This is an automated email from the ASF dual-hosted git repository.
fokko pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/iceberg-python.git
The following commit(s) were added to refs/heads/main by this push:
new 781096eb Add `table_exists` method to Catalog (#512)
781096eb is described below
commit 781096eb0c71fa540357e0e6e3b51104ad6469ee
Author: Anupam Saini <[email protected]>
AuthorDate: Thu Mar 14 07:06:45 2024 -0400
Add `table_exists` method to Catalog (#512)
* Add `table_exist` method to Catalog
* `table_exist` -> `table_exists`
* Add `table_exists` for RestCatalog
* Add doc
---
mkdocs/docs/api.md | 10 ++++++++++
pyiceberg/catalog/__init__.py | 7 +++++++
pyiceberg/catalog/rest.py | 8 ++++++++
tests/catalog/test_base.py | 11 +++++++++++
tests/catalog/test_rest.py | 20 ++++++++++++++++++++
5 files changed, 56 insertions(+)
diff --git a/mkdocs/docs/api.md b/mkdocs/docs/api.md
index 724a45c5..35b271ae 100644
--- a/mkdocs/docs/api.md
+++ b/mkdocs/docs/api.md
@@ -194,6 +194,16 @@ static_table = StaticTable.from_metadata(
The static-table is considered read-only.
+## Check if a table exists
+
+To check whether the `bids` table exists:
+
+```python
+catalog.table_exists("docs_example.bids")
+```
+
+Returns `True` if the table already exists.
+
## Write support
With PyIceberg 0.6.0 write support is added through Arrow. Let's consider an
Arrow Table:
diff --git a/pyiceberg/catalog/__init__.py b/pyiceberg/catalog/__init__.py
index 4f7b5a32..f2b46fcd 100644
--- a/pyiceberg/catalog/__init__.py
+++ b/pyiceberg/catalog/__init__.py
@@ -646,6 +646,13 @@ class Catalog(ABC):
delete_files(io, prev_metadata_files, PREVIOUS_METADATA)
delete_files(io, {table.metadata_location}, METADATA)
+ def table_exists(self, identifier: Union[str, Identifier]) -> bool:
+ try:
+ self.load_table(identifier)
+ return True
+ except NoSuchTableError:
+ return False
+
@staticmethod
def _write_metadata(metadata: TableMetadata, io: FileIO, metadata_path:
str) -> None:
ToOutputFile.table_metadata(metadata, io.new_output(metadata_path))
diff --git a/pyiceberg/catalog/rest.py b/pyiceberg/catalog/rest.py
index f67fe2c1..9f0d0544 100644
--- a/pyiceberg/catalog/rest.py
+++ b/pyiceberg/catalog/rest.py
@@ -717,3 +717,11 @@ class RestCatalog(Catalog):
updated=parsed_response.updated,
missing=parsed_response.missing,
)
+
+ @retry(**_RETRY_ARGS)
+ def table_exists(self, identifier: Union[str, Identifier]) -> bool:
+ identifier_tuple = self.identifier_to_tuple_without_catalog(identifier)
+ response = self._session.head(
+ self.url(Endpoints.load_table, prefixed=True,
**self._split_identifier_for_path(identifier_tuple))
+ )
+ return response.status_code == 200
diff --git a/tests/catalog/test_base.py b/tests/catalog/test_base.py
index 44c36a7d..5f78eb3b 100644
--- a/tests/catalog/test_base.py
+++ b/tests/catalog/test_base.py
@@ -413,6 +413,17 @@ def test_table_raises_error_on_table_not_found(catalog:
InMemoryCatalog) -> None
catalog.load_table(TEST_TABLE_IDENTIFIER)
+def test_table_exists(catalog: InMemoryCatalog) -> None:
+ # Given
+ given_catalog_has_a_table(catalog)
+ # Then
+ assert catalog.table_exists(TEST_TABLE_IDENTIFIER)
+
+
+def test_table_exists_on_table_not_found(catalog: InMemoryCatalog) -> None:
+ assert not catalog.table_exists(TEST_TABLE_IDENTIFIER)
+
+
def test_drop_table(catalog: InMemoryCatalog) -> None:
# Given
given_catalog_has_a_table(catalog)
diff --git a/tests/catalog/test_rest.py b/tests/catalog/test_rest.py
index 850e5f01..4956fffe 100644
--- a/tests/catalog/test_rest.py
+++ b/tests/catalog/test_rest.py
@@ -644,6 +644,26 @@ def test_load_table_404(rest_mock: Mocker) -> None:
assert "Table does not exist" in str(e.value)
+def test_table_exist_200(rest_mock: Mocker) -> None:
+ rest_mock.head(
+ f"{TEST_URI}v1/namespaces/fokko/tables/table",
+ status_code=200,
+ request_headers=TEST_HEADERS,
+ )
+ catalog = RestCatalog("rest", uri=TEST_URI, token=TEST_TOKEN)
+ assert catalog.table_exists(("fokko", "table"))
+
+
+def test_table_exist_500(rest_mock: Mocker) -> None:
+ rest_mock.head(
+ f"{TEST_URI}v1/namespaces/fokko/tables/table",
+ status_code=500,
+ request_headers=TEST_HEADERS,
+ )
+ catalog = RestCatalog("rest", uri=TEST_URI, token=TEST_TOKEN)
+ assert not catalog.table_exists(("fokko", "table"))
+
+
def test_drop_table_404(rest_mock: Mocker) -> None:
rest_mock.delete(
f"{TEST_URI}v1/namespaces/fokko/tables/does_not_exists",
