Sakurai/pdf.js
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Merge pull request #9435 from Rob--W/content-disposition-quoted-string-semi-space

Browse Source 
Support spaces and semicolons in filename (Content-Disposition)
This commit is contained in:
Tim van der Meij 2018-02-04 19:15:31 +01:00 committed by GitHub
commit 24f96e0f71
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 34 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
src/display/content_disposition.js
View File

@ -14,7 +14,7 @@
*/ */
// This getFilenameFromContentDispositionHeader function is adapted from // This getFilenameFromContentDispositionHeader function is adapted from
// https://github.com/Rob--W/open-in-browser/blob/9f5fcae11cf6d99c503a15894f22efdfcd2075b7/extension/content-disposition.js // https://github.com/Rob--W/open-in-browser/blob/7e2e35a38b8b4e981b11da7b2f01df0149049e92/extension/content-disposition.js
// with the following changes: // with the following changes:
// - Modified to conform to PDF.js's coding style. // - Modified to conform to PDF.js's coding style.
// - Support UTF-8 decoding when TextDecoder is unsupported. // - Support UTF-8 decoding when TextDecoder is unsupported.
@ -33,7 +33,7 @@ function getFilenameFromContentDispositionHeader(contentDisposition) {
let needsEncodingFixup = true; let needsEncodingFixup = true;
// filename*=ext-value ("ext-value" from RFC 5987, referenced by RFC 6266). // filename*=ext-value ("ext-value" from RFC 5987, referenced by RFC 6266).
let tmp = /(?:^|;)\s*filename\*\s*=\s*([^;\s]+)/i.exec(contentDisposition); let tmp = toParamRegExp('filename\\*', 'i').exec(contentDisposition);
if (tmp) { if (tmp) {
tmp = tmp[1]; tmp = tmp[1];
let filename = rfc2616unquote(tmp); let filename = rfc2616unquote(tmp);
@ -54,7 +54,7 @@ function getFilenameFromContentDispositionHeader(contentDisposition) {
} }
// filename=value (RFC 5987, section 4.1). // filename=value (RFC 5987, section 4.1).
tmp = /(?:^|;)\s*filename\s*=\s*([^;\s]+)/.exec(contentDisposition); tmp = toParamRegExp('filename', 'i').exec(contentDisposition);
if (tmp) { if (tmp) {
tmp = tmp[1]; tmp = tmp[1];
let filename = rfc2616unquote(tmp); let filename = rfc2616unquote(tmp);
@ -65,7 +65,17 @@ function getFilenameFromContentDispositionHeader(contentDisposition) {
// After this line there are only function declarations. We cannot put // After this line there are only function declarations. We cannot put
// "return" here for readability because babel would then drop the function // "return" here for readability because babel would then drop the function
// declarations... // declarations...
function toParamRegExp(attributePattern, flags) {
return new RegExp(
'(?:^|;)\\s*' + attributePattern + '\\s*=\\s*' +
// Captures: value = token | quoted-string
// (RFC 2616, section 3.6 and referenced by RFC 6266 4.1)
'(' +
'[^";\\s][^;\\s]*' +
'|' +
'"(?:[^"\\\\]|\\\\"?)+"?' +
')', flags);
}
function textdecode(encoding, value) { function textdecode(encoding, value) {
if (encoding) { if (encoding) {
if (!/^[^\x00-\xFF]+$/.test(value)) { if (!/^[^\x00-\xFF]+$/.test(value)) {
@ -102,7 +112,7 @@ function getFilenameFromContentDispositionHeader(contentDisposition) {
let matches = [], match; let matches = [], match;
// Iterate over all filename*n= and filename*n*= with n being an integer // Iterate over all filename*n= and filename*n*= with n being an integer
// of at least zero. Any non-zero number must not start with '0'. // of at least zero. Any non-zero number must not start with '0'.
let iter = /(?:^|;)\s*filename\*((?!0\d)\d+)(\*?)\s*=\s*([^;\s]+)/ig; let iter = toParamRegExp('filename\\*((?!0\\d)\\d+)(\\*?)', 'ig');
while ((match = iter.exec(contentDisposition)) !== null) { while ((match = iter.exec(contentDisposition)) !== null) {
let [, n, quot, part] = match; let [, n, quot, part] = match;
n = parseInt(n, 10); n = parseInt(n, 10);
@ -185,7 +195,7 @@ function getFilenameFromContentDispositionHeader(contentDisposition) {
// encoding = q or b // encoding = q or b
// encoded-text = any printable ASCII character other than ? or space. // encoded-text = any printable ASCII character other than ? or space.
// ... but Firefox permits ? and space. // ... but Firefox permits ? and space.
return value.replace(/=\?([\w\-]*)\?([QqBb])\?((?:[^?]|\?(?!=))*)\?=/g, return value.replace(/=\?([\w-]*)\?([QqBb])\?((?:[^?]|\?(?!=))*)\?=/g,
function(_, charset, encoding, text) { function(_, charset, encoding, text) {
if (encoding === 'q' || encoding === 'Q') { if (encoding === 'q' || encoding === 'Q') {
// RFC 2047 section 4.2. // RFC 2047 section 4.2.

18
test/unit/network_utils_spec.js
View File

@ -174,11 +174,29 @@ describe('network_utils', function() {
} }
})).toEqual('filename.pdf'); })).toEqual('filename.pdf');
expect(extractFilenameFromHeader((headerName) => {
if (headerName === 'Content-Disposition') {
return 'attachment; filename="filename.pdf and spaces.pdf"';
}
})).toEqual('filename.pdf and spaces.pdf');
expect(extractFilenameFromHeader((headerName) => {
if (headerName === 'Content-Disposition') {
return 'attachment; filename="tl;dr.pdf"';
}
})).toEqual('tl;dr.pdf');
expect(extractFilenameFromHeader((headerName) => { expect(extractFilenameFromHeader((headerName) => {
if (headerName === 'Content-Disposition') { if (headerName === 'Content-Disposition') {
return 'attachment; filename=filename.pdf'; return 'attachment; filename=filename.pdf';
} }
})).toEqual('filename.pdf'); })).toEqual('filename.pdf');
expect(extractFilenameFromHeader((headerName) => {
if (headerName === 'Content-Disposition') {
return 'attachment; filename=filename.pdf someotherparam';
}
})).toEqual('filename.pdf');
}); });
it('gets the filename from the response header (RFC 6266)', function() { it('gets the filename from the response header (RFC 6266)', function() {