mirrors/zotero
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages 1 Wiki Activity Actions

Split out a couple helper functions for PDF retrieval

Browse source
This commit is contained in:
Dan Stillman 2018-09-08 15:43:18 -04:00
parent c6a2af7ab5
commit b8db83af08
2 changed files with 40 additions and 22 deletions
Download patch file Download diff file Expand all files Collapse all files

39
chrome/content/zotero/xpcom/attachments.js
View file

@ -858,12 +858,8 @@ Zotero.Attachments = new function(){
Zotero.Utilities.Internal.saveURI(wbp, nsIURL, path, headers);
});
// If the file is supposed to be a PDF directory, fail if it's not
let sample = await Zotero.File.getContentsAsync(path, null, 1000);
if (options.isPDF && Zotero.MIME.sniffForMIMEType(sample) != 'application/pdf') {
Zotero.debug("Downloaded PDF was not a PDF", 2);
Zotero.debug(sample, 3);
throw new this.InvalidPDFException();
if (options.isPDF) {
await _enforcePDF(path);
}
}
catch (e) {
@ -878,6 +874,19 @@ Zotero.Attachments = new function(){
};
/**
* Make sure a file is a PDF
*/
async function _enforcePDF(path) {
var sample = await Zotero.File.getContentsAsync(path, null, 1000);
if (Zotero.MIME.sniffForMIMEType(sample) != 'application/pdf') {
Zotero.debug("Downloaded PDF was not a PDF", 2);
Zotero.debug(sample, 3);
throw new Zotero.Attachments.InvalidPDFException();
}
}
this.InvalidPDFException = function() {
this.message = "Downloaded PDF was not a PDF";
this.stack = new Error().stack;
@ -1301,26 +1310,12 @@ Zotero.Attachments = new function(){
if (contentType == 'application/pdf') {
Zotero.debug("URL resolves directly to PDF");
await Zotero.File.putContentsAsync(path, blob);
await _enforcePDF(path);
return { url: responseURL, props: urlResolver };
}
// Otherwise parse the Blob into a Document and translate that
else if (contentType.startsWith('text/html')) {
let charset = 'utf-8';
let matches = contentType.match(/charset=([a-z0-9\-_+])/i);
if (matches) {
charset = matches[1];
}
let responseText = await new Promise(function (resolve) {
let fr = new FileReader();
fr.addEventListener("loadend", function() {
resolve(fr.result);
});
fr.readAsText(blob, charset);
});
let parser = Components.classes["@mozilla.org/xmlextras/domparser;1"]
.createInstance(Components.interfaces.nsIDOMParser);
let doc = parser.parseFromString(responseText, 'text/html');
doc = Zotero.HTTP.wrapDocument(doc, responseURL);
let doc = await Zotero.Utilities.Internal.blobToHTMLDocument(blob, responseURL);
url = await Zotero.Utilities.Internal.getPDFFromDocument(doc);
}
}

23
chrome/content/zotero/xpcom/utilities_internal.js
View file

@ -712,6 +712,29 @@ Zotero.Utilities.Internal = {
},
/**
* Parse a Blob (e.g., as received from Zotero.HTTP.request()) into an HTML Document
*/
blobToHTMLDocument: async function (blob, url) {
var charset = null;
var matches = blob.type && blob.type.match(/charset=([a-z0-9\-_+])/i);
if (matches) {
charset = matches[1];
}
var responseText = await new Promise(function (resolve) {
let fr = new FileReader();
fr.addEventListener("loadend", function() {
resolve(fr.result);
});
fr.readAsText(blob, charset);
});
var parser = Components.classes["@mozilla.org/xmlextras/domparser;1"]
.createInstance(Components.interfaces.nsIDOMParser);
var doc = parser.parseFromString(responseText, 'text/html');
return Zotero.HTTP.wrapDocument(doc, url);
},
/**
* Converts Zotero.Item to a format expected by translators
* This is mostly the Zotero web API item JSON format, but with an attachments