Codeer.LowCode.Blazor 1.3.2

• net8.0

  • Codeer.LowCode.Blazor.Licensing (>= 1.3.0)
  • Microsoft.AspNetCore.Components.WebAssembly (>= 8.0.4)
  • Microsoft.AspNetCore.Http.Abstractions (>= 2.1.1)
  • Microsoft.AspNetCore.WebUtilities (>= 8.0.4)
  • Microsoft.CodeAnalysis.CSharp (>= 4.9.2)

Improved the design-time checks for scripts. Code that will always fail at runtime is now reported at design time.
     - Detects method, user-defined function, and constructor (new) calls whose argument types do not match any overload.
     - Detects if / while / for / conditional (?:) conditions whose type cannot be treated as a boolean (such as a module or a list).
     - Detects foreach over a type that is not enumerable (such as a number).
     - Detects indexing with [] on a type that is neither an array nor has an indexer.
     - Detects binary/unary operators used with incompatible types (e.g. arithmetic or comparison on a string or list, negating a non-boolean value).
     - Fixed false positives: assignments that the runtime converts at execution time (e.g. assigning a number or bool to a string variable, or a string to a number variable) are no longer reported as errors.
     Note: To avoid false positives, cases whose result depends on the actual value (e.g. converting a string to a number) and cases that cannot be determined statically (e.g. var or field values) are intentionally not reported.

     Removed an unused SQL parameter that was bound when generating "is null" / "is not null" search conditions. This is an internal cleanup with no change in behavior.

     Enabled indexing into string values in scripts (e.g. text[0] returns the character at that position).

     Added support for string interpolation ($"...") in scripts, including format (:fmt) and alignment (,N) specifiers (e.g. $"{name}: {price:N0}").

     Added support for implicit array creation (new[]{...}) in scripts. The element type is inferred from the values: all the same type uses that type, mixed numeric types use decimal[], and arrays containing null or mixed types use object[].

     Added filtering, limiting, and paging methods to ModuleSearcher used in scripts.
     - AddNotEqual(field, value): adds a "not equal" condition (passing null generates "is not null").
     - Limit(count): sets the maximum number of rows to fetch. Without it, all matching rows are fetched, so set it when searching tables that may contain many rows.
     - ExecutePage(pageIndex) / ExecuteRawPage(pageIndex): fetch the specified page (0-based) together with the total matching count in a single query, returned as ModulePageResult / ModuleDataPageResult (Items and TotalCount). Use them with Limit for pagers.
     - ExecuteFirstOrDefault(): returns the first matching module, or null when there is no match (handy for master/lookup references).
     - BatchModuleSearchResponse.GetBy(searcher): retrieves a batch result by the ModuleSearcher you passed to Execute, so you no longer need to track result indexes (passing a searcher not in the batch throws).

     Names that occupy the "type" namespace are now validated by the design check. A module or page-frame name that duplicates another such name (another module / page-frame, a ProCode component / module / list-element class name, or a framework type name; compared case-insensitively), or that is not a valid C# identifier, is now reported as a design error. This prevents name collisions that break type resolution in scripts.