         else:            
             data = request.FILES['ocr_file'].read().decode('utf-8')
 

+        if data is None:
+            return response.BadRequest().django_response('You must pass ocr_data or ocr_file.')
+

         if form.cleaned_data['generate_dc']:
             data = librarian.wrap_text(data, unicode(date.today()))
 
         if form.cleaned_data['generate_dc']:
             data = librarian.wrap_text(data, unicode(date.today()))
 


@@ -98,7 +113,7 @@ class LibraryHandler(BaseHandler):
         try:
             lock = lib.lock()            
             try:
         try:
             lock = lib.lock()            
             try:

-                print "DOCID", docid                
+                log.info("DOCID %s", docid)

                 doc = lib.document_create(docid)
                 # document created, but no content yet
 
                 doc = lib.document_create(docid)
                 # document created, but no content yet
 


@@ -106,9 +121,10 @@ class LibraryHandler(BaseHandler):
                     doc = doc.quickwrite('xml', data.encode('utf-8'),
                         '$AUTO$ XML data uploaded.', user=request.user.username)
                 except Exception,e:
                     doc = doc.quickwrite('xml', data.encode('utf-8'),
                         '$AUTO$ XML data uploaded.', user=request.user.username)
                 except Exception,e:

+                    import traceback

                     # rollback branch creation
                     lib._rollback()
                     # rollback branch creation
                     lib._rollback()

-                    raise LibraryException("Exception occured:" + repr(e))
+                    raise LibraryException(traceback.format_exc())

 
                 url = reverse('document_view', args=[doc.id])
 
 
                 url = reverse('document_view', args=[doc.id])
 


@@ -121,8 +137,9 @@ class LibraryHandler(BaseHandler):
             finally:
                 lock.release()
         except LibraryException, e:
             finally:
                 lock.release()
         except LibraryException, e:

+            import traceback

             return response.InternalError().django_response(\
             return response.InternalError().django_response(\

-                {'exception': repr(e) })                
+                {'exception': traceback.format_exc()} )

         except DocumentAlreadyExists:
             # Document is already there
             return response.EntityConflict().django_response(\
         except DocumentAlreadyExists:
             # Document is already there
             return response.EntityConflict().django_response(\


@@ -143,9 +160,9 @@ class BasicDocumentHandler(AnonymousBaseHandler):
 
         result = {
             'name': doc.id,
 
         result = {
             'name': doc.id,

-            'html_url': reverse('dochtml_view', args=[doc.id,doc.revision]),
-            'text_url': reverse('doctext_view', args=[doc.id,doc.revision]),
-            'dc_url': reverse('docdc_view', args=[doc.id,doc.revision]),
+            'html_url': reverse('dochtml_view', args=[doc.id]),
+            'text_url': reverse('doctext_view', args=[doc.id]),
+            'dc_url': reverse('docdc_view', args=[doc.id]),

             'public_revision': doc.revision,
         }
 
             'public_revision': doc.revision,
         }
 


@@ -161,22 +178,29 @@ class DocumentHandler(BaseHandler):
     @hglibrary
     def read(self, request, docid, lib):
         """Read document's meta data"""       
     @hglibrary
     def read(self, request, docid, lib):
         """Read document's meta data"""       

+        log.info(u"Read %s (%s)" % (docid, type(docid)) )

         try:
             doc = lib.document(docid)
             udoc = doc.take(request.user.username)
         try:
             doc = lib.document(docid)
             udoc = doc.take(request.user.username)

-        except RevisionNotFound:
-            return request.EnityNotFound().django_response()
+        except RevisionNotFound, e:
+            return response.EntityNotFound().django_response({
+                'exception': type(e), 'message': e.message,
+                'docid': docid })

 
         # is_shared = udoc.ancestorof(doc)
         # is_uptodate = is_shared or shared.ancestorof(document)
 
         result = {
             'name': udoc.id,
 
         # is_shared = udoc.ancestorof(doc)
         # is_uptodate = is_shared or shared.ancestorof(document)
 
         result = {
             'name': udoc.id,

-            'html_url': reverse('dochtml_view', args=[udoc.id,udoc.revision]),
-            'text_url': reverse('doctext_view', args=[udoc.id,udoc.revision]),
-            'dc_url': reverse('docdc_view', args=[udoc.id,udoc.revision]),
+            'html_url': reverse('dochtml_view', args=[udoc.id]),
+            'text_url': reverse('doctext_view', args=[udoc.id]),
+            'dc_url': reverse('docdc_view', args=[udoc.id]),
+            'gallery_url': reverse('docgallery_view', args=[udoc.id]),
+            'merge_url': reverse('docmerge_view', args=[udoc.id]),

