pyomnigraph 0.2.0__tar.gz → 0.2.2__tar.gz

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pyomnigraph
-Version: 0.2.0
+Version: 0.2.2
 Summary: pyomnigraph: Unified Python interface for multiple graph databases (SPARQL, Neo4j, Weaviate)
 Project-URL: Home, https://github.com/WolfgangFahl/omnigraph
 Project-URL: Documentation, https://wiki.bitplan.com/index.php/omnigraph
@@ -83,16 +83,16 @@ This leads to:
 
 ### Supported Triple Stores
 
-| Database | Status | Strengths |
-|----------|--------|-----------|
+| Database | Status | Notes |
+|----------|--------|-------|
 | **Apache Jena** | 🟢 Working | Robust, standards compliant |
 | **Blazegraph** | 🟢 Working | High performance, easy setup |
-| **GraphDB** | 🟢 Working | Enterprise features, reasoning - Free Version tested|
+| **Oxigraph** | 🟢 Working | Rust-based, embedded, fast |
 | **QLever** | 🟢 Working | Extremely fast queries |
-| **Virtuoso** | 🛑 In Progress | Mature, SQL integration |
-| **MillenniumDB** | 🛑 In Progress | Property Graph + RDF; SPARQL 1.1 (RDF) and MQL (PG); import-first workflow |
-| **Oxigraph** | 🛑 In Progress | Rust-based, embedded |
-| **Stardog** | 🛑 Planned | Knowledge graphs, reasoning |
+| **GraphDB** | 🛑 Disabled | License required (free version available) |
+| **MillenniumDB** | 🛑 Disabled | Import-first workflow, Property Graph + RDF |
+| **Virtuoso** | 🛑 Disabled | Permissions issue (#30) |
+| **Stardog** | 🛑 Planned | License required |
 
 Whether you're building a semantic web application, conducting research, or evaluating different triple stores, pyomnigraph eliminates the complexity of working with multiple graph database systems.
 
@@ -103,10 +103,11 @@ omnigraph omnigraph --list --include-inactive --doc-format  github
 | Active   | Name                                | Container Name       | Wikidata                                               | Image                                                                                            |   Port |   Test Port | Dataset   | User   |
 |----------|-------------------------------------|----------------------|--------------------------------------------------------|--------------------------------------------------------------------------------------------------|--------|-------------|-----------|--------|
 | 🟢️       | [blazegraph](http://localhost:9898) | blazegraph-omnigraph | [Q20127748](https://www.wikidata.org/wiki/Q20127748)   | [lyrasis/blazegraph:2.1.5](https://hub.docker.com/r/lyrasis/blazegraph)                          |   9898 |        7898 | kb        |        |
-| 🟢       | [graphdb](http://localhost:7200)    | graphdb-omnigraph    | [Q58425577](https://www.wikidata.org/wiki/Q58425577)   | [ontotext/graphdb:9.11.2-se](https://hub.docker.com/r/ontotext/graphdb)                          |   7200 |        7700 | repo1     |        |
 | 🟢️       | [jena](http://localhost:3030)       | jena-omnigraph       | [Q109376461](https://www.wikidata.org/wiki/Q109376461) | [stain/jena-fuseki:latest](https://hub.docker.com/r/stain/jena-fuseki)                           |   3030 |        7030 | ds        | admin  |
-| 🛑       | [oxigraph](http://localhost:7878)   | oxigraph-omnigraph   | [Q118980507](https://www.wikidata.org/wiki/Q118980507) | [oxigraph/oxigraph:latest](https://hub.docker.com/r/oxigraph/oxigraph)                           |   7878 |        7378 | default   |        |
+| 🟢️       | [oxigraph](http://localhost:7878)   | oxigraph-omnigraph   | [Q118980507](https://www.wikidata.org/wiki/Q118980507) | [oxigraph/oxigraph:latest](https://hub.docker.com/r/oxigraph/oxigraph)                           |   7878 |        7378 | default   |        |
 | 🟢️       | [qlever](http://localhost:7019)     | qlever-omnigraph     | [Q111016295](https://www.wikidata.org/wiki/Q111016295) | [adfreiburg/qlever:latest](https://hub.docker.com/r/adfreiburg/qlever)                           |   7019 |        7819 | olympics  |        |
+| 🛑       | [graphdb](http://localhost:7200)    | graphdb-omnigraph    | [Q58425577](https://www.wikidata.org/wiki/Q58425577)   | [ontotext/graphdb:11.2.0](https://hub.docker.com/r/ontotext/graphdb)                             |   7200 |        7700 | repo1     |        |
+| 🛑       | [millenniumdb](http://localhost:1234)| millenniumdb-omnigraph| [Q118954995](https://www.wikidata.org/wiki/Q118954995) | [imfd/millenniumdb](https://hub.docker.com/r/imfd/millenniumdb)                                  |   1234 |        1734 | default   |        |
 | 🛑       | [stardog](http://localhost:5820)    | stardog-omnigraph    | [Q91147741](https://www.wikidata.org/wiki/Q91147741)   | [stardog/stardog:latest](https://hub.docker.com/r/stardog/stardog)                               |   5820 |        5320 | mydb      | admin  |
 | 🛑       | [virtuoso](http://localhost:8890)   | virtuoso-omnigraph   | [Q7935239](https://www.wikidata.org/wiki/Q7935239)     | [openlink/virtuoso-opensource-7:latest](https://hub.docker.com/r/openlink/virtuoso-opensource-7) |   8890 |        8390 | KB        | dba    |
 

