Utility pro DAKOS import

seminar/dksdump/helpers.py

@@ -0,0 +1,22 @@
+from seminar import ovvpfile
+import os
+
+
+def read_all_tables(basedir):
+    tables = {}
+    for fn in os.listdir(basedir):
+        if fn.endswith('.csv'):
+            print "Reading %s ..." % (fn, )
+            with open(os.path.join(basedir, fn), 'r') as f:
+                o = ovvpfile.parse(f, with_headers=False)
+            tables[fn[:-4]] = o.rows
+            print "  %d lines, columns: %s" % (len(o.rows), ' '.join(o.columns), )
+    return tables
+
+def matchrows(tab, key, val):
+    return [r for r in tab if r[key]==val]
+
+def onerow(tab, key, val):
+    t = matchrows(tab, key, val)
+    assert(len(t) == 1)
+    return t[0]

seminar/ovvpfile.py

@@ -20,7 +20,7 @@ class OvvpFile(object):
     def to_string(self):
         return ''.join(self.to_lines())
 
-    def parse_from(self, source):
+    def parse_from(self, source, with_headers=True):
         "Parse data from file, string or line iterator, overwriting self"
         if isinstance(source, str) or isinstance(source, unicode):
             return self.parse_from(source.split('\n'))
@@ -29,15 +29,16 @@ class OvvpFile(object):
 
         # header
         self.headers = {}
-        for r in it:
+        if with_headers:
-            if isinstance(r, str):
+            for r in it:
-                r = r.decode('utf8')
+                if isinstance(r, str):
-            assert isinstance(r, unicode)
+                    r = r.decode('utf8')
-            r = r.rstrip('\n')
+                assert isinstance(r, unicode)
-            if r == u"":
+                r = r.rstrip('\n')
-                break
+                if r == u"":
-            k, v = r.split(u'\t', 1)
+                    break
-            self.headers[k] = v
+                k, v = r.split(u'\t', 1)
+                self.headers[k] = v
 
         # columns
         r = it.next()
@@ -61,7 +62,7 @@ class OvvpFile(object):
         
 
 
-def parse(source):
+def parse(source, with_headers=True):
     o = OvvpFile()
-    o.parse_from(source)
+    o.parse_from(source, with_headers=with_headers)
     return o