ACE — Morphology Processing Flowchart
This diagram shows the main forward path documented for ace_stem_tester.php and the ACE Analyzer. The Reverse Analyzer conceptually examines these layers in the opposite direction.
Documentation diagram only — it does not execute or modify the morphology engine.
1. Forward Morphology Processing Diagram
Scroll horizontally to inspect the complete SVG on a narrow screen.
2. Processing Stages — Reference
Root radicals, Form, tense/category and pronoun index.
ace_apply_extra_rules(): root-family adjustments.
build_combined_token(): pattern, radicals and inflections.
Perfect/imperfect suffixes and nominal declension.
Functional-word layers and assembly order.
Object and genitive pronoun attachments.
Combines segmented elements into a pronunciation-oriented intermediate representation.
latinToArabic() shown here as part of the documented forward flow.
getTenseTranslation() and getFunctionalWordTranslation().
Pronoun metadata, colored Arabic, Latin aid, IPA and translations.
3. Reverse Analyzer — Conceptual Direction
The Reverse Analyzer conceptually examines the surface Arabic token and identifies possible conjunctions, particles, article, stem, root, Form, tense and pronoun.
ACE architectural note: Arabic source tokens are authoritative. Latin/IPA serves as a pronunciation aid, not as the source for recovering Arabic morphology. Multiple structurally valid hypotheses are retained until an independent rule can exclude them.
Related documentation: Reverse Analyzer index (if installed at this path).