followthemoney 4.1.2__py3-none-any.whl → 4.2.1__py3-none-any.whl

followthemoney/__init__.py

@@ -9,7 +9,7 @@ from followthemoney.statement import Statement, StatementEntity, SE
 from followthemoney.dataset import Dataset, DefaultDataset, DS
 from followthemoney.util import set_model_locale
 
-__version__ = "4.1.2"
+__version__ = "4.2.1"
 
 # Data model singleton
 model = Model.instance()

followthemoney/entity.py

@@ -20,6 +20,21 @@ class ValueEntity(EntityProxy):
     applications should use this entity class as the base class.
     """
 
+    __slots__ = [
+        "schema",
+        "id",
+        "key_prefix",
+        "context",
+        "_properties",
+        "_size",
+        "_caption",
+        "datasets",
+        "referents",
+        "first_seen",
+        "last_seen",
+        "last_change",
+    ]
+
     def __init__(
         self,
         schema: Schema,

followthemoney/helpers.py

@@ -7,7 +7,7 @@
 # probably be the first place to break.
 from os.path import splitext
 from typing import Iterable, List, Optional, Set
-from normality import safe_filename
+from normality import safe_filename, squash_spaces
 from mimetypes import guess_extension
 from itertools import product
 from datetime import datetime, timedelta
@@ -148,23 +148,21 @@ def inline_names(entity: E, related: E) -> None:
 
 
 def combine_names(entity: E) -> E:
-    """This function will try to build names from name parts provided as part
-    of a person entity. This is of course impossible to do culturally correctly
-    for the whole planet at once, so it should be mostly used for internal-facing
-    (e.g. matching) processes."""
+    """Build a full names from name parts of a Person entity and add them as aliases.
+
+    This is of course impossible to do culturally correctly for the whole planet at
+    once, so it should be mostly used for internal-facing (e.g. matching) processes."""
     if entity.schema.is_a("Person"):
         first_names = entity.get("firstName")
-        second_names = entity.get("secondName")
-        second_names.append("")
-        middle_names = entity.get("middleName")
-        middle_names.append("")
-        father_names = entity.get("fatherName")
-        father_names.append("")
+        second_names = entity.get("secondName") + [""]
+        middle_names = entity.get("middleName") + [""]
+        father_names = entity.get("fatherName") + [""]
+        mother_names = entity.get("motherName") + [""]
         last_names = entity.get("lastName")
-        for (first, second, middle, father, last) in product(
-            first_names, second_names, middle_names, father_names, last_names
+        for (first, second, middle, father, mother, last) in product(
+            first_names, second_names, middle_names, father_names, mother_names, last_names
         ):
-            name = join_text(first, second, middle, father, last)
+            name = squash_spaces(" ".join([first, second, middle, father, mother, last]))
             if name is not None:
                 entity.add("alias", name)
 

followthemoney/property.py

@@ -1,3 +1,4 @@
+import re
 from banal import is_mapping, as_bool
 from typing import TYPE_CHECKING, Any, List, Optional, TypedDict
 
@@ -9,6 +10,18 @@ if TYPE_CHECKING:
     from followthemoney.schema import Schema
     from followthemoney.model import Model
 
+# Invalid property names.
+RESERVED = ["id", "caption", "schema", "schemata", "referents", "datasets"]
+PROP_NAME_RE = re.compile("^[a-z][a-zA-Z0-9]*$")
+
+
+def check_property_name(name: str) -> bool:
+    if name in RESERVED:
+        return False
+    if not PROP_NAME_RE.match(name):
+        return False
+    return True
+
 
 class ReverseSpec(TypedDict, total=False):
     name: str
@@ -66,9 +79,6 @@ class Property:
         "reverse",
     )
 
-    #: Invalid property names.
-    RESERVED = ["id", "caption", "schema", "schemata"]
-
     def __init__(self, schema: "Schema", name: str, data: PropertySpec) -> None:
         #: The schema which the property is defined for. This is always the
         #: most abstract schema that has this property, not the possible
@@ -77,11 +87,11 @@ class Property:
 
         #: Machine-readable name for this property.
         self.name = const(name)
+        if not check_property_name(self.name):
+            raise InvalidModel("Invalid name: %s" % self.name)
 
         #: Qualified property name, which also includes the schema name.
         self.qname = const("%s:%s" % (schema.name, self.name))
-        if self.name in self.RESERVED:
-            raise InvalidModel("Reserved name: %s" % self.name)
 
         self._hash = hash("<Property(%r)>" % self.qname)
 

followthemoney/schema/Risk.yaml

@@ -0,0 +1,43 @@
+Risk:
+  label: Risk
+  plural: Risks
+  description: "A risk associated with an entity"
+  extends:
+    - Interval
+  matchable: false
+  featured:
+    - entity
+    - country
+    - reason
+    - status
+    - startDate
+  required:
+    - entity
+  caption:
+    - reason
+    - status
+  properties:
+    entity:
+      label: "Entity"
+      reverse:
+        name: risks
+        label: "Risks"
+      type: entity
+      range: Thing
+    topics:
+      label: Topics
+      type: topic
+    status:
+      label: "Status"
+    duration:
+      label: "Duration"
+      type: number
+    reason:
+      label: "Reason"
+      type: text
+    country:
+      label: "Country"
+      type: country
+    listingDate:
+      label: "Listing date"
+      type: date

followthemoney/schema/Vessel.yaml

@@ -38,6 +38,9 @@ Vessel:
     grossRegisteredTonnage:
       label: Gross Registered Tonnage
       type: number
+    deadweightTonnage:
+      label: Deadweight Tonnage
+      type: number
     nameChangeDate:
       label: Date of Name Change
       type: date

followthemoney/statement/entity.py

@@ -2,6 +2,8 @@ from hashlib import sha1
 from collections.abc import Mapping
 from typing import Any, Dict, List, Optional, Set, Type
 from typing import Generator, Iterable, Tuple, TypeVar
+from rigour.langs import LangStr
+from rigour.names.pick import pick_lang_name
 
 from followthemoney.model import Model
 from followthemoney.exc import InvalidData
@@ -163,6 +165,7 @@ class StatementEntity(EntityProxy):
                 else:
                     self.last_change = max(self.last_change, stmt.first_seen)
         else:
+            self._caption = None
             if stmt.prop not in self._statements:
                 self._statements[stmt.prop] = set()
             self._statements[stmt.prop].add(stmt)
@@ -296,13 +299,17 @@ class StatementEntity(EntityProxy):
         prop_name = self._prop_name(prop, quiet=quiet)
         if prop_name is None or prop_name not in self._statements:
             return []
+        if prop_name in self.schema.caption:
+            self._caption = None
         return list({s.value for s in self._statements.pop(prop_name, [])})
 
     def remove(self, prop: P, value: str, quiet: bool = True) -> None:
         prop_name = self._prop_name(prop, quiet=quiet)
-        if prop_name is not None and prop_name in self._properties:
+        if prop_name is not None and prop_name in self._statements:
             stmts = {s for s in self._statements[prop_name] if s.value != value}
             self._statements[prop_name] = stmts
+            if prop_name in self.schema.caption:
+                self._caption = None
 
     def itervalues(self) -> Generator[Tuple[Property, str], None, None]:
         for name, statements in self._statements.items():
@@ -335,6 +342,34 @@ class StatementEntity(EntityProxy):
     def properties(self) -> Dict[str, List[str]]:
         return {p: list({s.value for s in vs}) for p, vs in self._statements.items()}
 
+    @property
+    def caption(self) -> str:
+        """The user-facing label to be used for this entity. This checks a list
+        of properties defined by the schema (caption) and returns the first
+        available value. If no caption is available, return the schema label.
+
+        This implementation prefers statements where the language property is that
+        of the preferred system language."""
+        if self._caption is None:
+            for prop_ in self.schema.caption:
+                stmts = self._statements.get(prop_)
+                if stmts is None:
+                    continue
+                prop = self.schema.properties[prop_]
+                if prop.type == registry.name and len(stmts) > 1:
+                    values = [LangStr(s.value, lang=s.lang) for s in stmts]
+                    name = pick_lang_name(values)
+                    if name is not None:
+                        self._caption = name
+                        break
+                else:
+                    for stmt in sorted(stmts):
+                        self._caption = stmt.value
+                        break
+            if self._caption is None:
+                self._caption = self.schema.label
+        return self._caption
+
     def iterprops(self) -> List[Property]:
         return [self.schema.properties[p] for p in self._statements.keys()]
 
