code-review-ai-cli 2.1.0__tar.gz → 2.2.1__tar.gz

code_review_ai_cli-2.2.1/PKG-INFO

@@ -0,0 +1,146 @@
+Metadata-Version: 2.4
+Name: code-review-ai-cli
+Version: 2.2.1
+Summary: Automated AI-powered code review CLI for Azure DevOps / TFS Pull Requests
+License: MIT
+Keywords: code-review,ai,azure-devops,tfs,pull-request,llm
+Classifier: Development Status :: 4 - Beta
+Classifier: Environment :: Console
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Software Development :: Quality Assurance
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+Requires-Dist: requests>=2.28.0
+Requires-Dist: PyYAML>=6.0
+Provides-Extra: bedrock
+Requires-Dist: boto3>=1.34.0; extra == "bedrock"
+Provides-Extra: openai
+Requires-Dist: openai>=1.0.0; extra == "openai"
+Provides-Extra: gemini
+Requires-Dist: google-generativeai>=0.3.0; extra == "gemini"
+Provides-Extra: claude
+Requires-Dist: anthropic>=0.18.0; extra == "claude"
+Provides-Extra: tree-sitter
+Requires-Dist: tree-sitter-languages>=1.10.2; extra == "tree-sitter"
+Provides-Extra: all
+Requires-Dist: boto3>=1.34.0; extra == "all"
+Requires-Dist: openai>=1.0.0; extra == "all"
+Requires-Dist: google-generativeai>=0.3.0; extra == "all"
+Requires-Dist: anthropic>=0.18.0; extra == "all"
+Requires-Dist: tree-sitter-languages>=1.10.2; extra == "all"
+Provides-Extra: dev
+Requires-Dist: pytest>=8.3.0; extra == "dev"
+Requires-Dist: pytest-cov>=5.0.0; extra == "dev"
+Requires-Dist: pytest-mock>=3.14.0; extra == "dev"
+
+# AI Code Review
+
+AI Code Review is an automated AI-powered code review CLI, compatible with various LLM providers, designed for integration with Azure DevOps, whether in the cloud or on-premises. It allows you to run a review on demand or as part of a pipeline triggered before a commit or when a pull request is created.
+
+## Features
+
+- AI Pull Request review (`pr-review`)
+- Structured PR comments (inline + general summary)
+- `dry-run` mode to validate without posting
+- PR listing with filters (`list-prs`)
+- Configuration exclusively via `config.yaml`
+- Providers LLM: OpenAI, Azure OpenAI, Gemini, Claude, Ollama, GitHub Copilot, AWS Bedrock
+- Diff filtering by **excluded path prefixes** (`excluded_paths`) and/or **file extensions** (`file_extensions_filter`)
+
+## Documentation
+
+- [Configuration Guide](docs/configuration.md) - Details on YAML parameters and settings.
+- [Commands Reference](docs/commands.md) - Complete list of CLI commands and arguments.
+- [Execution Flow](docs/execution_flow.md) - Mermaid diagram of the application's lifecycle.
+- [Troubleshooting](docs/troubleshooting.md) - Common issues and solutions.
+
+## Output Review Format
+
+![Review Output Format Example](/imgs/review_output.png)
+
+## Installation
+
+Install from PyPI:
+s
+```bash
+pip install code-review-ai-cli
+```
+
+After installing the package, generate ready-to-edit configuration files in your working directory:
+
+```bash
+ai-review init
+```
+
+This copies two bundled templates:
+
+- **`config.yaml`** — all available options with inline documentation
+- **`review_prompt.md`** — default review style rules, injected into every LLM prompt
+
+The tool looks for `config.yaml` in the **current working directory** at runtime. You can also pass a different path with `--config`:
+
+```bash
+ai-review pr-review --config ~/configs/ai-review.yaml
+```
+
+If you plan to run the test suite locally, also install development dependencies:
+
+```bash
+pip install "code-review-ai-cli[dev]"
+```
+
+## Minimal Example
+
+Example `config.yaml`:
+
+```yaml
+llm:
+  provider: openai
+  model: gpt-4o
+
+openai:
+  api_key: sk-xxxx
+
+tfs:
+  base_url: https://dev.azure.com/your-organization
+  project: ProjectName
+  pat: xxxxxxxxx
+  verify_ssl: true
+  # ca_bundle: C:/certs/corporate-root-ca.pem
+
+review:
+  verbosity: detailed
+  scope: diff_only
+  custom_prompt_file: review_prompt.md
+  # file limit sent to the LLM
+  max_diff_files: 50
+  # per-file limit
+  max_diff_lines: 2000
+  # extension allowlist (empty list = all files)
+  file_extensions_filter: [".cs", ".ts", ".py"]
+  # exclude path prefixes from the diff (applied before file_extensions_filter)
+  # leading slash is optional; add trailing slash for exact folder boundary
+  excluded_paths: ["libs/generated"]
+
+pr:
+  auto_post_comments: false
+  dry_run: false
+  comment_mode: structured
+
+output:
+  format: terminal
+  file: ""
+  color: true
+```
+
+## Available VS Code Tasks
+
+- `AI Review: Pull Request (Interactive)`
+- `AI Review: PR (Dry-Run)`
+- `AI Review: List Active PRs`
+- `AI Review: Interactive Mode`

