summaryrefslogtreecommitdiff
path: root/macros/luatex/generic/minim-xmp/minim-xmp.doc
blob: 01d2084f6ced937fae191d8f79e1f39d83378a20 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196

\input minim-doc.sty

\manual{minim-xmp}

\ifchapter \chapter Metadata

\else \startmetadata
    author   {Esger Renkema}
    title    {minim-xmp}
    date     {2021-06-01}
    version  {2021/1.0}
    keywords {LuaTeX; Plain TeX; XMP; metadata; PDF/A;}
    stopmetadata
\maketitle \fi

This package enables simple XMP (eXtensible Metadata Platform) packet inclusion 
and will automatically generate pdf/a extension schemas. Use it by saying 
⟦\input minim-xmp.tex⟧.
Thereafter, you can use ⟦*\setmetadata key {value}⟧ and ⟦*\getmetadata key⟧ for 
setting and retrieving document-level metadata values.

You do not need this package if you have your metadata ready-made in a separate 
file, for then you can simply say
\startelement{Code}
\startsinglecontentitem$$\halign{#\hfil\cr
⟦\immediate\pdfextension obj uncompressed⟧ \cr
\qquad⟦stream attr {/Type/Metadata /Subtype/XML}⟧ \cr
\qquad⟦file {your-file.xmp}⟧ \cr
⟦\pdfextension catalog⟧ \cr
\qquad⟦{/Metadata \pdffeedback lastobj 0 R}⟧\cr
}$$\stopsinglecontentitem
\vskip-\belowdisplayskip

\section Setting metadata

Metadata fields that contain (ordered or unordered) lists will be split on the 
⟦*\metadataseparator⟧ character; this is a semicolon by default.
Alternatively, you can just make multiple assignments: these will be 
concatenated.

Where applicable, language alternatives can be given like
⟦\setmetadata /de⟧ ⟦dc:title {...}⟧ or
⟦\setmetadata /{de_DE} dc:title {...}⟧.
Braces are necessary in the second case because the catcode of the underscore 
is not 11~or~12.
When no alternative is given, the value ⟦*x-default⟧ will be used.

Instead of using ⟦*\setmetadata⟧, multiple fields can be set in one go with 
⟦*\startmetadata⟧.
This way is particularly useful when assigning structured data to a key (see 
later on).
In this example, ⟦key1⟧ contains a normal value, ⟦key2⟧ language alternatives 
and ⟦key3⟧ structured data:

⟦\startmetadata
    key1 {...}
    key2 {... (default) ...}
        /alt1 {...}
        /alt2 {...}
    key3
        /field1 {...}
        /field2 {...}
    stopmetadata⟧

Since metadata values are read by lua as text, linebreaks and paragraphs are 
not preserved. You can work around this by saying
⟦{\def\par{\Uchar"A\Uchar"A} \setmetadata abstract {...}}⟧.

\section Getting metadata

Metadata values can be retrieved again with ⟦*\getmetadata key⟧. This command 
is fully expandible.

If the data is a list, it will be returned according to the current value of 
⟦\metadataseparator⟧.
If the data has language alternatives, the ⟦x-default⟧ value will be returned:
the alternatives are accessible by ⟦*\getmetadata /lang key⟧.

For structured types (discussed below), ⟦\getmetadata /field key⟧ will return 
the value of a single field and ⟦\getmetadata key⟧ will return all fields as
⟦/{field1} {value1} /{field2} {value2} ...⟧ (this can be used again as input to 
⟦\startmetadata⟧).

\section Supported metadata keys

Initially, the ⟦\setmetadata⟧ and ⟦\getmetadata⟧ recognise all pdf/a compatible 
fields in the ⟦pdf⟧, ⟦pdfaid⟧, ⟦dc⟧, ⟦xmp⟧, ⟦xmpMM⟧ and ⟦xmpRights⟧ namespaces.
Keys should be prefixed with their namespace, e.g. ⟦dc:creator⟧ or 
⟦xmp:CreatorTool⟧.
Note that the ⟦dc⟧ namespace has lower-case fields; field names are 
case-sensitive.

