Refactoring module structure and building (#15)

Author: yuzelin

.github/workflows/paimon-python-checks.yml

@@ -22,10 +22,11 @@ on:
   push:
   pull_request:
     paths-ignore:
-      - 'dev/**'
-      - 'java_based_implementation/paimon-python-java-bridge/**'
       - '**/*.md'
 
+env:
+  JDK_VERSION: 8
+
 concurrency:
   group: ${{ github.workflow }}-${{ github.event_name }}-${{ github.event.number || github.run_id }}
   cancel-in-progress: true
@@ -37,6 +38,11 @@ jobs:
     steps:
       - name: Checkout code
         uses: actions/checkout@v2
+      - name: Set up JDK ${{ env.JDK_VERSION }}
+        uses: actions/setup-java@v2
+        with:
+          java-version: ${{ env.JDK_VERSION }}
+          distribution: 'adopt'
       - name: Run lint-python.sh
         run: |
           chmod +x dev/lint-python.sh

.gitignore

@@ -15,4 +15,4 @@ dev/.stage.txt
 dev/download
 dev/log
 dist/
-paimon_python.egg-info/
+*.egg-info/

README.md

@@ -22,30 +22,28 @@ This repo is for Apache Paimon Python SDK.
 You can verify the setup by right-clicking on any file or folder in the flink-python project
 and running "External Tools" → "flake8".
 
-# Usage
-
-## Java-Based Implementation
+## Check
 
-We can use `py4j` to leverage Java code to read Paimon data. This section describes how to use this implementation.
+We provide script to check codes.
 
-### Set Environment Variables
+```shell
+./dev/lint-python.sh    # execute all checks
+./dev/lint-python.sh -h # run this to see more usages         
+```
 
-`py4j` need to access a JVM, so we should set JVM arguments (optional) and Java classpath. A convenient way is using
-`os` packages to set environment variables which only affect current process.
+## Build
 
-```python
-import os
+We provide script to build wheel.
 
-os.environ['PYPAIMON_JAVA_CLASSPATH'] = '/path/to/dependent_jars/*'
-os.environ['_PYPAIMON_JVM_ARGS'] = 'jvm_arg1 jvm_arg2 ...'
+```shell
+./dev/build-wheels.sh
 ```
 
-NOTE: the package has set paimon core and hadoop dependencies. If you just test in local or run code in hadoop, you doesn't
-need to set classpath. If you need other dependencies such as OSS/S3 filesystem jars, or special catalog which isn't implemented 
-in paimon core, please download jars and set classpath.
+The target wheel is under `dist/`
+
+# Usage
 
-# API Reference
-TODO
+See Apache Paimon Python API [Doc](https://paimon.apache.org/docs/master/program-api/python-api/).
 
 
 

paimon_python_api/__init__.py

@@ -15,3 +15,28 @@
 #  See the License for the specific language governing permissions and
 # limitations under the License.
 #################################################################################
+
+from .split import Split
+from .table_read import TableRead
+from .table_scan import TableScan, Plan
+from .read_builder import ReadBuilder
+from .commit_message import CommitMessage
+from .table_commit import BatchTableCommit
+from .table_write import BatchTableWrite
+from .write_builder import BatchWriteBuilder
+from .table import Table
+from .catalog import Catalog
+
+__all__ = [
+    'Split',
+    'TableRead',
+    'TableScan',
+    'Plan',
+    'ReadBuilder',
+    'CommitMessage',
+    'BatchTableCommit',
+    'BatchTableWrite',
+    'BatchWriteBuilder',
+    'Table',
+    'Catalog'
+]

paimon_python_api/catalog.py

@@ -17,7 +17,7 @@
 #################################################################################
 
 from abc import ABC, abstractmethod
-from paimon_python_api.table import Table
+from paimon_python_api import Table
 
 
 class Catalog(ABC):

paimon_python_api/read_builder.py

@@ -17,8 +17,7 @@
 #################################################################################
 
 from abc import ABC, abstractmethod
-from paimon_python_api.table_read import TableRead
-from paimon_python_api.table_scan import TableScan
+from paimon_python_api import TableRead, TableScan
 from typing import List
 
 

paimon_python_api/table.py

@@ -17,8 +17,7 @@
 #################################################################################
 
 from abc import ABC, abstractmethod
-from paimon_python_api.read_builder import ReadBuilder
-from paimon_python_api.write_builder import BatchWriteBuilder
+from paimon_python_api import ReadBuilder, BatchWriteBuilder
 
 
 class Table(ABC):

paimon_python_api/table_commit.py

@@ -17,7 +17,7 @@
 #################################################################################
 
 from abc import ABC, abstractmethod
-from paimon_python_api.commit_message import CommitMessage
+from paimon_python_api import CommitMessage
 from typing import List
 
 

paimon_python_api/table_read.py

@@ -16,15 +16,16 @@
 # limitations under the License.
 #################################################################################
 
+import pyarrow as pa
+
 from abc import ABC, abstractmethod
-from pyarrow import RecordBatchReader
-from paimon_python_api.split import Split
+from paimon_python_api import Split
 from typing import List
 
 
 class TableRead(ABC):
     """To read data from data splits."""
 
     @abstractmethod
-    def create_reader(self, splits: List[Split]) -> RecordBatchReader:
+    def create_reader(self, splits: List[Split]) -> pa.RecordBatchReader:
         """Return a reader containing batches of pyarrow format."""

paimon_python_api/table_scan.py

@@ -18,7 +18,7 @@
 
 from abc import ABC, abstractmethod
 from typing import List
-from paimon_python_api.split import Split
+from paimon_python_api import Split
 
 
 class TableScan(ABC):