--- cdli/cdli_files.py	2008/01/07 16:54:46	1.80.2.14
+++ cdli/cdli_files.py	2008/01/14 18:43:21	1.80.2.15
@@ -99,7 +99,7 @@ def formatAtfFullLineNum(txt, nolemma=Tr
     ret = []
     surf = ""
     col = ""
-    for line in txt.split("\n"):
+    for line in txt.splitlines():
         line = unicodify(line)
         if line and line[0] == '@':
             # surface or column
@@ -632,7 +632,7 @@ class CDLIBasketContainer(OrderedFolder)
                             ret+=str(object[0].getData())+"\n"
             elif current=="yes":
                             #search current object
-                            logging.info("crrent: %s"%object[1].getId().split(".")[0])
+                            logging.debug("current: %s"%object[1].getId().split(".")[0])
                             founds=self.CDLICatalog.search({'title':object[1].getId().split(".")[0]})
                             if len(founds)>0:      
                                 ret+=str(founds[0].getObject().getLastVersion().getData())+"\n"
@@ -1791,6 +1791,7 @@ def splitatf(fh,dir=None,ext=None):
     nf=None
     i=0
 
+    #ROC: why split \n first and then \r???
     if (type(fh) is StringType) or (type(fh) is UnicodeType):
         iter=fh.split("\n")
     else:
@@ -2191,7 +2192,7 @@ class CDLIRoot(Folder):
         # compile into regexp objects and escape parens
         wordlist = [re.compile(splitexp%re.escape(w)) for w in word.split(' ')]
             
-        for line in file.split("\n"):
+        for line in file.splitlines():
             for word in wordlist:
                 #logging.debug("showwordinfile: searching for %s in %s"%(word.pattern,ignoreable.sub('',line)))
                 if word.search(ignorable.sub('',line)):
@@ -2236,11 +2237,20 @@ class CDLIRoot(Folder):
         # split search terms again (for grapheme search with words)
         splitwords = dict(((w,self.splitter[indexName].process([w])) for w in words))
             
-        for line in file.split("\n"):
+        for line in file.linesplit():
             line = unicodify(line)
-            # ignore lemma lines
+            # ignore lemma and other lines
             if line.lstrip().startswith('#lem:'):
                 continue
+            # ignore p-num line
+            if line.startswith('&P'):
+                continue
+            # ignore version lines
+            if line.startswith('#version'):
+                continue
+            # ignore atf type lines
+            if line.startswith('#atf:'):
+                continue
 
             # first scan
             hitwords = []