code_review_ai_cli-2.2.1/README.md

@@ -0,0 +1,106 @@
+# AI Code Review
+
+AI Code Review is an automated AI-powered code review CLI, compatible with various LLM providers, designed for integration with Azure DevOps, whether in the cloud or on-premises. It allows you to run a review on demand or as part of a pipeline triggered before a commit or when a pull request is created.
+
+## Features
+
+- AI Pull Request review (`pr-review`)
+- Structured PR comments (inline + general summary)
+- `dry-run` mode to validate without posting
+- PR listing with filters (`list-prs`)
+- Configuration exclusively via `config.yaml`
+- Providers LLM: OpenAI, Azure OpenAI, Gemini, Claude, Ollama, GitHub Copilot, AWS Bedrock
+- Diff filtering by **excluded path prefixes** (`excluded_paths`) and/or **file extensions** (`file_extensions_filter`)
+
+## Documentation
+
+- [Configuration Guide](docs/configuration.md) - Details on YAML parameters and settings.
+- [Commands Reference](docs/commands.md) - Complete list of CLI commands and arguments.
+- [Execution Flow](docs/execution_flow.md) - Mermaid diagram of the application's lifecycle.
+- [Troubleshooting](docs/troubleshooting.md) - Common issues and solutions.
+
+## Output Review Format
+
+![Review Output Format Example](/imgs/review_output.png)
+
+## Installation
+
+Install from PyPI:
+s
+```bash
+pip install code-review-ai-cli
+```
+
+After installing the package, generate ready-to-edit configuration files in your working directory:
+
+```bash
+ai-review init
+```
+
+This copies two bundled templates:
+
+- **`config.yaml`** — all available options with inline documentation
+- **`review_prompt.md`** — default review style rules, injected into every LLM prompt
+
+The tool looks for `config.yaml` in the **current working directory** at runtime. You can also pass a different path with `--config`:
+
+```bash
+ai-review pr-review --config ~/configs/ai-review.yaml
+```
+
+If you plan to run the test suite locally, also install development dependencies:
+
+```bash
+pip install "code-review-ai-cli[dev]"
+```
+
+## Minimal Example
+
+Example `config.yaml`:
+
+```yaml
+llm:
+  provider: openai
+  model: gpt-4o
+
+openai:
+  api_key: sk-xxxx
+
+tfs:
+  base_url: https://dev.azure.com/your-organization
+  project: ProjectName
+  pat: xxxxxxxxx
+  verify_ssl: true
+  # ca_bundle: C:/certs/corporate-root-ca.pem
+
+review:
+  verbosity: detailed
+  scope: diff_only
+  custom_prompt_file: review_prompt.md
+  # file limit sent to the LLM
+  max_diff_files: 50
+  # per-file limit
+  max_diff_lines: 2000
+  # extension allowlist (empty list = all files)
+  file_extensions_filter: [".cs", ".ts", ".py"]
+  # exclude path prefixes from the diff (applied before file_extensions_filter)
+  # leading slash is optional; add trailing slash for exact folder boundary
+  excluded_paths: ["libs/generated"]
+
+pr:
+  auto_post_comments: false
+  dry_run: false
+  comment_mode: structured
+
+output:
+  format: terminal
+  file: ""
+  color: true
+```
+
+## Available VS Code Tasks
+
+- `AI Review: Pull Request (Interactive)`
+- `AI Review: PR (Dry-Run)`
+- `AI Review: List Active PRs`
+- `AI Review: Interactive Mode`

