1 files changed, 50 insertions, 0 deletions
diff --git a/lib/strutils.py b/lib/strutils.py
new file mode 100644
index 0000000..368d3d8
--- /dev/null
+++ b/lib/strutils.py
@@ -0,0 +1,50 @@
+
+#
+# String/unicode conversion utils.
+#
+    
+def safestr(s):
+    """
+    Safely corerce *anything* to a string. If the object can't be str'd, an
+    empty string will be returned.
+    
+    You can (and I do) use this for really crappy unicode handling, but it's
+    a bit like killing a mosquito with a bazooka.
+    """
+    if s is None:
+        return ""
+    if isinstance(s, unicode):
+        return s.encode('ascii', 'xmlcharrefreplace')
+    else:
+        try:
+            return str(s)
+        except:
+            return ""
+            
+def safeint(s):
+    """Like safestr(), but always returns an int. Returns 0 on failure."""
+    try:
+        return int(safestr(s))
+    except ValueError:
+        return 0
+
+
+def convertentity(m):
+    import htmlentitydefs
+    """Convert a HTML entity into normal string (ISO-8859-1)"""
+    if m.group(1)=='#':
+        try:
+            return chr(int(m.group(2)))
+        except ValueError:
+            return '&#%s;' % m.group(2)
+    try:
+        return htmlentitydefs.entitydefs[m.group(2)]
+    except KeyError:
+        return '&%s;' % m.group(2)
+
+def unquotehtml(s):
+    import re
+    """Convert a HTML quoted string into normal string (ISO-8859-1).
+   
+    Works with &#XX; and with &nbsp; &gt; etc."""
+    return re.sub(r'&(#?)(.+?);',convertentity,s)