codeant-ai-for-open-source[bot] commented on code in PR #40116: URL: https://github.com/apache/superset/pull/40116#discussion_r3240562866
########## docs/superset_relationship_design.md: ########## @@ -0,0 +1,639 @@ +# Dataset Relationship Engine — Design Document + +**Projeto:** hibi (fork Apache Superset) +**Data:** 2026-05-14 +**Autor:** Abacus AI Agent +**Status:** Proposta (Draft) + +--- + +## 1. Resumo Executivo + +Este documento descreve o design técnico para implementação de um **motor de relacionamentos entre datasets** no fork `hibi` do Apache Superset. O objetivo é permitir que dashboards combinem dados de múltiplos datasets — inclusive de **bancos de dados diferentes** — através de JOINs, drill-downs e propagação de filtros. + +### Problema + +Atualmente o Superset trata cada dataset como uma entidade isolada. Para combinar dados de fontes diferentes, o usuário precisa criar views SQL manualmente ou usar datasets virtuais. Não há suporte nativo para: + +- Relacionamentos declarativos entre datasets +- JOINs automáticos entre tabelas do mesmo banco +- Merge de dados entre bancos diferentes (cross-database) +- Drill-down hierárquico entre datasets relacionados +- Propagação automática de filtros entre charts com datasets diferentes + +### Solução Proposta + +Um engine de relacionamentos em 3 fases: + +1. **Phase 1 — Backend & Engine**: Novas tabelas, API e dual-mode JOIN engine +2. **Phase 2 — Frontend & Model View**: Canvas visual para modelagem de relacionamentos +3. **Phase 3 — Advanced Interactions**: Cross-filtering, drill-down e propagação de filtros + +--- + +## 2. Phase 1 — Backend & Engine + +### 2.1 Novas Tabelas de Banco de Dados + +#### Tabela `dataset_relationships` + +```sql +CREATE TABLE dataset_relationships ( + id INTEGER PRIMARY KEY AUTOINCREMENT, + uuid VARCHAR(36) UNIQUE NOT NULL, + + -- Source dataset + source_dataset_id INTEGER NOT NULL REFERENCES ab_datasets(id) ON DELETE CASCADE, + + -- Target dataset + target_dataset_id INTEGER NOT NULL REFERENCES ab_datasets(id) ON DELETE CASCADE, + Review Comment: **🟠 Architect Review — HIGH** The design uses inconsistent dataset table names: the initial DDL references `ab_datasets(id)` while the ORM and migration snippets use `tables.id`, conflicting with the actual Superset dataset model (`SqlaTable.__tablename__ = "tables"`). **Suggestion:** Normalize the design to reference the real dataset table (`tables.id` for `SqlaTable`) consistently across the DDL, ORM model, and migration examples to avoid implementing against a non-existent `ab_datasets` table. [Fix in Cursor](https://app.codeant.ai/fix-in-ide?tool=cursor&prompt=This%20is%20an%20%2A%2AArchitect%20%2F%20Logical%20Review%2A%2A%20comment%20left%20during%20a%20code%20review.%20These%20reviews%20are%20first-class%2C%20important%20findings%20%E2%80%94%20not%20optional%20suggestions.%20Do%20NOT%20dismiss%20this%20as%20a%20%27big%20architectural%20change%27%20just%20because%20the%20title%20says%20architect%20review%3B%20most%20of%20these%20can%20be%20resolved%20with%20a%20small%2C%20localized%20fix%20once%20the%20intent%20is%20understood.%0A%0A%2A%2APath%3A%2A%2A%20docs%2Fsuperset_relationship_design.md%0A%2A%2ALine%3A%2A%2A%2046%3A50%0A%2A%2AComment%3A%2A%2A%0A%09%2AHIGH%3A%20The%20design%20uses%20inconsistent%20dataset%20table%20names%3A%20the%20initial%20DDL%20references%20%60ab_datasets%28id%29%60%20while%20the%20ORM%20and%20migration%20snippets%20use%20%60tables.id%60%2C%20conflicting%20with%20the%20actual%20Superset%20dataset%20model%20%28%60SqlaTable.