Prepare version 0.2.11 (#210)

* add new api search_by_ids Signed-off-by: Yhz <[email protected]> * update example Signed-off-by: Yhz <[email protected]> * fix typo Signed-off-by: Yhz <[email protected]> * fix search by ids error Signed-off-by: Yhz <[email protected]> * add search_by_ids test case Signed-off-by: Yhz <[email protected]> * use connection pool Signed-off-by: yhz <[email protected]> * async api pass Signed-off-by: Yhz <[email protected]> * add async examples Signed-off-by: Yhz <[email protected]> * Fix TypeError raised in inserting vectors with async set true (#200) Signed-off-by: Yhz <[email protected]> * add timeout for async APIs Signed-off-by: Yhz <[email protected]> * add API close Signed-off-by: Yhz <[email protected]> * optimzie code Signed-off-by: Yhz <[email protected]> * remove terminate() in pool Signed-off-by: Yhz <[email protected]> * deprated connect Signed-off-by: Yhz <[email protected]> * get collection info failed when info is null (fix #201) Signed-off-by: yhz <[email protected]> * fix Signed-off-by: yhz <[email protected]> * update http handler Signed-off-by: yhz <[email protected]> * .. Signed-off-by: yhz <[email protected]> * filter -1 result Signed-off-by: Yhz <[email protected]> * Update APIs names Signed-off-by: yhz <[email protected]> * update future done not raise exception Signed-off-by: Yhz <[email protected]> * update Signed-off-by: yhz <[email protected]> * Fix excepted exceptions if async invoke with timeout (#207) Signed-off-by: yhz <[email protected]> * remove api compacity Signed-off-by: yhz <[email protected]> * add more params in pool Signed-off-by: yhz <[email protected]> * add pool try_connect and pre_ping params Signed-off-by: yhz <[email protected]> * remove parms recycle Signed-off-by: yhz <[email protected]> * update for http handler Signed-off-by: Yhz <[email protected]> * update example Signed-off-by: yhz <[email protected]> * update readme Signed-off-by: yhz <[email protected]> * update Signed-off-by: yhz <[email protected]> * rename project name Signed-off-by: yhz <[email protected]> * Prepare for v0.2.11 Signed-off-by: yhz <[email protected]> * update async example Signed-off-by: yhz <[email protected]>
milvus-io · May 15, 2020 · bb79982 · bb79982
1 parent 1631277
commit bb79982
Show file tree

Hide file tree

Showing 33 changed files with 3,090 additions and 906 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,7 +1,22 @@
-# pymilvus 0.2.11(TBD)
+# pymilvus 0.2.11(2020-05-15)
 
 ## Bug
 - \#190 - Fix rpc return error when error occur
+- \#200 - Fix TypeError raised in inserting vectors with async set true
+- \#201 - Fix get collection info failed when info is null
+- \#202 - Fix async 'NoneType' object has no attribute 'result'
+- \#207 - Fix excepted exceptions if async invoke with timeout
+
+## Improvement
+- \#175 - Remove connect API
+- \#203 - Add method 'is_done()' in Future class
+- \#205 - Filter search results which id is -1
+- \#206 - Update APIs names
+
+## New Feature
+- \#174 - Support connection pool
+- \#177 - Support async API
+- \#199 - Add API 'has_partition'
 
 
 # pymilvus 0.2.10(2020-04-15)

diff --git a/README.md b/README.md
@@ -6,7 +6,7 @@
 
 Python SDK for [Milvus](https://github.com/milvus-io/milvus). To contribute code to this project, please read our [contribution guidelines](https://github.com/milvus-io/milvus/blob/master/CONTRIBUTING.md) first.
 
-For detailed SDK documentation, refer to [API Documentation](https://milvus-io.github.io/milvus-sdk-python/pythondoc/v0.2.10/index.html).
+For detailed SDK documentation, refer to [API Documentation](https://milvus-io.github.io/milvus-sdk-python/pythondoc/v0.2.11/index.html).
 
 
 <!-- TOC -->
@@ -69,12 +69,13 @@ The following collection shows Milvus versions and recommended pymilvus versions
 | 0.7.0 | 0.2.8 |
 | 0.7.1 | 0.2.9 |
 | 0.8.0 | 0.2.10 |
+| 0.9.0 | 0.2.11 |
 
 
 You can install a specific version of pymilvus by:
 
 ```shell
-$ pip install pymilvus==0.2.10
+$ pip install pymilvus==0.2.11
 ```
 
 You can upgrade `pymilvus` to the latest version by:
@@ -98,12 +99,11 @@ Refer to [examples](/examples) for more example programs.
    >>> from milvus import Milvus, IndexType, MetricType, Status
    ```
 
-2. Connect to Milvus server by using one of the following methods:
+2. Create a client to Milvus server by using one of the following methods:
 
    ```python
    # Connect to Milvus server
    >>> client = Milvus(host='localhost', port='19530')
-   >>> client.connect()
    ```
 
    > Note: In the above code, default values are used for `host` and `port` parameters. Feel free to change them to the IP address and port you set for Milvus server.
@@ -148,11 +148,11 @@ You can split collections into partitions by partition tags for improved search
 >>> client.create_partition(collection_name='test01', partition_tag='tag01')
 ```
 
-Use `show_partitions()` to verify whether the partition is created.
+Use `list_partitions()` to verify whether the partition is created.
 
 ```python
 # Show partitions
->>> client.show_partitions(collection_name='test01')
+>>> client.list_partitions(collection_name='test01')
 ```
 
 ### Drop a partition
@@ -204,40 +204,34 @@ Use `show_partitions()` to verify whether the partition is created.
 
    ```python
    # Insert vectors
-   >>> client.insert(collection_name='test01', records=vectors)
+   >>> status, inserted_vector_ids = client.insert(collection_name='test01', records=vectors)
    ```
 
    Alternatively, you can also provide user-defined vector ids:
 
    ```python
    >>> vector_ids = [id for id in range(20)]
-   >>> client.insert(collection_name='test01', records=vectors, ids=vector_ids)
+   >>> status, inserted_vector_ids = client.insert(collection_name='test01', records=vectors, ids=vector_ids)
    ```
 
 ### Insert vectors in a partition
 
 ```python
->>> client.insert('test01', vectors, partition_tag="tag01")
+>>> status, inserted_vector_ids = client.insert('test01', vectors, partition_tag="tag01")
 ```
 
-To verify the vectors you have inserted, use `get_vector_by_id()`. Assume you have a vector with the following ID.
+To verify the vectors you have inserted, use `get_vector_by_id()`. Assume you have vector with the following ID.
 
 ```python
->>> status, vector = client.get_vector_by_id(collection_name='test01', vector_id=0)
+>>> status, vector = client.get_entity_by_id(collection_name='test01', ids=inserted_vector_ids[:10])
 ```
 
 ### Delete vectors by ID
 
-Assume you have some vectors with the following IDs:
-
-```python
->>> ids = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]
-```
-
 You can delete these vectors by:
 
 ```python
->>> client.delete_by_id('test01', ids)
+>>> client.delete_entity_by_id('test01', inserted_vector_ids[:10])
 ```
 
 ## Flush data in one or multiple collections to disk
@@ -285,8 +279,8 @@ A segment is a data file that Milvus automatically creates by merging inserted v
 
 > Note: If you do not specify `partition_tags`, Milvus searches the whole collection.
 
-## Disconnect from the Milvus server
+## close client
 