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/README.md

@@ -37,16 +37,16 @@ This leads to:
 
 ### Supported Triple Stores
 
-| Database | Status | Strengths |
-|----------|--------|-----------|
+| Database | Status | Notes |
+|----------|--------|-------|
 | **Apache Jena** | 🟢 Working | Robust, standards compliant |
 | **Blazegraph** | 🟢 Working | High performance, easy setup |
-| **GraphDB** | 🟢 Working | Enterprise features, reasoning - Free Version tested|
+| **Oxigraph** | 🟢 Working | Rust-based, embedded, fast |
 | **QLever** | 🟢 Working | Extremely fast queries |
-| **Virtuoso** | 🛑 In Progress | Mature, SQL integration |
-| **MillenniumDB** | 🛑 In Progress | Property Graph + RDF; SPARQL 1.1 (RDF) and MQL (PG); import-first workflow |
-| **Oxigraph** | 🛑 In Progress | Rust-based, embedded |
-| **Stardog** | 🛑 Planned | Knowledge graphs, reasoning |
+| **GraphDB** | 🛑 Disabled | License required (free version available) |
+| **MillenniumDB** | 🛑 Disabled | Import-first workflow, Property Graph + RDF |
+| **Virtuoso** | 🛑 Disabled | Permissions issue (#30) |
+| **Stardog** | 🛑 Planned | License required |
 
 Whether you're building a semantic web application, conducting research, or evaluating different triple stores, pyomnigraph eliminates the complexity of working with multiple graph database systems.
 