code_review_ai_cli-2.2.1/code_review_ai_cli.egg-info/PKG-INFO

@@ -0,0 +1,146 @@
+Metadata-Version: 2.4
+Name: code-review-ai-cli
+Version: 2.2.1
+Summary: Automated AI-powered code review CLI for Azure DevOps / TFS Pull Requests
+License: MIT
+Keywords: code-review,ai,azure-devops,tfs,pull-request,llm
+Classifier: Development Status :: 4 - Beta
+Classifier: Environment :: Console
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Software Development :: Quality Assurance
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+Requires-Dist: requests>=2.28.0
+Requires-Dist: PyYAML>=6.0
+Provides-Extra: bedrock
+Requires-Dist: boto3>=1.34.0; extra == "bedrock"
+Provides-Extra: openai
+Requires-Dist: openai>=1.0.0; extra == "openai"
+Provides-Extra: gemini
+Requires-Dist: google-generativeai>=0.3.0; extra == "gemini"
+Provides-Extra: claude
+Requires-Dist: anthropic>=0.18.0; extra == "claude"
+Provides-Extra: tree-sitter
+Requires-Dist: tree-sitter-languages>=1.10.2; extra == "tree-sitter"
+Provides-Extra: all
+Requires-Dist: boto3>=1.34.0; extra == "all"
+Requires-Dist: openai>=1.0.0; extra == "all"
+Requires-Dist: google-generativeai>=0.3.0; extra == "all"
+Requires-Dist: anthropic>=0.18.0; extra == "all"
+Requires-Dist: tree-sitter-languages>=1.10.2; extra == "all"
+Provides-Extra: dev
+Requires-Dist: pytest>=8.3.0; extra == "dev"
+Requires-Dist: pytest-cov>=5.0.0; extra == "dev"
+Requires-Dist: pytest-mock>=3.14.0; extra == "dev"
+
+# AI Code Review
+
+AI Code Review is an automated AI-powered code review CLI, compatible with various LLM providers, designed for integration with Azure DevOps, whether in the cloud or on-premises. It allows you to run a review on demand or as part of a pipeline triggered before a commit or when a pull request is created.
+
+## Features
+
+- AI Pull Request review (`pr-review`)
+- Structured PR comments (inline + general summary)
+- `dry-run` mode to validate without posting
+- PR listing with filters (`list-prs`)
+- Configuration exclusively via `config.yaml`
+- Providers LLM: OpenAI, Azure OpenAI, Gemini, Claude, Ollama, GitHub Copilot, AWS Bedrock
+- Diff filtering by **excluded path prefixes** (`excluded_paths`) and/or **file extensions** (`file_extensions_filter`)
+
+## Documentation
+
+- [Configuration Guide](docs/configuration.md) - Details on YAML parameters and settings.
+- [Commands Reference](docs/commands.md) - Complete list of CLI commands and arguments.
+- [Execution Flow](docs/execution_flow.md) - Mermaid diagram of the application's lifecycle.
+- [Troubleshooting](docs/troubleshooting.md) - Common issues and solutions.
+
+## Output Review Format
+
+![Review Output Format Example](/imgs/review_output.png)
+
+## Installation
+
+Install from PyPI:
+s
+```bash
+pip install code-review-ai-cli
+```
+
+After installing the package, generate ready-to-edit configuration files in your working directory:
+
+```bash
+ai-review init
+```
+
+This copies two bundled templates:
+
+- **`config.yaml`** — all available options with inline documentation
+- **`review_prompt.md`** — default review style rules, injected into every LLM prompt
+
+The tool looks for `config.yaml` in the **current working directory** at runtime. You can also pass a different path with `--config`:
+
+```bash
+ai-review pr-review --config ~/configs/ai-review.yaml
+```
+
+If you plan to run the test suite locally, also install development dependencies:
+
+```bash
+pip install "code-review-ai-cli[dev]"
+```
+
+## Minimal Example
+
+Example `config.yaml`:
+
+```yaml
+llm:
+  provider: openai
+  model: gpt-4o
+
+openai:
+  api_key: sk-xxxx
+
+tfs:
+  base_url: https://dev.azure.com/your-organization
+  project: ProjectName
+  pat: xxxxxxxxx
+  verify_ssl: true
+  # ca_bundle: C:/certs/corporate-root-ca.pem
+
+review:
+  verbosity: detailed
+  scope: diff_only
+  custom_prompt_file: review_prompt.md
+  # file limit sent to the LLM
+  max_diff_files: 50
+  # per-file limit
+  max_diff_lines: 2000
+  # extension allowlist (empty list = all files)
+  file_extensions_filter: [".cs", ".ts", ".py"]
+  # exclude path prefixes from the diff (applied before file_extensions_filter)
+  # leading slash is optional; add trailing slash for exact folder boundary
+  excluded_paths: ["libs/generated"]
+
+pr:
+  auto_post_comments: false
+  dry_run: false
+  comment_mode: structured
+
+output:
+  format: terminal
+  file: ""
+  color: true
+```
+
+## Available VS Code Tasks
+
+- `AI Review: Pull Request (Interactive)`
+- `AI Review: PR (Dry-Run)`
+- `AI Review: List Active PRs`
+- `AI Review: Interactive Mode`

