Fix in librarian
[wolnelektury.git] / apps / catalogue / utils.py
index 185f5fa..bcc5a0b 100644 (file)
@@ -2,30 +2,24 @@
 # This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later.
 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
 #
 # This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later.
 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
 #
-from __future__ import with_statement
-
+from collections import defaultdict
+import hashlib
 import random
 import re
 import time
 from base64 import urlsafe_b64encode
 
 import random
 import re
 import time
 from base64 import urlsafe_b64encode
 
-from django.http import HttpResponse, HttpResponseRedirect, Http404, HttpResponsePermanentRedirect
+from django.http import HttpResponse
 from django.core.files.uploadedfile import UploadedFile
 from django.core.files.uploadedfile import UploadedFile
-from django.core.files.base import File
 from django.core.files.storage import DefaultStorage
 from django.utils.encoding import force_unicode
 from django.core.files.storage import DefaultStorage
 from django.utils.encoding import force_unicode
-from django.utils.hashcompat import sha_constructor
 from django.conf import settings
 from django.conf import settings
-from celery.task import task
 from os import mkdir, path, unlink
 from errno import EEXIST, ENOENT
 from fcntl import flock, LOCK_EX
 from zipfile import ZipFile
 from os import mkdir, path, unlink
 from errno import EEXIST, ENOENT
 from fcntl import flock, LOCK_EX
 from zipfile import ZipFile
-from traceback import print_exc
 
 from reporting.utils import read_chunks
 
 from reporting.utils import read_chunks
-from celery.task import task
-import catalogue.models
 
 # Use the system (hardware-based) random number generator if it exists.
 if hasattr(random, 'SystemRandom'):
 
 # Use the system (hardware-based) random number generator if it exists.
 if hasattr(random, 'SystemRandom'):
@@ -36,16 +30,27 @@ MAX_SESSION_KEY = 18446744073709551616L     # 2 << 63
 
 
 def get_random_hash(seed):
 
 
 def get_random_hash(seed):
-    sha_digest = sha_constructor('%s%s%s%s' %
+    sha_digest = hashlib.sha1('%s%s%s%s' %
         (randrange(0, MAX_SESSION_KEY), time.time(), unicode(seed).encode('utf-8', 'replace'),
         settings.SECRET_KEY)).digest()
     return urlsafe_b64encode(sha_digest).replace('=', '').replace('_', '-').lower()
 
 
         (randrange(0, MAX_SESSION_KEY), time.time(), unicode(seed).encode('utf-8', 'replace'),
         settings.SECRET_KEY)).digest()
     return urlsafe_b64encode(sha_digest).replace('=', '').replace('_', '-').lower()
 
 
-def split_tags(tags):
-    result = {}
-    for tag in tags:
-        result.setdefault(tag.category, []).append(tag)
+def split_tags(*tag_lists):
+    if len(tag_lists) == 1:
+        result = defaultdict(list)
+        for tag in tag_lists[0]:
+            result[tag.category].append(tag)
+    else:
+        result = defaultdict(dict)
+        for tag_list in tag_lists:
+            for tag in tag_list:
+                try:
+                    result[tag.category][tag.pk].count += tag.count
+                except KeyError:
+                    result[tag.category][tag.pk] = tag
+        for k, v in result.items():
+            result[k] = sorted(v.values(), key=lambda tag: tag.sort_key)
     return result
 
 
     return result
 
 
@@ -79,12 +84,12 @@ class LockFile(object):
         try:
             unlink(self.lockname)
         except OSError as oe:
         try:
             unlink(self.lockname)
         except OSError as oe:
-            if oe.errno != oe.EEXIST:
+            if oe.errno != EEXIST:
                 raise oe
         self.lock.close()
 
 
                 raise oe
         self.lock.close()
 
 
-@task
+#@task
 def create_zip(paths, zip_slug):
     """
     Creates a zip in MEDIA_ROOT/zip directory containing files from path.
 def create_zip(paths, zip_slug):
     """
     Creates a zip in MEDIA_ROOT/zip directory containing files from path.
@@ -130,7 +135,7 @@ def remove_zip(zip_slug):
 class AttachmentHttpResponse(HttpResponse):
     """Response serving a file to be downloaded.
     """
 class AttachmentHttpResponse(HttpResponse):
     """Response serving a file to be downloaded.
     """
-    def __init__ (self, file_path, file_name, mimetype):
+    def __init__(self, file_path, file_name, mimetype):
         super(AttachmentHttpResponse, self).__init__(mimetype=mimetype)
         self['Content-Disposition'] = 'attachment; filename=%s' % file_name
         self.file_path = file_path
         super(AttachmentHttpResponse, self).__init__(mimetype=mimetype)
         self['Content-Disposition'] = 'attachment; filename=%s' % file_name
         self.file_path = file_path
@@ -140,41 +145,22 @@ class AttachmentHttpResponse(HttpResponse):
             for chunk in read_chunks(f):
                 self.write(chunk)
 
             for chunk in read_chunks(f):
                 self.write(chunk)
 
