Please provide more detail. You've removed the actual failing query from the error message you posted, which is the most critical part of it for debugging. I cannot reproduce it; possibly I can with deep code analysis, but that is time consuming.
Fixed in https://github.com/ocproducts/composr/commit/b5e01e52cc5e930eddec83bc7885c9f35ad8d57d
This is a class of bug that happens when the Comcode cache ("Comcode field precalculations" OTTOMH) is cleared. Composr has to update the parsed Comcode back into the database when it is missing, but if we have JOINed tables in the data row it tries to save alien data back into the row; we have to pre-filter out what is in the row. Usually we do it right, but there are 100s of cases in the code for us to get it wrong. I think I remember you having a similar issue before.
Ergh, sorry, yes. I accidentally triggered an undo in my editor before committing.
Fixed in https://github.com/ocproducts/composr/commit/2296fb769c01366b198c3a69eec3759266e08ac5
https://bosslife.org/errors/error.htm
This is a class of bug that happens when the Comcode cache ("Comcode field precalculations" OTTOMH) is cleared. Composr has to update the parsed Comcode back into the database when it is missing, but if we have JOINed tables in the data row it tries to save alien data back into the row; we have to pre-filter out what is in the row. Usually we do it right, but there are 100s of cases in the code for us to get it wrong. I think I remember you having a similar issue before.
Stack:
https://bosslife.org/errors/newerror.htm
Fixed in https://github.com/ocproducts/composr/commit/2296fb769c01366b198c3a69eec3759266e08ac5