function ExportXFDF(const FileName: string): Integer; overload;
function ExportXFDF(Stream: TStream): Integer; overload;
function ExportXFDFToStream(Stream: TStream): Integer;
ExportXFDF walks every page of the loaded document and produces an
ISO 19444-1 conformant XFDF document rooted in the
http://ns.adobe.com/xfdf/ namespace. The output consists of
a <fields> section and an <annots>
section, both emitted as UTF-8 XML.
For form fields, the method enumerates the widget annotations of every
page, reads each field value, and writes a <field>
element. Hierarchical (qualified) field names are split on the dot
separator and emitted as nested <field> elements so
that the parent/child structure round-trips. Checkboxes and radio buttons
export their export value (the appearance-state name that
represents the on state) as the field value. Signature fields are listed
but their value is never exported.
Value-presence flags keep an explicitly empty field value distinct from an omitted value, including fields nested more than two levels deep
For annotations, every non-widget, non-popup annotation on every page
is serialised. The exporter recognises 18 /Subtype values:
text, highlight, underline, strikeout, squiggly, line, circle, square,
caret, polygon, polyline, stamp, ink, freetext, fileattachment, sound,
link, and redact. Each is written as an element under
<annots> carrying its page index, bounding rectangle,
contents, title (author), name (NM), subject, modification
and creation dates, flags, and colour.
Encoding conventions follow ISO 19444-1: colours are written as
#RRGGBB hex strings; coordinate lists (rectangle, line
endpoints, polygon/polyline vertices, ink lists) are comma-separated
numbers; and annotation flags are emitted as a comma-separated list of
flag names rather than a raw bitmask.
For multi-line markup annotations, the coords attribute includes every quadrilateral in document order rather than only the first line
Annotation content uses the same presence semantics, so an explicit empty contents element survives export without turning an absent element into an empty one
PDFium itself exposes no XFDF API. ExportXFDF is implemented entirely in Pascal by PDFium Delphi Component on top of its own annotation and form-field data model, querying PDFium only for the per-annotation attribute values.
Returns the number of UTF-8 bytes written. The Stream
overload accepts a TStream to write into; the
FileName overload accepts a file path and creates/truncates
that file. ExportXFDFToStream is an alias for the
Stream overload.
CA) and geometric data (line endpoints, polygon/polyline vertices, ink lists) are exposed through high-level records that are read-only on the PDFium side, because PDFium has no SetNumberValue setter for annotation dictionaries. These attributes are preserved as-is during export, so they round-trip through XFDF even though they cannot be edited.ImportXFDF or consumed by Acrobat, and vice versa.
// Export all fields and annotations to an .xfdf file
BytesWritten := Pdf1.ExportXFDF('C:\Data\form.xfdf');
// Or to a stream
var
Stream: TFileStream;
begin
Stream := TFileStream.Create('C:\Data\form.xfdf', fmCreate);
try
Pdf1.ExportXFDF(Stream);
finally
Stream.Free;
end;
end;