             'user_revision': udoc.revision,
             'user_revision': udoc.revision,

-            'public_revision': doc.revision,            
+            'user_timestamp': udoc.revision.timestamp,
+            'public_revision': doc.revision,
+            'public_timestamp': doc.revision.timestamp,

         }       
 
         return result
         }       
 
         return result


@@ -189,23 +213,77 @@ class DocumentHandler(BaseHandler):
 #
 #
 class DocumentHTMLHandler(BaseHandler):
 #
 #
 class DocumentHTMLHandler(BaseHandler):

-    allowed_methods = ('GET', 'PUT')
+    allowed_methods = ('GET')

 
     @hglibrary
 
     @hglibrary

-    def read(self, request, docid, revision, lib):
+    def read(self, request, docid, lib):

         """Read document as html text"""
         try:
         """Read document as html text"""
         try:

+            revision = request.GET.get('revision', 'latest')
+

             if revision == 'latest':
                 document = lib.document(docid)
             else:
                 document = lib.document_for_rev(revision)
 
             if revision == 'latest':
                 document = lib.document(docid)
             else:
                 document = lib.document_for_rev(revision)
 

-            return librarian.html.transform(document.data('xml'), is_file=False)
-        except RevisionNotFound:
-            return response.EntityNotFound().django_response()
+            if document.id != docid:
+                return response.BadRequest().django_response({'reason': 'name-mismatch',
+                    'message': 'Provided revision refers, to document "%s", but provided "%s"' % (document.id, docid) })
+
+            return librarian.html.transform(document.data('xml'), is_file=False, \
+                parse_dublincore=False, stylesheet="partial",\
+                options={
+                    "with-paths": 'boolean(1)',                    
+                })
+                
+        except (EntryNotFound, RevisionNotFound), e:
+            return response.EntityNotFound().django_response({
+                'reason': 'not-found', 'message': e.message})
+        except librarian.ParseError, e:
+            return response.InternalError().django_response({
+                'reason': 'xml-parse-error', 'message': e.message })
+
+#
+# Image Gallery
+#
+
+class DocumentGalleryHandler(BaseHandler):
+    allowed_methods = ('GET')
+    
+    
+    def read(self, request, docid):
+        """Read meta-data about scans for gallery of this document."""
+        galleries = []
+        from urllib import quote

 
 

+        for assoc in GalleryForDocument.objects.filter(document=docid):
+            dirpath = os.path.join(settings.MEDIA_ROOT, assoc.subpath)

 
 

+            if not os.path.isdir(dirpath):
+                log.warn(u"[WARNING]: missing gallery %s", dirpath)
+                continue
+
+            gallery = {'name': assoc.name, 'pages': []}
+            
+            for file in os.listdir(dirpath):
+                if not isinstance(file, unicode):
+                    log.warn(u"File %r is gallery %r is not unicode. Ommiting."\
+                        % (file, dirpath) )
+                    continue
+                               
+                name, ext = os.path.splitext(os.path.basename(file))
+
+                if ext.lower() not in [u'.png', u'.jpeg', u'.jpg']:
+                    log.info(u"Ignoring: %s %s", name, ext)
+                    continue
+
+                url = settings.MEDIA_URL + assoc.subpath + u'/' + file;
+                gallery['pages'].append( quote(url.encode('utf-8')) )

 
 

+            gallery['pages'].sort()
+            galleries.append(gallery)
+
+        return galleries                      

 
 #
 # Document Text View
 
 #
 # Document Text View


@@ -214,32 +292,56 @@ class DocumentHTMLHandler(BaseHandler):
 XINCLUDE_REGEXP = r"""<(?:\w+:)?include\s+[^>]*?href=("|')wlrepo://(?P<link>[^\1]+?)\1\s*[^>]*?>"""
 #
 #
 XINCLUDE_REGEXP = r"""<(?:\w+:)?include\s+[^>]*?href=("|')wlrepo://(?P<link>[^\1]+?)\1\s*[^>]*?>"""
 #
 #

