file2html-ooxml
Version:
OOXML engine for file2html
132 lines (131 loc) • 6.54 kB
JavaScript
;
var __extends = (this && this.__extends) || (function () {
var extendStatics = Object.setPrototypeOf ||
({ __proto__: [] } instanceof Array && function (d, b) { d.__proto__ = b; }) ||
function (d, b) { for (var p in b) if (b.hasOwnProperty(p)) d[p] = b[p]; };
return function (d, b) {
extendStatics(d, b);
function __() { this.constructor = d; }
d.prototype = b === null ? Object.create(b) : (__.prototype = b.prototype, new __());
};
})();
Object.defineProperty(exports, "__esModule", { value: true });
var file2html = require("file2html");
var mime = require("file2html/lib/mime");
var errors_1 = require("file2html/lib/errors");
var sax_1 = require("file2html-xml-tools/lib/sax");
var file2html_archive_tools_1 = require("file2html-archive-tools");
var parse_core_props_1 = require("./parse-core-props");
var parse_document_content_1 = require("./word/parse-document-content");
var parse_document_styles_1 = require("./word/parse-document-styles");
var parse_document_relations_1 = require("./word/parse-document-relations");
var documentMimeType = mime.lookup('.docx');
// TODO: support Presentations and Spreadsheets
var supportedMimeTypes = [documentMimeType];
var OOXMLReader = (function (_super) {
__extends(OOXMLReader, _super);
function OOXMLReader() {
return _super !== null && _super.apply(this, arguments) || this;
}
OOXMLReader.prototype.read = function (_a) {
var fileInfo = _a.fileInfo;
var content = fileInfo.content;
var byteLength = content.byteLength;
return file2html_archive_tools_1.readArchive(content).then(function (archive) {
function getInvalidFileError(message) {
var archiveTree = Object.keys(archive.files || {}).join(',\n');
return Promise.reject(new Error(errors_1.errorsNamespace + ".invalidFile. Archive: [" + archiveTree + "]." + (message || '')));
}
var documentRelations = archive.file('_rels/.rels');
if (!documentRelations) {
return getInvalidFileError();
}
return documentRelations.async('string').then(function (documentRelationsContent) {
var corePropertiesEntryPath;
var documentEntryPath;
sax_1.parseXML(documentRelationsContent, {
onopentag: function (tagName, attrs) {
if (tagName === 'Relationship') {
var Type = attrs.Type, Target = attrs.Target;
if (!Type || !Target) {
return;
}
// `Target` might start from an excess slash symbol
var path = Target.replace(/^\//, '');
if (Type.indexOf('officeDocument') > 0) {
documentEntryPath = path;
}
else if (Type.indexOf('core-properties') > 0) {
corePropertiesEntryPath = path;
}
}
}
});
var documentEntry = documentEntryPath && archive.file(documentEntryPath);
if (!documentEntry) {
return getInvalidFileError("\ndocumentEntry not found,\n" + documentRelationsContent);
}
var documentFilename = documentEntryPath.split('/').pop();
var meta = Object.assign({
fileType: file2html.FileTypes.document,
mimeType: '',
name: '',
size: byteLength,
creator: '',
createdAt: '',
modifiedAt: ''
}, fileInfo.meta);
var queue = [];
var dataType = 'string';
var styles = '';
var content = '<div></div>';
var relations = {};
var relationsEntry = archive.file("word/_rels/" + documentFilename + ".rels");
if (relationsEntry) {
queue.push(relationsEntry.async(dataType).then(function (data) {
return parse_document_relations_1.default(data, archive).then(function (documentRelations) {
relations = documentRelations;
});
}));
}
return Promise.all(queue).then(function () {
var queue = [];
var coreProperties = (corePropertiesEntryPath &&
archive.file(corePropertiesEntryPath));
if (coreProperties) {
queue.push(coreProperties.async(dataType).then(function (data) { return parse_core_props_1.default(data, meta); }));
}
// is WordProcessingML
if (meta.mimeType === documentMimeType) {
var stylesEntry = archive.file('word/styles.xml');
if (stylesEntry) {
queue.push(stylesEntry.async(dataType).then(function (data) {
return parse_document_styles_1.default(data);
}).then(function (documentStyles) {
styles += '\n' + documentStyles;
}));
}
queue.push(documentEntry.async(dataType).then(function (data) {
return parse_document_content_1.default(data, {
relations: relations
});
}).then(function (data) {
styles += '\n' + data.styles;
content = data.content;
}));
}
return Promise.all(queue).then(function () { return new file2html.File({
meta: meta,
styles: "<style>" + styles + "</style>",
content: content
}); });
});
});
});
};
OOXMLReader.testFileMimeType = function (mimeType) {
return supportedMimeTypes.indexOf(mimeType) >= 0;
};
return OOXMLReader;
}(file2html.Reader));
exports.default = OOXMLReader;