mirror of
https://github.com/safishamsi/graphify.git
synced 2026-10-25 16:22:30 +00:00
Deepjyoti-Sarmah
01715accd5
A GO line with no `;` before it is read as an identifier, so the next
statement lands in an ERROR node and its table is lost. Guards cover
`;`-terminated batches, GO inside a block comment, a column named go,
and GO after a keyword line such as SET NOCOUNT ON.
(cherry picked from commit f00ecc42c9)
87 lines
2.8 KiB
Python
87 lines
2.8 KiB
Python
"""GO batch separators without `;` must not swallow the next statement."""
|
|
from __future__ import annotations
|
|
|
|
from pathlib import Path
|
|
|
|
import pytest
|
|
|
|
pytest.importorskip("tree_sitter_sql")
|
|
|
|
from graphify.extract import extract_sql # noqa: E402
|
|
|
|
TWO_TABLES = "CREATE TABLE dbo.A (id INT){sep}CREATE TABLE dbo.B (id INT){end}"
|
|
|
|
|
|
def _labels(path: Path) -> set[str]:
|
|
return {n["label"].lower() for n in extract_sql(path)["nodes"]}
|
|
|
|
|
|
def _run(tmp_path: Path, text: str) -> set[str]:
|
|
p = tmp_path / "t.sql"
|
|
p.write_bytes(text.encode())
|
|
return _labels(p)
|
|
|
|
|
|
# bug
|
|
@pytest.mark.parametrize("sep", [
|
|
"\nGO\n",
|
|
"\ngo\n",
|
|
"\nGO 5\n",
|
|
"\nGO -- next batch\n",
|
|
"\r\nGO\r\n",
|
|
"\n GO \n",
|
|
])
|
|
def test_go_separated_tables_are_both_found(tmp_path, sep):
|
|
labels = _run(tmp_path, TWO_TABLES.format(sep=sep, end="\nGO\n"))
|
|
assert {"dbo.a", "dbo.b"} <= labels
|
|
|
|
|
|
# bug
|
|
def test_go_separated_view_reads_its_table(tmp_path):
|
|
p = tmp_path / "t.sql"
|
|
p.write_text("CREATE TABLE dbo.A (id INT)\nGO\nCREATE VIEW dbo.V AS SELECT id FROM dbo.A\nGO\n")
|
|
r = extract_sql(p)
|
|
label = {n["id"]: n["label"].lower() for n in r["nodes"]}
|
|
triples = {(label.get(e["source"]), e["relation"], label.get(e["target"])) for e in r["edges"]}
|
|
assert ("dbo.v", "reads_from", "dbo.a") in triples
|
|
|
|
|
|
# bug
|
|
def test_second_table_keeps_its_own_line(tmp_path):
|
|
p = tmp_path / "t.sql"
|
|
p.write_text("CREATE TABLE dbo.A (id INT)\nGO\n\nCREATE TABLE dbo.B (id INT)\nGO\n")
|
|
nodes = {n["label"].lower(): n for n in extract_sql(p)["nodes"]}
|
|
assert "dbo.b" in nodes
|
|
assert nodes["dbo.b"].get("source_location") == "L4"
|
|
|
|
|
|
# guard
|
|
def test_semicolon_separated_tables_are_both_found(tmp_path):
|
|
labels = _run(tmp_path, TWO_TABLES.format(sep=";\nGO\n", end=";\nGO\n"))
|
|
assert {"dbo.a", "dbo.b"} <= labels
|
|
|
|
|
|
# guard
|
|
def test_go_inside_block_comment_is_not_a_separator(tmp_path):
|
|
text = "CREATE TABLE dbo.A (id INT);\n/*\nGO\nCREATE TABLE dbo.Ghost (id INT)\n*/\n"
|
|
labels = _run(tmp_path, text)
|
|
assert "dbo.a" in labels
|
|
assert "dbo.ghost" not in labels
|
|
|
|
|
|
# guard
|
|
def test_column_named_go_is_not_a_separator(tmp_path):
|
|
p = tmp_path / "t.sql"
|
|
p.write_text("CREATE TABLE dbo.A (id INT);\nCREATE VIEW dbo.V AS SELECT a.id,\n go\nFROM dbo.A a;\n")
|
|
r = extract_sql(p)
|
|
label = {n["id"]: n["label"].lower() for n in r["nodes"]}
|
|
triples = {(label.get(e["source"]), e["relation"], label.get(e["target"])) for e in r["edges"]}
|
|
assert ("dbo.v", "reads_from", "dbo.a") in triples
|
|
|
|
|
|
# guard
|
|
@pytest.mark.parametrize("prev", ["SET NOCOUNT ON", "SET ANSI_NULLS ON", "SET QUOTED_IDENTIFIER ON", "PRINT 'x' AS"])
|
|
def test_go_after_keyword_line_is_a_separator(tmp_path, prev):
|
|
labels = _run(tmp_path, f"{prev}\nGO\nCREATE TABLE dbo.A (id INT)\nGO\nCREATE TABLE dbo.B (id INT)\nGO\n")
|
|
assert {"dbo.a", "dbo.b"} <= labels
|