add pypi description

docs/description.md

@@ -0,0 +1,20 @@
+# UmakaParser
+
+This tool is to make a JSON file for [Umaka Viewer](https://umaka-viewer.dbcls.jp/).  
+​  
+First, you need to prepare a metadata file in the [SPARQL Builder Meatadata (SBM)](http://www.sparqlbuilder.org/doc/sbm_2015sep/) format.  
+If you want to get such a metadata file, [TripleDataProfiler](https://bitbucket.org/yayamamo/tripledataprofiler/src/master/) can generate it for a SPARQL endpoint.  
+​  
+Then, if you have ontology files for the target endpoint or RDF dataset, you need to make asset files by this tool as follows.  
+​  
+`umakaparser --build-index [--dist <Path to put an asset file>] <ontology files in Turtle>`  
+​  
+If you have ontology files only in RDF/XML, this tool converts them into those in Turtle as follows.  
+​  
+`umakaparser --convert <files in RDF/XML>`  
+​  
+Finally, this tool generates a JSON file that can be accepted by [Umaka Viewer](https://umaka-viewer.dbcls.jp/) as follows.  
+​  
+`umakaparser --build [--a <Path to asset files>|--d <Path to put a generated JSON file>] <an SBM file>`  
+​  
+The JSON file structure is [here](https://github.com/dbcls/umakaparser/wiki/Data-specification).

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "umakaparser"
-version = "0.1.8"
+version = "0.2.2"
 description = ""
 authors = ["DBCLS"]
 license = "MIT"
@@ -14,6 +14,7 @@ classifiers = [
     'Programming Language :: Python :: 3.9',
     'License :: OSI Approved :: MIT License',
 ]
+readme = "docs/description.md"
 
 [tool.poetry.dependencies]
 python = ">=3.7"