[gen] Added the possiblity to extract in a clean way text from XHTML field values.

gen/indexer.py

@@ -0,0 +1,49 @@
+'''This file defines code for extracting, from field values, the text to be
+   indexed.'''
+
+# ------------------------------------------------------------------------------
+from Products.ZCTextIndex.PipelineFactory import element_factory
+from appy.shared.xml_parser import XmlParser
+from appy.shared.utils import normalizeString
+
+# ------------------------------------------------------------------------------
+class XhtmlTextExtractor(XmlParser):
+    '''Extracts text from XHTML.'''
+    def startDocument(self):
+        XmlParser.startDocument(self)
+        self.res = []
+
+    def endDocument(self):
+        self.res = ' '.join(self.res)
+        return XmlParser.endDocument(self)
+
+    def characters(self, content):
+        c = normalizeString(content, usage='extractedText').strip().lower()
+        if len(c) > 1: self.res.append(c)
+        return self.env
+
+    # Do not raise exceptions when errors occur.
+    def error(self, error): pass
+    def fatalError(self, error): pass
+    def warning(self, error): pass
+
+# ------------------------------------------------------------------------------
+class XhtmlIndexer:
+    '''Extracts, from XHTML field values, the text to index.'''
+    def process(self, text):
+        # Wrap the XHTML chunk into a root tag, to get valid XML.
+        text = '<p>%s</p>' % text[0]
+        parser = XhtmlTextExtractor()
+        text = parser.parse(text)
+        res = text.split(' ')
+        # Remove tokens of a single char.
+        i = len(res)-1
+        while i > -1 :
+            if (len(res[i]) < 2) and not res[i].isdigit():
+                del res[i]
+            i -= 1
+        return res
+
+# ------------------------------------------------------------------------------
+element_factory.registerFactory('XHTML indexer', 'XHTML indexer', XhtmlIndexer)
+# ------------------------------------------------------------------------------

shared/utils.py

@@ -206,6 +206,9 @@ def normalizeString(s, usage='fileName'):
     # We work in unicode. Convert p_s to unicode if not unicode.
     if isinstance(s, str):           s = s.decode('utf-8')
     elif not isinstance(s, unicode): s = unicode(s)
+    if usage == 'extractedText':
+        # Replace single quotes with blanks.
+        s = s.replace("'", " ").replace(u'’', ' ')
     # Remove any special char like accents.
     s = unicodedata.normalize('NFKD', s).encode('ascii', 'ignore')
     # Remove any other char, depending on p_usage.

shared/xml_parser.py

@@ -19,8 +19,7 @@
 
 # ------------------------------------------------------------------------------
 import xml.sax, difflib, types, cgi
-from xml.sax.handler import ContentHandler, ErrorHandler, feature_external_ges,\
+from xml.sax.handler import ContentHandler, ErrorHandler, feature_external_ges
-                            property_interning_dict
 from xml.sax.xmlreader import InputSource
 from xml.sax import SAXParseException
 from appy.shared import UnicodeBuffer, xmlPrologue
@@ -170,6 +169,13 @@ class XmlParser(ContentHandler, ErrorHandler):
             # Put a question mark instead of raising an exception.
             self.characters('?')
 
+    # ErrorHandler methods ---------------------------------------------------
+    # Define methods below in your subclass if you want error handling that
+    # does not raise exceptions, but produces a partial result instead.
+    #def error(self, error): pass
+    #def fatalError(self, error): pass
+    #def warning(self, error): pass
+
     def parse(self, xml, source='string'):
         '''Parses a XML stream.
            * If p_source is "string", p_xml must be a string containing