From a7273c8efeea8ea0e97fa11cbfbe2c28de1ad0f6 Mon Sep 17 00:00:00 2001
From: Jonas Jenwald <jonas.jenwald@gmail.com>
Date: Wed, 27 Mar 2019 17:54:05 +0100
Subject: [PATCH] Avoid dispatching range requests to fetch PDF data that's
 already loaded with streaming (PR 10675 follow-up)

*Please note:* This patch purposely ignores `src/display/network.js`, since its support for progressive reading depends on the non-standard `moz-chunked-arraybuffer` responseType which is currently in the process of being removed.
---
 src/display/api.js              | 15 +++++++++++++++
 src/display/fetch_stream.js     |  7 +++++++
 src/display/node_stream.js      | 19 +++++++++++++------
 src/display/transport_stream.js | 16 ++++++++++++++--
 4 files changed, 49 insertions(+), 8 deletions(-)

diff --git a/src/display/api.js b/src/display/api.js
index 9c0a9ba0e..b27191ad5 100644
--- a/src/display/api.js
+++ b/src/display/api.js
@@ -1814,6 +1814,21 @@ class WorkerTransport {
       const rangeReader =
         this._networkStream.getRangeReader(data.begin, data.end);
 
+      // When streaming is enabled, it's possible that the data requested here
+      // has already been fetched via the `_fullRequestReader` implementation.
+      // However, given that the PDF data is loaded asynchronously on the
+      // main-thread and then sent via `postMessage` to the worker-thread,
+      // it may not have been available during parsing (hence the attempt to
+      // use range requests here).
+      //
+      // To avoid wasting time and resources here, we'll thus *not* dispatch
+      // range requests if the data was already loaded but has not been sent to
+      // the worker-thread yet (which will happen via the `_fullRequestReader`).
+      if (!rangeReader) {
+        sink.close();
+        return;
+      }
+
       sink.onPull = () => {
         rangeReader.read().then(function({ value, done, }) {
           if (done) {
diff --git a/src/display/fetch_stream.js b/src/display/fetch_stream.js
index 20ca3281f..7dcd3e537 100644
--- a/src/display/fetch_stream.js
+++ b/src/display/fetch_stream.js
@@ -42,6 +42,10 @@ class PDFFetchStream {
     this._rangeRequestReaders = [];
   }
 
+  get _progressiveDataLength() {
+    return (this._fullRequestReader ? this._fullRequestReader._loaded : 0);
+  }
+
   getFullReader() {
     assert(!this._fullRequestReader);
     this._fullRequestReader = new PDFFetchStreamReader(this);
@@ -49,6 +53,9 @@ class PDFFetchStream {
   }
 
   getRangeReader(begin, end) {
+    if (end <= this._progressiveDataLength) {
+      return null;
+    }
     let reader = new PDFFetchStreamRangeReader(this, begin, end);
     this._rangeRequestReaders.push(reader);
     return reader;
diff --git a/src/display/node_stream.js b/src/display/node_stream.js
index 5b744f867..a38b02bbd 100644
--- a/src/display/node_stream.js
+++ b/src/display/node_stream.js
@@ -54,19 +54,26 @@ class PDFNodeStream {
     this.isFsUrl = this.url.protocol === 'file:';
     this.httpHeaders = (this.isHttp && source.httpHeaders) || {};
 
-    this._fullRequest = null;
+    this._fullRequestReader = null;
     this._rangeRequestReaders = [];
   }
 
+  get _progressiveDataLength() {
+    return (this._fullRequestReader ? this._fullRequestReader._loaded : 0);
+  }
+
   getFullReader() {
-    assert(!this._fullRequest);
-    this._fullRequest = this.isFsUrl ?
+    assert(!this._fullRequestReader);
+    this._fullRequestReader = this.isFsUrl ?
       new PDFNodeStreamFsFullReader(this) :
       new PDFNodeStreamFullReader(this);
-    return this._fullRequest;
+    return this._fullRequestReader;
   }
 
   getRangeReader(start, end) {
+    if (end <= this._progressiveDataLength) {
+      return null;
+    }
     let rangeReader = this.isFsUrl ?
       new PDFNodeStreamFsRangeReader(this, start, end) :
       new PDFNodeStreamRangeReader(this, start, end);
@@ -75,8 +82,8 @@ class PDFNodeStream {
   }
 
   cancelAllRequests(reason) {
-    if (this._fullRequest) {
-      this._fullRequest.cancel(reason);
+    if (this._fullRequestReader) {
+      this._fullRequestReader.cancel(reason);
     }
 
     let readers = this._rangeRequestReaders.slice(0);
diff --git a/src/display/transport_stream.js b/src/display/transport_stream.js
index fe02746b7..2c055e51e 100644
--- a/src/display/transport_stream.js
+++ b/src/display/transport_stream.js
@@ -70,6 +70,10 @@ var PDFDataTransportStream = (function PDFDataTransportStreamClosure() {
        }
     },
 
+    get _progressiveDataLength() {
+      return (this._fullRequestReader ? this._fullRequestReader._loaded : 0);
+    },
+
     _onProgress: function PDFDataTransportStream_onDataProgress(evt) {
        if (this._rangeReaders.length > 0) {
          // Reporting to first range reader.
@@ -96,6 +100,9 @@ var PDFDataTransportStream = (function PDFDataTransportStreamClosure() {
     },
 
     getRangeReader: function PDFDataTransportStream_getRangeReader(begin, end) {
+      if (end <= this._progressiveDataLength) {
+        return null;
+      }
       var reader = new PDFDataTransportStreamRangeReader(this, begin, end);
       this._pdfDataRangeTransport.requestDataRange(begin, end);
       this._rangeReaders.push(reader);
@@ -121,6 +128,10 @@ var PDFDataTransportStream = (function PDFDataTransportStreamClosure() {
     this._done = false;
     this._filename = null;
     this._queuedChunks = queuedChunks || [];
+    this._loaded = 0;
+    for (const chunk of this._queuedChunks) {
+      this._loaded += chunk.byteLength;
+    }
     this._requests = [];
     this._headersReady = Promise.resolve();
     stream._fullRequestReader = this;
@@ -135,9 +146,10 @@ var PDFDataTransportStream = (function PDFDataTransportStreamClosure() {
       if (this._requests.length > 0) {
         var requestCapability = this._requests.shift();
         requestCapability.resolve({ value: chunk, done: false, });
-        return;
+      } else {
+        this._queuedChunks.push(chunk);
       }
-      this._queuedChunks.push(chunk);
+      this._loaded += chunk.byteLength;
     },
 
     get headersReady() {