__tablename__%20%3D%2 0%22tables%22%60%29.%0A%0AValidate%20the%20correctness%20of%20the%20flagged%20issue.%20If%20correct%2C%20How%20can%20I%20resolve%20this%3F%20If%20you%20propose%20a%20fix%2C%20implement%20it%20and%20please%20make%20it%20concise.%0AIf%20a%20suggested%20approach%20is%20provided%20above%2C%20use%20it%20as%20the%20authoritative%20instruction.%20If%20no%20explicit%20code%20suggestion%20is%20given%2C%20you%20MUST%20still%20draft%20and%20apply%20your%20own%20minimal%2C%20localized%20fix%20%E2%80%94%20do%20not%20punt%20back%20with%20%27no%20suggestion%20provided%2C%20review%20manually%27.%20Keep%20the%20change%20as%20small%20as%20possible%3A%20add%20a%20guard%20clause%2C%20gate%20on%20a%20loading%20state%2C%20reorder%20an%20await%2C%20wrap%20in%20a%20conditional%2C%20etc.%20Do%20not%20refactor%20surrounding%20code%20or%20expand%20scope%20beyond%20the%20finding.%0AOnce%20fix%20is%20implemented%2C%20also%20check%20other%20comments%20on%20the%20same%20PR%2C%20and%20ask%20user%20if%20the%20user% 20wants%20to%20fix%20the%20rest%20of%20the%20comments%20as%20well.%20if%20said%20yes%2C%20then%20fetch%20all%20the%20comments%20validate%20the%20correctness%20and%20implement%20a%20minimal%20fix%0A) | [Fix in VSCode Claude](https://app.codeant.ai/fix-in-ide?tool=vscode-claude&prompt=This%20is%20an%20%2A%2AArchitect%20%2F%20Logical%20Review%2A%2A%20comment%20left%20during%20a%20code%20review.%20These%20reviews%20are%20first-class%2C%20important%20findings%20%E2%80%94%20not%20optional%20suggestions.%20Do%20NOT%20dismiss%20this%20as%20a%20%27big%20architectural%20change%27%20just%20because%20the%20title%20says%20architect%20review%3B%20most%20of%20these%20can%20be%20resolved%20with%20a%20small%2C%20localized%20fix%20once%20the%20intent%20is%20understood.%0A%0A%2A%2APath%3A%2A%2A%20docs%2Fsuperset_relationship_design.md%0A%2A%2ALine%3A%2A%2A%2046%3A50%0A%2A%2AComment%3A%2A%2A%0A%09%2AHIGH%3A%20The%20design%20uses%20inconsistent%20dataset%20table%20names%3A%20the%20initial%20DDL%20refere nces%20%60ab_datasets%28id%29%60%20while%20the%20ORM%20and%20migration%20snippets%20use%20%60tables.id%60%2C%20conflicting%20with%20the%20actual%20Superset%20dataset%20model%20%28%60SqlaTable.__tablename__%20%3D%20%22tables%22%60%29.%0A%0AValidate%20the%20correctness%20of%20the%20flagged%20issue.%20If%20correct%2C%20How%20can%20I%20resolve%20this%3F%20If%20you%20propose%20a%20fix%2C%20implement%20it%20and%20please%20make%20it%20concise.%0AIf%20a%20suggested%20approach%20is%20provided%20above%2C%20use%20it%20as%20the%20authoritative%20instruction.%20If%20no%20explicit%20code%20suggestion%20is%20given%2C%20you%20MUST%20still%20draft%20and%20apply%20your%20own%20minimal%2C%20localized%20fix%20%E2%80%94%20do%20not%20punt%20back%20with%20%27no%20suggestion%20provided%2C%20review%20manually%27.%20Keep%20the%20change%20as%20small%20as%20possible%3A%20add%20a%20guard%20clause%2C%20gate%20on%20a%20loading%20state%2C%20reorder%20an%20await%2C%20wrap%20in%20a%20conditional%2C%20etc.%20Do%20not %20refactor%20surrounding%20code%20or%20expand%20scope%20beyond%20the%20finding.