data-prep-connector 0.2.3.dev0__tar.gz → 0.2.4__tar.gz

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/Makefile

@@ -32,8 +32,9 @@ venv::	pyproject.toml
 	$(PYTHON) -m venv venv
 	source venv/bin/activate; 	\
 	pip install --upgrade pip;	\
-	pip install -e .;		\
-	pip install pytest pytest-mock pytest-datadir pytest-cov moto==5.0.5 markupsafe==2.0.1
+	pip install uv;	\
+	uv pip install -e .;		\
+	uv pip install pytest pytest-mock pytest-datadir pytest-cov moto==5.0.5 markupsafe==2.0.1
 
 image::
 	@# Help: Placeholder does nothing for now.

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/PKG-INFO

@@ -1,13 +1,13 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.4
 Name: data_prep_connector
-Version: 0.2.3.dev0
+Version: 0.2.4
 Summary: Scalable and Compliant Web Crawler
 Author-email: Hiroya Matsubara <hmtbr@jp.ibm.com>
 License: Apache-2.0
 Keywords: data,data acquisition,crawler,web crawler,llm,generative,ai,fine-tuning,llmapps
-Requires-Python: <3.13,>=3.10
+Requires-Python: <3.14,>=3.10
 Description-Content-Type: text/markdown
-Requires-Dist: scrapy>=2.11.2
+Requires-Dist: scrapy==2.12.0
 Requires-Dist: pydantic>=2.8.1
 Requires-Dist: tldextract>=5.1.2
 Provides-Extra: dev

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/doc/overview.md

@@ -15,13 +15,13 @@ Features:
 ### From PyPI
 
 ```sh
-pip install data-prep-connector
+uv pip install data-prep-connector
 ```
 
 ### From Github
 
 ```sh
-pip install git+https://github.com/IBM/data-prep-kit.git@dev#subdirectory=data-connector-lib
+uv pip install git+https://github.com/data-prep-kit/data-prep-kit.git@dev#subdirectory=data-connector-lib
 ```
 
 ## Example usage

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/pyproject.toml

@@ -1,7 +1,7 @@
 [project]
 name = "data_prep_connector"
-version = "0.2.3.dev0"
-requires-python = ">=3.10,<3.13"
+version = "0.2.4"
+requires-python = ">=3.10,<3.14"
 keywords = [
     "data",
     "data acquisition",
@@ -18,18 +18,21 @@ license = { text = "Apache-2.0" }
 readme = { file = "README.md", content-type = "text/markdown" }
 authors = [{ name = "Hiroya Matsubara", email = "hmtbr@jp.ibm.com" }]
 dependencies = [
-    "scrapy>=2.11.2",
+    #lock to 2.12.0 as later versions has different requirements
+    #ERROR test/dpk_connector/core/test_sitemap_spider.py::test_parse - 
+    # RuntimeError: The installed reactor (twisted.internet.epollreactor.EPollReactor) does not match the requested one (twisted.internet.asyncioreactor.AsyncioSelectorReactor)
+    "scrapy==2.12.0",
     "pydantic>=2.8.1",
     "tldextract>=5.1.2",
 ]
 
 [project_urls]
-Repository = "https://github.com/IBM/data-prep-kit"
-Issues = "https://github.com/IBM/data-prep-kit/issues"
-Documentation = "https://ibm.github.io/data-prep-kit/"
+Repository = "https://github.com/data-prep-kit/data-prep-kit"
+Issues = "https://github.com/data-prep-kit/data-prep-kit/issues"
+Documentation = "https://data-prep-kit.github.io/data-prep-kit/"
 
 [build-system]
-requires = ["setuptools>=68.0.0", "wheel", "setuptools_scm[toml]>=7.1.0"]
+requires = ["setuptools>=68.0.0", "wheel", "setuptools_scm[toml]>=7.1.0,<=8.3.0"]
 build-backend = "setuptools.build_meta"
 
 [project.optional-dependencies]

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/src/data_prep_connector.egg-info/PKG-INFO

@@ -1,13 +1,13 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.4
 Name: data_prep_connector
-Version: 0.2.3.dev0
+Version: 0.2.4
 Summary: Scalable and Compliant Web Crawler
 Author-email: Hiroya Matsubara <hmtbr@jp.ibm.com>
 License: Apache-2.0
 Keywords: data,data acquisition,crawler,web crawler,llm,generative,ai,fine-tuning,llmapps
-Requires-Python: <3.13,>=3.10
+Requires-Python: <3.14,>=3.10
 Description-Content-Type: text/markdown
-Requires-Dist: scrapy>=2.11.2
+Requires-Dist: scrapy==2.12.0
 Requires-Dist: pydantic>=2.8.1
 Requires-Dist: tldextract>=5.1.2
 Provides-Extra: dev

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/src/data_prep_connector.egg-info/requires.txt

@@ -1,4 +1,4 @@
-scrapy>=2.11.2
+scrapy==2.12.0
 pydantic>=2.8.1
 tldextract>=5.1.2
 

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/src/dpk_connector/__init__.py

@@ -1,3 +1,4 @@
+# SPDX-License-Identifier: Apache-2.0
 # (C) Copyright IBM Corp. 2024.
 # Licensed under the Apache License, Version 2.0 (the “License”);
 # you may not use this file except in compliance with the License.

