base: Honor .gitignore for rsync

tmt/base.py

@@ -1796,7 +1796,7 @@ def _initialize_worktree(self) -> None:
  # Prepare worktree path and detect the source tree root
  assert self.workdir is not None # narrow type
  self.worktree = self.workdir / 'tree'
- tree_root = self.node.root
+ tree_root = Path(self.node.root) if self.node.root else None
 
  # Create an empty directory if there's no metadata tree
  if not tree_root:
@@ -1806,7 +1806,36 @@ def _initialize_worktree(self) -> None:
 
  # Sync metadata root to the worktree
  self.debug(f"Sync the worktree to '{self.worktree}'.", level=2)
- self.run(Command("rsync", "-ar", "--exclude", ".git", f"{tree_root}/", self.worktree))
+
+ ignore: list[Path] = [
+ Path('.git')
+ ]
+
+ # If we're in a git repository, honor .gitignore; xref
+ # https://stackoverflow.com/questions/13713101/rsync-exclude-according-to-gitignore-hgignore-svnignore-like-filter-c # noqa: E501
+ git_root = tmt.utils.git_root(fmf_root=tree_root, logger=self._logger)
+ if git_root:
+ ignore.extend(tmt.utils.git_ignore(root=git_root, logger=self._logger))
+
+ self.debug(
+ "Ignoring the following paths during worktree sync",
+ tmt.utils.format_value(ignore),
+ level=4)
+
+ with tempfile.NamedTemporaryFile(mode='w') as excludes_tempfile:
+ excludes_tempfile.write('\n'.join(str(path) for path in ignore))
+
+ # Make sure ignored paths are saved before telling rsync to use them.
+ # With Python 3.12, we could use `delete_on_false=False` and call `close()`.
+ excludes_tempfile.flush()
+
+ # Note: rsync doesn't use reflinks right now, so in the future it'd be even better to
+ # use e.g. `cp` but filtering out the above.
+ self.run(Command(
+ "rsync",
+ "-ar",
+ "--exclude-from", excludes_tempfile.name,
+ f"{tree_root}/", self.worktree))
 
  def _initialize_data_directory(self) -> None:
  """

tmt/utils/__init__.py

@@ -4524,6 +4524,31 @@ def git_add(*, path: Path, logger: tmt.log.Logger) -> None:
  raise GeneralError(f"Failed to add path '{path}' to git index.") from error
 
 
+def git_ignore(*, root: Path, logger: tmt.log.Logger) -> list[Path]:
+ """
+ Collect effective paths ignored by git.
+
+ :param root: path to the root of git repository.
+ :param logger: used for logging.
+ :returns: list of actual paths that would be ignored by git based on
+ its ``.gitignore`` files. If a whole directory is to be ignored,
+ it is listed as a directory path, not listing its content.
+ """
+
+ output = Command(
+ 'git',
+ 'ls-files',
+ # Consider standard git exclusion files
+ '--exclude-standard',
+ # List untracked files matching exclusion patterns
+ '-oi',
+ # If a whole directory is to be ignored, list only its name with a trailing slash
+ '--directory') \
+ .run(cwd=root, logger=logger)
+
+ return [Path(line.strip()) for line in output.stdout.splitlines()] if output.stdout else []
+
+
 def default_branch(
  *,
  repository: Path,