@@ -57,10 +57,11 @@ omnigraph omnigraph --list --include-inactive --doc-format  github
 | Active   | Name                                | Container Name       | Wikidata                                               | Image                                                                                            |   Port |   Test Port | Dataset   | User   |
 |----------|-------------------------------------|----------------------|--------------------------------------------------------|--------------------------------------------------------------------------------------------------|--------|-------------|-----------|--------|
 | 🟢️       | [blazegraph](http://localhost:9898) | blazegraph-omnigraph | [Q20127748](https://www.wikidata.org/wiki/Q20127748)   | [lyrasis/blazegraph:2.1.5](https://hub.docker.com/r/lyrasis/blazegraph)                          |   9898 |        7898 | kb        |        |
-| 🟢       | [graphdb](http://localhost:7200)    | graphdb-omnigraph    | [Q58425577](https://www.wikidata.org/wiki/Q58425577)   | [ontotext/graphdb:9.11.2-se](https://hub.docker.com/r/ontotext/graphdb)                          |   7200 |        7700 | repo1     |        |
 | 🟢️       | [jena](http://localhost:3030)       | jena-omnigraph       | [Q109376461](https://www.wikidata.org/wiki/Q109376461) | [stain/jena-fuseki:latest](https://hub.docker.com/r/stain/jena-fuseki)                           |   3030 |        7030 | ds        | admin  |
-| 🛑       | [oxigraph](http://localhost:7878)   | oxigraph-omnigraph   | [Q118980507](https://www.wikidata.org/wiki/Q118980507) | [oxigraph/oxigraph:latest](https://hub.docker.com/r/oxigraph/oxigraph)                           |   7878 |        7378 | default   |        |
+| 🟢️       | [oxigraph](http://localhost:7878)   | oxigraph-omnigraph   | [Q118980507](https://www.wikidata.org/wiki/Q118980507) | [oxigraph/oxigraph:latest](https://hub.docker.com/r/oxigraph/oxigraph)                           |   7878 |        7378 | default   |        |
 | 🟢️       | [qlever](http://localhost:7019)     | qlever-omnigraph     | [Q111016295](https://www.wikidata.org/wiki/Q111016295) | [adfreiburg/qlever:latest](https://hub.docker.com/r/adfreiburg/qlever)                           |   7019 |        7819 | olympics  |        |
+| 🛑       | [graphdb](http://localhost:7200)    | graphdb-omnigraph    | [Q58425577](https://www.wikidata.org/wiki/Q58425577)   | [ontotext/graphdb:11.2.0](https://hub.docker.com/r/ontotext/graphdb)                             |   7200 |        7700 | repo1     |        |
+| 🛑       | [millenniumdb](http://localhost:1234)| millenniumdb-omnigraph| [Q118954995](https://www.wikidata.org/wiki/Q118954995) | [imfd/millenniumdb](https://hub.docker.com/r/imfd/millenniumdb)                                  |   1234 |        1734 | default   |        |
 | 🛑       | [stardog](http://localhost:5820)    | stardog-omnigraph    | [Q91147741](https://www.wikidata.org/wiki/Q91147741)   | [stardog/stardog:latest](https://hub.docker.com/r/stardog/stardog)                               |   5820 |        5320 | mydb      | admin  |
 | 🛑       | [virtuoso](http://localhost:8890)   | virtuoso-omnigraph   | [Q7935239](https://www.wikidata.org/wiki/Q7935239)     | [openlink/virtuoso-opensource-7:latest](https://hub.docker.com/r/openlink/virtuoso-opensource-7) |   8890 |        8390 | KB        | dba    |
 

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/docs/index.md

@@ -1,4 +1,4 @@
-# omnigraph API Documentation
+# pyomnigraph API Documentation
 
 ::: omnigraph
     options:

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/mkdocs.yml

@@ -1,4 +1,4 @@
-site_name: omnigraph API Documentation
+site_name: pyomnigraph API Documentation
 theme:
   name: material
 plugins:

pyomnigraph-0.2.2/omnigraph/__init__.py

@@ -0,0 +1 @@
+__version__ = "0.2.2"

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/omnigraph/omnigraph_cmd.py

@@ -123,6 +123,10 @@ class OmnigraphCmd(BaseCmd):
         """
         if dataset is None:
             server.config.dumps_dir = self.ogp.examples_dir
+        elif dataset.rdf_file:
+            # If rdf_file is specified, set dumps_dir to the parent directory of the file
+            rdf_file_path = Path(dataset.rdf_file).expanduser().resolve()
+            server.config.dumps_dir = rdf_file_path.parent
         else:
             server.config.dumps_dir = self.ogp.dumps_dir / dataset.id
 
@@ -167,7 +171,8 @@ class OmnigraphCmd(BaseCmd):
         handled = False
         if cmds:
             for cmd in cmds:
-                if cmd == "load":
+                # issue #38: upload needs the per-dataset dumps_dir just like load
+                if cmd in ("load", "upload"):
                     cmd_iterator = self.load_iterator(server)
                 else:
                     cmd_iterator = iter([None])  # Single iteration

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/omnigraph/omniserver.py

@@ -70,6 +70,7 @@ class OmniServer:
             "start": lambda s: ServerCmd(title("start", s), s.start),
             "status": lambda s: ServerCmd(title("status", s), s.status_info),
             "stop": lambda s: ServerCmd(title("stop", s), s.stop),
+            "upload": lambda s: ServerCmd(title("native bulk upload dumps", s), s.upload_dump_files),
             "webui": lambda s: ServerCmd(title("webui", s), s.webui),
         }
         return server_cmds

