Smarten Punctuation option on epub download/update.

calibre-plugin/config.py

@@ -177,6 +177,7 @@ class ConfigWidget(QWidget):
         prefs['lookforurlinhtml'] = self.basic_tab.lookforurlinhtml.isChecked()
         prefs['checkforseriesurlid'] = self.basic_tab.checkforseriesurlid.isChecked()
         prefs['injectseries'] = self.basic_tab.injectseries.isChecked()
+        prefs['smarten_punctuation'] = self.basic_tab.smarten_punctuation.isChecked()
 
         if self.readinglist_tab:
             # lists
@@ -320,6 +321,11 @@ class BasicTab(QWidget):
         self.updateepubcover.setChecked(prefs['updateepubcover'])
         self.l.addWidget(self.updateepubcover)
 
+        self.smarten_punctuation = QCheckBox('Smarten Punctuation (EPUB only)',self)
+        self.smarten_punctuation.setToolTip("Run Smarten Punctuation from Calibre's Polish Book feature on each EPUB download and update.")
+        self.smarten_punctuation.setChecked(prefs['smarten_punctuation'])
+        self.l.addWidget(self.smarten_punctuation)
+
         cali_gb = groupbox = QGroupBox("Updating Calibre Options")
         self.l = QVBoxLayout()
         groupbox.setLayout(self.l)

calibre-plugin/dialogs.py

@@ -383,6 +383,7 @@ class AddNewDialog(SizePersistedDialog):
             'collision': unicode(self.collision.currentText()),
             'updatemeta': self.updatemeta.isChecked(),
             'updateepubcover': self.updateepubcover.isChecked(),
+            'smarten_punctuation':self.prefs['smarten_punctuation']
                 }
         
         if self.merge:
@@ -618,7 +619,8 @@ class UpdateExistingDialog(SizePersistedDialog):
     def __init__(self, gui, header, prefs, icon, books,
                  save_size_name='fanfictiondownloader_plugin:update list dialog'):
         SizePersistedDialog.__init__(self, gui, save_size_name)
-        
+
+        self.prefs = prefs
         self.setWindowTitle(header)
         self.setWindowIcon(icon)
         
@@ -736,6 +738,7 @@ class UpdateExistingDialog(SizePersistedDialog):
             'collision': unicode(self.collision.currentText()),
             'updatemeta': self.updatemeta.isChecked(),
             'updateepubcover': self.updateepubcover.isChecked(),
+            'smarten_punctuation':self.prefs['smarten_punctuation']
             }
 
 class StoryListTableWidget(QTableWidget):

calibre-plugin/jobs.py

@@ -49,7 +49,7 @@ def do_download_worker(book_list, options,
             args = ['calibre_plugins.fanfictiondownloader_plugin.jobs',
                     'do_download_for_worker',
                     (book,options)]
-            job = ParallelJob('arbitrary',
+            job = ParallelJob('arbitrary_n',
                               "url:(%s) id:(%s)"%(book['url'],book['calibre_id']),
                               done=None,
                               args=args)
@@ -84,7 +84,7 @@ def do_download_worker(book_list, options,
         book_id = job._book['calibre_id']
         #title = job._title
         count = count + 1
-        notification(float(count)/total, 'Downloaded Story')
+        notification(float(count)/total, '%d of %d stories finished downloading'%(count,total))
         # Add this job's output to the current log
         print('Logfile for book ID %s (%s)'%(book_id, job._book['title']))
         print(job.details)
@@ -106,7 +106,7 @@ def do_download_worker(book_list, options,
     # return the book list as the job result
     return book_list
 
-def do_download_for_worker(book,options):
+def do_download_for_worker(book,options,notification=lambda x,y:x):
     '''
     Child job, to extract isbn from formats for this specific book,
     when run as a worker job
@@ -206,6 +206,22 @@ def do_download_for_worker(book,options):
             book['comment'] = 'Update %s completed, added %s chapters for %s total.'%\
                 (options['fileform'],(urlchaptercount-chaptercount),urlchaptercount)
         
+        if options['smarten_punctuation'] and options['fileform'] == "epub":
+            # do smarten_punctuation from calibre's polish feature
+            from calibre.ebooks.oeb.polish.main import polish, ALL_OPTS
+            from calibre.utils.logging import Log
+            from collections import namedtuple
+        
+            data = {'smarten_punctuation':True}
+            opts = ALL_OPTS.copy()
+            opts.update(data)
+            O = namedtuple('Options', ' '.join(ALL_OPTS.iterkeys()))
+            opts = O(**opts)
+            
+            log = Log(level=Log.DEBUG)
+            # report = []
+            polish({outfile:outfile}, opts, log, print) # report.append
+        
     except NotGoingToDownload as d:
         book['good']=False
         book['comment']=unicode(d)
@@ -218,6 +234,6 @@ def do_download_for_worker(book,options):
         book['status'] = 'Error'
         print("Exception: %s:%s"%(book,unicode(e)))
         traceback.print_exc()
-
+        
     #time.sleep(10)
     return book

calibre-plugin/prefs.py

@@ -41,6 +41,7 @@ default_prefs['includeimages'] = False
 default_prefs['lookforurlinhtml'] = False
 default_prefs['checkforseriesurlid'] = True
 default_prefs['injectseries'] = False
+default_prefs['smarten_punctuation'] = False
 
 default_prefs['send_lists'] = ''
 default_prefs['read_lists'] = ''

fanficdownloader/adapters/adapter_test1.py

@@ -350,7 +350,7 @@ Don't&#8212e;ver&#8212d;o—that&#8212a;gain, 法 é
 <hr>
 horizontal rules
 <hr size=1 noshade>
-<p>Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.</p>
+<p>"Lorem ipsum dolor sit amet", consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore--et dolore magna aliqua. 'Ut enim ad minim veniam', quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.</p>
 <p>Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.</p>
 </div>
 '''