Skip to content

Improve sanitizing column headers and description for datastore import #4612

@stefan-korn

Description

@stefan-korn

User Story

There is already sanitizing of colum headers and column description done in the datastore import to align with MySQL requirements.

I would like to have two additional cases be handled by the sanitation:

  • put in a default string for an empty column name. If there is an empty column name, creating the MySQL table will fail because empty column name is not allowed.
  • the column name might be a "binary string", which is not allowed for the description. Sanitize the description from binary string.

While both cases might indicate an invalid file structure or an encoding problem, this is not necessarily the case. So imho it would be better to have the file imported in these cases, and let the users see for themselves if the datastore import is usable rather than just failing the import (users do not directly see the reason for failed import?). But is surely opinionated, so asking for your opinion.

Detailed Feature Description

Extend the sanitation calls in ImportJob

Problem and Motivation

No response

Target Audience

  • End Users
  • Developers
  • Administrators
  • Other

Possible Implementation

will create an MR.

Acceptance Criteria

No response

Metadata

Metadata

Assignees

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions