dump-things-pyclient 0.1.0__tar.gz → 0.1.2__tar.gz

{dump_things_pyclient-0.1.0 → dump_things_pyclient-0.1.2}/PKG-INFO

@@ -1,9 +1,11 @@
 Metadata-Version: 2.4
 Name: dump-things-pyclient
-Version: 0.1.0
+Version: 0.1.2
 Summary: A client library and some CLI command for dump-things-services
+Author-email: Christian Mönch <christian.moench@web.de>
 Requires-Python: >=3.11
 Description-Content-Type: text/markdown
+Requires-Dist: dump-things-service>=5.3.0
 Requires-Dist: requests>=2.32.5
 
 # Dump Things Python Client

dump_things_pyclient-0.1.2/dump_things_pyclient/commands/json2ttl.py

@@ -0,0 +1,64 @@
+from __future__ import annotations
+
+import argparse
+import json
+import re
+import sys
+
+from dump_things_service.converter import (
+    Format,
+    FormatConverter,
+)
+
+
+description = f"""Read JSON records from stdin and convert them to TTL
+
+This command reads one record per line, either JSON format or a JSON-string
+with a TTL-document from stdin, converts them to TTL or JSON and prints them
+to stdout.
+
+"""
+
+
+def main():
+    argument_parser = argparse.ArgumentParser(
+        description=description,
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+    )
+    argument_parser.add_argument('schema', help='URL of the schema that should be used')
+
+    arguments = argument_parser.parse_args()
+
+    print(f'Creating converter for schema {arguments.schema} ...', file=sys.stderr, end='', flush=True)
+    converter = FormatConverter(
+        arguments.schema,
+        input_format=Format.json,
+        output_format=Format.ttl,
+    )
+    print(' done', file=sys.stderr, flush=True)
+
+    error = False
+    for line in sys.stdin:
+        json_object = json.loads(line)
+
+        object_class = json_object.get('schema_type')
+        if object_class is None:
+            error = True
+            print(f'ERROR: No schema_type in {json_object}', file=sys.stderr, flush=True)
+            continue
+
+        class_name = re.search('([_A-Za-z0-9]*$)', object_class).group(0)
+
+        try:
+            ttl = converter.convert(json_object, class_name)
+        except ValueError as ve:
+            print(f'ERROR: conversion failed for {json_object}: {ve}', file=sys.stderr, flush=True)
+            continue
+
+        print(json.dumps(ttl))
+
+    return 1 if error else 0
+
+
+if __name__ == '__main__':
+    sys.exit(main())

dump_things_pyclient-0.1.2/dump_things_pyclient/commands/post_records.py

@@ -0,0 +1,59 @@
+from __future__ import annotations
+
+import argparse
+import json
+import os
+import sys
+
+from ..communicate import (
+    collection_write_record,
+    curated_write_record,
+)
+
+
+def main():
+    argument_parser = argparse.ArgumentParser()
+    argument_parser.add_argument('base_url')
+    argument_parser.add_argument('collection')
+    argument_parser.add_argument('cls', metavar='class')
+    argument_parser.add_argument('--curated', action='store_true', help='bypass inbox, requires curator token')
+
+    arguments = argument_parser.parse_args()
+
+    token = os.environ.get('DUMPTHINGS_TOKEN')
+    if token is None:
+        print(
+            'WARNING: environment variable DUMPTHINGS_TOKEN not set',
+            file=sys.stderr,
+            flush=True,
+        )
+
+    if arguments.curated:
+        write_record = curated_write_record
+    else:
+        write_record = collection_write_record
+
+    posted = False
+    for line in sys.stdin:
+        record = json.loads(line)
+        try:
+            write_record(
+                service_url=arguments.base_url,
+                collection=arguments.collection,
+                class_name=arguments.cls,
+                record=record,
+                token=token,
+            )
+        except Exception as e:
+            print(f'Error: {e}', file=sys.stderr, flush=True)
+        else:
+            posted = True
+            print('.', end='', flush=True)
+
+    if posted:
+        # final newline
+        print('')
+
+
+if __name__ == '__main__':
+    sys.exit(main())

{dump_things_pyclient-0.1.0 → dump_things_pyclient-0.1.2}/dump_things_pyclient/communicate.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import logging
+import re
 from itertools import count
 from typing import (
     Callable,
@@ -18,6 +19,7 @@ __all__ = [
     'JSON',
     'get_paginated',
     'get',
+    'collection_get_classes',
     'collection_delete_record',
     'collection_read_records',
     'collection_read_records_of_class',
@@ -99,6 +101,29 @@ def get(url: str,
     return _get_from_url(url, token, parameters)
 
