sadwitdastreetz commented on code in PR #683:
URL:
https://github.com/apache/incubator-hugegraph-toolchain/pull/683#discussion_r2459053335
##########
hugegraph-loader/src/main/java/org/apache/hugegraph/loader/HugeGraphLoader.java:
##########
@@ -153,11 +270,286 @@ private void createSchema() {
throw new LoadException("Failed to read schema file '%s'", e,
options.schema);
}
- groovyExecutor.execute(script, client);
+
+ if (!options.shorterIDConfigs.isEmpty()) {
+ for (ShortIdConfig config : options.shorterIDConfigs) {
+ PropertyKey propertyKey =
client.schema().propertyKey(config.getIdFieldName())
+ .ifNotExist()
+
.dataType(config.getIdFieldType())
+ .build();
+ client.schema().addPropertyKey(propertyKey);
+ }
+ groovyExecutor.execute(script, client);
+ List<VertexLabel> vertexLabels =
client.schema().getVertexLabels();
+ for (VertexLabel vertexLabel : vertexLabels) {
+ ShortIdConfig config;
+ if ((config =
options.getShortIdConfig(vertexLabel.name())) != null) {
+ config.setLabelID(vertexLabel.id());
+ IndexLabel indexLabel = client.schema()
+
.indexLabel(config.getVertexLabel() + "By" +
+
config.getIdFieldName())
+
.onV(config.getVertexLabel())
+
.by(config.getIdFieldName())
+ .secondary()
+ .ifNotExist()
+ .build();
+ client.schema().addIndexLabel(indexLabel);
+ }
+ }
+ } else {
+ groovyExecutor.execute(script, client);
+ }
+ }
+
+ // create schema for Graph Source
+ List<InputStruct> structs = this.mapping.structs();
+ for (InputStruct struct : structs) {
+ if (SourceType.GRAPH.equals(struct.input().type())) {
+ GraphSource graphSouce = (GraphSource) struct.input();
+ if (StringUtils.isEmpty(graphSouce.getPdPeers())) {
+ graphSouce.setPdPeers(this.options.pdPeers);
+ }
+ if (StringUtils.isEmpty(graphSouce.getMetaEndPoints())) {
+ graphSouce.setMetaEndPoints(this.options.metaEndPoints);
+ }
+ if (StringUtils.isEmpty(graphSouce.getCluster())) {
+ graphSouce.setCluster(this.options.cluster);
+ }
+ if (StringUtils.isEmpty(graphSouce.getUsername())) {
+ graphSouce.setUsername(this.options.username);
+ }
+ if (StringUtils.isEmpty(graphSouce.getPassword())) {
+ graphSouce.setPassword(this.options.password);
+ }
+
+ GraphSource graphSource = (GraphSource) struct.input();
+ createGraphSourceSchema(graphSource);
+ }
}
+
this.context.updateSchemaCache();
}
+ /**
+ * create schema like graphdb when source is graphdb;
+ *
+ * @param graphSource
+ */
+ private void createGraphSourceSchema(GraphSource graphSource) {
+
+ try (HugeClient sourceClient = graphSource.createHugeClient();
+ HugeClient client = HugeClientHolder.create(this.options, false))
{
+
+ createGraphSourceVertexLabel(sourceClient, client, graphSource);
+ createGraphSourceEdgeLabel(sourceClient, client, graphSource);
+ createGraphSourceIndexLabel(sourceClient, client, graphSource);
+ }
+ }
+
+ private void createGraphSourceVertexLabel(HugeClient sourceClient,
+ HugeClient targetClient,
+ GraphSource graphSource) {
+
+ sourceClient.assignGraph(graphSource.getGraphSpace(),
+ graphSource.getGraph());
+
+ // Create Vertex Schema
+ List<VertexLabel> vertexLabels = new ArrayList<>();
+ if (graphSource.getSelectedVertices() != null) {
+ List<String> selectedVertexLabels =
+ graphSource.getSelectedVertices()
+ .stream().map((des) -> des.getLabel())
+ .collect(Collectors.toList());
+
+ if (!CollectionUtils.isEmpty(selectedVertexLabels)) {
+ vertexLabels =
+ sourceClient.schema()
+ .getVertexLabels(selectedVertexLabels);
+ }
+ } else {
+ vertexLabels = sourceClient.schema().getVertexLabels();
+ }
+
+ Map<String, GraphSource.SeletedLabelDes> mapSelectedVertices
+ = new HashMap<>();
+ if (graphSource.getSelectedVertices() != null) {
+ for (GraphSource.SeletedLabelDes des :
+ graphSource.getSelectedVertices()) {
+ mapSelectedVertices.put(des.getLabel(), des);
+ }
+ }
+
+ for (VertexLabel label : vertexLabels) {
+ if (mapSelectedVertices.getOrDefault(label.name(),
+ null) != null) {
+ List<String> selectedProperties = mapSelectedVertices.get(
+ label.name()).getProperties();
+
+ if (selectedProperties != null) {
+ label.properties().clear();
+ label.properties().addAll(selectedProperties);
+ }
+ }
+ }
+
+ Map<String, GraphSource.IgnoredLabelDes> mapIgnoredVertices
+ = new HashMap<>();
+ if (graphSource.getIgnoredVertices() != null) {
+ for (GraphSource.IgnoredLabelDes des :
+ graphSource.getIgnoredVertices()) {
+ mapIgnoredVertices.put(des.getLabel(), des);
+ }
+ }
+
+ for (VertexLabel vertexLabel : vertexLabels) {
+ if (mapIgnoredVertices.containsKey(vertexLabel.name())) {
+ GraphSource.IgnoredLabelDes des
+ = mapIgnoredVertices.get(vertexLabel.name());
+
+ if (des.getProperties() != null) {
+ des.getProperties()
+ .forEach((p) -> vertexLabel.properties().remove(p));
+ }
+ }
+
+ Set<String> existedPKs =
+ targetClient.schema().getPropertyKeys().stream()
+ .map(pk ->
pk.name()).collect(Collectors.toSet());
+
+ for (String pkName : vertexLabel.properties()) {
+ PropertyKey pk = sourceClient.schema()
+ .getPropertyKey(pkName);
+ if (!existedPKs.contains(pk.name())) {
+ targetClient.schema().addPropertyKey(pk);
+ }
Review Comment:
The schema creation here is already idempotent — each propertyKey,
vertexLabel, and indexLabel uses .ifNotExist(), so re-running won’t cause
duplication or conflicts.
HugeGraph’s schema API doesn’t support transactions, so partial creation
rollback isn’t feasible.
Given that schema setup happens only once before data loading, this behavior
is acceptable.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
