sqlh 0.2.6__tar.gz → 0.2.7__tar.gz

{sqlh-0.2.6 → sqlh-0.2.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: sqlh
-Version: 0.2.6
+Version: 0.2.7
 Summary: A lightweight SQL lineage analysis library for tracking table dependencies in data pipelines
 Keywords: sql,lineage,data-pipeline,dag,dependency,database,etl,data-engineering
 Maintainer: Perry DU

{sqlh-0.2.6 → sqlh-0.2.7}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "sqlh"
-version = "0.2.6"
+version = "0.2.7"
 maintainers = [
   {name = "Perry DU", email = "duneite@gmail.com"}
 ]

{sqlh-0.2.6 → sqlh-0.2.7}/sqlh/__init__.py

@@ -11,9 +11,10 @@ from .utils import (
     search_related_tables,
     search_related_upstream_tables,
     visualize_dag,
+    table_count
 )
 
-__version__ = "0.2.6"
+__version__ = "0.2.7"
 
 __all__ = [
     "split_sql",
@@ -29,4 +30,5 @@ __all__ = [
     "search_related_tables",
     "search_command_json",
     "visualize_dag",
+    "table_count"
 ]

{sqlh-0.2.6 → sqlh-0.2.7}/sqlh/cli.py

@@ -19,6 +19,7 @@ from .utils import (
     search_related_tables,
     search_related_upstream_tables,
     visualize_dag,
+    table_count
 )
 
 
@@ -91,6 +92,15 @@ def arg_parse():
     web_parser.add_argument("--html-path", help="html file path for visualization", default=".")
     web_parser.add_argument("-h", "--help", action="help", default=argparse.SUPPRESS, help="show this help message")
 
+    # table-count 子命令
+    table_count_parser = subparsers.add_parser(
+        "table-count",
+        parents=[parent_parser],
+        help="count the number of tables",
+        add_help=False,
+    )
+    table_count_parser.add_argument("-t", "--table", help="table name to search")
+    table_count_parser.add_argument("-h", "--help", action="help", default=argparse.SUPPRESS, help="show this help message")
     return parser.parse_args()
 
 
@@ -151,3 +161,7 @@ def main():
         print(f"open web page: {html_file_path}")
         visualize_dag(get_all_dag(sql_stmt_str), template_type="dagre", filename=html_file_path)
         return
+    
+    elif args.command == "table-count":
+        for table, count in table_count(sql_stmt_str, args.table):
+            print(f"{table}: {count}")

{sqlh-0.2.6 → sqlh-0.2.7}/sqlh/tests/test_utils.py

@@ -26,7 +26,7 @@ def test_get_all_root_tables():
 
 
 def test_search_related_upstream_tables():
-    a = utils.search_related_upstream_tables(sql_stmt_str, "dws.dws_cy_cust_ltst_active_rec")
+    a = utils.search_related_upstream_tables(sql_stmt_str, "ods_hive.ods_order")
     if isinstance(a, Tuple):
         print(utils.list_command_text(a[0]))
     else:
@@ -34,7 +34,7 @@ def test_search_related_upstream_tables():
 
 
 def test_search_related_downstream_tables():
-    a = utils.search_related_downstream_tables(sql_stmt_str, "ods.ods_plr_dwm_hr_xy_shop_performance_all")
+    a = utils.search_related_downstream_tables(sql_stmt_str, "ods_hive.ods_order")
     if isinstance(a, Tuple):
         print(utils.list_command_text(a[0]))
     else:
@@ -42,7 +42,7 @@ def test_search_related_downstream_tables():
 
 
 def test_search_related_tables():
-    a = utils.search_related_tables(sql_stmt_str, "dim.dim_shopinfo")
+    a = utils.search_related_tables(sql_stmt_str, "ods_hive.ods_order")
     if isinstance(a, Tuple):
         print(utils.list_command_text(a[0]))
     else:
@@ -78,9 +78,15 @@ def test_search_command_json():
 
 
 def test_search_related_root_tables():
-    output = utils.search_related_root_tables(sql_stmt_str, "ods_hive.ods_ec_staff")
+    output = utils.search_related_root_tables(sql_stmt_str, "ods_hive.ods_order")
     # print(utils.list_command_text(output[0]))
     if isinstance(output, Tuple):
         print(utils.list_command_text(output[0]))
     else:
         print(output)
+
+
+def test_table_count():
+    print(utils.table_count(sql_stmt_str))
+    for table, count in utils.table_count(sql_stmt_str, 'ods_hive.ods_order'):
+        print(f"{table}: {count}")

{sqlh-0.2.6 → sqlh-0.2.7}/sqlh/utils.py

@@ -21,6 +21,7 @@ Example:
 
 import json
 import re
+from collections import Counter
 from pathlib import Path
 from typing import List, Literal, Tuple, Union
 
@@ -363,3 +364,15 @@ def search_command_text(search_result: SearchResult) -> str:
             return "- " + "\n- ".join(sorted(related_tables))
     else:
         return str(search_result)
+
+
+def table_count(sql_stmt_str: str, table_name: str | None = None):
+    tables = get_all_tables(sql_stmt_str)
+    counter = Counter()
+    if table_name:
+        counter[table_name] = sql_stmt_str.count(table_name.lower())
+        return [(table_name, counter[table_name])]
+    else:
+        for table in tables:
+            counter[table] += sql_stmt_str.count(table.lower())
+        return [(table, counter[table]) for table in sorted(counter, key=lambda x: counter[x], reverse=True)]