From 27791cc891c51b09386eae3e44db784dc4b72a3b Mon Sep 17 00:00:00 2001
From: m-holger <m.holger@qpdf.org>
Date: Sat, 2 Nov 2024 14:18:41 +0000
Subject: [PATCH] Refactor QPDF::isLinearized

---
 libqpdf/QPDF.cc               |  3 ++-
 libqpdf/QPDF_linearization.cc | 29 +++++++++++++----------------
 2 files changed, 15 insertions(+), 17 deletions(-)

diff --git a/libqpdf/QPDF.cc b/libqpdf/QPDF.cc
index ce42916..e56fc22 100644
--- a/libqpdf/QPDF.cc
+++ b/libqpdf/QPDF.cc
@@ -2764,7 +2764,8 @@ QPDF::pipeStreamData(
     try {
         auto buf = file->read(length, offset);
         if (buf.size() != length) {
-            throw damagedPDF(*file, "", offset + toO(buf.size()), "unexpected EOF reading stream data");
+            throw damagedPDF(
+                *file, "", offset + toO(buf.size()), "unexpected EOF reading stream data");
         }
         pipeline->write(buf.data(), length);
         attempted_finish = true;
diff --git a/libqpdf/QPDF_linearization.cc b/libqpdf/QPDF_linearization.cc
index b79959e..e02b19e 100644
--- a/libqpdf/QPDF_linearization.cc
+++ b/libqpdf/QPDF_linearization.cc
@@ -20,6 +20,7 @@
 #include <cstring>
 
 using namespace qpdf;
+using namespace std::literals;
 
 template <class T, class int_type>
 static void
@@ -97,26 +98,18 @@ QPDF::isLinearized()
 
     // The PDF spec says the linearization dictionary must be completely contained within the first
     // 1024 bytes of the file. Add a byte for a null terminator.
-    auto buffer =  m->file->read(1024, 0);
-
-    auto buf = buffer.data();
-    auto tbuf_size = buffer.size();
+    auto buffer = m->file->read(1024, 0);
     int lindict_obj = -1;
-    char* p = buf;
+    size_t pos = 0;
     while (lindict_obj == -1) {
         // Find a digit or end of buffer
-        while (((p - buf) < tbuf_size) && (!util::is_digit(*p))) {
-            ++p;
-        }
-        if (p - buf == tbuf_size) {
-            break;
+        pos = buffer.find_first_of("0123456789"sv, pos);
+        if (pos == std::string::npos) {
+            return false;
         }
         // Seek to the digit. Then skip over digits for a potential
         // next iteration.
-        m->file->seek(p - buf, SEEK_SET);
-        while (((p - buf) < tbuf_size) && util::is_digit(*p)) {
-            ++p;
-        }
+        m->file->seek(toO(pos), SEEK_SET);
 
         QPDFTokenizer::Token t1 = readToken(*m->file);
         if (t1.isInteger() && readToken(*m->file).isInteger() &&
@@ -124,13 +117,17 @@ QPDF::isLinearized()
             readToken(*m->file).getType() == QPDFTokenizer::tt_dict_open) {
             lindict_obj = toI(QUtil::string_to_ll(t1.getValue().c_str()));
         }
+        pos = buffer.find_first_not_of("0123456789"sv, pos);
+        if (pos == std::string::npos) {
+            return false;
+        }
     }
 
     if (lindict_obj <= 0) {
         return false;
     }
 
-    auto candidate = getObjectByID(lindict_obj, 0);
+    auto candidate = getObject(lindict_obj, 0);
     if (!candidate.isDictionary()) {
         return false;
     }
@@ -545,7 +542,7 @@ QPDF::maxEnd(ObjUser const& ou)
     }
     qpdf_offset_t end = 0;
     for (auto const& og: m->obj_user_to_objects[ou]) {
-        if (m->obj_cache.count(og) == 0) {
+        if (!m->obj_cache.count(og)) {
             stopOnError("unknown object referenced in object user table");
         }
         end = std::max(end, m->obj_cache[og].end_after_space);
--
libgit2 0.21.4