{code_review_ai_cli-2.1.0 → code_review_ai_cli-2.2.1}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "code-review-ai-cli"
-version = "2.1.0"
+version = "2.2.1"
 description = "Automated AI-powered code review CLI for Azure DevOps / TFS Pull Requests"
 readme = "README.md"
 requires-python = ">=3.10"

{code_review_ai_cli-2.1.0 → code_review_ai_cli-2.2.1}/src/ai_review.py

@@ -79,7 +79,7 @@ SCRIPT_DIR = os.path.dirname(os.path.abspath(__file__))
 PROJECT_ROOT = _ensure_project_root_on_path(__file__)
 
 from src.config import ReviewConfig, VALID_PROVIDERS  # noqa: E402
-from src.git_utils import GitUtils, GitError  # noqa: E402
+from src.git_utils import GitUtils  # noqa: E402
 from src.llm_client import LLMClient, LLMError  # noqa: E402
 from src.formatter import ReviewFormatter, Colors, save_output  # noqa: E402
 from src import __version__ as VERSION  # noqa: E402
@@ -426,6 +426,8 @@ def run_pr_review_workflow(args: argparse.Namespace, config: ReviewConfig,
             repo_name,
             pr_id,
             review_scope=config.review_scope,
+            excluded_paths=config.excluded_paths or None,
+            file_extensions_filter=config.file_extensions_filter or None,
         )
     except TFSError as exc:
         print(formatter.format_error(str(exc)))
@@ -451,17 +453,6 @@ def run_pr_review_workflow(args: argparse.Namespace, config: ReviewConfig,
     git_utils = GitUtils.__new__(GitUtils)
     git_utils.repo_path = os.getcwd()
 
-    # Filter extensions before limiting/truncating the diff sent to the LLM
-    if config.file_extensions_filter:
-        try:
-            diff = git_utils.filter_diff_by_extensions(
-                diff,
-                config.file_extensions_filter,
-            )
-        except GitError as exc:
-            print(formatter.format_warning(str(exc)))
-            return 0
-
     # Keep only added lines (+): ignore context and removed lines
     diff = git_utils.filter_diff_additions_only(diff)
     if not diff.strip():

{code_review_ai_cli-2.1.0 → code_review_ai_cli-2.2.1}/src/config.py

@@ -93,7 +93,8 @@ class ReviewConfig:
     max_scope_lines: int = 250              # Max lines in a single scope block before fallback to file_skeleton
     adaptive_context_lines: int = 10        # Context lines for Adaptive Diff fallback
     custom_prompt_file: str = "review_prompt.md"  # Markdown file with extra rules/context
-    file_extensions_filter: list = field(default_factory=list)
+    file_extensions_filter: list = field(default_factory=list)  # File extensions to include (empty = all)
+    excluded_paths: list = field(default_factory=list)  # Folder path segments to exclude from diff
 