%0AOnce%20fix%20is%20implemented%2C%20also%20check%20other%20comments%20on%20the%20same%20PR%2C%20and%20ask%20user%20if%20the%20user%20wants%20to%20fix%20the%20rest%20of%20the%20comments%20as%20well.%20if%20said%20yes%2C%20then%20fetch%20all%20the%20comments%20validate%20the%20correctness%20and%20implement%20a%20minimal%20fix%0A) *(Use Cmd/Ctrl + Click for best experience)* <details> <summary><b>Prompt for AI Agent 🤖 </b></summary> ```mdx This is an **Architect / Logical Review** comment left during a code review. These reviews are first-class, important findings — not optional suggestions. Do NOT dismiss this as a 'big architectural change' just because the title says architect review; most of these can be resolved with a small, localized fix once the intent is understood. **Path:** docs/superset_relationship_design.md **Line:** 46:50 **Comment:** *HIGH: The design uses inconsistent dataset table names: the initial DDL references `ab_datasets(id)` while the ORM and migration snippets use `tables.id`, conflicting with the actual Superset dataset model (`SqlaTable.__tablename__ = "tables"`). Validate the correctness of the flagged issue. If correct, How can I resolve this? If you propose a fix, implement it and please make it concise. If a suggested approach is provided above, use it as the authoritative instruction. If no explicit code suggestion is given, you MUST still draft and apply your own minimal, localized fix — do not punt back with 'no suggestion provided, review manually'. Keep the change as small as possible: add a guard clause, gate on a loading state, reorder an await, wrap in a conditional, etc. Do not refactor surrounding code or expand scope beyond the finding. Once fix is implemented, also check other comments on the same PR, and ask user if the user wants to fix the rest of the comments as well. if said yes, then fetch all the comments validate the correctness and implement a minimal fix ``` </details> ########## docs/superset_relationship_design.md: ########## @@ -0,0 +1,639 @@ +# Dataset Relationship Engine — Design Document + +**Projeto:** hibi (fork Apache Superset) +**Data:** 2026-05-14 +**Autor:** Abacus AI Agent +**Status:** Proposta (Draft) + +--- + +## 1. Resumo Executivo + +Este documento descreve o design técnico para implementação de um **motor de relacionamentos entre datasets** no fork `hibi` do Apache Superset. O objetivo é permitir que dashboards combinem dados de múltiplos datasets — inclusive de **bancos de dados diferentes** — através de JOINs, drill-downs e propagação de filtros. + +### Problema + +Atualmente o Superset trata cada dataset como uma entidade isolada. Para combinar dados de fontes diferentes, o usuário precisa criar views SQL manualmente ou usar datasets virtuais. Não há suporte nativo para: + +- Relacionamentos declarativos entre datasets +- JOINs automáticos entre tabelas do mesmo banco +- Merge de dados entre bancos diferentes (cross-database) +- Drill-down hierárquico entre datasets relacionados +- Propagação automática de filtros entre charts com datasets diferentes + +### Solução Proposta + +Um engine de relacionamentos em 3 fases: + +1. **Phase 1 — Backend & Engine**: Novas tabelas, API e dual-mode JOIN engine +2. **Phase 2 — Frontend & Model View**: Canvas visual para modelagem de relacionamentos +3. **Phase 3 — Advanced Interactions**: Cross-filtering, drill-down e