@@ -353,8 +388,8 @@ class StatementEntity(EntityProxy):
             raise InvalidData(msg % (self.id, e))
 
         if not isinstance(other, StatementEntity):
-            for prop, values in other._properties.items():
-                self.add(prop, values, cleaned=True, quiet=True)
+            for prop, value in other.itervalues():
+                self.unsafe_add(prop, value, cleaned=True, quiet=True)
             return self
         for stmt in other._iter_stmt():
             if self.id is not None:

followthemoney/types/topic.py

@@ -55,10 +55,14 @@ class TopicType(EnumType):
         "gov.judicial": _("Judicial branch of government"),
         "gov.security": _("Security services"),
         "gov.financial": _("Central banking and financial integrity"),
+        "gov.religion": _("Religious leadership"),
         "fin": _("Financial services"),
         "fin.bank": _("Bank"),
         "fin.fund": _("Fund"),
         "fin.adivsor": _("Financial advisor"),
+        "mare.detained": _("Maritime detention"),
+        "mare.shadow": _("Shadow fleet"),
+        "mare.sts": _("Ship-to-ship transfer"),
         "reg.action": _("Regulator action"),
         "reg.warn": _("Regulator warning"),
         "role.pep": _("Politician"),
@@ -87,5 +91,33 @@ class TopicType(EnumType):
         "poi": _("Person of interest"),
     }
 
+    RISKS = {
+        "corp.disqual",
+        "crime.boss",
+        "crime.fin",
+        "crime.fraud",
+        "crime.terror",
+        "crime.theft",
+        "crime.traffick",
+        "crime.war",
+        "crime",
+        "debarment",
+        "export.control",
+        "export.risk",
+        "poi",
+        "mare.detained",
+        "mare.shadow",
+        "reg.action",
+        "reg.warn",
+        "role.oligarch",
+        "role.pep",
+        "role.rca",
+        "sanction.counter",
+        "sanction.linked",
+        "sanction",
+        "wanted",
+    }
+    """A set of topics that imply a counterparty risk in business dealings."""
+
     def _locale_names(self, locale: Locale) -> EnumValues:
         return {k: gettext(v) for (k, v) in self._TOPICS.items()}

{followthemoney-4.1.2.dist-info → followthemoney-4.2.1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: followthemoney
-Version: 4.1.2
+Version: 4.2.1
 Summary: A data model for anti corruption data modeling and analysis.
 Project-URL: Documentation, https://followthemoney.tech/
 Project-URL: Repository, https://github.com/opensanctions/followthemoney.git
@@ -42,15 +42,15 @@ Requires-Dist: click<9.0.0,>=8.0
 Requires-Dist: networkx<3.5,>=2.5
 Requires-Dist: normality<4.0.0,>=3.0.1
 Requires-Dist: openpyxl<4.0.0,>=3.0.5
-Requires-Dist: orjson<4.0,>=3.10.18
+Requires-Dist: orjson<4.0,>=3.10.0
 Requires-Dist: phonenumbers<10.0.0,>=8.12.22
-Requires-Dist: prefixdate<1.0.0,>=0.4.0
-Requires-Dist: pydantic<3.0.0,>=2.11.7
+Requires-Dist: prefixdate<1.0.0,>=0.5.0
+Requires-Dist: pydantic<3.0.0,>=2.11.0
 Requires-Dist: pytz>=2021.1
 Requires-Dist: pyyaml<7.0.0,>=5.0.0
 Requires-Dist: rdflib<7.2.0,>=6.2.0
 Requires-Dist: requests<3.0.0,>=2.21.0
-Requires-Dist: rigour<2.0.0,>=1.2.0
+Requires-Dist: rigour<2.0.0,>=1.3.1
 Requires-Dist: sqlalchemy[mypy]<3.0.0,>=2.0.0
 Provides-Extra: dev
 Requires-Dist: build; extra == 'dev'

