mirror of
https://github.com/mozilla/pdf.js.git
synced 2025-04-20 23:28:06 +02:00
Makes PDF data reading Streams API friendly.
This commit is contained in:
Yury Delendik
parent
8cdb69634f
commit
0d591719d9
8 changed files with 1153 additions and 178 deletions
|
|
@ -8,6 +8,7 @@
|
|||
describe('api', function() {
|
||||
var basicApiUrl = combineUrl(window.location.href, '../pdfs/basicapi.pdf');
|
||||
var basicApiFileLength = 105779; // bytes
|
||||
var TEST_TIMEOUT = 20000;
|
||||
function waitsForPromiseResolved(promise, successCallback) {
|
||||
var resolved = false;
|
||||
promise.then(function(val) {
|
||||
|
|
@ -20,7 +21,7 @@ describe('api', function() {
|
|||
});
|
||||
waitsFor(function() {
|
||||
return resolved;
|
||||
}, 20000);
|
||||
}, TEST_TIMEOUT);
|
||||
}
|
||||
function waitsForPromiseRejected(promise, failureCallback) {
|
||||
var rejected = false;
|
||||
|
|
@ -34,7 +35,13 @@ describe('api', function() {
|
|||
});
|
||||
waitsFor(function() {
|
||||
return rejected;
|
||||
}, 20000);
|
||||
}, TEST_TIMEOUT);
|
||||
}
|
||||
function waitSome(callback) {
|
||||
var WAIT_TIMEOUT = 10;
|
||||
setTimeout(function () {
|
||||
callback();
|
||||
}, WAIT_TIMEOUT);
|
||||
}
|
||||
|
||||
describe('PDFJS', function() {
|
||||
|
|
@ -710,4 +717,86 @@ describe('api', function() {
|
|||
waitsForPromiseResolved(promiseDone, function() {});
|
||||
});
|
||||
});
|
||||
describe('PDFDataRangeTransport', function () {
|
||||
var pdfPath = combineUrl(window.location.href, '../pdfs/tracemonkey.pdf');
|
||||
var loadPromise;
|
||||
function getDocumentData() {
|
||||
if (loadPromise) {
|
||||
return loadPromise;
|
||||
}
|
||||
loadPromise = new Promise(function (resolve, reject) {
|
||||
var xhr = new XMLHttpRequest(pdfPath);
|
||||
xhr.open('GET', pdfPath);
|
||||
xhr.responseType = 'arraybuffer';
|
||||
xhr.onload = function () { resolve(new Uint8Array(xhr.response)); };
|
||||
xhr.onerror = function () { reject(new Error('PDF is not loaded')); };
|
||||
xhr.send();
|
||||
});
|
||||
return loadPromise;
|
||||
}
|
||||
it('should fetch document info and page using ranges', function () {
|
||||
var transport;
|
||||
var initialDataLength = 4000;
|
||||
var fetches = 0;
|
||||
var getDocumentPromise = getDocumentData().then(function (data) {
|
||||
var initialData = data.subarray(0, initialDataLength);
|
||||
transport = new PDFJS.PDFDataRangeTransport(data.length, initialData);
|
||||
transport.requestDataRange = function (begin, end) {
|
||||
fetches++;
|
||||
waitSome(function () {
|
||||
transport.onDataProgress(4000);
|
||||
transport.onDataRange(begin, data.subarray(begin, end));
|
||||
});
|
||||
};
|
||||
var loadingTask = PDFJS.getDocument(transport);
|
||||
return loadingTask.promise;
|
||||
});
|
||||
var pdfDocument;
|
||||
var getPagePromise = getDocumentPromise.then(function (pdfDocument_) {
|
||||
pdfDocument = pdfDocument_;
|
||||
var pagePromise = pdfDocument.getPage(10);
|
||||
return pagePromise;
|
||||
});
|
||||
|
||||
waitsForPromiseResolved(getPagePromise, function (page) {
|
||||
expect(pdfDocument.numPages).toEqual(14);
|
||||
expect(page.rotate).toEqual(0);
|
||||
expect(fetches).toBeGreaterThan(2);
|
||||
});
|
||||
});
|
||||
it('should fetch document info and page using range and streaming',
|
||||
function () {
|
||||
var transport;
|
||||
var initialDataLength = 4000;
|
||||
var fetches = 0;
|
||||
var getDocumentPromise = getDocumentData().then(function (data) {
|
||||
var initialData = data.subarray(0, initialDataLength);
|
||||
transport = new PDFJS.PDFDataRangeTransport(data.length, initialData);
|
||||
transport.requestDataRange = function (begin, end) {
|
||||
fetches++;
|
||||
if (fetches === 1) {
|
||||
// send rest of the data on first range request.
