update airbyte-ci bump-version to push more minimal changes (#34339)

I noticed that `airbyte-ci connectors bump-version ` changes more things than necessary. In particular, it reserialize json, which means we were losing comments and reordering the various keys. I moved to doing a search-and-replace. In addition, I also noticed that the changelog changes could be slightly improved: I'm not increasing the space for the version (so we can accomodate 0.21.53, or 0.125.2),  and adding a new line at the end of the file (which was flagged as a difference by `diff`

airbyte-ci/connectors/pipelines/README.md

@@ -547,6 +547,7 @@ E.G.: running `pytest` on a specific test folder:
 
 | Version | PR                                                         | Description                                                                                                       |
 | ------- | ---------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------- |
+| 3.5.3   | [#34339](https://github.com/airbytehq/airbyte/pull/34339)  | only do minimal changes on a connector version_bump                                                           |
 | 3.5.2   | [#34381](https://github.com/airbytehq/airbyte/pull/34381)  | Bind a sidecar docker host for `airbyte-ci test`                                                                  |
 | 3.5.1   | [#34321](https://github.com/airbytehq/airbyte/pull/34321)  | Upgrade to Dagger 0.9.6 .                                                                                         |
 | 3.5.0   | [#33313](https://github.com/airbytehq/airbyte/pull/33313)  | Pass extra params after Gradle tasks.                                                                             |

airbyte-ci/connectors/pipelines/pipelines/airbyte_ci/connectors/bump_version/pipeline.py

@@ -2,13 +2,13 @@
 # Copyright (c) 2023 Airbyte, Inc., all rights reserved.
 #
 import datetime
-from copy import deepcopy
 from typing import TYPE_CHECKING
 
 import semver
 from dagger import Container, Directory
 from pipelines.airbyte_ci.connectors.context import ConnectorContext
 from pipelines.airbyte_ci.connectors.reports import ConnectorReport, Report
+from pipelines.airbyte_ci.metadata.pipeline import MetadataValidation
 from pipelines.helpers import git
 from pipelines.helpers.connectors import metadata_change_helpers
 from pipelines.models.steps import Step, StepResult, StepStatus
@@ -87,7 +87,7 @@ def add_changelog_entry(self, og_doc_content: str) -> str:
         line_index_for_new_entry = self.find_line_index_for_new_entry(og_doc_content)
         new_entry = f"| {self.new_version} | {today} | [{self.pull_request_number}](https://github.com/airbytehq/airbyte/pull/{self.pull_request_number}) | {self.changelog_entry} |"
         lines.insert(line_index_for_new_entry, new_entry)
-        return "\n".join(lines)
+        return "\n".join(lines) + "\n"
 
 
 class BumpDockerImageTagInMetadata(Step):
@@ -105,17 +105,13 @@ def __init__(
         self.new_version = new_version
 
     @staticmethod
-    def get_metadata_with_bumped_version(previous_version: str, new_version: str, current_metadata: dict) -> dict:
-        updated_metadata = deepcopy(current_metadata)
-        updated_metadata["data"]["dockerImageTag"] = new_version
-        # Bump strict versions
-        if current_metadata["data"].get("registries", {}).get("cloud", {}).get("dockerImageTag") == previous_version:
-            updated_metadata["data"]["registries"]["cloud"]["dockerImageTag"] = new_version
-        return updated_metadata
+    def get_metadata_with_bumped_version(previous_version: str, new_version: str, metadata_str: str) -> str:
+        return metadata_str.replace("dockerImageTag: " + previous_version, "dockerImageTag: " + new_version)
 
     async def _run(self) -> StepResult:
         metadata_path = self.context.connector.metadata_file_path
         current_metadata = await metadata_change_helpers.get_current_metadata(self.repo_dir, metadata_path)
+        current_metadata_str = await metadata_change_helpers.get_current_metadata_str(self.repo_dir, metadata_path)
         current_version = metadata_change_helpers.get_current_version(current_metadata)
         if current_version is None:
             return StepResult(
@@ -124,11 +120,16 @@ async def _run(self) -> StepResult:
                 stdout="Can't retrieve the connector current version.",
                 output_artifact=self.repo_dir,
             )
-        updated_metadata = self.get_metadata_with_bumped_version(current_version, self.new_version, current_metadata)
-        repo_dir_with_updated_metadata = metadata_change_helpers.get_repo_dir_with_updated_metadata(
-            self.repo_dir, metadata_path, updated_metadata
+        updated_metadata_str = self.get_metadata_with_bumped_version(current_version, self.new_version, current_metadata_str)
+        repo_dir_with_updated_metadata = metadata_change_helpers.get_repo_dir_with_updated_metadata_str(
+            self.repo_dir, metadata_path, updated_metadata_str
         )
 
+        metadata_validation_results = await MetadataValidation(self.context).run()
+        # Exit early if the metadata file is invalid.
+        if metadata_validation_results.status is not StepStatus.SUCCESS:
+            return metadata_validation_results
+
         return StepResult(
             self,
             StepStatus.SUCCESS,

airbyte-ci/connectors/pipelines/pipelines/helpers/connectors/metadata_change_helpers.py

@@ -14,9 +14,17 @@ async def get_current_metadata(repo_dir: Directory, metadata_path: Path) -> dict
     return yaml.safe_load(await repo_dir.file(str(metadata_path)).contents())
 
 
+async def get_current_metadata_str(repo_dir: Directory, metadata_path: Path) -> str:
+    return await repo_dir.file(str(metadata_path)).contents()
+
+
 def get_repo_dir_with_updated_metadata(repo_dir: Directory, metadata_path: Path, updated_metadata: dict) -> Directory:
     return repo_dir.with_new_file(str(metadata_path), contents=yaml.safe_dump(updated_metadata))
 
 
+def get_repo_dir_with_updated_metadata_str(repo_dir: Directory, metadata_path: Path, updated_metadata_str: str) -> Directory:
+    return repo_dir.with_new_file(str(metadata_path), contents=updated_metadata_str)
+
+
 def get_current_version(current_metadata: dict) -> str:
     return current_metadata.get("data", {}).get("dockerImageTag")

airbyte-ci/connectors/pipelines/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "pipelines"
-version = "3.5.2"
+version = "3.5.3"
 description = "Packaged maintained by the connector operations team to perform CI for connectors' pipelines"
 authors = ["Airbyte <contact@airbyte.io>"]