{followthemoney-4.1.2.dist-info → followthemoney-4.2.1.dist-info}/RECORD

@@ -1,15 +1,15 @@
-followthemoney/__init__.py,sha256=1G8YlnRJL9_8pMYDUPl29Njf33pRdmat6kmg2elXBbs,856
+followthemoney/__init__.py,sha256=X-mG5wtVuhLqrHVWmz8NnB7mBFYuMwJj_wfYkL63n4M,856
 followthemoney/compare.py,sha256=bZlnj2VMoe67q4Lyq_VwS1a-EJnEK1kC8prbs8jyL9E,5774
-followthemoney/entity.py,sha256=hHY9yysn_iFTtXqcHg4hHhYfmgLw4prWul8rD1X82Y0,3184
+followthemoney/entity.py,sha256=bBiX7hNquXemS3vYCUHKtWI_IqX43Z6i8RQDbZ7gXsg,3449
 followthemoney/exc.py,sha256=GyMgwY4QVm87hLevDfV7gM1MJsDqfNCi_UQw7F_A8X8,858
 followthemoney/graph.py,sha256=7X1CGHGvmktS2LSZqld2iXWzG7B831eCNYyBqamqEJ8,10921
-followthemoney/helpers.py,sha256=Btb6BlHg_c-qCXZo-NP_LURKG-qu-QD3Fj1ev_c7Xic,7956
+followthemoney/helpers.py,sha256=EsneNJ5DZXHTPUYfXLxGESphsPwCtGiWv-71vvVBWus,7982
 followthemoney/messages.py,sha256=zUEa9CFecU8nRafIzhN6TKCh1kEihiIyIS1qr8PxY4g,806
 followthemoney/model.py,sha256=bWFVNa-DhYzc8BdSXBZdG2ev6Nh9uHx6i4tin8DvEEU,7374
 followthemoney/names.py,sha256=LODQqExKEHdH4z6Mmbhlm0KeKRzGcptaSWzYXZ7lONI,1120
 followthemoney/namespace.py,sha256=utggu9IGA8bhgEYom3OUB1KxkAJR_TrMNbY5MUF_db8,4536
 followthemoney/ontology.py,sha256=WWY_PYQGl5Ket4zZBuZglzQxD2Bh9UqHok6GJNNX7GA,3001
-followthemoney/property.py,sha256=RDTzTXJeeLFLptQL1_gr1S1T-vdDe-8MGMwsRaGQh0I,7665
+followthemoney/property.py,sha256=9qZ_o2iA-1llLMJ3O2hsW7c2XhkFU1YbvVqretGYUSA,7913
 followthemoney/proxy.py,sha256=LD4K1oPABXMX212UZxwLu7XOHRDyVBwTlqudTUsUZRQ,19619
 followthemoney/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 followthemoney/schema.py,sha256=WYnPE4Lego0pJHlojECEv0aO9Miw_YIvEb35HoDo4Zk,18087
@@ -97,6 +97,7 @@ followthemoney/schema/ProjectParticipant.yaml,sha256=xNehEu90uqUfboNouezhZQ8ZQLx
 followthemoney/schema/PublicBody.yaml,sha256=BNfLBqH1OapoEninAjWmqZx_n-G5QUnzzydW7300TiY,301
 followthemoney/schema/RealEstate.yaml,sha256=NWFHXqEHskYQN-kvQESZpu74nztShqoYSZEjZAr-DHM,1363
 followthemoney/schema/Representation.yaml,sha256=sCvFnUDQaElq2cqSB0rILcMYb2gaMZqlzxlHxyX9IGg,792
+followthemoney/schema/Risk.yaml,sha256=2BRVBqb6wiLHxb_V50P-YMAOhjC64UVHDyh5PASpCIA,728
 followthemoney/schema/Sanction.yaml,sha256=-3_NrTp1KPSsSMkttFzwo81zFKplx15w-9EmwPnE8-o,1278
 followthemoney/schema/Security.yaml,sha256=w8Och0cslWjHPAs60HZ6JarEXdIbqGlIbN1NlvgN_7Y,1212
 followthemoney/schema/Similar.yaml,sha256=gD8rZEaPQWzU-rEfsKdn62uEucF3KxYBcPMoSdnxvME,817
@@ -109,11 +110,11 @@ followthemoney/schema/UnknownLink.yaml,sha256=lneS_HZNgeLyJxwzWnLx0ZoyY3MXt99I_K
 followthemoney/schema/UserAccount.yaml,sha256=2bbPKNtt1R3zWSSkaq_SVzRPfFzX74kAxwtIxTymHA8,840