 ```python
->>> client.disconnect()
+>>> client.close()
 ```
diff --git a/conftest.py b/conftest.py
@@ -39,34 +39,35 @@ def connect(request, ghandler):
     port_default = default_http_port if handler == "HTTP" else default_grpc_port
     port = request.config.getoption("--port", default=port_default)
 
-    milvus = Milvus(handler=ghandler)
-    milvus.connect(host=ip, port=port)
+    client = Milvus(host=ip, port=port, handler=ghandler)
+    # milvus.connect()
 
     def fin():
         try:
-            milvus.disconnect()
+            client.close()
         except:
             pass
 
     request.addfinalizer(fin)
-    return milvus
+    return client
 
 
 @pytest.fixture(scope="module")
 def gcon(request, ghandler):
     ip = request.config.getoption("--ip")
     port = request.config.getoption("--port")
-    milvus = Milvus(host=ip, port=port, handler=ghandler)
+    client = Milvus(host=ip, port=port, handler=ghandler)
 
     def fin():
         try:
+            client.close()
             pass
         except Exception as e:
             print(e)
             pass
 
     request.addfinalizer(fin)
-    return milvus
+    return client
 
 
 @pytest.fixture(scope="module")
@@ -111,7 +112,7 @@ def gcollection(request, gcon):
     gcon.flush([table_name])
 
     def teardown():
-        status, table_names = gcon.show_collections()
+        status, table_names = gcon.list_collections()
         for name in table_names:
             gcon.drop_collection(name)
 

diff --git a/examples/example.py b/examples/example.py
@@ -2,20 +2,25 @@
 # create a vector collection,
 # insert 10 vectors, 
 # and execute a vector similarity search.
+
 import datetime
 import sys
 
 sys.path.append(".")
 import random
 import threading
 import time
+
+import numpy as np
+
 from milvus import Milvus, IndexType, MetricType, Status
 from milvus.client.abstract import TopKQueryResult
 
 # Milvus server IP address and port.
 # You may need to change _HOST and _PORT accordingly.
 _HOST = '127.0.0.1'
 _PORT = '19530'  # default value
+# _PORT = '19121'  # default http value
 
 # Vector parameters
 _DIM = 128  # dimension of vector
@@ -25,10 +30,9 @@
 
 def main():
     # Specify server addr when create milvus client instance
-    milvus = Milvus(_HOST, _PORT)
-
-    # Check if server is accessible
-    milvus.connect()
+    # milvus client instance maintain a connection pool, param
+    # `pool_size` specify the max connection num.
+    milvus = Milvus(_HOST, _PORT, pool_size=10)
 
     # Create collection demo_collection if it dosen't exist.
     collection_name = 'example_collection_'
@@ -45,22 +49,18 @@ def main():
         milvus.create_collection(param)
 
     # Show collections in Milvus server
-    _, collections = milvus.show_collections()
-
-    # present collection info
-    _, info = milvus.collection_info(collection_name)
-    print(info)
+    _, collections = milvus.list_collections()
 
     # Describe demo_collection
-    _, collection = milvus.describe_collection(collection_name)
+    _, collection = milvus.get_collection_info(collection_name)
     print(collection)
 
-    # 10000 vectors with 16 dimension
+    # 10000 vectors with 128 dimension
     # element per dimension is float32 type
     # vectors should be a 2-D array
     vectors = [[random.random() for _ in range(_DIM)] for _ in range(10000)]
     # You can also use numpy to generate random vectors:
-    #     `vectors = np.random.rand(10000, 16).astype(np.float32)`
+    #    vectors = np.random.rand(10000, _DIM).astype(np.float32)
 
     # Insert vectors into demo_collection, return status and vectors id list
     status, ids = milvus.insert(collection_name=collection_name, records=vectors)
@@ -69,7 +69,14 @@ def main():
     milvus.flush([collection_name])
 
     # Get demo_collection row count
-    status, result = milvus.count_collection(collection_name)
+    status, result = milvus.count_entities(collection_name)
+
+    # present collection statistics info
+    _, info = milvus.get_collection_stats(collection_name)
+    print(info)
+
+    # Obtain raw vectors by providing vector ids
+    status, result_vectors = milvus.get_entity_by_id(collection_name, ids[:10])
 
     # create index of vectors, search more rapidly
     index_param = {
@@ -83,7 +90,7 @@ def main():
     status = milvus.create_index(collection_name, IndexType.IVF_FLAT, index_param)
 
     # describe index, get information of index
-    status, index = milvus.describe_index(collection_name)
+    status, index = milvus.get_index_info(collection_name)
     print(index)
 
     # Use the top 10 vectors for similarity search