Merge pull request #12 from DerwenAI/rel/prep-1.2.1

prep for release 1.2.1

FORMAT.md

@@ -23,7 +23,6 @@ Graph data has two possible states:
   * _marshalled_: serialized and persisted in storage, i.e., "at rest"
   * _unmarshalled_: dynamic data structures in memory, i.e., "live"
 
-
 A node may be referenced either as a _source node_, which has directed edges, or as a _destination node_ which is the target of an edge.
 
 When a node from another partition is referenced as a _destination node_, then at least its "shadow" information (i.e., its unique symbol) gets included within the referencing partition. This is called a _shadow node_.
@@ -32,8 +31,11 @@ When a shadow node gets unmarshalled, that triggers an `asyncio` _future_ (calle
 
 
 ## Conventions: Nodes and Edges
+
 Records of type `Node` have always `"edge_id"` field set to `NOT_FOUND` value.
-Records of type `Edge` have always `"edge_id"` field set to an integer value greater or equal to`0` (type `pydantic.NonNegativeInt`).
+
+Records of type `Edge` have always `"edge_id"` field set to an integer value greater or equal to `0` (type `pydantic.NonNegativeInt`).
+
 
 ## Conventions: Missing Values, etc.
 

changelog.txt

@@ -1,5 +1,13 @@
 # `pynock` changelog
 
+## 1.2.1
+
+2022-10-11
+
+  * update dependencies, kudos @Mec-iS
+  * improve docs about `edge_id` requirements, kudos @Mec-iS
+
+
 ## 1.2.0
 
 2022-10-07

requirements.txt

@@ -1,8 +1,8 @@
 cloudpathlib >= 0.10
 icecream >= 2.1
-networkx>=2.8.7
+networkx >= 2.8.7
 pandas >= 1.4
-pydantic >= 1.10
 pyarrow >= 6.0
+pydantic >= 1.10
 rdflib >= 6.2
 typer[all] >= 0.6

setup.py

@@ -8,7 +8,7 @@
 import setuptools  # type: ignore
 
 
-VERSION = "1.2.0"
+VERSION = "1.2.1"
 
 DESCRIP = """
 A proposed standard `NOCK` for a Parquet format that supports efficient