acidmanic / meadow Goto Github PK

when a readChunk request has been made with null search id, it might result in database error. If (generated) scripts become more robust so that they can handle null search ids, it would be better solution than always null-protecting the request in client code.

Using meadow assistant tool, currently replaces all macros with their actual code. A marking standard can allow mat to recognize those macros which the developer wants to be replaced from those which the developer prefers to leave untouched. For example you can define marker words: "BY-FORCE", and "ALWAYS". Then add command arguments "forceFull (f)" and "weak (w)" to applyMacros command. This way

• the applyMacros command, when executed with forceFull argument, will only replace Macros which are marked "BY-FORCE",
  and leave unmarked and "ALWAYS" macros untouched.
• In the other hand, the applyMacros command, when executed with weak argument, can only replace macros which are marked "ALWAYS", unmarked macros and "BY-FORCE" macros will be left untouched.
• And if the applyMacros command get executed without any arguments, it will replace unmarked macros and also replace macros which are marked "ALWAYS".

Error messages are not being shown. it can be due to incorrect logLevel

This would make finding da's assembly easier while creating configurations for meadow

Currently the only way determine make a field's column size is annotations which are declared in Meadow.Framework. And it's not applicable on models that are not declared in the project. Providing another way to mark certain fields of certain type can solve this issue. This way FilterResults and SearchIndex tables can be adjusted to Meadow's preferences. It also can be used for other annotations like TreatAsLeaf.

full-tree code generators should decide for join method(left/right/inner) based on the field being mandatory (NotNull) or not (nullable), currently all fulltree view are created by inner joins which might lead to not returning data when a child node doesnt have any data inserted yet.

Save Code generators need to differentiate between AutoValued Id fields and just-unique id fields while generating the insertion section of the procedure