multiglossar-configuration-exclusions

MultiGlossar Configuration and Exclusions

MultiGlossar behavior is mostly controlled through addon settings. Correct configuration is essential to avoid over-replacement or performance issues.

Scope configuration (start/end tags)

Replacement is applied only within one configured document segment:

• glossar_starttag - regex/string for start marker (default around <body.*?>)
• glossar_endtag - regex/string for end marker (default </body>)

Projects can use explicit HTML comments in templates (for example <!--glossar_start--> and <!--glossar_stop-->) and configure these markers in settings.

Only one unique scope is expected. Ambiguous patterns can lead to missing replacements.

Ignored tags and classes

MultiGlossar has built-in ignored elements (for example a, h1-h6, figcaption, script, style, svg, dfn, exclude).

Additional excludes can be set via glossar_ignoretags:

• tag names, e.g. nav,aside,ul
• class selectors with dot prefix, e.g. .no-glossary,.teaser

During recursion, matching tags/classes are locked and skipped.

Article-level exclusions

Typical exclusion controls in boot.php include:

• current article is REDAXO notfound article -> always skip
• articles_exclude list (comma-separated article IDs)
• exclude_by_template (template ID list)
• exclude_by_meta_field + exclude_by_meta_condition (<0, =0, >0)

This is useful for legal pages, forms, and technical pages where glossary links are undesirable.

Multi-domain and YRewrite

If YRewrite is active, MultiGlossar can use domain-specific glossary article assignment:

• settings like article_<domainId> map each domain to its glossary article
• without YRewrite, fallback is generic article

Editorial term storage is not domain-separated; only output linking/context can differ by domain.

Cache behavior

Important options:

• use_cache - use MultiGlossar output cache
• use_turbocache - read cached replacement early (PACKAGES_INCLUDED)

Cache invalidation hooks are bound to article/category/slice events and glossary form saves. When integrating custom import/update scripts, clear or refresh caches accordingly.

Practical setup pattern

1. Define a narrow replacement scope around article content.
2. Add ignore tags/classes for navigation, UI components, and structured widgets.
3. Exclude legal/system pages by article ID or metainfo rule.
4. Validate per language and per domain.
5. Enable cache after functional behavior is confirmed.

Common pitfalls

• Using too broad start/end patterns, replacing content in headers/menus accidentally.
• Forgetting to exclude interactive UI containers, causing tooltip markup conflicts.
• Mixing article IDs from different environments without checking production IDs.
• Using metainfo exclusion with the wrong condition operator (<0, =0, >0).
• Enabling cache during debugging and interpreting stale HTML as parser failure.