|
||||
transport.onDataProgressiveRead(data.subarray(initialDataLength));
|
||||
}
|
||||
waitSome(function () {
|
||||
transport.onDataRange(begin, data.subarray(begin, end));
|
||||
});
|
||||
};
|
||||
var loadingTask = PDFJS.getDocument(transport);
|
||||
return loadingTask.promise;
|
||||
});
|
||||
var pdfDocument;
|
||||
var getPagePromise = getDocumentPromise.then(function (pdfDocument_) {
|
||||
pdfDocument = pdfDocument_;
|
||||
var pagePromise = pdfDocument.getPage(10);
|
||||
return pagePromise;
|
||||
});
|
||||
|
||||
waitsForPromiseResolved(getPagePromise, function (page) {
|
||||
expect(pdfDocument.numPages).toEqual(14);
|
||||
expect(page.rotate).toEqual(0);
|
||||
expect(fetches).toEqual(1);
|
||||
});
|
||||
});
|
||||
});
|
||||
});
|
||||
|
|
|
|||
169
test/unit/network_spec.js
Normal file
169
test/unit/network_spec.js
Normal file
|
|
@ -0,0 +1,169 @@
|
|||
/* globals expect, it, describe, waitsFor, combineUrl, PDFNetworkStream */
|
||||
|
||||
'use strict';
|
||||
|
||||
describe('network', function() {
|
||||
var pdf1 = combineUrl(window.location.href, '../pdfs/tracemonkey.pdf');
|
||||
var pdf1Length = 1016315;
|
||||
var pdf2 = combineUrl(window.location.href, '../pdfs/pdf.pdf');
|
||||
var pdf2Length = 32472771;
|
||||
|
||||
function waitsForPromiseResolved(promise, successCallback) {
|
||||
var TEST_TIMEOUT = 20000;
|
||||
var resolved = false;
|
||||
promise.then(function(val) {
|
||||
resolved = true;
|
||||
successCallback(val);
|
||||
},
|
||||
function(error) {
|
||||
// Shouldn't get here.
|
||||
expect(error).toEqual('the promise should not have been rejected');
|
||||
});
|
||||
waitsFor(function() {
|
||||
return resolved;
|
||||
}, TEST_TIMEOUT);
|
||||
}
|
||||
|
||||
it('read without stream and range', function() {
|
||||
var stream = new PDFNetworkStream({
|
||||
source: {
|
||||
url: pdf1,
|
||||
rangeChunkSize: 65536,
|
||||
disableStream: true,
|
||||
},
|
||||
disableRange: true
|
||||
});
|
||||
|
||||
var fullReader = stream.getFullReader();
|
||||
|
||||
var isStreamingSupported, isRangeSupported;
|
||||
var promise = fullReader.headersReady.then(function () {
|
||||
isStreamingSupported = fullReader.isStreamingSupported;
|
||||
isRangeSupported = fullReader.isRangeSupported;
|
||||
});
|
||||
|
||||
var len = 0, count = 0;
|
||||
var read = function () {
|
||||
return fullReader.read().then(function (result) {
|
||||
if (result.done) {
|
||||
return;
|
||||
}
|
||||
count++;
|
||||
len += result.value.byteLength;
|
||||
return read();
|
||||
});
|
||||
};
|
||||
|
||||
var readPromise = read();
|
||||
|
||||
waitsForPromiseResolved(readPromise, function (page) {
|
||||
expect(len).toEqual(pdf1Length);
|
||||
expect(count).toEqual(1);
|
||||
expect(isStreamingSupported).toEqual(false);
|
||||
expect(isRangeSupported).toEqual(false);
|
||||
});
|
||||
});
|
||||
|
||||
it('read with streaming', function() {
|
||||
var userAgent = window.navigator.userAgent;
|
||||
// The test is valid for FF only: the XHR has support of the
|
||||
// 'moz-chunked-array' response type.
|
||||
// TODO enable for other browsers, e.g. when fetch/streams API is supported.
|
||||
var m = /Mozilla\/5.0.*?rv:(\d+).*? Gecko/.exec(userAgent);
|
||||
if (!m || m[1] < 9) {
|
||||
return;
|
||||
}
|
||||
|
||||
var stream = new PDFNetworkStream({
|
||||
source: {
|
||||
url: pdf2,
|
||||
rangeChunkSize: 65536,
|
||||
disableStream: false,
|
||||
},
|
||||
disableRange: false
|
||||
});
|
||||
|
||||
var fullReader = stream.getFullReader();
|
||||
|
||||
var isStreamingSupported, isRangeSupported;
|
||||
var promise = fullReader.headersReady.then(function () {
|
||||
isStreamingSupported = fullReader.isStreamingSupported;
|
||||
isRangeSupported = fullReader.isRangeSupported;
|
||||
});
|
||||
|
||||
var len = 0, count = 0;
|
||||
var read = function () {
|
||||
return fullReader.read().then(function (result) {
|
||||
if (result.done) {
|
||||
return;
|
||||
}
|
||||
count++;
|
||||
len += result.value.byteLength;
|
||||
return read();
|
||||
});
|
||||
};
|
||||
|
||||
var readPromise = read();
|
||||
|
||||
waitsForPromiseResolved(readPromise, function (page) {
|
||||
expect(len).toEqual(pdf2Length);
|
||||
expect(count).toBeGreaterThan(1);
|
||||
expect(isStreamingSupported).toEqual(true);
|
||||
});
|
||||
});
|
||||
|
||||
it('read custom ranges', function () {
|
||||
// We don't test on browsers that don't support range request, so
|
||||
// requiring this test to pass.
