fix #6: allow update existing table or skipping it by perform no action on existing data

Author: chfw

pyexcel_io/__init__.py

@@ -21,7 +21,7 @@
 )
 from .csvbook import CSVBook, CSVWriter
 from .csvzipbook import CSVZipWriter, CSVZipBook
-from .sqlbook import SQLBookReader, SQLBookWriter
+from .sqlbook import SQLBookReader, SQLBookWriter, PyexcelSQLSkipRowException
 from .djangobook import DjangoBookReader, DjangoBookWriter
 from ._compact import is_string, BytesIO, StringIO, isstream, OrderedDict, PY2
 from .constants import (

pyexcel_io/sqlbook.py

@@ -10,7 +10,8 @@
 from ._compact import OrderedDict
 from .constants import (
     MESSAGE_INVALID_PARAMETERS,
-    MESSAGE_EMPTY_ARRAY
+    MESSAGE_EMPTY_ARRAY,
+    MESSAGE_IGNORE_ROW
 )
 from .base import (
     BookReaderBase,
@@ -23,6 +24,14 @@
 )
 
 
+class PyexcelSQLSkipRowException(Exception):
+    """
+    Raised this exception to skipping a row
+    while data import
+    """
+    pass
+
+
 class SQLTableReader(SheetReaderBase):
     """Read a table
     """
@@ -83,21 +92,28 @@ def write_row(self, array):
             print(MESSAGE_EMPTY_ARRAY)
         else:
             new_array = swap_empty_string_for_none(array)
-            self._write_row(new_array)
+            try:
+                self._write_row(new_array)
+            except PyexcelSQLSkipRowException:
+                print(MESSAGE_IGNORE_ROW)
+                print(new_array)
 
     def _write_row(self, array):
         row = dict(zip(self.column_names, array))
+        obj = None
         if self.initializer:
-            o = self.initializer(row)
-        else:
-            o = self.table()
+            # allow initinalizer to return None
+            # if skipping is needed
+            obj = self.initializer(row)
+        if obj is None:
+            obj = self.table()
             for name in self.column_names:
                 if self.mapdict is not None:
                     key = self.mapdict[name]
                 else:
                     key = name
-                setattr(o, key, row[name])
-        self.session.add(o)
+                setattr(obj, key, row[name])
+        self.session.add(obj)
 
     def close(self):
         if self.auto_commit:

tests/test_sql_book.py

@@ -8,7 +8,8 @@
     SQLBookReader,
     SQLBookWriter,
     from_query_sets,
-    OrderedDict
+    OrderedDict,
+    PyexcelSQLSkipRowException
 )
 from pyexcel_io.sqlbook import SQLTableReader, SQLTableWriter
 from sqlalchemy.orm import relationship, backref
@@ -121,6 +122,72 @@ def test_one_table(self):
         assert results == self.results
         mysession.close()
 
+    def test_update_existing_row(self):
+        mysession = Session()
+        # write existing data
+        writer = SQLTableWriter(mysession,
+                                [Pyexcel,self.data[0], None, None])
+        writer.write_array(self.data[1:])
+        writer.close()
+        query_sets=mysession.query(Pyexcel).all()
+        results = from_query_sets(self.data[0], query_sets)
+        assert results == self.results
+        # update data using custom initializer
+        update_data = [
+            ['birth', 'id', 'name', 'weight'],
+            [datetime.date(2014, 11, 11), 0, 'Adam_E', 12.25],
+            [datetime.date(2014, 11, 12), 1, 'Smith_E', 11.25]
+        ]
+        updated_results = [
+            ['birth', 'id', 'name', 'weight'],
+            ['2014-11-11', 0, 'Adam_E', 12.25],
+            ['2014-11-12', 1, 'Smith_E', 11.25]
+        ]
+        def row_updater(row):
+            an_instance = mysession.query(Pyexcel).get(row['id'])
+            if an_instance is None:
+                an_instance = Pyexcel()
+            for name in row.keys():
+                setattr(an_instance, name, row[name])
+            return an_instance
+        writer = SQLTableWriter(mysession,
+                                [Pyexcel,update_data[0], None, row_updater])
+        writer.write_array(update_data[1:])
+        writer.close()
+        query_sets=mysession.query(Pyexcel).all()
+        results = from_query_sets(self.data[0], query_sets)
+        assert results == updated_results
+        mysession.close()
+
+    def test_skipping_rows_if_data_exist(self):
+        mysession = Session()
+        # write existing data
+        writer = SQLTableWriter(mysession,
+                                [Pyexcel,self.data[0], None, None])
+        writer.write_array(self.data[1:])
+        writer.close()
+        query_sets=mysession.query(Pyexcel).all()
+        results = from_query_sets(self.data[0], query_sets)
+        assert results == self.results
+        # update data using custom initializer
+        update_data = [
+            ['birth', 'id', 'name', 'weight'],
+            [datetime.date(2014, 11, 11), 0, 'Adam_E', 12.25],
+            [datetime.date(2014, 11, 12), 1, 'Smith_E', 11.25]
+        ]
+        def row_updater(row):
+            an_instance = mysession.query(Pyexcel).get(row['id'])
+            if an_instance is not None:
+                raise PyexcelSQLSkipRowException()
+        writer = SQLTableWriter(mysession,
+                                [Pyexcel,update_data[0], None, row_updater])
+        writer.write_array(update_data[1:])
+        writer.close()
+        query_sets=mysession.query(Pyexcel).all()
+        results = from_query_sets(self.data[0], query_sets)
+        assert results == self.results
+        mysession.close()
+
     def test_one_table_with_empty_rows(self):
         mysession = Session()
         data = [