DB query in File::find($filepath) should be case-sensitive

[micschk]

• There's two files (My-FILE.jpg and My-file.jpg) on a case-sensitive filesystem (as any webserver should be).
• NowFile::find('My-file.jpg) will (wrongly) return the My-FILE.jpg record if that file/record was created earlier.
• This is because File::find queries the DB on filename/path in the default case-insensitive manner.

Proposed fix: make File::find() case-sensitive by applying the:case searchfilter on the Name query;

    public static function find($filename)
    {
        // ...
        $item = File::get()->filter([
                'Name:case' => $part,
                'ParentID' => $parentID
            ])->first();
        // ...
    }

[michalkleiner]

Makes sense, though I wonder how many files would suddenly become missing...

[dhensby]

I think a safer implementation would be to not allow duplicate file names (by case) - I'm surprised this is even possible 🤔

[micschk]

Makes sense, though I wonder how many files would suddenly become missing...

None should go missing (unless there's somehow a discrepancy between filenames in the DB and on the filesys, but that could be considered a data-integrity issue which should be fixed anyway/elsewhere?). If however issues like files gone missing are indeed to be expected maybe we could fallback to case-insensitive query if no records are found via case-sensitive?

[micschk]

I think a safer implementation would be to not allow duplicate file names (by case) - I'm surprised this is even possible 🤔

Safer implementation possibly, but not a correction of this issue imo :-)

This situation is possible when loading files into File dataobjects manually (for example when checking for existing files via file_exists() instead of File::find()). Even if the decision would be to disallow duplicate filenames eg when uploaded via the CMS, still the current result of File::find() would be incorrect.

Also I did notice what I think is at least one occurence of this issue in CMS-uploaded files as well.

[dhensby]

Historically I know case-sensitivity (including database sensitivity) has been considered quite a bit around files. We should try to find the historical decisions around this.

We need to remember that we are a framework that should be striving to provide interoperability regardless of the case-sensitivity of the filesystem being used. That must be the priority, if that means not allowing files to share names regardless of case, then that's what we should do (if that's decided to be the best approach).