Inconsistent material classification for CF/GF filament variants

[Icezaza2543]

Hi, I noticed something while checking some older filament entries.

Some products have a very specific material name, but the structured "material" field only keeps the base polymer.

For example:

• "Fiberon™ PA12-CF10" → "material: "PA""
• "Fiberon™ PA6-CF20" → "material: "PA""
• "Fiberon™ PA6-GF25" → "material: "PA""

There are similar cases with PETG-CF, PLA-CF, TPU hardness variants, and probably others.

The problem is that this removes useful information from the structured data.

For example, these are all quite different products:

• "PA6-CF5"
• "PA6-CF10"
• "PA6-CF20"
• "PA6-GF25"

But if they are stored as:

{
"material": "PA6",
"fill": "carbon fiber"
}

then "CF5", "CF10", and "CF20" all become indistinguishable from each other.

The same applies to glass-fiber percentages.

Because of that, I think the full formulation probably needs to stay in the "material" field, for example:

{
"material": "PA6-CF20",
"fill": "carbon fiber"
}

or:

{
"material": "PA6-GF25",
"fill": "glass fiber"
}

That way the exact material is preserved, while "fill" can still be useful as a broader category.

This also matters for filtering. A user looking specifically for "PETG-CF", "PA6-CF20", or "TPU-95A" should ideally not get the same classification as ordinary PETG, generic PA, or generic TPU.

The schema already seems to allow more specific material names such as "ABS-CF" and "PC+ABS-CF10", so I wanted to check whether keeping the full formulation in "material" is the intended convention before correcting older records.

I ran into this while cleaning up older entries in my community-maintained SpoolmanDB fork, and there seem to be enough affected records that having one consistent rule would help a lot.

Related to #32, but this is mainly about preserving the actual material variant in existing data rather than changing the overall schema.