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/omnigraph/rdf_dataset.py

@@ -8,6 +8,7 @@ from dataclasses import dataclass, field
 from typing import Dict, Optional
 
 from basemkit.yamlable import lod_storable
+from lodstorage.params import Params
 from lodstorage.query import Query
 from lodstorage.sparql import SPARQL
 
@@ -19,38 +20,80 @@ class RdfDataset:
     """
 
     name: str  # Human-readable dataset name
-    base_url: str  # Base URL e.g. for tryit
-    endpoint_url: str  # SPARQL endpoint URL
+    base_url: Optional[str] = None  # Base URL e.g. for tryit
+    endpoint_url: Optional[str] = None  # SPARQL endpoint URL
     description: Optional[str] = None  # Optional dataset description
     database: Optional[str] = "jena"  # the database type of the endpoint
     expected_solutions: Optional[int] = None  # Expected number of solutions
     select_pattern: str = "?s ?p ?o"  # Basic Graph Pattern for queries
     construct_template: Optional[str] = field(default="?s ?p ?o")
+    params: Optional[Dict[str, str]] = None  # npq parameters for {{ name }} templates (issue #36)
     prefix_sets: Optional[list] = field(default_factory=list)
     active: Optional[bool] = False
+    rdf_file: Optional[str] = None  # Path to local RDF file (alternative to endpoint_url)
     # fields to be configured by post_init
     id: Optional[str] = field(default=None)
     count_query: Optional[Query] = field(default=None)
     select_query: Optional[Query] = field(default=None)
     sparql: Optional[SPARQL] = field(default=None)
 
+    def apply_params(self, params_dict: Optional[Dict[str, str]] = None) -> None:
+        """
+        Apply npq parameters to select_pattern and construct_template.
+
+        {{ name }} templates are replaced with the values of the merged
+        params: the dataset's own params overridden by the given ones.
+        Substitution always starts from the raw templates, so CLI overrides
+        can be re-applied after YAML defaults. The blanket Params audit is
+        off - it rejects quoted VALUES lists (see issue #36); values stem
+        from local configuration and CLI.
+
+        Args:
+            params_dict: additional parameter values overriding self.params
+        """
+        # preserve the raw templates on first call
+        if not hasattr(self, "raw_templates"):
+            self.raw_templates = {
+                "select_pattern": self.select_pattern,
+                "construct_template": self.construct_template,
+            }
+        merged = dict(self.params) if self.params else {}
+        if params_dict:
+            merged.update(params_dict)
+        if merged:
+            for attr, raw_template in self.raw_templates.items():
+                if raw_template:
+                    params = Params(raw_template, with_audit=False)
+                    if params.has_params:
+                        params.set(merged)
+                        setattr(self, attr, params.apply_parameters())
+
+    def build_queries(self) -> None:
+        """
+        (Re)build count_query and select_query from the current select_pattern.
+        Only initializes SPARQL-related fields if endpoint_url is provided.
+        """
+        if self.endpoint_url:
+            self.count_query = Query(
+                name=f"{self.name}_count",
+                query=f"SELECT (COUNT(*) AS ?count) WHERE {{ {self.select_pattern} }}",
+                endpoint=self.endpoint_url,
+                description=f"Count query for {self.name}",
+            )
+            self.select_query = Query(
+                name=f"{self.name}_select",
+                query=f"SELECT * WHERE {{ {self.select_pattern} }}",
+                endpoint=self.endpoint_url,
+                description=f"Select query for {self.name}",
+            )
+            self.sparql = SPARQL(self.endpoint_url)
+
     def __post_init__(self):
         """
-        Generate count_query and construct_pattern from select_pattern.
+        Apply npq params and generate the queries from select_pattern.
         """
-        self.count_query = Query(
-            name=f"{self.name}_count",
-            query=f"SELECT (COUNT(*) AS ?count) WHERE {{ {self.select_pattern} }}",
-            endpoint=self.endpoint_url,
-            description=f"Count query for {self.name}",
-        )
-        self.select_query = Query(
-            name=f"{self.name}_select",
-            query=f"SELECT * WHERE {{ {self.select_pattern} }}",
-            endpoint=self.endpoint_url,
-            description=f"Select query for {self.name}",
-        )
-        self.sparql = SPARQL(self.endpoint_url)
+        self.apply_params()
+        self.build_queries()
 