     # --- PR Review ----------------------------------------------------
     auto_post_comments: bool = False        # Post comments automatically
@@ -222,6 +223,7 @@ class ReviewConfig:
             "adaptive_context_lines": ("review", "adaptive_context_lines"),
             "custom_prompt_file": ("review", "custom_prompt_file"),
             "file_extensions_filter": ("review", "file_extensions_filter"),
+            "excluded_paths": ("review", "excluded_paths"),
             # PR
             "auto_post_comments": ("pr", "auto_post_comments"),
             "dry_run": ("pr", "dry_run"),

{code_review_ai_cli-2.1.0 → code_review_ai_cli-2.2.1}/src/git_utils.py

@@ -189,44 +189,6 @@ class GitUtils:
         )
         return limited, True, omitted_files
 
-    def filter_diff_by_extensions(self, diff: str, extensions: list[str]) -> str:
-        """
-        Filters the diff to include only files with specific extensions.
-        
-        Args:
-            diff: The full diff.
-            extensions: List of extensions (e.g., ['.py', '.js', '.cs']).
-        """
-        if not extensions:
-            return diff
-
-        filtered_sections = []
-        current_section: list[str] = []
-        include_section = False
-
-        for line in diff.split("\n"):
-            if line.startswith("diff --git"):
-                # Save previous section if applicable
-                if include_section and current_section:
-                    filtered_sections.append("\n".join(current_section))
-                current_section = [line]
-                # Check if file has an allowed extension
-                file_path = line.split(" b/")[-1] if " b/" in line else ""
-                include_section = any(file_path.endswith(ext) for ext in extensions)
-            else:
-                current_section.append(line)
-
-        # Last section
-        if include_section and current_section:
-            filtered_sections.append("\n".join(current_section))
-
-        result = "\n".join(filtered_sections)
-        if not result.strip():
-            raise GitError(
-                f"After filtering by extensions {extensions}, no changes remain."
-            )
-        return result
-
     def truncate_diff(self, diff: str, max_lines: int = 2000) -> tuple[str, bool]:
         """
         Kept for compatibility: applies per-file truncation when

{code_review_ai_cli-2.1.0 → code_review_ai_cli-2.2.1}/src/prompts/config.yaml.template

@@ -13,7 +13,7 @@
 llm:
   provider: openai
 
-  # Model to use (empty = provider default)
+# Model to use (empty = provider default)
   # Examples:
   #   openai:  gpt-4o, gpt-4-turbo, gpt-4
   #   gemini:  gemini-1.5-pro, gemini-pro, gemini-2.0-flash
@@ -91,7 +91,7 @@ review:
   max_diff_lines: 4500
 
   # Max changed logical blocks for Enclosing Scopes strategy (diff_only mode)
-  max_scope_blocks: 3
+  max_scope_blocks: 10
 
   # Max lines in a single enclosing scope before falling back to file_skeleton
   # (prevents sending entire large classes to the LLM)
@@ -100,6 +100,12 @@ review:
   # Surrounding lines to keep for Adaptive Diff fallback (diff_only mode)
   adaptive_context_lines: 10
 
+  # Exclude paths from the diff by prefix (empty list = no exclusions).
+  # Files whose path starts with a listed prefix are excluded.
+  # Matching is case-insensitive.
+  # Example: ["IoT/_input/packages", "libs/generated"]
+  excluded_paths: []
+
   # Filter by file extensions (empty list = all files)
   # Example: [".cs", ".ts"]
   file_extensions_filter: []

{code_review_ai_cli-2.1.0 → code_review_ai_cli-2.2.1}/src/tfs_client.py

@@ -290,18 +290,23 @@ class TFSClient:
 
         files = []
         for change in changes.get("changeEntries", []):
-            item = change.get("item", {})
+            # `item` can be null for deleted files; use `or {}` as a safe fallback.
+            item = change.get("item") or {}
             if item.get("isFolder"):
                 continue
+            # `path` can be null for deleted files; fall back to `originalPath`.
+            path = item.get("path") or change.get("originalPath") or ""
             files.append({
-                "path": item.get("path", ""),
+                "path": path,
                 "change_type": change.get("changeType", "unknown"),
-                "original_path": change.get("originalPath", ""),
+                "original_path": change.get("originalPath") or path,
             })
         return files
 
     def get_pull_request_diff(self, repository: str, pr_id: int,
-                              review_scope: str = "diff_only") -> str:
+                              review_scope: str = "diff_only",
+                              excluded_paths: list[str] | None = None,
+                              file_extensions_filter: list[str] | None = None) -> str:
         """
         Gets the diff of a specific Pull Request.
 
