Open kaurov opened 1 year ago
Thanks for your suggestions, @kaurov. We will put them on our todo list. You can speed up processing of your requested points by writing to info@blackbit.de - we will send you an offer and then you can decide which points shall be implemented and in which order.
One note about the attribute mapping hints:
Here, you can add your existing hint 'Field * is used as key field. Consider adding an index in the field definition with the 'Indexed' checkbox to improve performance' -- which user can see only if open import logs, currently. This is already visible as a mouseover hint when you move the mouse over the key icon of a key field:
And about:
Please show not only for currenly loaded file (first) but all unique values of all XML files from folder provided as source for dataport. I mean following columns: incoming data 'Example result' 'Parsed' data after processing or PHP Exception
Currently you can select another raw data item used for preview by clicking inside the Example result
column:
Of course you do not see all possible values in one sight this way but can only manually test certain raw data items. For this reason we plan a dry-run. At the end of a dry-run you would see an overview of all the imported objects and what data changed. Sadly until now we did not find a client who was willing to pay for that...
Please store plugin_pimdataport->sourceconfig container in dataport.json as JSON, not serialised string. The reason is -- that string is of HUGE size and developer commit it as generated without reading it, so error can be easily made.
This is an important point for comparing it in Git (or with the integrated versioning tool). Think I will implement this shortly...
Concerning
Also, Customer provided values like following which DataDirector cannot process itself, I need to prepare in func: '- 52 F' -- space after minus is not supported '12 turns per minute' -- measurement unit cannot have spaces even is fexists in Pimcore. why?
I have enhanced quantity value recognition when provided as a string. From version 3.3 the following works:
Can you please improve DataDirector to make it safier for support:
We need unit-tests for callback-functions used for attributes on 'Attribute Mapping' Tab. It will allow us to write assets wich result to expect for which incoming string. The design can be similar to the form for callback function. Ideally, to show alist of all unique values provided in the right panel so developer can copy them for asserts. The reason for it: I made a typo in callback function for Path attribute, which cause error only during import of some xml files.
if ($params['value'] = 'somevalue') {} // '=' instead of '=='
Yes, I am guilty for that. But I would like to know about problems in my data and my import, not to break products inheritance. Email is not sent for such cases because no php error or measurements units error happened. Also, Customer provided values like following which DataDirector cannot process itself, I need to prepare in func: '- 52 F' -- space after minus is not supported '12 turns per minute' -- measurement unit cannot have spaces even is fexists in Pimcore. why? Such errors can at least be seen in Email, but beter to see in 'Attributes Mapping' TabPlease show not only for currenly loaded file (first) but all unique values of all XML files from folder provided as source for dataport. I mean following columns: incoming data 'Example result' 'Parsed' data after processing or PHP Exception
Please store plugin_pimdataport->sourceconfig container in dataport.json as JSON, not serialised string.
The reason is -- that string is of HUGE size and developer commit it as generated without reading it, so error can be easily made.
Please add hints for attributes in 'Attributes Mapping' Tab, right near red buttn 'delete'. This (i) button can show onHover hints like 'Maximum data lenth provided for this field is 47 characters. We recommend you to decrease this field lenth from 190 to 60 characters.' Here, you can add your existing hint 'Field * is used as key field. Consider adding an index in the field definition with the 'Indexed' checkbox to improve performance' -- which user can see only if open import logs, currently.
Perhaps you will introduce data-tests called by Maintenance Jobs, to test at least that amount of products is usual, and children product samples are inheriting parent product. This is on the last position because we developers can do it in commandline command.