Archive Functions

• size: original file size
• last-modified: timestamp, formatted as xs:dateTime
• compressed-size: compressed file size

<archive:entry size="1840" last-modified="2024-03-20T03:30:32" compressed-size="672">
  doc/index.html
</archive:entry>

sum(archive:entries(file:read-binary('zip.zip'))/@size)

{
  "format": "zip",
  "algorithm": "deflate"
}

let $archive := file:read-binary("documents.zip")
for $entry in archive:entries($archive)[ends-with(., '.txt')]
return archive:extract-text($archive, $entry)

let $archive  := file:read-binary('archive.zip')
let $entries  := archive:entries($archive)
let $contents := archive:extract-binary($archive)
return for-each-pair($entries, $contents, fn($entry, $content) {
  file:create-dir(replace($entry, "[^/]+$", "")),
  file:write-binary($entry, $content)
})

The $entries argument contains metadata. Its items may be of type xs:string, representing the name of the file, or element(archive:entry), containing the name as string value and additional, optional attributes:

• last-modified: timestamp, specified as xs:dateTime (default: current time)
• compression-level: 0–9, 0 = uncompressed (default: 8)
• encoding: for textual entries (default: UTF-8)

<archive:entry last-modified='2011-11-11T11:11:11' compression-level='8' encoding='US-ASCII'>
  hello.txt
</archive:entry>

The $contents must have one of the following types:

• Items of type xs:string are treated as text.
• Items of type xs:base64Binary or xs:hexBinary are treated as binaries.
• In the case of updates (see below), an empty array indicates that an entry is to be deleted.

The following $options are available:

archive:create(<archive:entry>file.txt</archive:entry>, 'Hello World')

let $path  := 'audio/'
let $files := file:list($path, true(), '*.mp3')
let $zip   := archive:create($files, $files ! file:read-binary($path || .))
return file:write-binary('mp3.zip', $zip)

let $archive := archive:update('source.zip', 'delete-me.txt', [ ])
return file:write-binary('target.zip', $archive)

declare variable $input  := "HelloWorld.docx";
declare variable $output := "HelloUniverse.docx";
declare variable $doc    := "word/document.xml";

let $xml := parse-xml(archive:extract-text($input, $doc))
let $updated := $xml update {
  replace value of node .//*[text() = "HELLO WORLD!"] with "HELLO UNIVERSE!"
}
let $archive := archive:update($input, $updated, serialize($updated))
return file:write-binary($output, $archive)

let $zip := file:read-binary('old.zip')
let $entries := archive:entries($zip)[matches(., '\.x?html?$', 'i')]
return file:write-binary('new.zip', archive:delete($zip, $entries))

• recursive: parse all files recursively (default: true; ignored if entries are specified via the last argument).
• root-dir: use name of supplied directory as archive root directory (default: false).

let $zip := archive:create-from('/home/user/')
return file:write-binary('archive.zip', $zip)

archive:extract-to('.', 'archive.zip')

let $files := file:children('music')[ends-with(., 'mp3')]
return archive:write(
  'music.zip',
  ('info.txt', $files ! file:name(.)),
  ('Archive with MP3 files', $files ! file:read-binary(.))
)

• The format of $entries and $contents is the same as for archive:create.
• If the path points to a remote resource or a GZIP archive, it is rejected.
• Custom compression levels are ignored.

archive:refresh(
  'archive.zip',
  ('readme.txt', 'changelog.txt'),
  ('Read this info carefully: ...', 'These are the latest changes: ...')
)

archive:refresh('archive.zip', 'delete-me.txt', [])