attachment list for links + bad links in red

[redakcja.git] / apps / catalogue / xml_tools.py

diff --git a/apps/catalogue/xml_tools.py b/apps/catalogue/xml_tools.py
index 242714b..abab7f2 100644 (file)
--- a/apps/catalogue/xml_tools.py
+++ b/apps/catalogue/xml_tools.py
@@ -2,6 +2,7 @@
 from copy import deepcopy
 import re
 
+from django.utils.encoding import force_str
 from lxml import etree
 from catalogue.constants import TRIM_BEGIN, TRIM_END, MASTERS
 
@@ -14,7 +15,7 @@ class ParseError(BaseException):
 
 
 def _trim(text, trim_begin=True, trim_end=True):
-    """ 
+    """
         Cut off everything before RE_TRIM_BEGIN and after RE_TRIM_END, so
         that eg. one big XML file can be compiled from many small XML files.
     """
@@ -26,7 +27,7 @@ def _trim(text, trim_begin=True, trim_end=True):
 
 
 def compile_text(parts):
-    """ 
+    """
         Compiles full text from an iterable of parts,
         trimming where applicable.
     """
@@ -103,7 +104,7 @@ def split_xml(text):
         name_elem = deepcopy(element)
         for tag in 'extra', 'motyw', 'pa', 'pe', 'pr', 'pt', 'uwaga':
             for a in name_elem.findall('.//' + tag):
-                a.text=''
+                a.text = ''
                 del a[:]
         name = etree.tostring(name_elem, method='text', encoding='utf-8').strip()
 
@@ -123,15 +124,11 @@ def split_xml(text):
             while parent[0] is not element:
                 del parent[0]
             element, parent = parent, parent.getparent()
-        chunks[:0] = [[name,
-            unicode(etree.tostring(copied, encoding='utf-8'), 'utf-8')
-            ]]
+        chunks[:0] = [[name, unicode(etree.tostring(copied, encoding='utf-8'), 'utf-8')]]
 
         parts = src.findall('.//naglowek_rozdzial')
 
-    chunks[:0] = [[u'początek',
-        unicode(etree.tostring(src, encoding='utf-8'), 'utf-8')
-        ]]
+    chunks[:0] = [[u'początek', unicode(etree.tostring(src, encoding='utf-8'), 'utf-8')]]
 
     for ch in chunks[1:]:
         ch[1] = add_trim_begin(ch[1])
@@ -139,3 +136,105 @@ def split_xml(text):
         ch[1] = add_trim_end(ch[1])
 
     return chunks
+
+
+def wl2_to_wl1(wl2_xml, slug):
+    from lxml import etree
+    import re
+    from StringIO import StringIO
+    from urllib import unquote
+    import os.path
+    from django.conf import settings
+    from fnpdjango.utils.text.slughifi import slughifi
+    from librarian import ParseError, DCNS
+
+    def _register_function(f):
+        """ Register extension function with lxml """
+        ns = etree.FunctionNamespace('http://wolnelektury.pl/functions')
+        ns[f.__name__] = f
+        return f
+
+    @_register_function
+    def slugify(context, text):
+        """Remove unneeded whitespace from beginning and end"""
+        if isinstance(text, list):
+            text = ''.join(text)
+        return slughifi(text)
+
+    @_register_function
+    def rmext(context, text):
+        if isinstance(text, list):
+            text = ''.join(text)
+        text = unquote(text)
+        if '.' in text:
+            name, ext = text.rsplit('.', 1)
+            if ext.lower() in ('doc', 'docx', 'odt', 'pdf', 'jpg', 'jpeg'):
+                text = name
+        return text
+
+    t = etree.parse(os.path.join(settings.PROJECT_ROOT, 'xslt/wl2to1.xslt'))
+    ft = wl2_xml.replace(' ', ' ')
+    f2 = StringIO(ft)
+    i1 = etree.parse(f2)
+
+    for sect in i1.findall('//section'):
+        if sect[0].text and sect[0].text.strip() == u'Przebieg zajęć':
+            # Prostujemy.
+            first = sect.find('section')
+            subs = first.findall('.//section')
+            for sub in subs:
+                sect.append(sub)
+            break
+    else:
+        dc_type = i1.findall('//dc:type', namespaces={'dc': DCNS.uri})
+        if dc_type and dc_type[0] in ('course', 'synthetic'):
+            raise ParseError('Brak przebiegu')
+
+    i1.getroot().attrib['redslug'] = slug
+    i1.getroot().attrib['wlslug'] = slug  # THIS!
+    w1t = i1.xslt(t)
+    for h in w1t.findall('//aktywnosc/opis'):
+        if len(h) == 0:
+            raise ParseError('Pusty element aktywnosc/opis')
+        # FIXME assumption that every lesson has at most 9 parts
+        if not h[0].text or not re.match(r'\d\.\s', h[0].text):
+            raise ParseError('Niepoprawny nagłówek (aktywnosc/opis): %s' % repr(h[0].text))
+        h[0].text = h[0].text[3:]
+    return etree.tostring(w1t, encoding='utf-8')
+
+
+EXCEPTIONS = [
+    ('div', 'img'),
+    ('div', 'video'),
+    ('div', 'table.cell'),
+]
+
+
+def remove_element(element):
+    parent = element.getparent()
+    tail = element.tail
+    if tail:
+        prev = element.getprevious()
+        if prev is not None:
+            prev.tail = (prev.tail or '') + tail
+        else:
+            parent.text = (parent.text or '') + tail
+    parent.remove(element)
+
+
+def remove_empty_elements(xml):
+    try:
+        tree = etree.fromstring(force_str(xml.replace(' ', u'\xa0')))
+    except SyntaxError:
+        return None
+    changed = False
+    another_loop = True
+    while another_loop:
+        another_loop = False
+        for element in tree.findall('.//*'):
+            if (not element.text or not element.text.strip()) and len(element) == 0:
+                if (element.tag, element.attrib.get('class')) not in EXCEPTIONS:
+                    remove_element(element)
+                    changed = True
+                    another_loop = True
+    return etree.tostring(tree, encoding=unicode) if changed else None