feat Nextcloudclient

README.md

@@ -221,8 +221,100 @@ If using vault authentication, make sure the token file is available in the cont
 docker run --rm -v $(pwd):/data dbpedia/databus-python-client download https://databus.dbpedia.org/dbpedia-enterprise/live-fusion-snapshots/fusion/2025-08-23/fusion_props=all_subjectns=commons-wikimedia-org_vocab=all.ttl.gz --token vault-token.dat
 ```
 
-## Module Usage
 
+### Upload-and-deploy command
+```bash
+databusclient upload-and-deploy --help
+```
+```text
+Usage: databusclient upload-and-deploy [OPTIONS] [FILES]...
+
+  Upload files to Nextcloud and deploy to DBpedia Databus.
+
+Arguments:
+  FILES...  files in the form of List[path], where every path must exist locally, which will be uploaded and deployed
+
+Options:
+  --webdav-url TEXT   WebDAV URL (e.g.,
+                      https://cloud.example.com/remote.php/webdav)
+  --remote TEXT       rclone remote name (e.g., 'nextcloud')
+  --path TEXT         Remote path on Nextcloud (e.g., 'datasets/mydataset')
+  --no-upload         Skip file upload and use existing metadata
+  --metadata PATH     Path to metadata JSON file (required if --no-upload is
+                      used)
+  --version-id TEXT   Target databus version/dataset identifier of the form <h
+                      ttps://databus.dbpedia.org/$ACCOUNT/$GROUP/$ARTIFACT/$VE
+                      RSION>  [required]
+  --title TEXT        Dataset title  [required]
+  --abstract TEXT     Dataset abstract max 200 chars  [required]
+  --description TEXT  Dataset description  [required]
+  --license TEXT      License (see dalicc.net)  [required]
+  --apikey TEXT       API key  [required]
+  --help              Show this message and exit.
+```
+The script uploads all given files and all files in the given folders to the given remote.
+Then registers them on the databus.
+
+
+#### Example of using upload-and-deploy command
+
+```bash
+databusclient upload-and-deploy \
+--webdav-url https://cloud.scadsai.uni-leipzig.de/remote.php/webdav \
+--remote scads-nextcloud \
+--path test \
+--version-id https://databus.org/user/dataset/version/1.0 \
+--title "Test Dataset" \
+--abstract "This is a short abstract of the test dataset." \
+--description "This dataset was uploaded for testing the Nextcloud → Databus deployment pipeline." \
+--license https://dalicc.net/licenselibrary/Apache-2.0 \
+--apikey "API-KEY" \
+/home/test \
+/home/test_folder/test
+```
+
+
+### deploy-with-metadata command
+```bash
+databusclient deploy-with-metadata --help
+```
+```text
+Usage: databusclient deploy-with-metadata [OPTIONS]
+
+  Deploy to DBpedia Databus using metadata json file.
+
+Options:
+  --metadata PATH     Path to metadata JSON file  [required]
+  --version-id TEXT   Target databus version/dataset identifier of the form <h
+                      ttps://databus.dbpedia.org/$ACCOUNT/$GROUP/$ARTIFACT/$VE
+                      RSION>  [required]
+  --title TEXT        Dataset title  [required]
+  --abstract TEXT     Dataset abstract max 200 chars  [required]
+  --description TEXT  Dataset description  [required]
+  --license TEXT      License (see dalicc.net)  [required]
+  --apikey TEXT       API key  [required]
+  --help              Show this message and exit.
+```
+
+Use the metadata.json file (see [databusclient/metadata.json](databusclient/metadata.json)) to list all files which should be added to the databus.
+The script registers all files on the databus.
+
+
+#### Examples of using deploy-with-metadata command
+
+```bash
+databusclient deploy-with-metadata \
+  --metadata /home/metadata.json \
+  --version-id https://databus.org/user/dataset/version/1.0 \
+  --title "Test Dataset" \
+  --abstract "This is a short abstract of the test dataset." \
+  --description "This dataset was uploaded for testing the Nextcloud → Databus deployment pipeline." \
+  --license https://dalicc.net/licenselibrary/Apache-2.0 \
+  --apikey "API-KEY"
+```
+
+
+## Module Usage
 ### Step 1: Create lists of distributions for the dataset
 
 ```python

databusclient/cli.py

@@ -1,8 +1,11 @@
 #!/usr/bin/env python3
+import json
+
 import click
 from typing import List
 from databusclient import client
 
