From 3093cc1eab7c9d6197553f6a1b2577696198271b Mon Sep 17 00:00:00 2001
From: Alexander Schlemmer <alexander@mail-schlemmer.de>
Date: Thu, 16 Dec 2021 10:50:14 +0100
Subject: [PATCH] FIX: identifiables used as references in properties of
identifiables are now retrieved
---
provenance.yml | 82 -------------------------
src/newcrawler/crawl.py | 2 +-
src/newcrawler/identifiable_adapters.py | 31 ++++++++--
3 files changed, 28 insertions(+), 87 deletions(-)
delete mode 100644 provenance.yml
diff --git a/provenance.yml b/provenance.yml
deleted file mode 100644
index 20eb0021..00000000
--- a/provenance.yml
+++ /dev/null
@@ -1,82 +0,0 @@
-!!python/object/apply:collections.defaultdict
-args:
-- !!python/name:newcrawler.crawl.%3Clambda%3E ''
-dictitems:
- Measurement:
- date: !!python/tuple
- - - SimulationData
- - 2020_climate-model-predict
- - '2020-02-01'
- - - SimulationData
- - project_dir
- - measurement
- description: !!python/tuple
- - - SimulationData
- - 2020_climate-model-predict
- - '2020-02-01'
- - README.md
- - description
- - - SimulationData
- - project_dir
- - measurement
- - README
- - description
- identifier: !!python/tuple
- - - SimulationData
- - 2020_climate-model-predict
- - '2020-02-01'
- - - SimulationData
- - project_dir
- - measurement
- project: !!python/tuple
- - - SimulationData
- - 2020_climate-model-predict
- - '2020-02-01'
- - - SimulationData
- - project_dir
- - measurement
- responsible: !!python/tuple
- - - SimulationData
- - 2020_climate-model-predict
- - '2020-02-01'
- - README.md
- - responsible
- - - SimulationData
- - project_dir
- - measurement
- - README
- - responsible_single
- Person:
- first_name: !!python/tuple
- - - SimulationData
- - 2020_climate-model-predict
- - '2020-02-01'
- - README.md
- - responsible
- - - SimulationData
- - project_dir
- - measurement
- - README
- - responsible_single
- last_name: !!python/tuple
- - - SimulationData
- - 2020_climate-model-predict
- - '2020-02-01'
- - README.md
- - responsible
- - - SimulationData
- - project_dir
- - measurement
- - README
- - responsible_single
- Project:
- date: !!python/tuple
- - - SimulationData
- - 2020_climate-model-predict
- - - SimulationData
- - project_dir
- identifier: !!python/tuple
- - - SimulationData
- - 2020_climate-model-predict
- - - SimulationData
- - project_dir
diff --git a/src/newcrawler/crawl.py b/src/newcrawler/crawl.py
index f4c15fee..6b31f31a 100644
--- a/src/newcrawler/crawl.py
+++ b/src/newcrawler/crawl.py
@@ -226,7 +226,7 @@ class Crawler(object):
comp = compare_entities(record, identifiable)
identical = True
for j in range(2):
- # TODO: should be implemented elsewhere
+ # TODO: should be implemented elsewhere
for label in ("properties", "parents"):
if len(comp[j][label]) > 0:
identical = False
diff --git a/src/newcrawler/identifiable_adapters.py b/src/newcrawler/identifiable_adapters.py
index 00c8597a..f306a27c 100644
--- a/src/newcrawler/identifiable_adapters.py
+++ b/src/newcrawler/identifiable_adapters.py
@@ -27,6 +27,7 @@ import caosdb as db
from abc import abstractmethod
from .utils import get_value, has_parent
+
class IdentifiableAdapter(object):
"""
Base class for identifiable adapters.
@@ -40,8 +41,8 @@ class IdentifiableAdapter(object):
- Identifiable is the concrete identifiable, e.g. the Record based on
the registered identifiable with all the values filled in.
- - Identified record is the result of retrieving a record based on the identifiable
- from the database.
+ - Identified record is the result of retrieving a record based on the
+ identifiable from the database.
General question to clarify:
Do we want to support multiple identifiables per RecordType?
@@ -212,6 +213,10 @@ class LocalStorageIdentifiableAdapter(IdentifiableAdapter):
def check_record(self, record: db.Record, identifiable: db.Record):
"""
+ Check for a record from the local storage (named "record") if it is
+ the identified record for an identifiable which was created by
+ a run of the crawler.
+
Naming of the parameters could be confusing:
record is the record from the local database to check against.
identifiable is the record that was created during the crawler run.
@@ -222,13 +227,31 @@ class LocalStorageIdentifiableAdapter(IdentifiableAdapter):
return False
for prop in identifiable.properties:
prop_record = record.get_property(prop.name)
+ if prop_record is None:
+ return False
+
# if prop is an entity, it needs to be resolved first.
# there are two different cases:
# a) prop_record.value has a registered identifiable:
# in this case, fetch the identifiable and set the value accordingly
+ if isinstance(prop.value, db.Entity):
+ registered = self.get_registered_identifiable(prop.value)
+
+ if registered is None:
+ raise NotImplementedError("Non-identifiable references cannot"
+ " be used as properties in identifiables.")
+
+ value_identifiable = self.retrieve_identified_record(prop.value)
+ if value_identifiable is None:
+ raise RuntimeError("The identifiable which is used as property"
+ " here has to be inserted first.")
+
+ if value_identifiable.id is None:
+ raise RuntimeError("The entity has not been assigned an ID.")
+
+ prop.value = value_identifiable.id
+
- if prop_record is None:
- return False
if prop.value != prop_record.value:
if prop.name == "project":
breakpoint()
--
GitLab