Vignette CMS Export format
This format reads and writes the export/import XML produced by the
Vignette content management system's vgnexport tool. A document
contains one or more <importContentInstance> blocks, each holding a
set of <attribute name="…"> payloads for one content item in one
locale. The reader walks every instance and extracts the payloads
whose attribute names are listed in the parts configuration; the
remaining XML structure is preserved.
By default the reader works bilingually: source and target instances
are paired through the value of the source-id attribute, and only
paired instances reach the translatable stream. In monolingual mode
every instance is extracted independently. Encoding is forced to
UTF-8 on read and write, matching the upstream filter. A payload may
itself contain HTML; for attributes configured with the okf_html
sub-filter the reader decodes HTML entities and strips an outer <p>
wrapper, while every other configuration treats the payload as a
single literal block.
Configuration is supplied on the collection's format: block in a
.kapi project, with keys that mirror the Okapi parameter names:
partsNames — a comma-separated list of attribute names to extract
(defaults to the standard SMC attribute set); partsConfigurations —
a comma-separated list of sub-filter ids paired by index with
partsNames, where okf_html decodes HTML and any other value
(including default) means no sub-filtering; sourceId — the
attribute name holding the source-id used to pair instances (default
SOURCE_ID); localeId — the attribute name holding the locale
identifier (default LOCALE_ID); monolingual — when true, extract
every instance independently instead of pairing; and useCDATA —
when true, wrap written payloads in a CDATA section (write-side only).
This format has no configurable parameters.
Examples
Extract specific attributes monolingually
Override the parts list and extract every instance independently, decoding the body attribute as HTML.
format:
name: vignette
config:
partsNames: "SMCCONTENT-TITLE, SMCCONTENT-BODY"
partsConfigurations: "default, okf_html"
monolingual: trueProcessing notes
One block per extracted attribute payload;
partsNamesandpartsConfigurationsare paired by list index.Source and target instances pair via the
sourceIdattribute value unlessmonolingualis enabled.The
useCDATAoption affects only the writer; it wraps emitted payloads in a CDATA section and has no effect on reading.
Limitations
In bilingual mode (the default), only instances that pair through the source-id attribute are extracted; unpaired instances are skipped.
The parts configuration recognises only
okf_htmlas a sub-filter; every other value, includingdefault, is treated as no sub-filtering.Input and output encoding are forced to UTF-8.
← Back to the Format Reference