+from nextcloudclient import upload
 
 @click.group()
 def app():
@@ -36,6 +39,78 @@ def deploy(version_id, title, abstract, description, license_url, apikey, distri
     client.deploy(dataid=dataid, api_key=apikey)
 
 
+@app.command()
+@click.option(
+    "--metadata", "metadata_file",
+    required=True,
+    type=click.Path(exists=True),
+    help="Path to metadata JSON file",
+)
+@click.option(
+    "--version-id", "version_id",
+    required=True,
+    help="Target databus version/dataset identifier of the form "
+         "<https://databus.dbpedia.org/$ACCOUNT/$GROUP/$ARTIFACT/$VERSION>",
+)
+@click.option("--title", required=True, help="Dataset title")
+@click.option("--abstract", required=True, help="Dataset abstract max 200 chars")
+@click.option("--description", required=True, help="Dataset description")
+@click.option("--license", "license_url", required=True, help="License (see dalicc.net)")
+@click.option("--apikey", required=True, help="API key")
+def deploy_with_metadata(metadata_file, version_id, title, abstract, description, license_url, apikey):
+    """
+    Deploy to DBpedia Databus using metadata json file.
+    """
+
+    with open(metadata_file, 'r') as f:
+        metadata = json.load(f)
+
+    client.deploy_from_metadata(metadata, version_id, title, abstract, description, license_url, apikey)
+
+
+@app.command()
+@click.option(
+    "--webdav-url", "webdav_url",
+    required=True,
+    help="WebDAV URL (e.g., https://cloud.example.com/remote.php/webdav)",
+)
+@click.option(
+    "--remote",
+    required=True,
+    help="rclone remote name (e.g., 'nextcloud')",
+)
+@click.option(
+    "--path",
+    required=True,
+    help="Remote path on Nextcloud (e.g., 'datasets/mydataset')",
+)
+@click.option(
+    "--version-id", "version_id",
+    required=True,
+    help="Target databus version/dataset identifier of the form "
+         "<https://databus.dbpedia.org/$ACCOUNT/$GROUP/$ARTIFACT/$VERSION>",
+)
+@click.option("--title", required=True, help="Dataset title")
+@click.option("--abstract", required=True, help="Dataset abstract max 200 chars")
+@click.option("--description", required=True, help="Dataset description")
+@click.option("--license", "license_url", required=True, help="License (see dalicc.net)")
+@click.option("--apikey", required=True, help="API key")
+@click.argument(
+    "files",
+    nargs=-1,
+    type=click.Path(exists=True),
+)
+def upload_and_deploy(webdav_url, remote, path, version_id, title, abstract, description, license_url, apikey,
+                      files: List[str]):
+    """
+    Upload files to Nextcloud and deploy to DBpedia Databus.
+    """
+
+    click.echo(f"Uploading data to nextcloud: {remote}")
+    metadata = upload.upload_to_nextcloud(files, remote, path, webdav_url)
+    client.deploy_from_metadata(metadata, version_id, title, abstract, description, license_url, apikey)
+
+
 @app.command()
 @click.argument("databusuris", nargs=-1, required=True)
 @click.option("--localdir", help="Local databus folder (if not given, databus folder structure is created in current working directory)")

databusclient/client.py

@@ -205,6 +205,79 @@ def create_distribution(
 
     return f"{url}|{meta_string}"
 
