Abbreviation Files
The Abbreviation File Collector supports one documentable thing: an authored Markdown abbreviation dictionary. It retains the original file and converts its entries into source-owned SKOS concepts.
[//]: # (Shared abbreviation meanings)
*[EKG]: Enterprise Knowledge Graph
*[API]: Application Programming Interface
Select the file from the project configuration:
@prefix ga: <https://ekgf.org/ontology/grapharch#> .
<urn:project> ga:collector <urn:project:abbreviations> .
<urn:project:abbreviations> a ga:CollectorUse ;
ga:collectorType ga:AbbreviationFileCollector ;
ga:source "abbreviations.md" .
The path is relative to the project root. The file may live outside the content
directory, but must remain inside the project. Every nonblank line is one
*[TERM]: Definition declaration or a [//]: # comment. Definitions are plain
text. Duplicate terms, including identical duplicates, are errors with line
numbers; GraphArch does not select a meaning according to source order.
The original UTF-8 bytes, comments, line endings, source path, and revision are
retained in the documentation graph. Derived concepts carry notation, preferred
label, definition, scheme membership, and source-fragment provenance. The
project selects the generated scheme through ga:abbreviationScheme.
The collector accepts at most 128 KiB, 1,024 entries, 128 bytes per term, and 4,096 bytes per definition. Invalid syntax, invalid UTF-8, empty dictionaries, and paths escaping the project fail collection. HTML and PDF can render after the source tree is deleted; renderers read the pinned graph.
See abbreviation presentation for HTML explanations, code exclusions, PDF behavior, and direct RDF dictionaries.