+#
+

 class DocumentTextHandler(BaseHandler):
 class DocumentTextHandler(BaseHandler):

-    allowed_methods = ('GET', 'PUT')
+    allowed_methods = ('GET', 'POST')

 
     @hglibrary
 
     @hglibrary

-    def read(self, request, docid, revision, lib):
-        """Read document as raw text"""               
+    def read(self, request, docid, lib):
+        """Read document as raw text"""
+        revision = request.GET.get('revision', 'latest')
+        part = request.GET.get('part', False)
+        

         try:
             if revision == 'latest':
                 document = lib.document(docid)
             else:
                 document = lib.document_for_rev(revision)
         try:
             if revision == 'latest':
                 document = lib.document(docid)
             else:
                 document = lib.document_for_rev(revision)

+
+            if document.id != docid:
+                return response.BadRequest().django_response({'reason': 'name-mismatch',
+                    'message': 'Provided revision is not valid for this document'})

             
             # TODO: some finer-grained access control
             
             # TODO: some finer-grained access control

-            return document.data('xml')
-        except RevisionNotFound:
-            return response.EntityNotFound().django_response()
+            if part is False:
+                # we're done :)
+                return document.data('xml')
+            else:
+                xdoc = parser.WLDocument.from_string(document.data('xml'))
+                ptext = xdoc.part_as_text(part)
+
+                if ptext is None:
+                    return response.EntityNotFound().django_response({
+                      'reason': 'no-part-in-document'                     
+                    })
+
+                return ptext
+        except librarian.ParseError:
+            return response.EntityNotFound().django_response({
+                'reason': 'invalid-document-state',
+                'exception': type(e), 'message': e.message
+            })
+        except (EntryNotFound, RevisionNotFound), e:
+            return response.EntityNotFound().django_response({
+                'reason': 'not-found',
+                'exception': type(e), 'message': e.message
+            })   

 
     @hglibrary
 
     @hglibrary

-    def update(self, request, docid, revision, lib):
+    def create(self, request, docid, lib):

         try:
         try:

-            data = request.PUT['contents']            
-
-            if request.PUT.has_key('message'):
-                msg = u"$USER$ " + request.PUT['message']
-            else:
-                msg = u"$AUTO$ XML content update."
+            revision = request.POST['revision']

 
             current = lib.document(docid, request.user.username)
             orig = lib.document_for_rev(revision)
 
             current = lib.document(docid, request.user.username)
             orig = lib.document_for_rev(revision)


@@ -250,10 +352,39 @@ class DocumentTextHandler(BaseHandler):
                         "provided_revision": orig.revision,
                         "latest_revision": current.revision })
 
                         "provided_revision": orig.revision,
                         "latest_revision": current.revision })
 

+            if request.POST.has_key('message'):
+                msg = u"$USER$ " + request.POST['message']
+            else:
+                msg = u"$AUTO$ XML content update."
+
+            if request.POST.has_key('contents'):
+                data = request.POST['contents']
+            else:
+                if not request.POST.has_key('chunks'):
+                    # bad request
+                    return response.BadRequest().django_response({'reason': 'invalid-arguments',
+                        'message': 'No contents nor chunks specified.'})
+
+                    # TODO: validate
+                parts = json.loads(request.POST['chunks'])                    
+                xdoc = parser.WLDocument.from_string(current.data('xml'))
+                   
+                errors = xdoc.merge_chunks(parts)
+
+                if len(errors):
+                    return response.EntityConflict().django_response({
+                            "reason": "invalid-chunks",
+                            "message": "Unable to merge following parts into the document: %s " % ",".join(errors)
+                    })
+
+                data = xdoc.serialize()
+

             # try to find any Xinclude tags
             includes = [m.groupdict()['link'] for m in (re.finditer(\
                 XINCLUDE_REGEXP, data, flags=re.UNICODE) or []) ]
 
             # try to find any Xinclude tags
             includes = [m.groupdict()['link'] for m in (re.finditer(\
                 XINCLUDE_REGEXP, data, flags=re.UNICODE) or []) ]
 

