diff --git a/swh/loader/mercurial/tests/data/build.py b/swh/loader/mercurial/tests/data/build.py
new file mode 100755
--- /dev/null
+++ b/swh/loader/mercurial/tests/data/build.py
@@ -0,0 +1,187 @@
+#!/usr/bin/env python
+# Copyright (C) 2018-2020  The Software Heritage developers
+# See the AUTHORS file at the top-level directory of this distribution
+# License: GNU General Public License version 3, or any later version
+# See top-level LICENSE file for more information
+
+import json
+import shutil
+import subprocess
+from datetime import datetime
+from pathlib import Path
+
+import click
+
+
+def abort(message):
+    """Abort the script with a message."""
+    click.echo(message, err=True)
+    click.get_current_context().abort()
+
+
+def backup(path: Path):
+    """Rename an existing path."""
+    click.echo(f"Creating backup of {path}")
+    now = datetime.now()
+    backup_path = path.with_suffix(f"{path.suffix}.bak.{now:%Y%m%d%H%M%S}")
+    path.rename(backup_path)
+    click.echo(f"Backup created: {str(backup_path)!r}")
+
+
+def _build_repository(script: str) -> Path:
+    """Build a repository from a bash script."""
+    script_path = Path(script).absolute()
+
+    if not script_path.exists():
+        abort(f"Path {script_path!r} does not exists.")
+
+    if script_path.suffix != ".sh":
+        abort(f"Wrong suffix: {script_path.suffix!r}. Expected: '.sh'")
+
+    repository_path = script_path.with_suffix("")
+
+    if repository_path.exists():
+        backup(repository_path)
+
+    click.echo(f"Running build script: {str(script_path)!r}")
+    subprocess.call(
+        ["bash", "-euo", "pipefail", script_path], env={"HG_REPO": str(repository_path)}
+    )
+
+    return repository_path
+
+
+def _build_json(source: str) -> Path:
+    if source.endswith(".tgz"):
+        archive_path = Path(source).absolute()
+        repository_path = archive_path.with_suffix("")
+
+        if repository_path.exists():
+            backup(repository_path)
+
+        subprocess.call(["tar", "-xf", archive_path], cwd=archive_path.parent)
+    elif source.endswith(".sh"):
+        repository_path = _build_repository(source)
+    else:
+        repository_path = Path(source).absolute()
+
+    click.echo(f"Extracting object identities: {str(repository_path)!r}")
+    output = subprocess.check_output(["swh-hg-identify", "all"], cwd=repository_path)
+    lines = output.decode().splitlines()
+
+    directory_swhids = []
+    revision_swhids = []
+    release_swhids = []
+
+    for line in lines:
+        uri, _ = line.split("\t")
+        _, _, swhid_type, swhid = uri.split(":")
+        if swhid_type == "dir":
+            directory_swhids.append(swhid)
+        elif swhid_type == "rev":
+            revision_swhids.append(swhid)
+        elif swhid_type == "rel":
+            release_swhids.append(swhid)
+        elif swhid_type == "snp":
+            snapshot_swhid = swhid
+        else:
+            abort(f"{line!r} unknown type {swhid_type!r}")
+
+    json_path = repository_path.with_suffix(".json")
+
+    if json_path.exists():
+        backup(json_path)
+
+    click.echo(f"Creating object identities file: {str(json_path)!r}")
+    json_path.write_text(
+        json.dumps(
+            {
+                "directories": directory_swhids,
+                "revisions": revision_swhids,
+                "releases": release_swhids,
+                "snapshot": snapshot_swhid,
+            }
+        )
+    )
+
+    return json_path
+
+
+@click.group()
+def main():
+    """Build example repositories archive from bash scripts."""
+
+
+@main.command("repository")
+@click.argument("script")
+def build_repository(script: str):
+    """Build a repository.
+
+    SCRIPT must be is a bash script with a `.sh` suffix
+
+    The generated repository will have the same path minor the `.sh` suffix.
+
+    The script will be executed in an already initialized repository.
+    So it only need to execute commands to populate the repository.
+    """
+    _build_repository(script)
+
+
+@main.command("json")
+@click.argument("source")
+def build_json(source: str):
+    """Build a json file of object identities.
+
+    SOURCE can be a script as required by the `repository` command
+    (see repository --help), a repository archive, or an existing repository.
+
+    The produced file will have the source path the `.json` suffix.
+    """
+    _build_json(source)
+
+
+@main.command("archive")
+@click.option(
+    "--clean", "-c", default=False, is_flag=True, help="Remove created artifacts",
+)
+@click.argument("source")
+def build_archive(source: str, clean: bool = False):
+    """Build a repository archive.
+
+    SOURCE can be a script as required by the `repository` command
+    (see repository --help), or an existing repository.
+
+    The produced archive will have the source path with the `.tgz` suffix.
+    It will contain the repository along with the json file of object identities.
+    """
+    if source.endswith(".sh"):
+        repository_path = _build_repository(source)
+    else:
+        repository_path = Path(source).absolute()
+        if not (repository_path / ".hg").exists():
+            abort(f"{str(repository_path)!r} is not a Mercurial repository")
+
+    json_path = _build_json(str(repository_path))
+
+    archive_path = repository_path.with_suffix(".tgz")
+    if archive_path.exists():
+        backup(archive_path)
+
+    subprocess.call(
+        [
+            "tar",
+            "-cf",
+            archive_path.relative_to(archive_path.parent),
+            repository_path.relative_to(archive_path.parent),
+            json_path.relative_to(archive_path.parent),
+        ],
+        cwd=archive_path.parent,
+    )
+
+    if clean:
+        shutil.rmtree(repository_path)
+        json_path.unlink()
+
+
+if __name__ == "__main__":
+    main()