@@ -317,16 +322,28 @@ class TFSClient:
         with ``+``, producing a pseudo-diff that represents the complete file
         as added content.
 
+        Files are filtered **before** any HTTP content request is made, so
+        excluded or extension-mismatched files incur no network cost.
+
         Args:
             repository: Repository name.
             pr_id: Pull Request ID.
             review_scope: ``"diff_only"`` (default) or ``"full_code"``.
+            excluded_paths: List of path prefixes to skip (case-insensitive,
+                leading slash optional). Files whose path starts with any
+                prefix are excluded before content is fetched.
+                Example: ``["/Libs/", "/IoT/_input/packages"]``.
+            file_extensions_filter: Allowlist of file extensions
+                (case-insensitive). When non-empty, only files whose path ends
+                with one of the listed extensions are processed.
+                Example: ``[".cs", ".ts"]``.
 
         Returns:
             Diff as text, ready to be passed to the LLM client.
 
         Raises:
-            TFSError: If the PR has no iterations or contains no file changes.
+            TFSError: If the PR has no iterations or contains no file changes
+                after filtering.
         """
         # Get PR details
         path = f"git/repositories/{repository}/pullrequests/{pr_id}"
@@ -352,17 +369,48 @@ class TFSClient:
 
         review_scope = (review_scope or "diff_only").lower()
 
+        # Normalise filters once before the loop (forward slashes, lower-case).
+        # excluded_paths: strip optional leading slash so "Libs/" == "/Libs/".
+        _excl_prefixes = [
+            p.replace("\\", "/").lstrip("/").lower()
+            for p in (excluded_paths or [])
+        ]
+        _ext_filter = [e.lower() for e in (file_extensions_filter or [])]
+
         # Build diff from the changes
         diff_parts = []
         for change in changes.get("changeEntries", []):
-            item = change.get("item", {})
+            # `item` can be null for deleted files; use `or {}` as a safe fallback.
+            item = change.get("item") or {}
             change_type = change.get("changeType", "unknown")
-            file_path = item.get("path", "unknown")
-            original_path = change.get("originalPath") or file_path
 
             if item.get("isFolder"):
                 continue
 
+            # Deleted files are skipped from the LLM diff.
+            # They are still reported in the terminal summary via _get_pr_changed_files.
+            if change_type == "delete":
+                continue
+
+            # `path` can be null; fall back to `originalPath` (safe for add/edit/rename).
+            file_path = item.get("path") or change.get("originalPath") or "unknown"
+            original_path = change.get("originalPath") or file_path
+
+            # --- Early-exit filters (no HTTP request made for skipped files) ---
+            # Normalise once: strip leading slash, lower-case, forward slashes.
+            norm_path = file_path.replace("\\", "/").lstrip("/").lower()
+
+            if _excl_prefixes and any(
+                norm_path.startswith(excl) for excl in _excl_prefixes
+            ):
+                continue
+
+            if _ext_filter and not any(
+                norm_path.endswith(ext) for ext in _ext_filter
+            ):
+                continue
+            # ------------------------------------------------------------------
+
             if review_scope == "full_code":
                 diff_parts.extend(
                     self._build_full_code_diff_part(
@@ -388,7 +436,7 @@ class TFSClient:
             diff_parts.append("")
 
         if not diff_parts:
-            raise TFSError(f"PR #{pr_id} contains no file changes.")
+            raise TFSError(f"PR #{pr_id} contains no file changes after filtering.")
 
         return "\n".join(diff_parts)
 
@@ -463,7 +511,8 @@ class TFSClient:
             file_path: Server path of the new version of the file.
             original_path: Server path of the old version (differs on renames).
             change_type: TFS change type string (``"edit"``, ``"add"``,
-                ``"rename"``, ``"delete"``, …).
+                ``"rename"``). ``"delete"`` entries are filtered out
+                by the caller before this method is invoked.
             source_branch: Source branch ref (feature branch — new version).
             target_branch: Target branch ref (base branch — old version).