Incorrect LVS netlists created with xschem symbol updates.

[d-m-bailey]

This applies to #607.

It is my understanding that spiceprefix is used to generate valid spice file and that removing it may cause problems.

For example, if someone assigns the name Decap1 to a capacitor, without the spiceprefix, this would be output as

Decap1 ...

which is a diode. I don't think we want the user to be responsible for ensuring the prefix matches the device model.

With a spiceprefix of C, this would be output as

CDecap1 ...

The problem that you might be facing is that the simulation (ngspice) and LVS (klayout) netlists may be expecting different device names. ngspice simulation expects device names that start with X that match the primitive model subcircuits (line 51 for cap_cmim), while klayout expects device names that start with C. Incidentally, the current magic tech expects that many devices that start with X.

One solution is to use a separate lvsprefix for LVS.
This results in the misinterpretation of devices as explained above. Adding an lvsprefix variable to the symbol would solve the problem.

See #607 (review)