-@task
-def async_build_pdf(book_id, customizations, file_name):
-    """
-    A celery task to generate pdf files.
-    Accepts the same args as Book.build_pdf, but with book id as first parameter
-    instead of Book instance
-    """
-    try:
-        book = catalogue.models.Book.objects.get(id=book_id)
-        print "will gen %s" % DefaultStorage().path(file_name)
-        if not DefaultStorage().exists(file_name):
-            book.build_pdf(customizations=customizations, file_name=file_name)
-        print "done."
-    except Exception, e:
-        print "Error during pdf creation: %s" % e
-        print_exc
-        raise e
-
-
 class MultiQuerySet(object):
     def __init__(self, *args, **kwargs):
         self.querysets = args
         self._count = None
 class MultiQuerySet(object):
     def __init__(self, *args, **kwargs):
         self.querysets = args
         self._count = None
-    
+
     def count(self):
         if not self._count:
             self._count = sum(len(qs) for qs in self.querysets)
         return self._count
     def count(self):
         if not self._count:
             self._count = sum(len(qs) for qs in self.querysets)
         return self._count
-    
+
     def __len__(self):
         return self.count()
     def __len__(self):
         return self.count()
-        
+
     def __getitem__(self, item):
         try:
     def __getitem__(self, item):
         try:
-            indices = (offset, stop, step) = item.indices(self.count())
+            (offset, stop, step) = item.indices(self.count())
         except AttributeError:
             # it's not a slice - make it one
             return self[item : item + 1][0]
         except AttributeError:
             # it's not a slice - make it one
             return self[item : item + 1][0]
@@ -192,6 +178,55 @@ class MultiQuerySet(object):
                     stop = total_len - len(items)
                     continue
 
                     stop = total_len - len(items)
                     continue
 
+class SortedMultiQuerySet(MultiQuerySet):
+    def __init__(self, *args, **kwargs):
+        self.order_by = kwargs.pop('order_by', None)
+        self.sortfn = kwargs.pop('sortfn', None)
+        if self.order_by is not None:
+            self.sortfn = lambda a, b: cmp((getattr(a, f) for f in self.order_by),
+                                           (getattr(b, f) for f in self.order_by))
+        super(SortedMultiQuerySet, self).__init__(*args, **kwargs)
+
+    def __getitem__(self, item):
+        sort_heads = [0] * len(self.querysets)
+        try:
+            (offset, stop, step) = item.indices(self.count())
+        except AttributeError:
+            # it's not a slice - make it one
+            return self[item : item + 1][0]
+        items = []
+        total_len = stop - offset
+        skipped = 0
+        i_s = range(len(sort_heads))
+
+        while len(items) < total_len:
+            candidate = None
+            candidate_i = None
+            for i in i_s:
+                def get_next():
+                    return self.querysets[i][sort_heads[i]]
+                try:
+                    if candidate is None:
+                        candidate = get_next()
+                        candidate_i = i
+                    else:
+                        competitor = get_next()
+                        if self.sortfn(candidate, competitor) > 0:
+                            candidate = competitor
+                            candidate_i = i
+                except IndexError:
+                    continue # continue next sort_head
+            # we have no more elements:
+            if candidate is None:
+                break
+            sort_heads[candidate_i] += 1
+            if skipped < offset:
+                skipped += 1
+                continue # continue next item
+            items.append(candidate)
+
+        return items
+
 
 def truncate_html_words(s, num, end_text='...'):
     """Truncates HTML to a certain number of words (not counting tags and
 
 def truncate_html_words(s, num, end_text='...'):
     """Truncates HTML to a certain number of words (not counting tags and
@@ -260,3 +295,64 @@ def truncate_html_words(s, num, end_text='...'):
         out += '</%s>' % tag
     # Return string
     return out
         out += '</%s>' % tag
     # Return string
     return out
+
+
+def customizations_hash(customizations):
+    customizations.sort()
+    return hash(tuple(customizations))
+
+
+def get_customized_pdf_path(book, customizations):
+    """
+    Returns a MEDIA_ROOT relative path for a customized pdf. The name will contain a hash of customization options.
+    """
+    h = customizations_hash(customizations)
+    return 'book/%s/%s-custom-%s.pdf' % (book.slug, book.slug, h)
+
+
+def clear_custom_pdf(book):
+    """
+    Returns a list of paths to generated customized pdf of a book
+    """
+    from waiter.utils import clear_cache
+    clear_cache('book/%s' % book.slug)
+
+
+class AppSettings(object):
+    """Allows specyfying custom settings for an app, with default values.
+
+    Just subclass, set some properties and instantiate with a prefix.
+    Getting a SETTING from an instance will check for prefix_SETTING
+    in project settings if set, else take the default. The value will be
+    then filtered through _more_SETTING method, if there is one.
+
+    """
+    def __init__(self, prefix):
+        self._prefix = prefix
+
+    def __getattribute__(self, name):
+        if name.startswith('_'):
+            return object.__getattribute__(self, name)
+        value = getattr(settings,
+                         "%s_%s" % (self._prefix, name),
+                         object.__getattribute__(self, name))
+        more = "_more_%s" % name
+        if hasattr(self, more):
+            value = getattr(self, more)(value)
+        return value
+
+
+def trim_query_log(trim_to=25):
+    """
+connection.queries includes all SQL statements -- INSERTs, UPDATES, SELECTs, etc. Each time your app hits the database, the query will be recorded.
+This can sometimes occupy lots of memory, so trim it here a bit.
+    """
+    if settings.DEBUG:
+        from django.db import connection
+        connection.queries = trim_to > 0 \
+            and connection.queries[-trim_to:] \
+            or []
+
+
+def delete_from_cache_by_language(cache, key_template):
+    cache.delete_many([key_template % lc for lc, ln in settings.LANGUAGES])