open-agreements
Version:
Open-source legal template filling CLI and library
93 lines • 3.92 kB
JavaScript
import AdmZip from 'adm-zip';
const XML_NAMESPACE = 'http://www.w3.org/XML/1998/namespace';
/** Set xml:space without duplicating a parsed namespaced attribute. */
export function preserveXmlSpace(element) {
for (let index = element.attributes.length - 1; index >= 0; index--) {
const attribute = element.attributes.item(index);
if (attribute && (attribute.name === 'xml:space' ||
(attribute.namespaceURI === XML_NAMESPACE && attribute.localName === 'space'))) {
element.removeAttributeNode(attribute);
}
}
element.setAttributeNS(XML_NAMESPACE, 'xml:space', 'preserve');
}
/**
* Enumerate all OOXML text parts present in a DOCX zip.
*/
export function enumerateTextParts(zip) {
const entries = zip.getEntries().map((e) => e.entryName);
const headerPattern = /^word\/header\d+\.xml$/;
const footerPattern = /^word\/footer\d+\.xml$/;
return {
document: entries.includes('word/document.xml') ? 'word/document.xml' : null,
headers: entries.filter((e) => headerPattern.test(e)).sort(),
footers: entries.filter((e) => footerPattern.test(e)).sort(),
endnotes: entries.includes('word/endnotes.xml') ? 'word/endnotes.xml' : null,
footnotes: entries.includes('word/footnotes.xml') ? 'word/footnotes.xml' : null,
comments: entries.filter((e) => /^word\/comments(?:Extended|Extensible)?\.xml$/.test(e)).sort(),
};
}
/**
* Return a flat list of part names that should be processed by the patcher,
* cleaner (for footnote refs and paragraph patterns), verifier, and scanner.
*
* Excludes footnotes.xml because it has separator/continuationSeparator logic
* that requires special handling in the cleaner.
*/
export function getGeneralTextPartNames(parts) {
const names = [];
if (parts.document)
names.push(parts.document);
names.push(...parts.headers);
names.push(...parts.footers);
if (parts.endnotes)
names.push(parts.endnotes);
return names;
}
/**
* Return every text part, footnotes included.
*
* `getGeneralTextPartNames()` excludes `word/footnotes.xml` because the
* *cleaner* has to special-case its separator / continuationSeparator
* paragraphs. That exclusion is about mutation. Read-only passes — detecting
* which fields the source owns a sigil for, and scanning rendered output for
* doubling artifacts — have no such constraint, and a document whose footnote
* renders `8%%` while the verifier reports success is worse than one that
* fails loudly. Use this list for inspection; use the general list for
* anything that rewrites parts.
*/
export function getAllTextPartNames(parts) {
const names = getGeneralTextPartNames(parts);
if (parts.footnotes)
names.push(parts.footnotes);
names.push(...parts.comments);
return names;
}
/**
* Copy OPC parts from one zip to another, excluding directory entries.
*
* OOXML packages are a flat set of parts; directory entries such as `word/`
* are JSZip/zip container artifacts and can make Word repair the document.
*
* `getData` may return a replacement Buffer for an entry, or `null` to drop
* the entry from the destination entirely (used for orphaned media parts).
*/
export function copyEntriesSkippingDirs(sourceZip, destinationZip, getData = (_entryName, entryData) => entryData) {
for (const entry of sourceZip.getEntries()) {
if (entry.isDirectory || entry.entryName.endsWith('/'))
continue;
const data = getData(entry.entryName, entry.getData());
if (data === null)
continue;
destinationZip.addFile(entry.entryName, data);
}
}
/**
* Rebuild a zip from scratch while dropping directory entries.
*/
export function rezipWithoutDirEntries(zip) {
const outZip = new AdmZip();
copyEntriesSkippingDirs(zip, outZip);
return outZip;
}
//# sourceMappingURL=ooxml-parts.js.map