mirror of
https://github.com/qpdf/qpdf.git
synced 2025-01-22 22:58:33 +00:00
Remove redundant code in QPDFParser::parse and parseRemainder
Also, fix test cases.
This commit is contained in:
m-holger
parent
5a1bf035f9
commit
172cc61305
@ -21,7 +21,6 @@
|
||||
|
||||
#include <memory>
|
||||
|
||||
|
||||
QPDFObjectHandle
|
||||
QPDFParser::parse(bool& empty, bool content_stream)
|
||||
{
|
||||
@ -30,327 +29,110 @@ QPDFParser::parse(bool& empty, bool content_stream)
|
||||
// effect of reading the object and changing the file pointer. If you do this, it will cause a
|
||||
// logic error to be thrown from QPDF::inParse().
|
||||
|
||||
const static std::shared_ptr<QPDFObject> null_oh = QPDF_Null::create();
|
||||
QPDF::ParseGuard pg(context);
|
||||
|
||||
empty = false;
|
||||
|
||||
std::shared_ptr<QPDFObject> object;
|
||||
bool set_offset = false;
|
||||
stack.clear();
|
||||
stack.emplace_back(input, st_top);
|
||||
frame = &stack.back();
|
||||
object = nullptr;
|
||||
|
||||
// std::vector<StackFrame> stack{{input, st_top}};
|
||||
stack.clear(); // NEW
|
||||
stack.emplace_back(input, st_top); // NEW
|
||||
bool done = false;
|
||||
bool b_contents = false;
|
||||
bool is_null = false;
|
||||
frame = &stack.back(); // CHANGED
|
||||
if (!tokenizer.nextToken(*input, object_description)) {
|
||||
warn(tokenizer.getErrorMessage());
|
||||
}
|
||||
|
||||
while (!done) {
|
||||
bool indirect_ref = false;
|
||||
is_null = false;
|
||||
object = nullptr;
|
||||
set_offset = false;
|
||||
|
||||
if (!tokenizer.nextToken(*input, object_description)) {
|
||||
warn(tokenizer.getErrorMessage());
|
||||
switch (tokenizer.getType()) {
|
||||
case QPDFTokenizer::tt_eof:
|
||||
if (content_stream) {
|
||||
// In content stream mode, leave object uninitialized to indicate EOF
|
||||
return {};
|
||||
}
|
||||
++good_count; // optimistically
|
||||
QTC::TC("qpdf", "QPDFParser eof in parse");
|
||||
warn("unexpected EOF");
|
||||
return {QPDF_Null::create()};
|
||||
|
||||
switch (tokenizer.getType()) {
|
||||
case QPDFTokenizer::tt_eof:
|
||||
if (stack.size() > 1) {
|
||||
warn("parse error while reading object");
|
||||
}
|
||||
case QPDFTokenizer::tt_bad:
|
||||
QTC::TC("qpdf", "QPDFParser bad token in parse");
|
||||
return {QPDF_Null::create()};
|
||||
|
||||
case QPDFTokenizer::tt_brace_open:
|
||||
case QPDFTokenizer::tt_brace_close:
|
||||
QTC::TC("qpdf", "QPDFParser bad brace");
|
||||
warn("treating unexpected brace token as null");
|
||||
return {QPDF_Null::create()};
|
||||
|
||||
case QPDFTokenizer::tt_array_close:
|
||||
QTC::TC("qpdf", "QPDFParser bad array close");
|
||||
warn("treating unexpected array close token as null");
|
||||
return {QPDF_Null::create()};
|
||||
|
||||
case QPDFTokenizer::tt_dict_close:
|
||||
QTC::TC("qpdf", "QPDFParser bad dictionary close");
|
||||
warn("unexpected dictionary close token");
|
||||
return {QPDF_Null::create()};
|
||||
|
||||
case QPDFTokenizer::tt_array_open:
|
||||
case QPDFTokenizer::tt_dict_open:
|
||||
stack.emplace_back(
|
||||
input,
|
||||
(tokenizer.getType() == QPDFTokenizer::tt_array_open) ? st_array : st_dictionary);
|
||||
return parseRemainder(content_stream);
|
||||
|
||||
case QPDFTokenizer::tt_bool:
|
||||
object = QPDF_Bool::create((tokenizer.getValue() == "true"));
|
||||
break;
|
||||
|
||||
case QPDFTokenizer::tt_null:
|
||||
return {QPDF_Null::create()};
|
||||
|
||||
case QPDFTokenizer::tt_integer:
|
||||
object = QPDF_Integer::create(QUtil::string_to_ll(tokenizer.getValue().c_str()));
|
||||
break;
|
||||
|
||||
case QPDFTokenizer::tt_real:
|
||||
object = QPDF_Real::create(tokenizer.getValue());
|
||||
break;
|
||||
|
||||
case QPDFTokenizer::tt_name:
|
||||
object = QPDF_Name::create(tokenizer.getValue());
|
||||
break;
|
||||
|
||||
case QPDFTokenizer::tt_word:
|
||||
{
|
||||
auto const& value = tokenizer.getValue();
|
||||
if (content_stream) {
|
||||
// In content stream mode, leave object uninitialized to indicate EOF
|
||||
return {};
|
||||
}
|
||||
// QTC::TC("qpdf", "QPDFParser eof in parse");
|
||||
warn("unexpected EOF");
|
||||
return {QPDF_Null::create()};
|
||||
|
||||
case QPDFTokenizer::tt_bad:
|
||||
// QTC::TC("qpdf", "QPDFParser bad token in parse");
|
||||
if (tooManyBadTokens()) {
|
||||
return {QPDF_Null::create()};
|
||||
}
|
||||
is_null = true;
|
||||
break;
|
||||
|
||||
case QPDFTokenizer::tt_brace_open:
|
||||
case QPDFTokenizer::tt_brace_close:
|
||||
// QTC::TC("qpdf", "QPDFParser bad brace");
|
||||
warn("treating unexpected brace token as null");
|
||||
if (tooManyBadTokens()) {
|
||||
return {QPDF_Null::create()};
|
||||
}
|
||||
is_null = true;
|
||||
break;
|
||||
|
||||
case QPDFTokenizer::tt_array_close:
|
||||
if (frame->state == st_array) {
|
||||
if (stack.size() < 2) {
|
||||
throw std::logic_error("QPDFParser::parseInternal: st_stop encountered with "
|
||||
"insufficient elements in stack");
|
||||
}
|
||||
object = QPDF_Array::create(std::move(frame->olist), frame->null_count > 100);
|
||||
setDescription(object, frame->offset - 1);
|
||||
// The `offset` points to the next of "[". Set the rewind offset to point to the
|
||||
// beginning of "[". This has been explicitly tested with whitespace surrounding the
|
||||
// array start delimiter. getLastOffset points to the array end token and therefore
|
||||
// can't be used here.