propagação de filtros + +--- + +## 2. Phase 1 — Backend & Engine + +### 2.1 Novas Tabelas de Banco de Dados + +#### Tabela `dataset_relationships` + +```sql +CREATE TABLE dataset_relationships ( + id INTEGER PRIMARY KEY AUTOINCREMENT, + uuid VARCHAR(36) UNIQUE NOT NULL, + + -- Source dataset + source_dataset_id INTEGER NOT NULL REFERENCES ab_datasets(id) ON DELETE CASCADE, + + -- Target dataset + target_dataset_id INTEGER NOT NULL REFERENCES ab_datasets(id) ON DELETE CASCADE, + + -- Relationship metadata + relationship_type VARCHAR(20) NOT NULL DEFAULT 'many_to_one', + -- one_to_one, one_to_many, many_to_one, many_to_many + + join_type VARCHAR(10) NOT NULL DEFAULT 'LEFT', + -- LEFT, INNER, RIGHT, FULL + + -- Cross-database flag (computed on save) + is_cross_database BOOLEAN NOT NULL DEFAULT FALSE, + + -- Descriptive + name VARCHAR(256), + description TEXT, + + -- Audit + created_by_fk INTEGER REFERENCES ab_user(id), + changed_by_fk INTEGER REFERENCES ab_user(id), + created_on DATETIME DEFAULT CURRENT_TIMESTAMP, + changed_on DATETIME DEFAULT CURRENT_TIMESTAMP, + + CONSTRAINT uq_dataset_relationship UNIQUE (source_dataset_id, target_dataset_id) +); +``` + +#### Tabela `dataset_relationship_columns` + +```sql +CREATE TABLE dataset_relationship_columns ( + id INTEGER PRIMARY KEY AUTOINCREMENT, + + relationship_id INTEGER NOT NULL + REFERENCES dataset_relationships(id) ON DELETE CASCADE, + + -- Column mapping + source_column VARCHAR(256) NOT NULL, + target_column VARCHAR(256) NOT NULL, + + -- Operator for join condition (default =) + operator VARCHAR(10) NOT NULL DEFAULT '=', + + -- Order of this column pair in multi-column joins + ordinal INTEGER NOT NULL DEFAULT 0, + + CONSTRAINT uq_rel_column_pair UNIQUE (relationship_id, source_column, target_column) +); +``` + +### 2.2 SQLAlchemy Models + +```python +# superset/models/dataset_relationship.py + +class DatasetRelationship(Model, AuditMixinNullable): + """Represents a declared relationship between two datasets.""" + + __tablename__ = "dataset_relationships" + + id = Column(Integer, primary_key=True, autoincrement=True) + uuid = Column(String(36), unique=True, nullable=False, default=generate_uuid) + + source_dataset_id = Column(Integer, ForeignKey("tables.id", ondelete="CASCADE"), nullable=False) + target_dataset_id = Column(Integer, ForeignKey("tables.id", ondelete="CASCADE"), nullable=False) + + relationship_type = Column(String(20), nullable=False, default="many_to_one") + join_type = Column(String(10), nullable=False, default="LEFT") + is_cross_database = Column(Boolean, nullable=False, default=False) + + name = Column(String(256)) + description = Column(Text) + + # Relationships + source_dataset = relationship("SqlaTable", foreign_keys=[source_dataset_id]) + target_dataset = relationship("SqlaTable", foreign_keys=[target_dataset_id]) + column_mappings = relationship( + "DatasetRelationshipColumn", + back_populates="relationship", + cascade="all, delete-orphan", + order_by="DatasetRelationshipColumn.ordinal", + ) + + +class DatasetRelationshipColumn(Model): + """Column pair mapping within a dataset relationship.""" + + __tablename__ = "dataset_relationship_columns" + + id = Column(Integer, primary_key=True, autoincrement=True) + relationship_id = Column( + Integer, + ForeignKey("dataset_relationships.id", ondelete="CASCADE"), + nullable=False, + ) + + source_column = Column(String(256), nullable=False) + target_column = Column(String(256), nullable=False) + operator = Column(String(10), nullable=False, default="=") + ordinal = Column(Integer, nullable=False, default=0) + + relationship = relationship("DatasetRelationship", back_populates="column_mappings") +``` + +### 2.3 Dual-Mode JOIN Engine + +O engine opera em dois modos, determinados automaticamente pelo flag `is_cross_database`: + +#### Modo 1 — Same-Database (SQL JOIN) + +Quando ambos os datasets estão no mesmo banco de dados, o engine injeta cláusulas JOIN diretamente na query SQL gerada. + +**Ponto de injeção:** `superset/models/helpers.py` → `ExploreMixin.get_sqla_query()` + +```python +# Pseudocódigo da modificação em get_sqla_query() +def get_sqla_query(self, query_obj, ...): + qry = sa.select(columns).select_from(self.get_from_clause()) + + # === NEW: Inject relationship JOINs === + if query_obj.relationships: + for rel in query_obj.relationships: + if not rel.is_cross_database: + target_table = rel.target_dataset.get_sqla_table() + join_condition = and_( + *[ + getattr(source_table.c, cm.source_column) + == getattr(target_table.c, cm.target_column) + for cm in rel.column_mappings + ] + ) + qry = qry.join( + target_table, + join_condition, + isouter=(rel.join_type == "LEFT"), + ) + # === END NEW === + + # ... resto do pipeline de filtros, groupby, orderby ... + return qry +``` + +#### Modo 2 — Cross-Database (Pandas Merge) + +Quando os datasets estão em bancos diferentes, o engine executa queries separadas e faz merge dos DataFrames resultantes no nível da aplicação. + +**Ponto de injeção:** `superset/models/helpers.py` → `ExploreMixin.get_query_result()` + +```python +# Pseudocódigo da modificação em get_query_result() +def get_query_result(self, query_obj, ...): + # Executa query principal + primary_result = self.query(query_obj) + primary_df = primary_result.df + + # === NEW: Cross-database merges === + if query_obj.relationships: + for rel in query_obj.relationships: + if rel.is_cross_database: + # Proteção de memória + if len(primary_df) > RELATIONSHIP_MAX_MERGE_ROWS: + raise RelationshipMergeError( + f"Primary dataset exceeds {RELATIONSHIP_MAX_MERGE_ROWS} rows limit" + ) + + # Executa query no dataset target + target_result = rel.target_dataset.query( + _build_target_query(rel, primary_df) + ) + target_df = target_result.df + + if len(target_df) > RELATIONSHIP_MAX_MERGE_ROWS: + raise RelationshipMergeError( + f"Target dataset exceeds {RELATIONSHIP_MAX_MERGE_ROWS} rows limit" + ) + + # Merge dos DataFrames + primary_df = primary_df.merge( + target_df, + left_on=[cm.source_column for cm in rel.column_mappings], + right_on=[cm.target_column for cm in rel.column_mappings], + how=rel.join_type.lower(), Review Comment: **🟠 Architect Review — HIGH** JOIN semantics in the design are inconsistent with the declared `join_type` options: in same-database mode only `LEFT` vs non-`LEFT` is honored (so `RIGHT` and `FULL` behave as `INNER`), and in cross-database mode `join_type` is passed directly to `pandas.DataFrame.merge(how=...)