+def create_distributions_from_metadata(metadata: List[Dict[str, Union[str, int]]]) -> List[str]:
+    """
+    Create distributions from metadata entries.
+
+    Parameters
+    ----------
+    metadata : List[Dict[str, Union[str, int]]]
+        List of metadata entries, each containing:
+        - filename: str - Name of the file
+        - checksum: str - SHA-256 hex digest (64 characters)
+        - size: int - File size in bytes (positive integer)
+        - url: str - Download URL for the file
+
+    Returns
+    -------
+    List[str]
+        List of distribution identifier strings for use with create_dataset
+    """
+    distributions = []
+    counter = 0
+    for entry in metadata:
+        # Validate required keys
+        required_keys = ["filename", "checksum", "size", "url"]
+        missing_keys = [key for key in required_keys if key not in entry]
+        if missing_keys:
+            raise ValueError(f"Metadata entry missing required keys: {missing_keys}")
+
+        filename = entry["filename"]
+        checksum = entry["checksum"]
+        size = entry["size"]
+        if not isinstance(size, int) or size <= 0:
+            raise ValueError(f"Invalid size for {filename}: expected positive integer, got {size}")
+        url = entry["url"]
+        # Validate SHA-256 hex digest (64 hex chars)
+        if not isinstance(checksum, str) or len(checksum) != 64 or not all(
+            c in '0123456789abcdefABCDEF' for c in checksum):
+                raise ValueError(f"Invalid checksum for {filename}")
+        # Known compression extensions
+        COMPRESSION_EXTS = {"gz", "bz2", "xz", "zip", "7z", "tar", "lz", "zst"}
+
+        parts = filename.split(".")
+        if len(parts) == 1:
+            file_format = "none"
+            compression = "none"
+        elif len(parts) == 2:
+            file_format = parts[-1]
+            compression = "none"
+        else:
+            # Check if last part is a known compression
+
+            if parts[-1] in COMPRESSION_EXTS:
+                compression = parts[-1]
+                # Handle compound extensions like .tar.gz
+                if len(parts) > 2 and parts[-2] in COMPRESSION_EXTS:
+                    file_format = parts[-3] if len(parts) > 3 else "file"
+                else:
+                    file_format = parts[-2]
+            else:
+                file_format = parts[-1]
+                compression = "none"
+
+        distributions.append(
+            create_distribution(
+                url=url,
+                cvs={"count": f"{counter}"},
+                file_format=file_format,
+                compression=compression,
+                sha256_length_tuple=(checksum, size)
+            )
+        )
+        counter += 1
+    return distributions
+
 
 def create_dataset(
     version_id: str,
@@ -393,6 +466,55 @@ def deploy(
         print(resp.text)
 
 
+def deploy_from_metadata(
+    metadata: List[Dict[str, Union[str, int]]],
+    version_id: str,
+    title: str,
+    abstract: str,
+    description: str,
+    license_url: str,
+    apikey: str
+) -> None:
+    """
+    Deploy a dataset from metadata entries.
+
+    Parameters
+    ----------
+    metadata : List[Dict[str, Union[str, int]]]
+        List of file metadata entries (see create_distributions_from_metadata)
+    version_id : str
+        Dataset version ID in the form $DATABUS_BASE/$ACCOUNT/$GROUP/$ARTIFACT/$VERSION
+    title : str
+        Dataset title
+    abstract : str
+        Short description of the dataset
+    description : str
+        Long description (Markdown supported)
+    license_url : str
+        License URI
+    apikey : str
+        API key for authentication
+    """
+    distributions = create_distributions_from_metadata(metadata)
+
+    dataset = create_dataset(
+        version_id=version_id,
+        title=title,
+        abstract=abstract,
+        description=description,
+        license_url=license_url,
+        distributions=distributions
+    )
+
+    print(f"Deploying dataset version: {version_id}")
+    deploy(dataset, apikey)
+
+    print(f"Successfully deployed to {version_id}")
+    print(f"Deployed {len(metadata)} file(s):")
+    for entry in metadata:
+        print(f"  - {entry['filename']}")
+
+
 def __download_file__(url, filename, vault_token_file=None, auth_url=None, client_id=None) -> None:
     """
     Download a file from the internet with a progress bar using tqdm.
@@ -635,7 +757,7 @@ def __download_list__(urls: List[str],
 def __get_databus_id_parts__(uri: str) -> Tuple[Optional[str], Optional[str], Optional[str], Optional[str], Optional[str], Optional[str]]:
     uri = uri.removeprefix("https://").removeprefix("http://")
     parts = uri.strip("/").split("/")
-    parts += [None] * (6 - len(parts))  # pad with None if less than 6 parts
+    parts += [None] * (6 - len(parts))  # pad fwith None if less than 6 parts
     return tuple(parts[:6])  # return only the first 6 parts
 
 

databusclient/metadata.json

@@ -0,0 +1,14 @@
+[
+  {
+    "filename": "example.ttl",
+    "checksum": "0929436d44bba110fc7578c138ed770ae9f548e195d19c2f00d813cca24b9f39",
+    "size": 12345,
+    "url": "https://cloud.example.com/remote.php/webdav/datasets/mydataset/example.ttl"
+  },
+  {
+    "filename": "example.csv.gz",
+    "checksum": "2238acdd7cf6bc8d9c9963a9f6014051c754bf8a04aacc5cb10448e2da72c537",
+    "size": 54321,
+    "url": "https://cloud.example.com/remote.php/webdav/datasets/mydataset/example.csv.gz"
+  }
+]