Because the precise details of the above metadata namespaces can be confusing, 
you might want use one of the aliases
⟦author⟧ (dc:creator), ⟦title⟧ (dc:title), ⟦date⟧ (dc:date and xmp:CreateDate), 
⟦language⟧ (dc:language), ⟦keywords⟧ (dc:subject and pdf:Keywords), ⟦publisher⟧ 
(dc:publisher), ⟦abstract⟧ (dc:description), ⟦copyright⟧ (dc:rights), ⟦version⟧ 
(xmpMM:VersionID) and ⟦identifier⟧ (dc:identifier).
New aliases can be defined in the ⟦aliases⟧ table of the lua module.


\section Adding new keys and schemas

\directlua{require('minim-xmp').add_namespace(
  'Example namespace', 'colours',
  'http://example.com/minim/colours/', {
    Favourite = { 'Colour', 'The author’s favourite colour' },
  }, {
    Colour = { 'RGB Colour', {
      R = { 'Integer', 'Red component' },
      G = { 'Integer', 'Green component' },
      B = { 'Integer', 'Blue component' }
    }, prefix = 'c' },
  })}
\startmetadata colours:Favourite /R 0 /G 0 /B 255 stopmetadata

New metadata namespaces can be added in the following way:

⟦require('minim-xmp').add_namespace(
  'Example namespace', 'colours',
  'http://example.com/minim/colours/', {
    -- metadata keys
    Favourite = { 'Colour', 'The author’s favourite colour' },
  }, {
    -- value types
    Colour = { 'RGB Colour', {
      R = { 'Integer', 'Red component' },
      G = { 'Integer', 'Green component' },
      B = { 'Integer', 'Blue component' }
    }, prefix = 'c' },
  })⟧

This will setup a namespace with prefix ⟦colours⟧ and one key: ⟦Favourite⟧, of 
type ⟦Colour⟧. That value type happens to be a structured type with three 
fields, which are also described.
You can now use this namespace as

⟦\startmetadata colours:Favourite /R 0 /G 0 /B 255 stopmetadata⟧

or the equivalent but more verbose

⟦\setmetadata/R colours:Favourite 0
\setmetadata/G colours:Favourite 0
\setmetadata/B colours:Favourite 255⟧

after which the generated XMP code will be

⟦  <rdf:Description rdf:about=""
      xmlns:colours="http://example.com/minim/colours/"
      xmlns:c="http://example.com/minim/colours/">
    <colours:Favourite rdf:parseType="Resource">
      <c:B>255</c:B>
      <c:G>0</c:G>
      <c:R>0</c:R>
    </colours:Favourite>
  </rdf:Description>⟧

You probably will not need defining your own value types,
so in most cases you should omit the fifth argument to ⟦add_namespace⟧.
If you do define a new value type, you can specify its prefix if it differs 
from the namespace prefix (as is done above) and likewise its ⟦uri⟧ identifier 
if it differs from the namespace URI.

List types can be given as ⟦'Bag TypeName'⟧ or ⟦'Seq TypeName'⟧; language 
alternatives as ⟦'Lang Alt'⟧. All other types will be treated as ⟦'Text'⟧, 
though for ⟦'Boolean'⟧, ⟦'Integer'⟧ and ⟦'Date'⟧ some validation is performed 
when setting values.

Additional metadata value type handling can be defined in the ⟦getters⟧ and 
⟦setters⟧ tables of the lua module.
Appropriate entries to those tables will be generated automatically for new 
structured types (which is why you could set the colour like we did above).
Value types without fields, however, will be stored and retrieved as if they 
were ⟦Text⟧ until you provide another way.


\section Generated XMP

All metadata will be written to the PDF file uncompressed.

The ⟦*\metadatamodification⟧ setting controls whether XMP packets will be 
marked as read-only (value~0; default) or writeable (value~1).
Writeable XMP packets will be padded with about 2\thinspace kB of whitespace.

If the document-level metadata contains values in the ⟦pdfaid⟧ namespace, 
metadata extension schemas will be appended to the document-level metadata 
packet automatically.
These extension schema’s will only include keys that have been set somewhere, 
though they need not have been set in the document-level metadata.
No extension schemas are generated for the built-in namespaces, as they are 
already included in the pdf/a standards.


\endmanual