`, so `FULL` becomes `"full"`, which is not a valid `how` value. **Suggestion:** Constrain `join_type` to the subset both engines can correctly implement or define explicit per-engine mappings (e.g., map `FULL` to a proper SQL full outer join and to `how="outer"` in pandas) so that the same `join_type` yields consistent, valid behavior in both same-DB and cross-DB paths. [Fix in Cursor](https://app.codeant.ai/fix-in-ide?tool=cursor&prompt=This%20is%20an%20%2A%2AArchitect%20%2F%20Logical%20Review%2A%2A%20comment%20left%20during%20a%20code%20review.%20These%20reviews%20are%20first-class%2C%20important%20findings%20%E2%80%94%20not%20optional%20suggestions.%20Do%20NOT%20dismiss%20this%20as%20a%20%27big%20architectural%20change%27%20just%20because%20the%20title%20says%20architect%20review%3B%20most%20of%20these%20can%20be%20resolved%20with%20a%20small%2C%20localized%20fix%20once%20the%20intent%20is%20understood.%0A%0A%2A%2APath%3A%2A%2A%20docs%2Fsuperset_relationship_design.md%0A%2A%2ALine%3A%2A%2A%20179%3A229%0A%2A%2AComment%3A%2A%2A%0A%09%2AHIGH%3A%20JOIN%20semantics%20in%20the%20design%20are%20inconsistent%20with%20the%20declared%20%60join_type%60%20options%3A%20in%20same-database%20mode%20only%20%60LEFT%60%20vs%20non-%60LEFT%60%20is%20honored%20%28so%20%60RIGHT%60%20and%20%60FULL%60%20behave%20as%20%60INNER%60%29%2C%20and%20in%20cross-database%20mo de%20%60join_type%60%20is%20passed%20directly%20to%20%60pandas.DataFrame.merge%28how%3D...%29%60%2C%20so%20%60FULL%60%20becomes%20%60%22full%22%60%2C%20which%20is%20not%20a%20valid%20%60how%60%20value.%0A%0AValidate%20the%20correctness%20of%20the%20flagged%20issue.%20If%20correct%2C%20How%20can%20I%20resolve%20this%3F%20If%20you%20propose%20a%20fix%2C%20implement%20it%20and%20please%20make%20it%20concise.%0AIf%20a%20suggested%20approach%20is%20provided%20above%2C%20use%20it%20as%20the%20authoritative%20instruction.%20If%20no%20explicit%20code%20suggestion%20is%20given%2C%20you%20MUST%20still%20draft%20and%20apply%20your%20own%20minimal%2C%20localized%20fix%20%E2%80%94%20do%20not%20punt%20back%20with%20%27no%20suggestion%20provided%2C%20review%20manually%27.%20Keep%20the%20change%20as%20small%20as%20possible%3A%20add%20a%20guard%20clause%2C%20gate%20on%20a%20loading%20state%2C%20reorder%20an%20await%2C%20wrap%20in%20a%20conditional%2C%20etc.%20Do%20not%20refactor%20surrounding%20code %20or%20expand%20scope%20beyond%20the%20finding.%0AOnce%20fix%20is%20implemented%2C%20also%20check%20other%20comments%20on%20the%20same%20PR%2C%20and%20ask%20user%20if%20the%20user%20wants%20to%20fix%20the%20rest%20of%20the%20comments%20as%20well.%20if%20said%20yes%2C%20then%20fetch%20all%20the%20comments%20validate%20the%20correctness%20and%20implement%20a%20minimal%20fix%0A) | [Fix in VSCode Claude](https://app.codeant.ai/fix-in-ide?tool=vscode-claude&prompt=This%20is%20an%20%2A%2AArchitect%20%2F%20Logical%20Review%2A%2A%20comment%20left%20during%20a%20code%20review.%20These%20reviews%20are%20first-class%2C%20important%20findings%20%E2%80%94%20not%20optional%20suggestions.%20Do%20NOT%20dismiss%20this%20as%20a%20%27big%20architectural%20change%27%20just%20because%20the%20title%20says%20architect%20review%3B%20most%20of%20these%20can%20be%20resolved%20with%20a%20small%2C%20localized%20fix%20once%20the%20intent%20is%20understood.%0A%0A%2A%2APath%3A%2A%2A%20docs%2Fsuperset_relationshi p_design.md%0A%2A%2ALine%3A%2A%2A%20179%3A229%0A%2A%2AComment%3A%2A%2A%0A%09%2AHIGH%3A%20JOIN%20semantics%20in%20the%20design%20are%20inconsistent%20with%20the%20declared%20%60join_type%60%20options%3A%20in%20same-database%20mode%20only%20%60LEFT%60%20vs%20non-%60LEFT%60%20is%20honored%20%28so%20%60RIGHT%60%20and%20%60FULL%60%20behave%20as%20%60INNER%60%29%2C%20and%20in%20cross-database%20mode%20%60join_type%60%20is%20passed%20directly%20to%20%60pandas.DataFrame.merge%28how%3D...