Answer the question
In order to leave comments, you need to log in
Are there languages for describing the terms of reference for generating complex reports in tabular form?
A large number of reports are generated on the database. Reports from small to large. "Terms of Reference" for the rules for calculating cell values can take 50 pages or more. Tasks are written, almost formally (for lines .., graphs ... details, such and such, such and such conditions, such and such exceptions, ...). It is impossible to write an automatic parser due to the formal approach in the description of the conditions. I think so, in order not to reinvent the wheel, that there are already ways or languages for writing such tasks, according to which queries are then built automatically. Can you tell me where to look? Because when "tasks" come in this form, it takes a lot of time to study it, it is difficult to understand what has changed, what has been added.
Answer the question
In order to leave comments, you need to log in
SQL was originally designed for managers.
Not for programmers.
It is quite suitable for use by the most ordinary users.
https://www.quora.com/What-are-some-good-visual-SQ...
There are actually very, very many of them...
But here the question arises -
will users formally enough approach the task statement so
that all these formal tools can be used automatically.
I think it's easier for them to blame it on the programmers.
In addition to finding and agreeing on the use of some description language, it will also be necessary to train directors in the use of this language. This second task can be much more unbearable than the first.
As for languages for describing/modeling reporting forms, there are no well-known standards. All this is always tied to a specific BI tool and built into its graphical IDE, with which these reports are created: MS SSRS , Crystal Reports, Cogons BI, Oracle BI, etc.
In simple cases, you could do with the usual UML with the addition of OCL, but is that enough for you?
If you try to solve another problem - to see changes in the TOR for the reporting form, then you do not need a language, but a tool for maintaining requirements (or an ALM system).
Strange task. Seems a little implausible.
Before you talk about languages, try to formalize your problem itself.
Vsevolod Vladimirov would definitely pay attention to the last phrase " it takes a lot of time to study it, it's hard to understand what has changed, what has been added ." With a high probability, you have problems with the task / bug tracker. He is not. As there is no system for storing jobs and versioning solutions. Consequently, you cannot trace the history of changes and major design loads on the formal structures and applied algorithms. It follows from this that no one uses much when describing
By describing the work structures, you would have concise and concise definitions that could be used in the assignment. For example, lists, trees, temporary tables, selections, static selections, stacks, links. You can always apply to data
. Never had balances at the "end of period"? Unique. Search for duplicates. Finding the first occurrence. Sorting.
It is necessary to constantly classify reports and structure them using at least basic taxonomic principles. Structures and hierarchies allow you to quickly search for similar ones, which in general can already be a solution to the problem, because. replacing a condition is not the same as recreating a database query.
After each (or each group) report development cycle, it would be good to increase (try to do so) the abstraction level of the problem description. After all, your goal is ultimately automation. And what could be nicer than the exact answer to the request: "Upload the most active customers of each first quarter for 10 years."
It turns out that the best tool is again a pencil.
Didn't find what you were looking for?
Ask your questionAsk a Question
731 491 924 answers to any question