     @property
     def full_name(self):
@@ -66,6 +109,8 @@ class RdfDataset:
             Number of solutions available from the count query
         """
         count = self.sparql.getValue(self.count_query.query, "count")
+        # issue #37: getValue returns the raw literal which may be a str
+        count = int(count)
         return count
 
     def getTryItUrl(self, database: str = "blazegraph") -> str:

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/omnigraph/rdfdump_cmd.py

@@ -10,6 +10,8 @@ import os
 import webbrowser
 from argparse import ArgumentParser, Namespace
 
+from basemkit.argparse_action import StoreDictKeyPair
+
 from omnigraph.basecmd import BaseCmd
 from omnigraph.rdf_dataset import RdfDataset, RdfDatasets
 from omnigraph.rdfdump import RdfDumpDownloader
@@ -61,6 +63,13 @@ class RdfDumpCmd(BaseCmd):
             default=None,
             help="Maximum number of solutions/triples to download (uses dataset expected_solutions if not specified)",
         )
+        parser.add_argument(
+            "--param",
+            action=StoreDictKeyPair,
+            dest="params",
+            metavar="KEY=VALUE,KEY=VALUE...",
+            help="npq parameter values overriding the dataset's params (issue #36)",
+        )
         parser.add_argument("--no-progress", action="store_true", help="Disable progress bar")
         parser.add_argument("--output-path", default=".", help="Path for dump files")
         parser.add_argument("--tryit", action="store_true", help="open the try it! URL [default: %(default)s]")
@@ -97,6 +106,11 @@ class RdfDumpCmd(BaseCmd):
         """
         super().handle_args(args)
         datasets = self.datasets
+        if getattr(args, "params", None):
+            # apply CLI npq parameter overrides and rebuild the queries
+            for dataset in datasets.values():
+                dataset.apply_params(args.params)
+                dataset.build_queries()
         if self.args.about:
             self.about()
 

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/omnigraph/resources/examples/datasets.yaml

@@ -65,4 +65,40 @@ datasets:
         ?pName ?p4Name ?o4Name
     # Optional dataset description
     description: "W2306 postal code coordinates"
-    active: false
+    active: false
+  papers_of_my_network_wikidata:
+    # npq (named parameterized query) example - issue #36
+    # ORCID is the join key; {{ orcids }} is the VALUES scaling knob;
+    # endpoint is late-bound - own copy answers completely and fastest,
+    # see the endpoint matrix in the issue
+    name: "Papers of my network (wikidata)"
+    base_url: "https://qlever-api.wikidata.dbis.rwth-aachen.de"
+    endpoint_url: "https://qlever-api.wikidata.dbis.rwth-aachen.de/"
+    database: "qlever"
+    select_pattern: |
+      VALUES ?orcid { {{ orcids }} }
+      ?author <http://www.wikidata.org/prop/direct/P496> ?orcid .
+      ?work <http://www.wikidata.org/prop/direct/P50> ?author .
+      ?work ?p ?o
+    construct_template: "?work ?p ?o"
+    params:
+      orcids: '"0000-0002-0821-6995" "0000-0003-2533-6363" "0000-0001-9879-3827" "0000-0001-6324-7164"'
+    description: "Papers authored by the co-author network, seeded by ORCIDs"
+    active: false
+  papers_of_my_network_dblp:
+    # npq companion - same {{ orcids }} param drives the dblp graph
+    name: "Papers of my network (dblp)"
+    base_url: "https://sparql.dblp.org"
+    endpoint_url: "https://sparql.dblp.org/sparql"
+    database: "qlever"
+    select_pattern: |
+      VALUES ?orcid { {{ orcids }} }
+      BIND(IRI(CONCAT("https://orcid.org/", ?orcid)) AS ?orcid_uri)
+      ?author <https://dblp.org/rdf/schema#orcid> ?orcid_uri .
+      ?paper <https://dblp.org/rdf/schema#authoredBy> ?author .
+      ?paper ?p ?o
+    construct_template: "?paper ?p ?o"
+    params:
+      orcids: '"0000-0002-0821-6995" "0000-0003-2533-6363" "0000-0001-9879-3827" "0000-0001-6324-7164"'
+    description: "Papers authored by the co-author network in dblp, same ORCID seed"
+    active: false