%29%60%2C%20so%20%60FULL%60%20becomes%20%60%22full%22%60%2C%20which%20is%20not%20a%20valid%20%60how%60%20value.%0A%0AValidate%20the%20correctness%20of%20the%20flagged%20issue.%20If%20correct%2C%20How%20can%20I%20resolve%20this%3F%20If%20you%20propose%20a%20fix%2C%20implement%20it%20and%20please%20make%20it%20concise.%0AIf%20a%20suggested%20approach%20is%20provided%20above%2C%20use%20it%20as%20the%20authoritative%20instruction.%20If%20no%20explicit%20code%20suggestion%20is%20given%2C%20you%20MUST%20still%20dr aft%20and%20apply%20your%20own%20minimal%2C%20localized%20fix%20%E2%80%94%20do%20not%20punt%20back%20with%20%27no%20suggestion%20provided%2C%20review%20manually%27.%20Keep%20the%20change%20as%20small%20as%20possible%3A%20add%20a%20guard%20clause%2C%20gate%20on%20a%20loading%20state%2C%20reorder%20an%20await%2C%20wrap%20in%20a%20conditional%2C%20etc.%20Do%20not%20refactor%20surrounding%20code%20or%20expand%20scope%20beyond%20the%20finding.%0AOnce%20fix%20is%20implemented%2C%20also%20check%20other%20comments%20on%20the%20same%20PR%2C%20and%20ask%20user%20if%20the%20user%20wants%20to%20fix%20the%20rest%20of%20the%20comments%20as%20well.%20if%20said%20yes%2C%20then%20fetch%20all%20the%20comments%20validate%20the%20correctness%20and%20implement%20a%20minimal%20fix%0A) *(Use Cmd/Ctrl + Click for best experience)* <details> <summary><b>Prompt for AI Agent 🤖 </b></summary> ```mdx This is an **Architect / Logical Review** comment left during a code review. These reviews are first-class, important findings — not optional suggestions. Do NOT dismiss this as a 'big architectural change' just because the title says architect review; most of these can be resolved with a small, localized fix once the intent is understood. **Path:** docs/superset_relationship_design.md **Line:** 179:229 **Comment:** *HIGH: JOIN semantics in the design are inconsistent with the declared `join_type` options: in same-database mode only `LEFT` vs non-`LEFT` is honored (so `RIGHT` and `FULL` behave as `INNER`), and in cross-database mode `join_type` is passed directly to `pandas.DataFrame.merge(how=...)`, so `FULL` becomes `"full"`, which is not a valid `how` value. Validate the correctness of the flagged issue. If correct, How can I resolve this? If you propose a fix, implement it and please make it concise. If a suggested approach is provided above, use it as the authoritative instruction. If no explicit code suggestion is given, you MUST still draft and apply your own minimal, localized fix — do not punt back with 'no suggestion provided, review manually'. Keep the change as small as possible: add a guard clause, gate on a loading state, reorder an await, wrap in a conditional, etc. Do not refactor surrounding code or expand scope beyond the finding. Once fix is implemented, also check other comments on the same PR, and ask user if the user wants to fix the rest of the comments as well. if said yes, then fetch all the comments validate the correctness and implement a minimal fix ``` </details> -- 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]
