Document Metadata

The <Metadata> section inside <Document> sets PDF information fields (author, title, dates, etc.) that are embedded in the output file.

Example

<Document>
  <Metadata>
    <DocumentTitle>Annual Report 2025</DocumentTitle>
    <Author>Docraft Team</Author>
    <Subject>Financial summary</Subject>
    <Keywords>finance, report, 2025</Keywords>
    <Creator>Docraft 1.0</Creator>
    <Producer>Docraft/libharu</Producer>
    <CreationDate>
      <Year>2025</Year>
      <Month>12</Month>
      <Day>31</Day>
    </CreationDate>
    <AutoKeywords enabled="true" max_keywords="15"
                  min_length="4" language="en"/>
  </Metadata>
  <Body>
    <!-- content -->
  </Body>
</Document>

Metadata Fields

Element	Description
`<DocumentTitle>`	PDF title string.
`<Author>`	Author name.
`<Creator>`	Application that created the content.
`<Producer>`	Application that produced the PDF.
`<Subject>`	Document subject.
`<Keywords>`	Comma-separated keyword list.
`<Trapped>`	Trapping flag (`True`, `False`, `Unknown`).
`<GtsPdfx>`	GTS_PDFXVersion string.

Date Elements

<CreationDate> and <ModificationDate> contain sub-elements:

Element	Type	Description
`<Year>`	int	Four-digit year.
`<Month>`	int	Month (1–12).
`<Day>`	int	Day (1–31).
`<Hour>`	int	Hour (0–23).
`<Minutes>`	int	Minutes (0–59).
`<Seconds>`	int	Seconds (0–59).
`<Ind>`	char	UTC indicator (`+`, `-`, or `Z`).
`<OffHour>`	int	UTC offset hours.
`<OffMinutes>`	int	UTC offset minutes.

Automatic Keywords

Enable <AutoKeywords> to have the library automatically extract keywords from text nodes using term-frequency analysis.

Attribute	Type	Description
`enabled`	bool	Enable/disable extraction.
`max_keywords`	int	Maximum number of keywords (default `10`).
`min_length`	int	Minimum word length (default `4`).
`language`	string	Stop-word languages, comma-separated (`en,it,fr,de,es`).