epub converter: add footnotes, list processing

src/epub_converter.py

@@ -13,9 +13,17 @@ from ebooklib.utils import debug
 
 from src.data_objects import ChapterItem, NavPoint
 from src.html_epub_preprocessor import unwrap_structural_tags, get_tags_between_ids, prepare_title_and_content, \
-    preprocess_image
+    preprocess_image, preprocess_footnotes
 
 
+# epub3 examples:
+# https://github.com/IDPF/epub3-samples
+# specification:
+# https://idpf.github.io/epub-vocabs/structure/
+# footnotes:
+# http://www.theheratik.net/books/tech-epub/chapter-8/
+# http://kb.daisy.org/publishing/docs/html/epub-type.html
+# todo: http://kb.daisy.org/publishing/docs/html/notes.html
 # todo: https://docs.python.org/3/howto/unicode.html
 
 
@@ -34,6 +42,10 @@ class EpubBookAdapter:
 
         self.id_anchor_exist_in_nav_points = False
         self.href2soup_html: Dict[str, BeautifulSoup] = self.build_href2soup_content()
+        self.footnotes = []
+        for href in self.href2soup_html:
+            self.footnotes.extend(preprocess_footnotes(self.href2soup_html[href], self.href2soup_html,
+                                                       noteref_attr_name='data-type'))
         # если в content.opf есть в spine toc атрибут  -> можно найти ncx файл -> из него достать navMap
         # если его там нет, пробуют искать nav tag в manifest -> EpubNav. это у epub3 (не тестировалось todo)
         self.href2ids = defaultdict(list)
@@ -71,8 +83,6 @@ class EpubBookAdapter:
 
     def build_adjacency_list_from_toc(self, element, lvl=0):
         # use book.toc as a root
-        # todo: read _create_section in get_nav
-        # todo: try list on hrefs, extra info in another db
 
         if isinstance(element, Link):
             # todo: check if link exists
@@ -210,7 +220,8 @@ if __name__ == "__main__":
     l = [x.to_dict() for x in top_level_chapters]
 
     tmp = {
-        "content": l
+        "content": l,
+        "footnotes": adapter.footnotes
     }
 
     with codecs.open('tmp.json', 'w', encoding='utf-8') as f: