corysanin/tubestation
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

correctly detect when a microsummary page isn't HTML or XML, and explicitly don't handle multipart content (multipart/x-mixed-replace), since there's no good way to determine which part of it is the last part when downloading it via XMLHttpRequest

Browse Source 
bug=346624
r=mconnor
This commit is contained in:
myk@mozilla.org
2007-01-14 05:38:23 +00:00
parent a3c949b0ee
commit 0a477ca063
1 changed files with 18 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

26
browser/components/microsummaries/src/nsMicrosummaryService.js
View File

@@ -1071,7 +1071,7 @@ Microsummary.prototype = {
},
_handlePageLoad: function MS__handlePageLoad(resource) {
if (!resource.isXML && !resource.contentType == "text/html")
if (!resource.isXML && resource.contentType != "text/html")
throw("page is neither HTML nor XML");
this.pageContent = resource.content;
@@ -1798,10 +1798,16 @@ MicrosummaryResource.prototype = {
if (this._self._httpAuthFailed) {
// Technically the request succeeded, but we treat it as a failure,
// since we aren't able to handle HTTP authentication. So we abort
// just like errorHandler does for other kinds of load failures.
// since we aren't able to handle HTTP authentication.
LOG(this._self.uri.spec + " load failed; HTTP auth required");
try { this._self.destroy() }
try { this._self._handleError(event) }
finally { this._self = null }
}
else if (event.target.channel.contentType == "multipart/x-mixed-replace") {
// Technically the request succeeded, but we treat it as a failure,
// since we aren't able to handle multipart content.
LOG(this._self.uri.spec + " load failed; contains multipart content");
try { this._self._handleError(event) }
finally { this._self = null }
}
else {
@@ -1812,9 +1818,6 @@ MicrosummaryResource.prototype = {
}
};
// At the moment the only thing we do after a load error is destroy
// ourselves to prevent memory leaks, but we should ultimately notify
// the caller and let it handle the error if it wants to.
var errorHandler = {
_self: this,
handleEvent: function MSR_errorHandler_handleEvent(event) {
@@ -1822,7 +1825,7 @@ MicrosummaryResource.prototype = {
this._self._loadTimer.cancel();
LOG(this._self.uri.spec + " load failed");
try { this._self.destroy() }
try { this._self._handleError(event) }
finally { this._self = null }
}
};
@@ -1891,6 +1894,13 @@ MicrosummaryResource.prototype = {
}
},
_handleError: function MSR__handleError(event) {
// At the moment the only thing we do after a load error is destroy
// ourselves to prevent memory leaks, but we should ultimately notify
// the caller and let it handle the error if it wants to.
this.destroy();
},
/**
* Parse a string of HTML text. Used by _load() when it retrieves HTML.
* We do this via hidden XUL iframes, which according to bz is the best way