|
||||
var rangeSize = 32768;
|
||||
var stream = new PDFNetworkStream({
|
||||
source: {
|
||||
url: pdf1,
|
||||
length: pdf1Length,
|
||||
rangeChunkSize: rangeSize,
|
||||
disableStream: true,
|
||||
},
|
||||
disableRange: false
|
||||
});
|
||||
|
||||
var fullReader = stream.getFullReader();
|
||||
|
||||
var isStreamingSupported, isRangeSupported, fullReaderCancelled;
|
||||
var promise = fullReader.headersReady.then(function () {
|
||||
isStreamingSupported = fullReader.isStreamingSupported;
|
||||
isRangeSupported = fullReader.isRangeSupported;
|
||||
// we shall be able to close the full reader without issues
|
||||
fullReader.cancel('Don\'t need full reader');
|
||||
fullReaderCancelled = true;
|
||||
});
|
||||
|
||||
// Skipping fullReader results, requesting something from the PDF end.
|
||||
var tailSize = (pdf1Length % rangeSize) || rangeSize;
|
||||
|
||||
var range1Reader = stream.getRangeReader(pdf1Length - tailSize - rangeSize,
|
||||
pdf1Length - tailSize);
|
||||
var range2Reader = stream.getRangeReader(pdf1Length - tailSize, pdf1Length);
|
||||
|
||||
var result1 = {value: 0}, result2 = {value: 0};
|
||||
var read = function (reader, lenResult) {
|
||||
return reader.read().then(function (result) {
|
||||
if (result.done) {
|
||||
return;
|
||||
}
|
||||
lenResult.value += result.value.byteLength;
|
||||
return read(reader, lenResult);
|
||||
});
|
||||
};
|
||||
|
||||
var readPromises = Promise.all([read(range1Reader, result1),
|
||||
read(range2Reader, result2),
|
||||
promise]);
|
||||
|
||||
waitsForPromiseResolved(readPromises, function (page) {
|
||||
expect(result1.value).toEqual(rangeSize);
|
||||
expect(result2.value).toEqual(tailSize);
|
||||
expect(isRangeSupported).toEqual(true);
|
||||
expect(fullReaderCancelled).toEqual(true);
|
||||
});
|
||||
});
|
||||
});
|
||||
|
|
@ -36,6 +36,7 @@
|
|||
<script src="util_spec.js"></script>
|
||||
<script src="cmap_spec.js"></script>
|
||||
<script src="annotation_layer_spec.js"></script>
|
||||
<script src="network_spec.js"></script>
|
||||
|
||||
<script>
|
||||
'use strict';
|
||||
|
|
@ -48,11 +49,12 @@
|
|||
'pdfjs/core/annotation', 'pdfjs/core/crypto', 'pdfjs/core/stream',
|
||||
'pdfjs/core/fonts', 'pdfjs/core/ps_parser', 'pdfjs/core/function',
|
||||
'pdfjs/core/parser', 'pdfjs/core/evaluator', 'pdfjs/core/cmap',
|
||||
'pdfjs/core/worker', 'pdfjs/display/api', 'pdfjs/display/metadata'],
|
||||
'pdfjs/core/worker', 'pdfjs/core/network', 'pdfjs/display/api',
|
||||
'pdfjs/display/metadata'],
|
||||
function (sharedUtil, sharedGlobal, corePrimitives, coreAnnotation,
|
||||
coreCrypto, coreStream, coreFonts, corePsParser, coreFunction,
|
||||
coreParser, coreEvaluator, coreCMap, coreWorker, displayAPI,
|
||||
displayMetadata) {
|
||||
coreParser, coreEvaluator, coreCMap, coreWorker,
|
||||
coreNetwork, displayAPI, displayMetadata) {
|
||||
|
||||
pdfjsLibs = {
|
||||
sharedUtil: sharedUtil,
|
||||
|
|
@ -68,6 +70,7 @@
|
|||
coreEvaluator: coreEvaluator,
|
||||
coreCMap: coreCMap,
|
||||
coreWorker: coreWorker,
|
||||
coreNetwork: coreNetwork,
|
||||
displayAPI: displayAPI,
|
||||
displayMetadata: displayMetadata
|
||||
};
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue