fix epub for weird corner case

[librarian.git] / librarian / functions.py

diff --git a/librarian/functions.py b/librarian/functions.py
index 8427ba8..75e2911 100644 (file)
--- a/librarian/functions.py
+++ b/librarian/functions.py
@@ -6,14 +6,18 @@
 from lxml import etree
 import re
 
+from librarian.dcparser import Person
+from librarian import get_resource
+
+
 def _register_function(f):
     """ Register extension function with lxml """
     ns = etree.FunctionNamespace('http://wolnelektury.pl/functions')
     ns[f.__name__] = f
 
 
-def reg_substitute_entities(): 
-    ENTITY_SUBSTITUTIONS = [
+def reg_substitute_entities():
+    entity_substitutions = [
         (u'---', u'—'),
         (u'--', u'–'),
         (u'...', u'…'),
@@ -25,7 +29,7 @@ def reg_substitute_entities():
         """XPath extension function converting all entites in passed text."""
         if isinstance(text, list):
             text = ''.join(text)
-        for entity, substitutution in ENTITY_SUBSTITUTIONS:
+        for entity, substitutution in entity_substitutions:
             text = text.replace(entity, substitutution)
         return text
 
@@ -68,9 +72,9 @@ def reg_wrap_words():
             text = ''.join(text)
         if not wrapping:
             return text
-    
+
         words = re.split(r'\s', text)
-    
+
         line_length = 0
         lines = [[]]
         for word in words:
@@ -89,7 +93,56 @@ def reg_person_name():
         """ Converts "Name, Forename" to "Forename Name" """
         if isinstance(text, list):
             text = ''.join(text)
-        return ' '.join([t.strip() for t in text.split(',', 1)[::-1]])
+        return Person.from_text(text).readable()
     _register_function(person_name)
 
 
+def reg_texcommand():
+    def texcommand(context, text):
+        """Remove non-letters"""
+        if isinstance(text, list):
+            text = ''.join(text)
+        return re.sub(r'[^a-zA-Z]', '', text).strip()
+    _register_function(texcommand)
+
+
+def reg_lang_code_3to2():
+    def lang_code_3to2(context, text):
+        """Convert 3-letter language code to 2-letter code"""
+        result = ''
+        text = ''.join(text)
+        with open(get_resource('res/ISO-639-2_8859-1.txt'), 'rb') as f:
+            for line in f:
+                list = line.strip().split('|')
+                if list[0] == text:
+                    result = list[2]
+        if result == '':
+            return text
+        else:
+            return result
+    _register_function(lang_code_3to2)
+
+
+def mathml_latex(context, trees):
+    from librarian.embeds.mathml import MathML
+    text = MathML(trees[0]).to_latex().data
+    # Remove invisible multiplications, they produce unwanted spaces.
+    text = text.replace(u'\u2062', '')
+    return text
+
+
+def reg_mathml_latex():
+    _register_function(mathml_latex)
+
+
+def reg_mathml_epub(zipf):
+    from librarian.embeds.mathml import MathML
+
+    def mathml(context, trees):
+        data = MathML(trees[0]).to_latex().to_png().data
+        name = "math%d.png" % mathml.count
+        mathml.count += 1
+        zipf.writestr('OPS/' + name, data)
+        return name
+    mathml.count = 0
+    _register_function(mathml)