From d4700102934db36be8e18aac32b5e3cc9c5608b4 Mon Sep 17 00:00:00 2001
From: Jonas Jenwald <jonas.jenwald@gmail.com>
Date: Sat, 15 Oct 2022 11:55:37 +0200
Subject: [PATCH] Re-factor the PDF version parsing in the worker-thread
Part of this is very old code, and back when support for parsing the catalog-version was added things became less clear (in my opinion).
Hence this patch tries to improve things, by e.g. validating the header- and catalog-version separately.
---
src/core/catalog.js | 13 ++++++++-----
src/core/core_utils.js | 3 +++
src/core/document.js | 35 ++++++++++++++---------------------
3 files changed, 25 insertions(+), 26 deletions(-)
diff --git a/src/core/catalog.js b/src/core/catalog.js
index 561d6a9be..c3911c352 100644
--- a/src/core/catalog.js
+++ b/src/core/catalog.js
@@ -16,6 +16,7 @@
import {
collectActions,
MissingDataException,
+ PDF_VERSION_REGEXP,
recoverJsURL,
toRomanNumerals,
XRefEntryException,
@@ -84,11 +85,13 @@ class Catalog {
get version() {
const version = this._catDict.get("Version");
- return shadow(
- this,
- "version",
- version instanceof Name ? version.name : null
- );
+ if (version instanceof Name) {
+ if (PDF_VERSION_REGEXP.test(version.name)) {
+ return shadow(this, "version", version.name);
+ }
+ warn(`Invalid PDF catalog version: ${version.name}`);
+ }
+ return shadow(this, "version", null);
}
get lang() {
diff --git a/src/core/core_utils.js b/src/core/core_utils.js
index 8b4ed0e83..77e7f3df1 100644
--- a/src/core/core_utils.js
+++ b/src/core/core_utils.js
@@ -26,6 +26,8 @@ import {
import { Dict, isName, Ref, RefSet } from "./primitives.js";
import { BaseStream } from "./base_stream.js";
+const PDF_VERSION_REGEXP = /^[1-9]\.\d$/;
+
function getLookupTableFactory(initializer) {
let lookup;
return function () {
@@ -585,6 +587,7 @@ export {
numberToString,
ParserEOFException,
parseXFAPath,
+ PDF_VERSION_REGEXP,
readInt8,
readUint16,
readUint32,
diff --git a/src/core/document.js b/src/core/document.js
index 6fd413fdd..37a1597c4 100644
--- a/src/core/document.js
+++ b/src/core/document.js
@@ -37,6 +37,7 @@ import {
getNewAnnotationsMap,
isWhiteSpace,
MissingDataException,
+ PDF_VERSION_REGEXP,
validateCSSFont,
XRefEntryException,
XRefParseException,
@@ -712,8 +713,6 @@ const FINGERPRINT_FIRST_BYTES = 1024;
const EMPTY_FINGERPRINT =
"\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00";
-const PDF_HEADER_VERSION_REGEXP = /^[1-9]\.\d$/;
-
function find(stream, signature, limit = 1024, backwards = false) {
if (
typeof PDFJSDev === "undefined" ||
@@ -818,14 +817,6 @@ class PDFDocument {
parse(recoveryMode) {
this.xref.parse(recoveryMode);
this.catalog = new Catalog(this.pdfManager, this.xref);
-
- // The `checkHeader` method is called before this method and parses the
- // version from the header. The specification states in section 7.5.2
- // that the version from the catalog, if present, should overwrite the
- // version from the header.
- if (this.catalog.version) {
- this._version = this.catalog.version;
- }
}
get linearization() {
@@ -911,8 +902,11 @@ class PDFDocument {
) {
version += String.fromCharCode(ch);
}
- if (!this._version) {
+
+ if (PDF_VERSION_REGEXP.test(version)) {
this._version = version;
+ } else {
+ warn(`Invalid PDF header version: ${version}`);
}
}
@@ -1260,6 +1254,14 @@ class PDFDocument {
: null;
}
+ /**
+ * The specification states in section 7.5.2 that the version from
+ * the catalog, if present, should overwrite the version from the header.
+ */
+ get version() {
+ return this.catalog.version || this._version;
+ }
+
get formInfo() {
const formInfo = {
hasFields: false,
@@ -1307,17 +1309,8 @@ class PDFDocument {
}
get documentInfo() {
- let version = this._version;
- if (
- typeof version !== "string" ||
- !PDF_HEADER_VERSION_REGEXP.test(version)
- ) {
- warn(`Invalid PDF header version number: ${version}`);
- version = null;
- }
-
const docInfo = {
- PDFFormatVersion: version,
+ PDFFormatVersion: this.version,
Language: this.catalog.lang,
EncryptFilterName: this.xref.encrypt
? this.xref.encrypt.filterName