Use shorter build_key

Fixes #611.

conda-store-server/conda_store_server/alembic/versions/64a749e87619_add_build_key_version.py

@@ -0,0 +1,28 @@
+"""add build_key_version
+
+Revision ID: 64a749e87619
+Revises: b387747ca9b7
+Create Date: 2023-11-05 00:59:57.132192
+
+"""
+import sqlalchemy as sa
+from alembic import op
+
+# revision identifiers, used by Alembic.
+revision = "64a749e87619"
+down_revision = "b387747ca9b7"
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    op.add_column(
+        "build",
+        sa.Column(
+            "build_key_version", sa.Integer(), nullable=False, server_default="1"
+        ),
+    )
+
+
+def downgrade():
+    op.drop_column("build", "build_key_version")

conda-store-server/conda_store_server/orm.py

@@ -178,6 +178,14 @@ class Build(Base):
     started_on = Column(DateTime, default=None)
     ended_on = Column(DateTime, default=None)
     deleted_on = Column(DateTime, default=None)
+    build_key_version = Column(Integer, default=2, nullable=False)
+
+    @validates("build_key_version")
+    def validate_build_key_version(self, key, build_key_version):
+        if build_key_version not in [1, 2]:
+            raise ValueError(f"invalid build_key_version={build_key_version}")
+
+        return build_key_version
 
     build_artifacts = relationship(
         "BuildArtifact", back_populates="build", cascade="all, delete-orphan"
@@ -225,15 +233,29 @@ def build_key(self):
         The build key should be a key that allows for the environment
         build to be easily identified and found in the database.
         """
-        datetime_format = "%Y%m%d-%H%M%S-%f"
-        return f"{self.specification.sha256}-{self.scheduled_on.strftime(datetime_format)}-{self.id}-{self.specification.name}"
+        if self.build_key_version == 1:
+            datetime_format = "%Y%m%d-%H%M%S-%f"
+            return f"{self.specification.sha256}-{self.scheduled_on.strftime(datetime_format)}-{self.id}-{self.specification.name}"
+        elif self.build_key_version == 2:
+            hash = self.specification.sha256[:4]
+            timestamp = int(self.scheduled_on.timestamp())
+            id = self.id
+            name = self.specification.name[:16]
+            return f"{hash}-{timestamp}-{id}-{name}"
+        else:
+            raise ValueError(f"invalid build key version: {self.build_key_version}")
 
     @staticmethod
     def parse_build_key(key):
         parts = key.split("-")
-        if len(parts) < 5:
-            return None
-        return int(parts[4])  # build_id
+        # Note: cannot rely on the number of dashes to differentiate between
+        # versions because name can contain dashes. Instead, this relies on the
+        # hash size to infer the format. The name is the last field, so indexing
+        # to find the id is okay.
+        if key[4] == "-":  # v2
+            return int(parts[2])  # build_id
+        else:  # v1
+            return int(parts[4])  # build_id
 
     @property
     def log_key(self):

conda-store-server/tests/test_actions.py

@@ -1,4 +1,5 @@
 import asyncio
+import datetime
 import pathlib
 import re
 import sys
@@ -224,17 +225,20 @@ def test_add_lockfile_packages(
 
 
 @pytest.mark.parametrize(
-    "is_legacy_build",
+    "is_legacy_build, build_key_version",
     [
-        False,
-        True,
+        (False, 0),  # invalid
+        (False, 1),  # long (legacy)
+        (False, 2),  # short (default)
+        (True, 1),  # build_key_version doesn't matter because there's no lockfile
     ],
 )
 def test_api_get_build_lockfile(
-    request, conda_store, db, simple_specification_with_pip, conda_prefix, is_legacy_build
+    request, conda_store, db, simple_specification_with_pip, conda_prefix, is_legacy_build, build_key_version
 ):
     # initializes data needed to get the lockfile
     specification = simple_specification_with_pip
+    specification.name = "this-is-a-long-environment-name-to-test-truncation"
     namespace = "pytest"
 
     class MyAuthentication(DummyAuthentication):
@@ -250,6 +254,16 @@ def authorize_request(self, *args, **kwargs):
     )
     db.commit()
     build = api.get_build(db, build_id=build_id)
+    # makes this more visible in the lockfile
+    build_id = 12345678
+    build.id = build_id
+    if build_key_version is 0:  # invalid
+        with pytest.raises(ValueError, match=r"invalid build_key_version=0"):
+            build.build_key_version = build_key_version
+        return  # invalid, nothing more to test
+    build.build_key_version = build_key_version
+    # makes sure the timestamp in build_key is always the same
+    build.scheduled_on = datetime.datetime(2023, 11, 5, 3, 54, 10, 510258)
     environment = api.get_environment(db, namespace=namespace)
 
     # adds packages (returned in the lockfile)
@@ -293,12 +307,21 @@ def authorize_request(self, *args, **kwargs):
         assert re.match("http.*//.*tar.bz2#.*", lines[2]) is not None
     else:
         # new build: redirects to lockfile generated by conda-lock
-        lockfile_url_pattern = (
-            "lockfile/"
-            "89e5a99aa094689b7aafc66c47987fa186e08f9d619a02ab1a469d0759da3b8b-"
-            ".*-test.yml"
-        )
+        def lockfile_url(build_key):
+            return f"lockfile/{build_key}.yml"
+        if build_key_version is 1:
+            build_key = (
+                "7a0c9317530e3732a25f22c2017a881dcd6f84ff85c96a609210168deea280ef-"
+                "20231105-035410-510258-12345678-this-is-a-long-environment-name-to-test-truncation"
+            )
+        elif build_key_version is 2:
+            build_key = "7a0c-1699156450-12345678-this-is-a-long-e"
+        else:
+            raise ValueError(f"unexpected build_key_version: {build_key_version}")
         assert type(res) is RedirectResponse
         assert key == res.headers['location']
-        assert re.match(lockfile_url_pattern, res.headers['location']) is not None
+        assert build.build_key == build_key
+        assert build.parse_build_key(build_key) == 12345678
+        assert lockfile_url(build_key) == build.conda_lock_key
+        assert lockfile_url(build_key) == res.headers['location']
         assert res.status_code == 307