#1032: epubs for virtualo

[librarian.git] / librarian / pdf.py

diff --git a/librarian/pdf.py b/librarian/pdf.py
index 5debb68..1989239 100644 (file)
--- a/librarian/pdf.py
+++ b/librarian/pdf.py
@@ -21,7 +21,7 @@ from lxml.etree import XMLSyntaxError, XSLTApplyError
 
 from librarian.dcparser import Person
 from librarian.parser import WLDocument
-from librarian import ParseError, DCNS
+from librarian import ParseError, DCNS, get_resource
 from librarian import functions
 
 
@@ -36,7 +36,7 @@ STYLESHEETS = {
 }
 
 
-def insert_tags(doc, split_re, tagname):
+def insert_tags(doc, split_re, tagname, exclude=None):
     """ inserts <tagname> for every occurence of `split_re' in text nodes in the `doc' tree 
 
     >>> t = etree.fromstring('<a><b>A-B-C</b>X-Y-Z</a>');
@@ -46,6 +46,8 @@ def insert_tags(doc, split_re, tagname):
     """
 
     for elem in doc.iter(tag=etree.Element):
+        if exclude and elem.tag in exclude:
+            continue
         if elem.text:
             chunks = split_re.split(elem.text)
             while len(chunks) > 1:
@@ -67,13 +69,17 @@ def insert_tags(doc, split_re, tagname):
 def substitute_hyphens(doc):
     insert_tags(doc, 
                 re.compile("(?<=[^-\s])-(?=[^-\s])"),
-                "dywiz")
+                "dywiz",
+                exclude=[DCNS("identifier.url"), DCNS("rights.license")]
+                )
 
 
 def fix_hanging(doc):
     insert_tags(doc, 
                 re.compile("(?<=\s\w)\s+"),
-                "nbsp")
+                "nbsp",
+                exclude=[DCNS("identifier.url"), DCNS("rights.license")]
+                )
 
 
 def move_motifs_inside(doc):
@@ -131,9 +137,6 @@ def parse_creator(doc):
         creator.getparent().insert(0, creator_parsed)
 
 
-def get_resource(path):
-    return os.path.join(os.path.dirname(__file__), path)
-
 def get_stylesheet(name):
     return get_resource(STYLESHEETS[name])
 
@@ -276,9 +279,11 @@ def load_including_children(provider, slug=None, uri=None, file_path=None):
     else:
         raise ValueError('Neither slug, URI nor file path provided for a book.')
 
-    document = WLDocument.from_file(f, True,
-        parse_dublincore=True,
-        preserve_lines=False)
+    text = f.read().decode('utf-8')
+    text = re.sub(ur"([\u0400-\u04ff]+)", ur"<alien>\1</alien>", text)
+
+    document = WLDocument.from_string(text, True,
+        parse_dublincore=True)
 
     f.close()