validate chargeBoxId for WS connections (#1526)

[goekay]

No description provided.

[juherr]

I think the new regex is too restrictive and will break instances.

[goekay]

I think the new regex is too restrictive and will break instances.

fair enough, and i was fearing this as well.

new suggestion: instead of whitelisting what is allowed in the regex, i could blacklist what should not appear in the regex. i am thinking something like [^=/()<>]* i.e. every character with the exception of =/()<> is allowed.

[juherr]

The blacklist solution sounds better.

But even if I'll be surprised that a charger uses one of this chars somewhere you need a way to block the upgrade if a charger uses one of them in the database.
Otherwise some users will lost the charger without any way to change its configuration.

[goekay]

But even if I'll be surprised that a charger uses one of this chars somewhere you need a way to block the upgrade if a charger uses one of them in the database.
Otherwise some users will lost the charger without any way to change its configuration.

i wonder how big that percentage is. if we are about to lose a minority (for ex 0.1%) of installations, i would be okay to make that sacrifice.

i can put the new logic behind a config, but i then i would like to have it enabled by default. installations that have problems can still opt out, while accepting the risks.

[juherr]

It can be enabled by default but I think it must be possible to disable it or block the upgrade if one station doesn't respect the convention.
Otherwise the hardware will be totally lost with no way to restore it.

[juherr]

LGTM except the ui fix which should be extended everywhere a displayed data comes from a station (maybe in another pr?).

[goekay]

yes, there will be another issue and another PR.