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/omnigraph/resources/examples/servers.yaml

@@ -14,6 +14,7 @@ servers:
     port: 9898
     test_port: 7898
     host: "localhost"
+    # docker_bind: "0.0.0.0"  # Optional: Uncomment to allow external access (default: "127.0.0.1" for localhost-only, recommended for security)
     prefix_sets: ["rdf", "blazegraph","gov","wikidata"]
     active: true
   graphdb:

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/omnigraph/server_config.py

@@ -161,6 +161,7 @@ class ServerConfig:
     has_license: bool = False
     protocol: str = "http"
     host: str = "localhost"
+    docker_bind: str = "127.0.0.1"  # Docker port bind address (default: localhost-only for security)
     rdf_format: str = "turtle"
     auth_user: Optional[str] = None
     auth_password: Optional[str] = None

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/omnigraph/servers/blazegraph.py

@@ -8,6 +8,8 @@ from dataclasses import dataclass
 from pathlib import Path
 import re
 import os
+import shutil
+
 from omnigraph.server_config import ServerLifecycleState, ServerStatus
 from omnigraph.sparql_server import ServerConfig, ServerEnv, SparqlServer
 
@@ -42,7 +44,7 @@ class BlazegraphConfig(ServerConfig):
             f"docker run -d --name {self.container_name} "
             f"-e BLAZEGRAPH_UID={os.getuid()} "
             f"-e BLAZEGRAPH_GID={os.getgid()} "
-            f"-p {self.port}:8080 "
+            f"-p {self.docker_bind}:{self.port}:8080 "
             f"-v {data_dir}/RWStore.properties:/RWStore.properties "
             f"-v {data_dir}:/data "
             f"{self.image}"
@@ -94,6 +96,68 @@ class Blazegraph(SparqlServer):
             rwstore_path.write_text(props)
 
 
+    def get_dataloader_xml(self, container_path: str) -> str:
+        """
+        Build the DataLoader servlet properties XML for the given
+        container-side file or directory path.
+
+        Args:
+            container_path: path as seen inside the container (mounted /data)
+
+        Returns:
+            the properties XML document
+        """
+        xml = f"""<?xml version="1.0" encoding="UTF-8" standalone="no"?>
+<!DOCTYPE properties SYSTEM "http://java.sun.com/dtd/properties.dtd">
+<properties>
+  <entry key="namespace">{self.config.dataset}</entry>
+  <entry key="propertyFile">/RWStore.properties</entry>
+  <entry key="fileOrDirs">{container_path}</entry>
+  <entry key="-durableQueues">true</entry>
+</properties>"""
+        return xml
+
+    def upload_dump_files(self, file_pattern: str = None) -> int:
+        """
+        Bulk-load dump files via Blazegraph's REST DataLoader servlet, which
+        reads the files from the server's own filesystem (the mounted /data
+        directory) instead of pushing them through HTTP request bodies.
+
+        Args:
+            file_pattern: Glob pattern for dump files
+
+        Returns:
+            Number of files loaded successfully
+        """
+        container_name = self.config.container_name
+        files = self.get_dump_files(file_pattern)
+        loaded_count = 0
+        if not files:
+            self.log.log("⚠️", container_name, f"No dump files found for pattern: {file_pattern}")
+        else:
+            # stage the dumps under data_dir so the container sees them at /data/dumps
+            stage_dir = Path(self.config.base_data_dir) / "dumps"
+            stage_dir.mkdir(parents=True, exist_ok=True)
+            for file in files:
+                target = stage_dir / file.name
+                if not target.exists():
+                    shutil.copy2(file, target)
+            xml = self.get_dataloader_xml("/data/dumps")
+            response = self.make_request(
+                "POST",
+                self.config.dataloader_url,
+                headers={"Content-Type": "application/xml"},
+                data=xml,
+                timeout=self.config.upload_timeout,
+            )
+            if response.success:
+                loaded_count = len(files)
+                self.log.log("✅", container_name, f"DataLoader loaded {loaded_count} file(s)")
+            else:
+                error_msg = str(response.error) if response.error else f"HTTP {response.response.status_code}"
+                self.log.log("❌", container_name, f"DataLoader failed: {error_msg}")
+        return loaded_count
+
     def status(self) -> ServerStatus:
         """
         Get server status information.

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/omnigraph/servers/graphdb.py

@@ -60,7 +60,7 @@ class GraphDBConfig(ServerConfig):
         # 4. Construct Command
         docker_run_command = (
             f"docker run {env_str} -d --name {self.container_name} "
-            f"-p 127.0.0.1:{self.port}:7200 "
+            f"-p {self.docker_bind}:{self.port}:7200 "
             f"-v {data_dir}:/opt/graphdb/home "
             f"{target_image}"
         )

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/omnigraph/servers/jena.py

@@ -7,6 +7,8 @@ Apache Jena SPARQL support
 """
 
 from dataclasses import dataclass
+from pathlib import Path
+from typing import List
 
 from omnigraph.server_config import ServerLifecycleState, ServerStatus
 from omnigraph.sparql_server import ServerConfig, ServerEnv, SparqlServer
@@ -48,7 +50,7 @@ class JenaConfig(ServerConfig):
             env = f"{env} -e ADMIN_PASSWORD={self.auth_password}"
         docker_run_command = (
             f"docker run {self.docker_user_flag} {env} -d --name {self.container_name} "
-            f"-p {self.port}:3030 "
+            f"-p {self.docker_bind}:{self.port}:3030 "
             f"-v {data_dir}:/fuseki "
             f"{self.image}"
         )
@@ -96,7 +98,65 @@ class Jena(SparqlServer):
         Returns:
             Tuple of (response, exception)
         """
-        result,error=self.execute_update_query_with_post(update_query)
-        return result,error
+        result, error = self.execute_update_query_with_post(update_query)
+        return result, error
+
+    def get_tdbloader_command(self, files: List[Path]) -> str:
+        """
+        Build the tdb2.tdbloader docker command for the given dump files.
+
+        The loader classes ship inside fuseki-server.jar, so the server's own
+        image is reused with an overridden entrypoint. The TDB2 dataset is
+        single-writer - the server must be stopped while the loader runs.
+
+        Args:
+            files: dump files to load (from the dumps directory)
+
+        Returns:
+            the docker run command string
+        """
+        loc = f"/fuseki/databases/{self.config.dataset}"
+        dumps_dir = Path(self.config.dumps_dir)
+        file_args = " ".join(f"/dumps/{file.name}" for file in files)
+        command = (
+            f"docker run --rm {self.config.docker_user_flag} --entrypoint java "
+            f"-v {self.config.base_data_dir}:/fuseki "
+            f"-v {dumps_dir}:/dumps "
+            f"{self.config.image} "
+            f"-cp /jena-fuseki/fuseki-server.jar tdb2.tdbloader "
+            f"--loc {loc} {file_args}"
+        )
+        return command
+
+    def upload_dump_files(self, file_pattern: str = None) -> int:
+        """
+        Bulk-load dump files with tdb2.tdbloader directly into the TDB2
+        database files - the native path for the file-backed store, avoiding
+        the HTTP single-POST transaction that faults the mmap'd node table
+        (issue #25).
+
+        Args:
+            file_pattern: Glob pattern for dump files
+
+        Returns:
+            Number of files loaded successfully
+        """
+        container_name = self.config.container_name
+        files = self.get_dump_files(file_pattern)
+        loaded_count = 0
+        if not files:
+            self.log.log("⚠️", container_name, f"No dump files found for pattern: {file_pattern}")
+        else:
+            self.stop()
+            loader_cmd = self.get_tdbloader_command(files)
+            shell_result = self.run_shell_command(
+                loader_cmd,
+                success_msg=f"tdb2.tdbloader loaded {len(files)} file(s)",
+                error_msg="tdb2.tdbloader failed",
+            )
+            if shell_result.success:
+                loaded_count = len(files)
+            self.start()
+        return loaded_count
 
 

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/omnigraph/servers/millenniumdb.py