 
+def collection_get_classes(service_url: str,
+                           collection: str,
+                           ) -> Generator[str, None, None]:
+    """Read classes that are supported by the collection
+
+    Get the name of the classes that are known in the collection. If the
+    collection does not exist on the server, no class names are returned.
+
+    :param service_url: the base URL of the service, i.e., the URL up to
+           `/<collection>/...` or `/server`
+    :param collection: the name of the collection
+
+    :return: a generator yielding names of the supported classes
+    """
+    service_url = f'{service_url[:-1]}' if service_url.endswith('/') else service_url
+    matcher = re.compile(f'/{collection}/record/([A-Z][_a-zA-Z0-9]*)$')
+    open_api_spec = _get_from_url(service_url + '/openapi.json', None)
+    for path in open_api_spec['paths']:
+        match = matcher.match(path)
+        if match:
+            yield match.group(1)
+
+
 def collection_read_record_with_pid(service_url: str,
                                     collection: str,
                                     pid: str,
@@ -236,11 +261,12 @@ def collection_write_record(
             than one record due to inlined-relations extraction. The individual
             records might have annotations added
     """
+    _check_format_value(format)
     return _post_to_url(
         url=_build_url(service_url, collection, f'record/{class_name}'),
         token=token,
-        json=record,
-        params={'format': format})
+        params={'format': format},
+        **(dict(json=record) if format == 'json' else dict(data=record)))
 
 
 def collection_validate_record(
@@ -269,12 +295,12 @@ def collection_validate_record(
 
     :return: True
     """
-    service_url = f'{service_url[:-1]}' if service_url.endswith('/') else service_url
+    _check_format_value(format)
     return _post_to_url(
         url=_build_url(service_url, collection, f'validate/{class_name}'),
         token=token,
-        json=record,
-        params={'format': format})
+        params={'format': format},
+        **(dict(json=record) if format == 'json' else dict(data=record)))
 
 
 def collection_delete_record(
@@ -660,10 +686,10 @@ def _get_from_url(url: str,
 
 def _post_to_url(url: str,
                  token: str | None,
-                 json: JSON,
                  params: dict[str, str] | None = None,
+                 **kwargs,
                  ) -> JSON:
-    return _do_request(requests.post, url, token, params, json=json)
+    return _do_request(requests.post, url, token, params, **kwargs)
 
 
 def _delete_url(url: str,
@@ -717,3 +743,8 @@ def _get_page(url_base: str,
     parameters['page'] = first_page
     parameters['size'] = page_size
     return _get_from_url(url_base, token, parameters)
+
+
+def _check_format_value(format: str) -> None:
+    if format not in ('json', 'ttl'):
+        raise ValueError('Format must be either "json" or "ttl"')

{dump_things_pyclient-0.1.0 → dump_things_pyclient-0.1.2}/dump_things_pyclient.egg-info/PKG-INFO

@@ -1,9 +1,11 @@
 Metadata-Version: 2.4
 Name: dump-things-pyclient
-Version: 0.1.0
+Version: 0.1.2
 Summary: A client library and some CLI command for dump-things-services
+Author-email: Christian Mönch <christian.moench@web.de>
 Requires-Python: >=3.11
 Description-Content-Type: text/markdown
+Requires-Dist: dump-things-service>=5.3.0
 Requires-Dist: requests>=2.32.5
 
 # Dump Things Python Client

{dump_things_pyclient-0.1.0 → dump_things_pyclient-0.1.2}/dump_things_pyclient.egg-info/SOURCES.txt

@@ -11,4 +11,6 @@ dump_things_pyclient.egg-info/top_level.txt
 dump_things_pyclient/commands/__init__.py
 dump_things_pyclient/commands/auto_curate.py
 dump_things_pyclient/commands/get_records.py
+dump_things_pyclient/commands/json2ttl.py
+dump_things_pyclient/commands/post_records.py
 dump_things_pyclient/commands/read_pages.py

{dump_things_pyclient-0.1.0 → dump_things_pyclient-0.1.2}/dump_things_pyclient.egg-info/entry_points.txt

@@ -1,4 +1,6 @@
 [console_scripts]
 auto-curate = dump_things_pyclient.commands.auto_curate:main
 get-records = dump_things_pyclient.commands.get_records:main
+json2ttl = dump_things_pyclient.commands.json2ttl:main
+post-records = dump_things_pyclient.commands.post_records:main
 read-pages = dump_things_pyclient.commands.read_pages:main

dump_things_pyclient-0.1.2/dump_things_pyclient.egg-info/requires.txt

@@ -0,0 +1,2 @@
+dump-things-service>=5.3.0
+requests>=2.32.5

{dump_things_pyclient-0.1.0 → dump_things_pyclient-0.1.2}/pyproject.toml

@@ -1,10 +1,14 @@
 [project]
 name = "dump-things-pyclient"
-version = "0.1.0"
+version = "0.1.2"
 description = "A client library and some CLI command for dump-things-services"
 readme = "README.md"
 requires-python = ">=3.11"
+authors = [
+    {name="Christian Mönch", email="christian.moench@web.de"},
+]
 dependencies = [
+    "dump-things-service>=5.3.0",
     "requests>=2.32.5",
 ]
 
@@ -17,3 +21,5 @@ tests = [
 auto-curate = "dump_things_pyclient.commands.auto_curate:main"
 read-pages = "dump_things_pyclient.commands.read_pages:main"
 get-records = "dump_things_pyclient.commands.get_records:main"
+json2ttl = "dump_things_pyclient.commands.json2ttl:main"
+post-records = "dump_things_pyclient.commands.post_records:main"

dump_things_pyclient-0.1.0/dump_things_pyclient.egg-info/requires.txt

@@ -1 +0,0 @@
-requests>=2.32.5