Merge pull request zarr-developers#1 from d-v-b/zarr-python-dtypes

fill out numpy `U` dtype

Author: jhamman

data-types/datetime64/README.md

@@ -0,0 +1,52 @@
+# `fixed_length_utf32` data type
+
+This document defines a data type for fixed-length Unicode strings encoded using [UTF-32](https://www.unicode.org/versions/Unicode5.0.0/appC.pdf#M9.19040.HeadingAppendix.C2.Encoding.Forms.in.ISOIEC.10646). UTF-32, also known as UCS4, is an encoding of Unicode strings that allocates 4 bytes to each Unicode code point.
+
+"Fixed length" as used here means that the `fixed_length_utf32` data type is parametrized by a integral length, which sets a fixed length for every scalar belonging to that data type.
+
+### Name
+
+The name of this data type is the string `"fixed_length_utf32"`
+
+### Configuration
+
+This data type requires a configuration. The configuration for this data type is a JSON object with the following fields:
+
+| field name | type | required | notes |
+|------------|----------|---|---|
+| `"length_bytes"` | integer | yes | The number MUST represent an integer divisible by 4 in the inclusive range `[0, 2147483644]` |
+
+> Note: the maximum length of 2147483644 was chosen to match the semantics of the [NumPy `"U"` data type](https://numpy.org/devdocs/reference/arrays.scalars.html#numpy.str_), which as of this writing has a maximum length in code points of 536870911, i.e. 2147483644 / 4.
+
+> Note: given a particular `fixed_length_utf32` data type, the length of an array element in Unicode code points is the value of the `length_bytes` field divided by 4.
+
+### Examples
+
+```json
+{
+  "name": "fixed_length_utf32",
+  "configuration" : {
+    "length_bytes": 4
+  }
+}
+```
+
+## Fill value representation
+
+The value of the `fill_value` metadata key must be a string. When encoded in UTF-32, the fill value MUST have a length in bytes equal to the value of the `length_bytes` specified in the `configuration` of this data type.
+
+## Codec compatibility
+
+This data type is compatible with any codec that supports arrays with fixed-sized data types.
+
+## Notes
+
+This data type is designed for NumPy compatibility. UTF-32 is not a good fit for many applications that need to model arrays of strings, as real string datasets are often composed of variable-length strings. A variable-length string data type should be preferred in these cases.
+
+## Change log
+
+No changes yet.
+
+## Current maintainers
+
+* [zarr-python core development team](https://github.com/orgs/zarr-developers/teams/python-core-devs)

data-types/datetime64/schema.json

@@ -5,7 +5,7 @@
       "type": "object",
       "properties": {
         "name": {
-          "const": "fixed-length-ucs4"
+          "const": "fixed_length_utf32"
         },
         "configuration": {
           "type": "object",
@@ -14,13 +14,13 @@
               "type": "integer"
             }
           },
-          "required": ["length_bits"],
+          "required": ["length_bytes"],
           "additionalProperties": false
         }
       },
       "required": ["name", "configuration"],
       "additionalProperties": false
     },
-    { "const": "fixed-length-ucs4" }
+    { "const": "fixed_length_utf32" }
   ]
 }