imbajin opened a new pull request, #839:
URL: https://github.com/apache/hugegraph-toolchain/pull/839

   ## Purpose of the PR
   
   Hubble already targets Java 17, but legacy date handling, duplicated file 
previews and mutable empty query results still cause avoidable failures. This 
cleanup fixes those paths while retaining the existing Server compatibility 
boundaries and metadata format.
   
   ## Main Changes
   
   - Replace mutable date formatters with explicit UTC+08:00 `java.time` 
handling; correct month/week boundaries, day-end task counts and 
locale-sensitive growth rates.
   - Share CSV/TEXT preview parsing and honor the selected charset, quoting, 
empty cells and skipped lines. Initial upload stays tolerant until the user can 
choose the correct charset.
   - Validate the complete account CSV before writes and close file resources; 
retain the existing CSV parser semantics. Share Gremlin/Cypher result assembly 
and create independent empty views.
   - Remove retired license code and redundant POM overrides; use Spring Boot's 
managed Caffeine version. Make the DNS failure test independent of local 
resolver behavior.
   
   ```mermaid
   flowchart LR
       A[Upload bytes] --> B[Initial preview]
       B --> C[Choose charset]
       C --> D[Re-read header and sample]
   ```
   
   ## Verifying these changes
   
   - [x] Already covered by existing tests, with focused regressions added for 
dates, previews, account CSV and query results.
   - [x] JDK 17 backend suite, Checkstyle and complete Hubble 
distribution/license/SDK audits passed; the merged Operations UI tests passed.
   - [x] Real Server 1.8/RocksDB and packaged Hubble: schema creation, 
CSV-to-Loader import with direct Server readback, Gremlin counts, Cypher 
scalar/vertex/edge views and persisted history.
   - [x] Chrome: authenticated and anonymous flows, graph/table/JSON views, 
nonempty → empty → nonempty results, successful asynchronous count, account 
visibility and Server node metrics.
   - [x] Real HTTP: GB18030 upload followed by charset selection, quoted/empty 
preview cells, malformed account CSV rejected before writes, and normal-user 
batch denial.
   
   Validation limits: the locked Server has no batch-account creation route, so 
valid/header-only batch creation returns HTTP 405 and is not claimed as 
verified. PD/Store, external Elasticsearch and released Server 1.7/1.5 runtime 
matrices were not rerun.
   
   ## Does this PR potentially affect the following parts?
   
   - [ ] Nope
   - [x] Dependencies (Caffeine follows the Boot BOM; distribution license 
audit passed)
   - [x] Modify configurations (remove the unused license-only `server.id` 
option)
   - [ ] The public API
   - [x] Other affects (file preview correctness, date calculations and 
isolated query results)
   
   ## Documentation Status
   
   - [x] `Doc - Done`
   
   Org working PR: https://github.com/hugegraph/hugegraph-toolchain/pull/85. 
Both PRs use the same org branch.
   


-- 
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]

Reply via email to