 followthemoney/schema/Value.yaml,sha256=cNHTCtakMTXDW0Qpb6ArZodi9rMJ-Ebvp1WsOIRRzw4,310
 followthemoney/schema/Vehicle.yaml,sha256=Ypl4A5HJFOZfZh3DK0ewN-hyJuCMcovR0mPNddIZrOA,1051
-followthemoney/schema/Vessel.yaml,sha256=nFaUJ_0BzFJstvog1iDvwV9DHKHr9ky4DLb1NZGGh1E,1096
+followthemoney/schema/Vessel.yaml,sha256=zWHUfSK8g6Pz58ZyCaK0AFJ4u_UHjEIUGC4c_7oS11Y,1170
 followthemoney/schema/Video.yaml,sha256=LY3DYMWTHXiAhL0hxBCNCz50cp2sPbUlEhhig5Fbjos,327
 followthemoney/schema/Workbook.yaml,sha256=iikWPElz4klA7SkWH7eae6xqhbkMCIP_3zdeXzFEMU0,354
 followthemoney/statement/__init__.py,sha256=7m2VUCAuqNZXIY0WFJRFkw5UG14QuxATL4f_xbqKwhw,633
-followthemoney/statement/entity.py,sha256=MKHGmFeDwcW2lTbAeKSdU53YwDuoLm5iKy4roeb8_lo,16172
+followthemoney/statement/entity.py,sha256=r01enw01FbSVguvI8WcsSNYoYG-KWPLNNTNwbe7J7I0,17676
 followthemoney/statement/serialize.py,sha256=9eXzQ1biR2mSxWRID5C7xDdku4b4ZImHeRJ53yLZ0yo,7225
 followthemoney/statement/statement.py,sha256=Ae-EYuzS8S12BkaRqrvMuI1C7YwlRKa5C_pTBELyNMM,8029
 followthemoney/statement/util.py,sha256=B-ozuRc1TWvpop52873Pqt5OPj8H6uk4KyRJLfAhr10,780
@@ -158,10 +159,10 @@ followthemoney/types/name.py,sha256=ZWGDebv01qByh_yBYOVoS3Edlm3_JVPShQMklKc6ZOA,
 followthemoney/types/number.py,sha256=OdVuHDd4IYIIHhx_317JKeMjBAGtsJ2TAcxoZKZ4MkY,3948
 followthemoney/types/phone.py,sha256=r8uRqWinS0CYnYBTs405k5gO4jeatUDgjdzzijoMKJE,3811
 followthemoney/types/string.py,sha256=fqyTauAm4mNnNaoH-yH087RBbNh-G5ZZUO3awTGQUUg,1230
-followthemoney/types/topic.py,sha256=CS5IoI8gm4MSVxfV6K4mGd20_tT1SaKMkcOt_ObSsAg,3678
+followthemoney/types/topic.py,sha256=Mi0Gx0m3bDeTmyuvM6jdRMqv81O03U4eI99R13KGu2Y,4503
 followthemoney/types/url.py,sha256=QFpS_JIV8unFHuh_uGv22SWUUkocBoOpzLsAJWom_gI,1455
-followthemoney-4.1.2.dist-info/METADATA,sha256=Hy34w4IjkrPyJ2O0nPRhUcamZ3oiN7h_kbGYdyIUnnI,6748
-followthemoney-4.1.2.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-followthemoney-4.1.2.dist-info/entry_points.txt,sha256=caoFTlf213jhg5sz3TNSofutjUTzaKtWATuSIdd9Cps,653
-followthemoney-4.1.2.dist-info/licenses/LICENSE,sha256=H6_EVXisnJC0-18CjXIaqrBSFq_VH3OnS7u3dccOv6g,1148
-followthemoney-4.1.2.dist-info/RECORD,,
+followthemoney-4.2.1.dist-info/METADATA,sha256=M-Hlc_ugS8WclQ3QM9L43YCL4qfegF9FhsGvrTwzWS8,6747
+followthemoney-4.2.1.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+followthemoney-4.2.1.dist-info/entry_points.txt,sha256=caoFTlf213jhg5sz3TNSofutjUTzaKtWATuSIdd9Cps,653
+followthemoney-4.2.1.dist-info/licenses/LICENSE,sha256=H6_EVXisnJC0-18CjXIaqrBSFq_VH3OnS7u3dccOv6g,1148
+followthemoney-4.2.1.dist-info/RECORD,,