Bug with line break in column name props generation + align renameToCamelCase and normalization behavior

[AndreiKingsley]

Bug:

• In DataFrame renameToCamelCase API
• in compiler plugin (should be 😄)
• in NameNormalizer.Companion.default
• in ValidFieldName.Companion#of
• jupyter codegen

Might be other places, need to inspect.

[koperagen]

Changing org.jetbrains.kotlinx.dataframe.codeGen.ValidFieldName.Companion#of should fix it for both compiler plugin and jupyter codegen
NameNormalizer.Companion.default serves different purpose: in KSP and Gradle plugins, when schema is generated from data, it renames extension properties to camel case (old logic) but keeps column names in df intact. I think now that we have renameToCamelCase support in compiler plugin, it's somewhat outdated feature. Anyway doesn't work in jupyter notebooks - there we get names "as is", except ValidFieldName.Companion#of fixes names that otherwise won't compile

So to summarize
ValidFieldName.Companion#of -> fixes invalid symbols in identifiers created from data (string arguments, or infered from CSV), used by jupyter and compiler plugin
NameNormalizer -> stylistic feature that made sense for KSP/Gradle