You can search the forums by clicking the magnifying glass to the far right in the top banner (Right next to "Support"). I just clicked it and searched for "excel" and found several forum threads. One of them concerned linked documents and ended with Phil saying pretty much what you stated -- can't really be done.
I regularly search for threads related to what I have posted about Foxpro thus far and have had no trouble finding older threads on the topic.
You mention that all indexes update when you use Postman, but only some indexes update when you use...what? What tool or programming language are you using to perform the API request that doesn't work correctly? Just a straight URL call? How are the changed indexes being assembled? JSON or XML? Do you have an example?
In my experience, changing indexes either works completely or does not work at all (returning an error). However, I am not sure what happens if you have field-level security in place (our security structure has been kept pretty simple -- no field-level permissions). Do you have field-level permissions set up by Role or User? And if so, are you using different credentials in your Postman call compared to other API calls you are performing?
It would be nice for a DocuWare tech to weigh in and answer whether or not it is possible that only some indexes change when such a request is made. What happens if the user performing the request is locked out of certain fields? Will the index change be partially completed or will the request error out unless all fields can be changed successfully?
Really just a shot in the dark, but glad the clue about Sections helped. Now I will just have to remember that if and when I try to do integrations involving document editing... *smile*
I have never edited a document and then tried to put it back in DocuWare, but this is an example of doing it in .NET:
Are you using .NET or straight HTTP calls? What version of DocuWare? Cloud or on-premise?
One thing I notice is that editing operations involves downloading the content of a document (as "Sections"), which appears to be different from downloading the whole document. It is also an asynchronous process.
I am not sure what the .NET calls translate to in terms of the URLs to use... Looking at Fiddler as I edit a document, it looks like one of the relevant resources is:
Like I said, Sections vs Documents -- not totally sure what the difference is, though.
I assume you mean you have a file cabinet containing vendor-related documents, and one of the indexes on those documents is the vendor name? So, when a vendor name changes, you want all the existing vendor name indexes to change to the new name?
If that is the scenario, I would need to understand how DocuWare would be made aware that a vendor name had changed? Unless there is some database trigger or automated lookup, there is no way for DocuWare to know a vendor name has changed in whatever system you are referring to (ERP system? Purchase Order system?).
At this point, it would seem the only way to change all existing documents would be to bring up all documents under the old name and then bulk-change the vendor name on documents in the result list.
If you have a more sophisticated DocuWare module such as Workflow or AutoIndex, I would guess you could perform an external lookup to try to catch vendor name changes and then have the indexing process change old indexes. Even that process, though, would need to know the old and new name to perform a change.
None of this even gets into the fact that you will then have documents in your system where the vendor name on the document does not match the vendor name in the index. But I assume you are operating under business rules as agreed upon at the DocuWare site..
I got this to work, similar to what Gerardo lists (but does not allow lowercase letters):
Let's try to check the fields next time before they get re-written as properly indexed text. We can grab the text from SQL Server then analyze it character by character in Foxpro to see what the ASCII values are.
If we decide we need to write a scrubber routine we can do that as an automated task...
Great! Now when I forget how it will be on the forums for me to find. *smile* The life you support may be your own!
Add indicators for beginning and end of line, otherwise the expression just thinks you are starting a new string on the same line. This seemed to make an online regex checker work better when I tested:
That is, same as before, but with a caret at the front and dollar sign at the end. That terminates front and back, so something of 0 characters or 19 characters should no longer be considered a match.
If that doesn't work, then yes, I would start to wonder if there is a bug, or if perhaps the regex validator in DW isn't quite standard. All of the other help documentation would indicate that it is.
It look slike the following expression should be close:
That is, a period, open curly bracket, 1, comma, 18, and closing curly bracket.
The period means any character except line-feed, and the range of numbers in curly brackets means the character can occur that many times (so, effectively a lower and upper bound on length in this scenario).
If the DocuWare help isn't helping, any tutorial on regular expressions should be of assistance. Figuring out regex stuff can be a challenge... Just google for whatever issue you are trying to solve (including keywords "regular expression" or "regex"), and it can help zero in on what you are trying to do.