Add keep_html_attrs and replace_tags_with_spans features.

calibre-plugin/plugin-defaults.ini

@@ -266,12 +266,32 @@ keep_summary_html:true
 
 ## If set true, any style attributes on tags in the story HTML will be
 ## kept.  Useful for keeping extra colors & formatting from original.
+## DEPRECATED - use keep_html_attrs (or add_to_keep_html_attrs) instead.
 #keep_style_attr: false
 
 ## If set true, any title attributes on tags in the story HTML will be
 ## kept.  Some sites(AO3) use <a title="Translation">Foreign language</a>
+## DEPRECATED - use keep_html_attrs (or add_to_keep_html_attrs) instead.
 #keep_title_attr: false
 
+## Some attributes cause problems for EBook readers.  By default,
+## FanFicFare will remove all attributes except the ones specified
+## from all tags.  (The only exception is that <img> tags will also
+## keep src, alt and longdesc attributes.)
+## Example: To add 'style', 'title' and 'align' to the list to keep,
+## in your personal.ini [defaults] put:
+## add_to_keep_html_attrs:,style,title,align
+keep_html_attrs:href,name,class,id
+
+## Tags listed here will be replaced with <span class="tagname">.
+## For example: <u>underlined text</u> becomes
+## <span class="u">underlined text</span>
+## Note that the output_css should contain the class .u, .big, etc for
+## the spans to be useful.
+## This feature is for replacing old tags deprecated/removed in newer
+## HTML and EPUB standards.
+replace_tags_with_spans:u,big,small
+
 ## If a chapter range was given, use this pattern for the book title.
 ## replace_metadata and include/exclude will be applied *after* this.
 ## Set to empty value to disable.
@@ -507,6 +527,8 @@ output_css:
  .smcap    {font-variant: small-caps;}
  .u        {text-decoration: underline;}
  .bold     {font-weight: bold;}
+ .big { font-size: larger; }
+ .small { font-size: smaller; }
 
 [txt]
 ## Add URLs since there aren't links.
@@ -587,6 +609,8 @@ output_css:
  .smcap    {font-variant: small-caps;}
  .u        {text-decoration: underline;}
  .bold     {font-weight: bold;}
+ .big { font-size: larger; }
+ .small { font-size: smaller; }
 
 ## include images from img tags in the body and summary of
 ## stories.  Images will be converted to jpg for size if possible.

fanficfare/adapters/base_adapter.py

@@ -561,11 +561,13 @@ class BaseSiteAdapter(Configurable):
         if not fetch:
             fetch=self._fetchUrlRaw
 
-        acceptable_attributes = ['href','name','class','id']
+        acceptable_attributes = self.getConfigList('keep_html_attrs',['href','name','class','id'])
+        
         if self.getConfig("keep_style_attr"):
             acceptable_attributes.append('style')
         if self.getConfig("keep_title_attr"):
             acceptable_attributes.append('title')
+            
         #print("include_images:"+self.getConfig('include_images'))
         if self.getConfig('include_images'):
             acceptable_attributes.extend(('src','alt','longdesc'))
@@ -590,7 +592,7 @@ class BaseSiteAdapter(Configurable):
                 # these are not acceptable strict XHTML.  But we do already have 
                 # CSS classes of the same names defined
                 if t and hasattr(t,'name') and t.name is not None:
-                    if t.name in ('u'):
+                    if t.name in self.getConfigList('replace_tags_with_spans',['u']):
                         t['class']=t.name
                         t.name='span'
                     if t.name in ('center'):

fanficfare/configurable.py

@@ -111,6 +111,8 @@ def get_valid_list_entries():
                  'authorId',
                  'authorUrl',
                  'lastupdate',
+                 'keep_html_attrs',
+                 'replace_tags_with_spans',
                  ])
 
 boollist=['true','false']
@@ -273,6 +275,8 @@ def get_valid_keywords():
                  'join_string_authorHTML',
                  'keep_style_attr',
                  'keep_title_attr',
+                 'keep_html_attrs',
+                 'replace_tags_with_spans',
                  'keep_summary_html',
                  'logpage_end',
                  'logpage_entries',
@@ -449,15 +453,18 @@ class Configuration(ConfigParser.SafeConfigParser):
         return val
 
