Standardize documentation to use INSERT with comma-separated patterns #17
Conversation
Updates all documentation to follow ISO GQL standard and best practices: - Replace CREATE keyword with INSERT (ISO GQL standard) - Use comma-separated patterns for multiple node insertions instead of separate statements (single atomic transaction vs multiple) - Add clarifying comments explaining single vs multiple node patterns - Remove unnecessary variable bindings when not used Files updated: - Main README.md and all binding READMEs (Java, Python, SDK) - Documentation guides (Quick Start, Getting Started With GQL) - Add comprehensive JSON format test suite (json_format_tests.rs) - Add debug output to pattern_tests.rs for NEXT clause investigation Technical rationale: - CREATE and INSERT are synonyms (parser.rs:3933-3936) - Comma-separated pattern is more efficient (1 transaction vs N) - Parser supports comma-separated syntax (parser.rs:3937) - Atomic all-or-nothing semantics
There was a problem hiding this comment.
We have 13 out of 15 tests failing due to:
Error: RuntimeError("No graph context available. Use SESSION SET GRAPH or specify FROM clause.").
The test code appears to have issues with how it sets up graph context for CLI queries. The query_with_context helper prepends SESSION SET commands, but individual tests also try to use FROM clause which seems to conflict.
There was a problem hiding this comment.
On line #804 we have #ignore added for the test forWITH and NEXT clause composition, due to which this test is being skipped. And looks like we only have eprintln debug statements added as a change to the test.
So should this test be included in this update, especially since it is being ignored and only prints additional debug statements anyway ?
Updates all documentation to follow ISO GQL standard and best practices:
Files updated:
Technical rationale: