Skip to main content

On-Prem Notifications

Follow new updates and improvements to CodeScene.

Code Health version change: JavaScript/TypeScript 6.5

The following changes have been made:

  • Fixed an issue in the JavaScript/TypeScript analysis where regex literals containing escaped slashes or backslashes (e.g. /\/items$/ or /\\/g) were misparsed. This could cause function boundaries to merge, leading to incorrect function ranges and complexity results.

Scores can therefore shift slightly even though your code has not changed.

Because of this change, the stored Code Health results for the affected languages need to be recalculated. This may cause the next analysis of your projects to take longer to complete than usual. Following analyses will be back to normal.

Code Health version change: PHP 6.3

The following changes have been made:

  • * Fix: PHP constructors are now checked against `constructor_max_arguments` instead of `function_max_arguments`. Previously the constructor limit had no effect when it was set higher than the function limit.

Scores can therefore shift slightly even though your code has not changed.

Because of this change, the stored Code Health results for the affected languages need to be recalculated. This may cause the next analysis of your projects to take longer to complete than usual. Following analyses will be back to normal.

Code Health version change: Clojure 6.1

The following changes have been made:

  • Clojure: no more false Large Method findings on documented data vars

    A def with a docstring and a data value, such as (def name "docstring" [...]) with a long vector, map or set, was analysed as a function. This gave false findings like Large Method on constant tables. These forms are now treated as plain vars, the same as when there is no docstring. Functions defined with defn, defn- and fn are analysed as before.

    Code Health scores may change: Clojure files with long documented defs can score higher. Cached Clojure scores are recalculated on the next analysis, which can take longer on large Clojure codebases.

Scores can therefore shift slightly even though your code has not changed.

Because of this change, the stored Code Health results for the affected languages need to be recalculated. This may cause the next analysis of your projects to take longer to complete than usual. Following analyses will be back to normal.

Earlier updates