+            log.info("INCLUDES: %s", includes)
+

             # TODO: provide useful routines to make this simpler
             def xml_update_action(lib, resolve):
                 try:
             # TODO: provide useful routines to make this simpler
             def xml_update_action(lib, resolve):
                 try:


@@ -268,13 +399,15 @@ class DocumentTextHandler(BaseHandler):
                     f.write(json.dumps(includes))
                     f.close()
 
                     f.write(json.dumps(includes))
                     f.close()
 

+                    lib._fileadd(resolve('parts'))
+

                     # update the parts cache
                     PartCache.update_cache(docid, current.owner,\
                         stored_includes, includes)
 
                 # now that the parts are ok, write xml
                 f = lib._fileopen(resolve('xml'), 'w+')
                     # update the parts cache
                     PartCache.update_cache(docid, current.owner,\
                         stored_includes, includes)
 
                 # now that the parts are ok, write xml
                 f = lib._fileopen(resolve('xml'), 'w+')

-                f.write(data)
+                f.write(data.encode('utf-8'))

                 f.close()
 
             ndoc = None
                 f.close()
 
             ndoc = None


@@ -287,14 +420,17 @@ class DocumentTextHandler(BaseHandler):
                     "document": ndoc.id,
                     "subview": "xml",
                     "previous_revision": current.revision,
                     "document": ndoc.id,
                     "subview": "xml",
                     "previous_revision": current.revision,

-                    "updated_revision": ndoc.revision,
-                    "url": reverse("doctext_view", args=[ndoc.id, ndoc.revision])
+                    "revision": ndoc.revision,
+                    'timestamp': ndoc.revision.timestamp,
+                    "url": reverse("doctext_view", args=[ndoc.id])

                 })
             except Exception, e:
                 if ndoc: lib._rollback()
                 raise e        
         except RevisionNotFound, e:
                 })
             except Exception, e:
                 if ndoc: lib._rollback()
                 raise e        
         except RevisionNotFound, e:

-            return response.EntityNotFound().django_response(e)
+            return response.EntityNotFound(mimetype="text/plain").\
+                django_response(e.message)
+

 
 #
 # Dublin Core handlers
 
 #
 # Dublin Core handlers


@@ -302,27 +438,37 @@ class DocumentTextHandler(BaseHandler):
 # @requires librarian
 #
 class DocumentDublinCoreHandler(BaseHandler):
 # @requires librarian
 #
 class DocumentDublinCoreHandler(BaseHandler):

-    allowed_methods = ('GET', 'PUT')
+    allowed_methods = ('GET', 'POST')

 
     @hglibrary
 
     @hglibrary

-    def read(self, request, docid, revision, lib):
+    def read(self, request, docid, lib):

         """Read document as raw text"""        
         try:
         """Read document as raw text"""        
         try:

+            revision = request.GET.get('revision', 'latest')
+

             if revision == 'latest':
                 doc = lib.document(docid)
             else:
                 doc = lib.document_for_rev(revision)
             if revision == 'latest':
                 doc = lib.document(docid)
             else:
                 doc = lib.document_for_rev(revision)

+
+
+            if document.id != docid:
+                return response.BadRequest().django_response({'reason': 'name-mismatch',
+                    'message': 'Provided revision is not valid for this document'})

             
             bookinfo = dcparser.BookInfo.from_string(doc.data('xml'))
             return bookinfo.serialize()
             
             bookinfo = dcparser.BookInfo.from_string(doc.data('xml'))
             return bookinfo.serialize()

-        except RevisionNotFound:
-            return response.EntityNotFound().django_response()
+        except (EntryNotFound, RevisionNotFound), e:
+            return response.EntityNotFound().django_response({
+                'exception': type(e), 'message': e.message})

 
     @hglibrary
 
     @hglibrary

-    def update(self, request, docid, revision, lib):
+    def create(self, request, docid, lib):

         try:
         try:

-            bi_json = request.PUT['contents']            
-            if request.PUT.has_key('message'):
+            bi_json = request.POST['contents']
+            revision = request.POST['revision']
+            
+            if request.POST.has_key('message'):

                 msg = u"$USER$ " + request.PUT['message']
             else:
                 msg = u"$AUTO$ Dublin core update."
                 msg = u"$USER$ " + request.PUT['message']
             else:
                 msg = u"$AUTO$ Dublin core update."


@@ -350,16 +496,16 @@ class DocumentDublinCoreHandler(BaseHandler):
                     "document": ndoc.id,
                     "subview": "dc",
                     "previous_revision": current.revision,
                     "document": ndoc.id,
                     "subview": "dc",
                     "previous_revision": current.revision,

-                    "updated_revision": ndoc.revision
+                    "revision": ndoc.revision,
+                    'timestamp': ndoc.revision.timestamp,
+                    "url": reverse("docdc_view", args=[ndoc.id])

                 }
             except Exception, e:
                 }
             except Exception, e:

-                lib._rollback()
+                if ndoc: lib._rollback()

                 raise e
         except RevisionNotFound:
             return response.EntityNotFound().django_response()
 
                 raise e
         except RevisionNotFound:
             return response.EntityNotFound().django_response()
 

-
-

 class MergeHandler(BaseHandler):
     allowed_methods = ('POST',)
 
 class MergeHandler(BaseHandler):
     allowed_methods = ('POST',)
 


@@ -401,21 +547,25 @@ class MergeHandler(BaseHandler):
                     "provided": target_rev,
                     "latest": udoc.revision })
 
                     "provided": target_rev,
                     "latest": udoc.revision })
 

-        if not request.user.has_perm('explorer.book.can_share'):
+        if not request.user.has_perm('explorer.document.can_share'):

             # User is not permitted to make a merge, right away
             # So we instead create a pull request in the database
             # User is not permitted to make a merge, right away
             # So we instead create a pull request in the database

-            prq = PullRequest(
-                comitter=request.user,
-                document=docid,
-                source_revision = str(udoc.revision),
-                status="N",
-                comment = form.cleaned_data['comment'] or '$AUTO$ Document shared.'
-            )
-
-            prq.save()
-            return response.RequestAccepted().django_response(\
-                ticket_status=prq.status, \
-                ticket_uri=reverse("pullrequest_view", args=[prq.id]) )
+            try:
+                prq, created = PullRequest.get_or_create(                                        
+                    source_revision = str(udoc.revision),
+                    defaults = {
+                        'comitter': request.user,
+                        'document': docid,
+                        'status': "N",
+                        'comment': form.cleaned_data['message'] or '$AUTO$ Document shared.',
+                    }
+                )
+
+                return response.RequestAccepted().django_response(\
+                    ticket_status=prq.status, \
+                    ticket_uri=reverse("pullrequest_view", args=[prq.id]) )
+            except IntegrityError, e:
+                return response.InternalError().django_response()

 
         if form.cleaned_data['type'] == 'update':
             # update is always performed from the file branch
 
         if form.cleaned_data['type'] == 'update':
             # update is always performed from the file branch


@@ -423,19 +573,22 @@ class MergeHandler(BaseHandler):
             success, changed = udoc.update(request.user.username)
 
         if form.cleaned_data['type'] == 'share':
             success, changed = udoc.update(request.user.username)
 
         if form.cleaned_data['type'] == 'share':

-            success, changed = udoc.share(form.cleaned_data['comment'])
+            success, changed = udoc.share(form.cleaned_data['message'])

 
         if not success:
 
         if not success:

-            return response.EntityConflict().django_response()
+            return response.EntityConflict().django_response({
+                'reason': 'merge-failure',
+            })

 
         if not changed:
             return response.SuccessNoContent().django_response()
 
 
         if not changed:
             return response.SuccessNoContent().django_response()
 

-        new_udoc = udoc.latest()
+        nudoc = udoc.latest()

 
         return response.SuccessAllOk().django_response({
 
         return response.SuccessAllOk().django_response({

-            "name": udoc.id,
+            "name": nudoc.id,

             "parent_user_resivion": udoc.revision,
             "parent_revision": doc.revision,
             "parent_user_resivion": udoc.revision,
             "parent_revision": doc.revision,

-            "revision": udoc.revision,
-        })
\ No newline at end of file
+            "revision": nudoc.revision,
+            'timestamp': nudoc.revision.timestamp,
+        })