add clone for file iterator

usc-isi-i2 · Apr 19, 2017 · 727dbdc · 727dbdc
1 parent 3bd1769
commit 727dbdc
Show file tree

Hide file tree

Showing 2 changed files with 19 additions and 2 deletions.
diff --git a/examples/file_iter.py b/examples/file_iter.py
@@ -14,3 +14,10 @@
                             type='csv', id_column='id', value_columns=['content']) # field_names=['xx', 'yy']
 for id, value in iter:
     print id, value
+
+iter2 = iter.copy()
+iter3 = iter.copy()
+for id, value in iter2:
+    print id, value
+for id, value in iter3:
+    print id, value
diff --git a/rltk/file_iterator.py b/rltk/file_iterator.py
@@ -6,13 +6,17 @@
 
 class FileIterator(object):
 
+    _file_path = None
+    _type = None
+    _kwargs = {}
     _file_handler = None
     _count = 0
-    _type = None
 
     def __init__(self, file_path, type='text', **kwargs):
-        self._file_handler = open(file_path, 'r')
+        self._file_path = file_path
         self._type = type
+        self._kwargs = kwargs
+        self._file_handler = open(file_path, 'r')
 
         if type == 'json_line':
             # pre-compile json path, raise exception if not exists
@@ -30,6 +34,12 @@ def __init__(self, file_path, type='text', **kwargs):
         else: # text
             self._id_prefix = hashlib.md5(file_path).hexdigest()[:6]
 
+    def __copy__(self):
+        return FileIterator(self._file_path, self._type, **self._kwargs)
+
+    def copy(self):
+        return self.__copy__()
+
     def next(self):
         """
         Returns: