octoleo/qpdf
2
1
Fork 0
mirror of https://github.com/qpdf/qpdf.git synced 2024-12-22 10:58:58 +00:00
Code Issues Releases Wiki Activity

In QPDF::readToken change input parameter to InputSource&

Browse Source
This commit is contained in:
m-holger 2024-08-22 00:30:18 +01:00
parent 20edfb3f91
commit 258343fcc9
3 changed files with 23 additions and 22 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
include/qpdf/QPDF.hh
View File

@ -1061,7 +1061,7 @@ class QPDF
QPDFObjectHandle readObjectInStream(std::shared_ptr<InputSource>& input, int obj); QPDFObjectHandle readObjectInStream(std::shared_ptr<InputSource>& input, int obj);
size_t recoverStreamLength( size_t recoverStreamLength(
std::shared_ptr<InputSource> input, QPDFObjGen const& og, qpdf_offset_t stream_offset); std::shared_ptr<InputSource> input, QPDFObjGen const& og, qpdf_offset_t stream_offset);
QPDFTokenizer::Token readToken(std::shared_ptr<InputSource>, size_t max_len = 0); QPDFTokenizer::Token readToken(InputSource&, size_t max_len = 0);
QPDFObjectHandle readObjectAtOffset( QPDFObjectHandle readObjectAtOffset(
bool attempt_recovery, bool attempt_recovery,

36
libqpdf/QPDF.cc
View File

@ -419,7 +419,7 @@ QPDF::findHeader()
bool bool
QPDF::findStartxref() QPDF::findStartxref()
{ {
if (readToken(m->file).isWord("startxref") && readToken(m->file).isInteger()) { if (readToken(*m->file).isWord("startxref") && readToken(*m->file).isInteger()) {
// Position in front of offset token // Position in front of offset token
m->file->seek(m->file->getLastOffset(), SEEK_SET); m->file->seek(m->file->getLastOffset(), SEEK_SET);
return true; return true;
@ -457,7 +457,7 @@ QPDF::parse(char const* password)
PatternFinder sf(*this, &QPDF::findStartxref); PatternFinder sf(*this, &QPDF::findStartxref);
qpdf_offset_t xref_offset = 0; qpdf_offset_t xref_offset = 0;
if (m->file->findLast("startxref", start_offset, 0, sf)) { if (m->file->findLast("startxref", start_offset, 0, sf)) {
xref_offset = QUtil::string_to_ll(readToken(m->file).getValue().c_str()); xref_offset = QUtil::string_to_ll(readToken(*m->file).getValue().c_str());
} }
try { try {
@ -575,12 +575,12 @@ QPDF::reconstruct_xref(QPDFExc& e)
// Don't allow very long tokens here during recovery. All the interesting tokens are covered. // Don't allow very long tokens here during recovery. All the interesting tokens are covered.
static size_t const MAX_LEN = 10; static size_t const MAX_LEN = 10;
while (m->file->tell() < eof) { while (m->file->tell() < eof) {
QPDFTokenizer::Token t1 = readToken(m->file, MAX_LEN); QPDFTokenizer::Token t1 = readToken(*m->file, MAX_LEN);
qpdf_offset_t token_start = m->file->tell() - toO(t1.getValue().length()); qpdf_offset_t token_start = m->file->tell() - toO(t1.getValue().length());
if (t1.isInteger()) { if (t1.isInteger()) {
auto pos = m->file->tell(); auto pos = m->file->tell();
QPDFTokenizer::Token t2 = readToken(m->file, MAX_LEN); QPDFTokenizer::Token t2 = readToken(*m->file, MAX_LEN);
if ((t2.isInteger()) && (readToken(m->file, MAX_LEN).isWord("obj"))) { if ((t2.isInteger()) && (readToken(*m->file, MAX_LEN).isWord("obj"))) {
int obj = QUtil::string_to_int(t1.getValue().c_str()); int obj = QUtil::string_to_int(t1.getValue().c_str());
int gen = QUtil::string_to_int(t2.getValue().c_str()); int gen = QUtil::string_to_int(t2.getValue().c_str());
if (obj <= m->xref_table_max_id) { if (obj <= m->xref_table_max_id) {
@ -989,7 +989,7 @@ QPDF::read_xrefTable(qpdf_offset_t xref_offset)
} }
} }
qpdf_offset_t pos = m->file->tell(); qpdf_offset_t pos = m->file->tell();
if (readToken(m->file).isWord("trailer")) { if (readToken(*m->file).isWord("trailer")) {
break; break;
} else { } else {
m->file->seek(pos, SEEK_SET); m->file->seek(pos, SEEK_SET);
@ -1470,7 +1470,7 @@ QPDF::readTrailer()
// Nothing in the PDF spec appears to allow empty objects, but they have been encountered in // Nothing in the PDF spec appears to allow empty objects, but they have been encountered in
// actual PDF files and Adobe Reader appears to ignore them. // actual PDF files and Adobe Reader appears to ignore them.
warn(damagedPDF("trailer", "empty object treated as null")); warn(damagedPDF("trailer", "empty object treated as null"));
} else if (object.isDictionary() && readToken(m->file).isWord("stream")) { } else if (object.isDictionary() && readToken(*m->file).isWord("stream")) {
warn(damagedPDF("trailer", m->file->tell(), "stream keyword found in trailer")); warn(damagedPDF("trailer", m->file->tell(), "stream keyword found in trailer"));
} }
// Override last_offset so that it points to the beginning of the object we just read // Override last_offset so that it points to the beginning of the object we just read
@ -1496,10 +1496,10 @@ QPDF::readObject(std::string const& description, QPDFObjGen og)
warn(damagedPDF(*m->file, m->file->getLastOffset(), "empty object treated as null")); warn(damagedPDF(*m->file, m->file->getLastOffset(), "empty object treated as null"));
return object; return object;
} }
auto token = readToken(m->file); auto token = readToken(*m->file);
if (object.isDictionary() && token.isWord("stream")) { if (object.isDictionary() && token.isWord("stream")) {
readStream(object, og, offset); readStream(object, og, offset);
token = readToken(m->file); token = readToken(*m->file);
} }
if (!token.isWord("endobj")) { if (!token.isWord("endobj")) {
QTC::TC("qpdf", "QPDF err expected endobj"); QTC::TC("qpdf", "QPDF err expected endobj");
@ -1535,7 +1535,7 @@ QPDF::readStream(QPDFObjectHandle& object, QPDFObjGen og, qpdf_offset_t offset)
// Seek in two steps to avoid potential integer overflow // Seek in two steps to avoid potential integer overflow
m->file->seek(stream_offset, SEEK_SET); m->file->seek(stream_offset, SEEK_SET);
m->file->seek(toO(length), SEEK_CUR); m->file->seek(toO(length), SEEK_CUR);
if (!readToken(m->file).isWord("endstream")) { if (!readToken(*m->file).isWord("endstream")) {
QTC::TC("qpdf", "QPDF missing endstream"); QTC::TC("qpdf", "QPDF missing endstream");
throw damagedPDF("expected endstream"); throw damagedPDF("expected endstream");
} }
@ -1622,7 +1622,7 @@ bool
QPDF::findEndstream() QPDF::findEndstream()
{ {
// Find endstream or endobj. Position the input at that token. // Find endstream or endobj. Position the input at that token.
auto t = readToken(m->file, 20); auto t = readToken(*m->file, 20);
if (t.isWord("endobj") || t.isWord("endstream")) { if (t.isWord("endobj") || t.isWord("endstream")) {
m->file->seek(m->file->getLastOffset(), SEEK_SET); m->file->seek(m->file->getLastOffset(), SEEK_SET);
return true; return true;
@ -1642,7 +1642,7 @@ QPDF::recoverStreamLength(
if (m->file->findFirst("end", stream_offset, 0, ef)) { if (m->file->findFirst("end", stream_offset, 0, ef)) {
length = toS(m->file->tell() - stream_offset); length = toS(m->file->tell() - stream_offset);
// Reread endstream but, if it was endobj, don't skip that. // Reread endstream but, if it was endobj, don't skip that.
QPDFTokenizer::Token t = readToken(m->file); QPDFTokenizer::Token t = readToken(*m->file);
if (t.getValue() == "endobj") { if (t.getValue() == "endobj") {
m->file->seek(m->file->getLastOffset(), SEEK_SET); m->file->seek(m->file->getLastOffset(), SEEK_SET);
} }
@ -1685,7 +1685,7 @@ QPDF::recoverStreamLength(
} }
QPDFTokenizer::Token QPDFTokenizer::Token
QPDF::readToken(std::shared_ptr<InputSource> input, size_t max_len) QPDF::readToken(InputSource& input, size_t max_len)
{ {
return m->tokenizer.readToken(input, m->last_object_description, true, max_len); return m->tokenizer.readToken(input, m->last_object_description, true, max_len);
} }
@ -1728,20 +1728,20 @@ QPDF::readObjectAtOffset(
m->file->seek(offset, SEEK_SET); m->file->seek(offset, SEEK_SET);
try { try {
QPDFTokenizer::Token tobjid = readToken(m->file); QPDFTokenizer::Token tobjid = readToken(*m->file);
bool objidok = tobjid.isInteger(); bool objidok = tobjid.isInteger();
QTC::TC("qpdf", "QPDF check objid", objidok ? 1 : 0); QTC::TC("qpdf", "QPDF check objid", objidok ? 1 : 0);
if (!objidok) { if (!objidok) {
QTC::TC("qpdf", "QPDF expected n n obj"); QTC::TC("qpdf", "QPDF expected n n obj");
throw damagedPDF(offset, "expected n n obj"); throw damagedPDF(offset, "expected n n obj");
} }
QPDFTokenizer::Token tgen = readToken(m->file); QPDFTokenizer::Token tgen = readToken(*m->file);
bool genok = tgen.isInteger(); bool genok = tgen.isInteger();
QTC::TC("qpdf", "QPDF check generation", genok ? 1 : 0); QTC::TC("qpdf", "QPDF check generation", genok ? 1 : 0);
if (!genok) { if (!genok) {
throw damagedPDF(offset, "expected n n obj"); throw damagedPDF(offset, "expected n n obj");
} }
QPDFTokenizer::Token tobj = readToken(m->file); QPDFTokenizer::Token tobj = readToken(*m->file);
bool objok = tobj.isWord("obj"); bool objok = tobj.isWord("obj");
QTC::TC("qpdf", "QPDF check obj", objok ? 1 : 0); QTC::TC("qpdf", "QPDF check obj", objok ? 1 : 0);
@ -1954,8 +1954,8 @@ QPDF::resolveObjectsInStream(int obj_stream_number)
bp.get())); bp.get()));
for (int i = 0; i < n; ++i) { for (int i = 0; i < n; ++i) {
QPDFTokenizer::Token tnum = readToken(input); QPDFTokenizer::Token tnum = readToken(*input);
QPDFTokenizer::Token toffset = readToken(input); QPDFTokenizer::Token toffset = readToken(*input);
if (!(tnum.isInteger() && toffset.isInteger())) { if (!(tnum.isInteger() && toffset.isInteger())) {
throw damagedPDF( throw damagedPDF(
*input, *input,

7
libqpdf/QPDF_linearization.cc
View File

@ -118,9 +118,10 @@ QPDF::isLinearized()
++p; ++p;
} }
QPDFTokenizer::Token t1 = readToken(m->file); QPDFTokenizer::Token t1 = readToken(*m->file);
if (t1.isInteger() && readToken(m->file).isInteger() && readToken(m->file).isWord("obj") && if (t1.isInteger() && readToken(*m->file).isInteger() &&
(readToken(m->file).getType() == QPDFTokenizer::tt_dict_open)) { readToken(*m->file).isWord("obj") &&
readToken(*m->file).getType() == QPDFTokenizer::tt_dict_open) {
lindict_obj = toI(QUtil::string_to_ll(t1.getValue().c_str())); lindict_obj = toI(QUtil::string_to_ll(t1.getValue().c_str()));
} }
} }