     # split and strip each.
-    def get_config_list(self, sections, key):
+    def get_config_list(self, sections, key, default=[]):
         vlist = re.split(r'(?<!\\),',self.get_config(sections,key)) # don't split on \,
         vlist = filter( lambda x : x !='', [ v.strip().replace('\,',',') for v in vlist ])
         #print "vlist("+key+"):"+str(vlist)
-        return vlist        
+        if not vlist:
+            return default
+        else:
+            return vlist
     
     # used by adapters & writers, non-convention naming style
-    def getConfigList(self, key):
-        return self.get_config_list(self.sectionslist, key)
+    def getConfigList(self, key, default=[]):
+        return self.get_config_list(self.sectionslist, key, default)
 
     # Moved here for test_config.
     def get_generate_cover_settings(self):
@@ -690,8 +697,8 @@ class Configurable(object):
     def get_config(self, sections, key, default=""):
         return self.configuration.get_config(sections,key,default)
 
-    def getConfigList(self, key):
-        return self.configuration.getConfigList(key)
+    def getConfigList(self, key, default=[]):
+        return self.configuration.getConfigList(key,default)
 
     def get_config_list(self, sections, key):
         return self.configuration.get_config_list(sections,key)

fanficfare/defaults.ini

@@ -275,12 +275,32 @@ keep_summary_html:true
 
 ## If set true, any style attributes on tags in the story HTML will be
 ## kept.  Useful for keeping extra colors & formatting from original.
+## DEPRECATED - use keep_html_attrs (or add_to_keep_html_attrs) instead.
 #keep_style_attr: false
 
 ## If set true, any title attributes on tags in the story HTML will be
 ## kept.  Some sites(AO3) use <a title="Translation">Foreign language</a>
+## DEPRECATED - use keep_html_attrs (or add_to_keep_html_attrs) instead.
 #keep_title_attr: false
 
+## Some attributes cause problems for EBook readers.  By default,
+## FanFicFare will remove all attributes except the ones specified
+## from all tags.  (The only exception is that <img> tags will also
+## keep src, alt and longdesc attributes.)
+## Example: To add 'style', 'title' and 'align' to the list to keep,
+## in your personal.ini [defaults] put:
+## add_to_keep_html_attrs:,style,title,align
+keep_html_attrs:href,name,class,id
+
+## Tags listed here will be replaced with <span class="tagname">.
+## For example: <u>underlined text</u> becomes
+## <span class="u">underlined text</span>
+## Note that the output_css should contain the class .u, .big, etc for
+## the spans to be useful.
+## This feature is for replacing old tags deprecated/removed in newer
+## HTML and EPUB standards.
+replace_tags_with_spans:u,big,small
+
 ## If a chapter range was given, use this pattern for the book title.
 ## replace_metadata and include/exclude will be applied *after* this.
 ## Set to empty value to disable.
@@ -511,6 +531,8 @@ output_css:
  .smcap    {font-variant: small-caps;}
  .u        {text-decoration: underline;}
  .bold     {font-weight: bold;}
+ .big { font-size: larger; }
+ .small { font-size: smaller; }
 
 [txt]
 ## Add URLs since there aren't links.
@@ -592,6 +614,8 @@ output_css:
  .smcap    {font-variant: small-caps;}
  .u        {text-decoration: underline;}
  .bold     {font-weight: bold;}
+ .big { font-size: larger; }
+ .small { font-size: smaller; }
 
 ## include images from img tags in the body and summary of
 ## stories.  Images will be converted to jpg for size if possible.