@@ -62,8 +62,8 @@ class MillenniumDBConfig(ServerConfig):
 
         docker_run_command = (
             f"docker run -d --name {self.container_name} "
-            f"-p {self.sparql_port}:1234 "
-            f"-p {self.web_port}:4321 "
+            f"-p {self.docker_bind}:{self.sparql_port}:1234 "
+            f"-p {self.docker_bind}:{self.web_port}:4321 "
             f"-v {data_dir}:/data "
             f"{self.image} "
             f"mdb-server {db_path}"

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/omnigraph/servers/oxigraph.py

@@ -47,7 +47,7 @@ class OxigraphConfig(ServerConfig):
         """
         docker_run_command = (
             f"docker run {self.docker_user_flag} -d --name {self.container_name} "
-            f"-p {self.port}:7878 "
+            f"-p {self.docker_bind}:{self.port}:7878 "
             f"-v {data_dir}:/data "
             f"{self.image} serve --bind 0.0.0.0:7878 --location /data"
         )

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/omnigraph/servers/qlever.py

@@ -5,6 +5,7 @@ Created on 2025-05-28
 """
 
 import os
+import shutil
 from configparser import ConfigParser, ExtendedInterpolation
 from dataclasses import dataclass
 from pathlib import Path
@@ -238,6 +239,71 @@ class QLever(SparqlServer):
 
         return None
 
+    def get_index_commands(self, files: List[Path]) -> List[str]:
+        """
+        Build the qlever CLI commands to (re)index the given dump files.
+
+        QLever loads bulk data by building its index from files - there is no
+        native HTTP bulk-write path; the SPARQL INSERT route is only suitable
+        for small increments.
+
+        Args:
+            files: dump files staged in the data directory
+
+        Returns:
+            list of shell commands to run in the data directory
+        """
+        commands = [
+            "qlever stop",
+            "qlever index --overwrite-existing",
+            f"qlever start --server-container {self.config.container_name}",
+        ]
+        return commands
+
+    def upload_dump_files(self, file_pattern: str = None) -> int:
+        """
+        Bulk-load dump files by rebuilding the QLever index from them -
+        the native path; the turtle-to-INSERT conversion of upload_request
+        is unsuitable for bulk data.
+
+        Args:
+            file_pattern: Glob pattern for dump files
+
+        Returns:
+            Number of files loaded successfully
+        """
+        container_name = self.config.container_name
+        files = self.get_dump_files(file_pattern)
+        loaded_count = 0
+        if not files:
+            self.log.log("⚠️", container_name, f"No dump files found for pattern: {file_pattern}")
+            return loaded_count
+        data_dir = Path(self.config.data_dir)
+        qleverfile_path = data_dir / "Qleverfile"
+        qlever_file = QLeverfile.ofFile(qleverfile_path)
+        if qlever_file is None:
+            self.log.log("❌", container_name, f"no Qleverfile in {data_dir} - run start (setup-config) first")
+            return loaded_count
+        # stage the dumps in the data directory and register them as input files
+        for file in files:
+            target = data_dir / file.name
+            if not target.exists():
+                shutil.copy2(file, target)
+        input_files = " ".join(file.name for file in files)
+        qlever_file.set("index", "INPUT_FILES", input_files)
+        qlever_file.save()
+        ok = True
+        for command in self.get_index_commands(files):
+            shell_result = self.run_shell_command(f"cd {data_dir};{command}")
+            if not shell_result.success:
+                self.log.log("❌", container_name, f"failed: {command}")
+                ok = False
+                break
+        if ok:
+            loaded_count = len(files)
+            self.log.log("✅", container_name, f"index rebuilt from {loaded_count} file(s)")
+        return loaded_count
+
     def upload_request(self, file_content: bytes) -> Response:
         """Upload request for QLever using SPARQL INSERT statements."""
         turtle_data = file_content.decode("utf-8")

{pyomnigraph-0.2.0 → pyomnigraph-0.2.2}/omnigraph/servers/stardog.py

@@ -62,7 +62,7 @@ class StardogConfig(ServerConfig):
         # 4. Construct Command using target_image
         docker_run_command = (
             f"docker run {self.docker_user_flag}{env_str} -d --name {self.container_name} "
-            f"-p {self.port}:5820 "
+            f"-p {self.docker_bind}:{self.port}:5820 "
             f"-v {data_dir}:/var/opt/stardog "
             f"{target_image}"
         )