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/src/dpk_connector/core/__init__.py

@@ -1,3 +1,4 @@
+# SPDX-License-Identifier: Apache-2.0
 # (C) Copyright IBM Corp. 2024.
 # Licensed under the Apache License, Version 2.0 (the “License”);
 # you may not use this file except in compliance with the License.

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/src/dpk_connector/core/crawler.py

@@ -1,3 +1,4 @@
+# SPDX-License-Identifier: Apache-2.0
 # (C) Copyright IBM Corp. 2024.
 # Licensed under the Apache License, Version 2.0 (the “License”);
 # you may not use this file except in compliance with the License.

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/src/dpk_connector/core/item.py

@@ -1,3 +1,4 @@
+# SPDX-License-Identifier: Apache-2.0
 # (C) Copyright IBM Corp. 2024.
 # Licensed under the Apache License, Version 2.0 (the “License”);
 # you may not use this file except in compliance with the License.

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/src/dpk_connector/core/logging.py

@@ -1,3 +1,4 @@
+# SPDX-License-Identifier: Apache-2.0
 # (C) Copyright IBM Corp. 2024.
 # Licensed under the Apache License, Version 2.0 (the “License”);
 # you may not use this file except in compliance with the License.

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/src/dpk_connector/core/middlewares.py

@@ -1,3 +1,4 @@
+# SPDX-License-Identifier: Apache-2.0
 # (C) Copyright IBM Corp. 2024.
 # Licensed under the Apache License, Version 2.0 (the “License”);
 # you may not use this file except in compliance with the License.

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/src/dpk_connector/core/pipelines.py

@@ -1,3 +1,4 @@
+# SPDX-License-Identifier: Apache-2.0
 # (C) Copyright IBM Corp. 2024.
 # Licensed under the Apache License, Version 2.0 (the “License”);
 # you may not use this file except in compliance with the License.

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/src/dpk_connector/core/settings.py

@@ -1,3 +1,4 @@
+# SPDX-License-Identifier: Apache-2.0
 # (C) Copyright IBM Corp. 2024.
 # Licensed under the Apache License, Version 2.0 (the “License”);
 # you may not use this file except in compliance with the License.

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/src/dpk_connector/core/spiders/__init__.py

@@ -1,3 +1,4 @@
+# SPDX-License-Identifier: Apache-2.0
 # (C) Copyright IBM Corp. 2024.
 # Licensed under the Apache License, Version 2.0 (the “License”);
 # you may not use this file except in compliance with the License.

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/src/dpk_connector/core/spiders/sitemap.py

@@ -1,3 +1,4 @@
+# SPDX-License-Identifier: Apache-2.0
 # (C) Copyright IBM Corp. 2024.
 # Licensed under the Apache License, Version 2.0 (the “License”);
 # you may not use this file except in compliance with the License.

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/src/dpk_connector/core/utils.py

@@ -1,3 +1,4 @@
+# SPDX-License-Identifier: Apache-2.0
 # (C) Copyright IBM Corp. 2024.
 # Licensed under the Apache License, Version 2.0 (the “License”);
 # you may not use this file except in compliance with the License.

data_prep_connector-0.2.4/test/dpk_connector/core/__init__.py

@@ -0,0 +1 @@
+# SPDX-License-Identifier: Apache-2.0

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/test/dpk_connector/core/test_crawler.py

@@ -1,3 +1,4 @@
+# SPDX-License-Identifier: Apache-2.0
 # (C) Copyright IBM Corp. 2024.
 # Licensed under the Apache License, Version 2.0 (the “License”);
 # you may not use this file except in compliance with the License.

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/test/dpk_connector/core/test_middlewares.py

@@ -1,3 +1,4 @@
+# SPDX-License-Identifier: Apache-2.0
 # (C) Copyright IBM Corp. 2024.
 # Licensed under the Apache License, Version 2.0 (the “License”);
 # you may not use this file except in compliance with the License.

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/test/dpk_connector/core/test_sitemap_spider.py

@@ -1,3 +1,4 @@
+# SPDX-License-Identifier: Apache-2.0
 # (C) Copyright IBM Corp. 2024.
 # Licensed under the Apache License, Version 2.0 (the “License”);
 # you may not use this file except in compliance with the License.

{data_prep_connector-0.2.3.dev0 → data_prep_connector-0.2.4}/test/dpk_connector/core/test_utils.py

@@ -1,3 +1,4 @@
+# SPDX-License-Identifier: Apache-2.0
 # (C) Copyright IBM Corp. 2024.
 # Licensed under the Apache License, Version 2.0 (the “License”);
 # you may not use this file except in compliance with the License.