Extend structured-content extraction

POM Translate exposes focused parser filters:

• pom_translate_parser_container_tags;
• pom_translate_parser_inline_tags;
• pom_translate_parser_translatable_attributes;
• pom_translate_parser_skip_tags.

Return normalized tag or attribute arrays and preserve existing values. Container tags divide textual blocks; inline tags become protected structure; translatable attributes expose selected human text; skipped tags remain untouched.

Choose the correct category. Treating a layout container as inline or exposing technical attributes can produce unstable skeletons and broken placeholders.

For proprietary JSON or serialized content, these HTML filters are insufficient. Register a known metadata surface and implement schema-aware extraction instead of recursively translating every string.

Test valid and malformed markup, nested components, repeated text, source edits, and target fallbacks before shipping the integration.