Add value for handlepid field on publish to coe repository #133.

nomad/coe_repo/__init__.py

@@ -41,11 +41,12 @@ To load an entity from the database use :data:`nomad.infrastructure.repository_d
 .. autoclass:: Calc
     :members:
     :undoc-members:
+.. autofunction:: create_handle
 .. autoclass:: DataSet
     :members:
     :undoc-members:
 """
 
 from .user import User, ensure_test_user, admin_user, LoginException
-from .calc import Calc, DataSet
+from .calc import Calc, DataSet, create_handle
 from .upload import UploadMetaData, Upload

nomad/coe_repo/calc.py

@@ -19,7 +19,7 @@ from sqlalchemy.orm import relationship, aliased
 from sqlalchemy.sql.expression import literal
 from datetime import datetime
 
-from nomad import infrastructure, utils
+from nomad import infrastructure, utils, config
 from nomad.datamodel import DFTCalcWithMetadata
 
 from . import base
@@ -29,6 +29,25 @@ from .base import Base, calc_citation_association, ownership, co_authorship, sha
     CodeVersion, StructRatio, UserMetaData
 
 
+handle_base = '0123456789abcdefghijklmnopqrstuvwxyz'
+
+
+def create_handle(pid: int) -> str:
+    """
+    Create a handle for the given pid. The pid is an autoincrement number. The handle
+    a 'base32' encoded string of that number. Therefore, its string representation is a
+    little shorter. The handle is prefixed with the configured handle prefix.
+    """
+
+    value = pid
+    result = ''
+    while value > 0:
+        result += handle_base[value & 31]
+        value = value >> 5
+
+    return config.repository_db.handle_prefix + result[::-1]
+
+
 class PublishContext:
     """
     Utilities necessary during adding calculations to the repo db.
@@ -58,6 +77,7 @@ class Calc(Base):
     __tablename__ = 'calculations'
 
     coe_calc_id = Column('calc_id', Integer, primary_key=True, autoincrement=True)
+    handlepid = Column(String)
     origin_id = Column(Integer, ForeignKey('uploads.upload_id'))
     upload = relationship('Upload', lazy='joined')
     checksum = Column(String)

nomad/coe_repo/upload.py

@@ -54,7 +54,7 @@ from sqlalchemy import exc as sa_exc
 from nomad import utils, infrastructure, config
 from nomad.datamodel import UploadWithMetadata, DFTCalcWithMetadata
 
-from .calc import Calc, PublishContext
+from .calc import Calc, PublishContext, create_handle
 from .base import Base
 from .user import User
 
@@ -186,6 +186,7 @@ class Upload(Base):  # type: ignore
                 # reuse the cache for the whole transaction to profit from repeating
                 # star schema entries for users, ds, topics, etc.
                 context = PublishContext(upload_id=upload.upload_id)
+                coe_calcs = []
                 for calc in upload.calcs:
                     has_calcs = True
                     coe_calc = Calc(
@@ -193,6 +194,7 @@ class Upload(Base):  # type: ignore
                         checksum=calc.calc_id,
                         upload=coe_upload)
                     repo_db.add(coe_calc)
+                    coe_calcs.append(coe_calc)
 
                     coe_calc.apply_calc_with_metadata(
                         cast(DFTCalcWithMetadata, calc), context=context)
@@ -203,6 +205,11 @@ class Upload(Base):  # type: ignore
 
                 result = None
                 if has_calcs:
+                    repo_db.flush()
+                    for coe_calc in coe_calcs:
+                        coe_calc.handlepid = create_handle(coe_calc.coe_calc_id)
+                    logger.debug('created all handlepids')
+
                     repo_db.commit()
                     logger.info('committed publish transaction')
                     result = coe_upload

nomad/config.py

@@ -97,7 +97,8 @@ repository_db = NomadConfig(
     port=5432,
     dbname='nomad_fairdi_repo_db',
     user='postgres',
-    password='nomad'
+    password='nomad',
+    handle_prefix='21.11132/'
 )
 
 mongo = NomadConfig(

tests/test_coe_repo.py

@@ -17,7 +17,7 @@ from typing import cast
 from passlib.hash import bcrypt
 from datetime import datetime
 
-from nomad.coe_repo import User, Calc, Upload
+from nomad.coe_repo import User, Calc, Upload, create_handle
 from nomad.coe_repo.calc import PublishContext
 from nomad import processing, parsing, datamodel
 
@@ -65,15 +65,18 @@ def assert_coe_upload(upload_id: str, upload: datamodel.UploadWithMetadata = Non
             if user_metadata is not None:
                 calc.apply_user_metadata(user_metadata)
 
-            assert_coe_calc(coe_calc, cast(datamodel.DFTCalcWithMetadata, calc))
+            assert_coe_calc(coe_calc, cast(datamodel.DFTCalcWithMetadata, calc), has_handle=True)
 
         if upload is not None and upload.upload_time is not None:
             assert coe_upload.created.isoformat()[:26] == upload.upload_time.isoformat()
 
 
-def assert_coe_calc(coe_calc: Calc, calc: datamodel.DFTCalcWithMetadata):
+def assert_coe_calc(coe_calc: Calc, calc: datamodel.DFTCalcWithMetadata, has_handle: bool = False):
     if calc.pid is not None:
         assert coe_calc.pid == calc.pid
+    elif has_handle:
+        assert coe_calc.pid is not None
+        assert create_handle(coe_calc.pid) == coe_calc.handlepid
 
     # calc data
     assert len(coe_calc.files) == len(calc.files)