|
||||
set_offset = true;
|
||||
stack.pop_back();
|
||||
frame = &stack.back();
|
||||
} else {
|
||||
// QTC::TC("qpdf", "QPDFParser bad array close");
|
||||
warn("treating unexpected array close token as null");
|
||||
if (tooManyBadTokens()) {
|
||||
return {QPDF_Null::create()};
|
||||
}
|
||||
is_null = true;
|
||||
}
|
||||
break;
|
||||
|
||||
case QPDFTokenizer::tt_dict_close:
|
||||
if (frame->state == st_dictionary) {
|
||||
if (stack.size() < 2) {
|
||||
throw std::logic_error("QPDFParser::parseInternal: st_stop encountered with "
|
||||
"insufficient elements in stack");
|
||||
}
|
||||
|
||||
// Convert list to map. Alternating elements are keys. Attempt to recover more or
|
||||
// less gracefully from invalid dictionaries.
|
||||
std::set<std::string> names;
|
||||
for (auto& obj: frame->olist) {
|
||||
if (obj) {
|
||||
if (obj->getTypeCode() == ::ot_name) {
|
||||
names.insert(obj->getStringValue());
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
std::map<std::string, QPDFObjectHandle> dict;
|
||||
int next_fake_key = 1;
|
||||
for (auto iter = frame->olist.begin(); iter != frame->olist.end();) {
|
||||
// Calculate key.
|
||||
std::string key;
|
||||
if (*iter && (*iter)->getTypeCode() == ::ot_name) {
|
||||
key = (*iter)->getStringValue();
|
||||
++iter;
|
||||
} else {
|
||||
for (bool found_fake = false; !found_fake;) {
|
||||
key = "/QPDFFake" + std::to_string(next_fake_key++);
|
||||
found_fake = (names.count(key) == 0);
|
||||
// QTC::TC("qpdf", "QPDFParser found fake", (found_fake ? 0 : 1));
|
||||
}
|
||||
warn(
|
||||
frame->offset,
|
||||
"expected dictionary key but found non-name object; inserting key " +
|
||||
key);
|
||||
}
|
||||
if (dict.count(key) > 0) {
|
||||
// QTC::TC("qpdf", "QPDFParser duplicate dict key");
|
||||
warn(
|
||||
frame->offset,
|
||||
"dictionary has duplicated key " + key +
|
||||
"; last occurrence overrides earlier ones");
|
||||
}
|
||||
|
||||
// Calculate value.
|
||||
std::shared_ptr<QPDFObject> val;
|
||||
if (iter != frame->olist.end()) {
|
||||
val = *iter;
|
||||
++iter;
|
||||
} else {
|
||||
// QTC::TC("qpdf", "QPDFParser no val for last key");
|
||||
warn(
|
||||
frame->offset,
|
||||
"dictionary ended prematurely; using null as value for last key");
|
||||
val = QPDF_Null::create();
|
||||
}
|
||||
|
||||
dict[std::move(key)] = std::move(val);
|
||||
}
|
||||
if (!frame->contents_string.empty() && dict.count("/Type") &&
|
||||
dict["/Type"].isNameAndEquals("/Sig") && dict.count("/ByteRange") &&
|
||||
dict.count("/Contents") && dict["/Contents"].isString()) {
|
||||
dict["/Contents"] = QPDFObjectHandle::newString(frame->contents_string);
|
||||
dict["/Contents"].setParsedOffset(frame->contents_offset);
|
||||
}
|
||||
object = QPDF_Dictionary::create(std::move(dict));
|
||||
setDescription(object, frame->offset - 2);
|
||||
// The `offset` points to the next of "<<". Set the rewind offset to point to the
|
||||
// beginning of "<<". This has been explicitly tested with whitespace surrounding
|
||||
// the dictionary start delimiter. getLastOffset points to the dictionary end token
|
||||
// and therefore can't be used here.
|
||||
set_offset = true;
|
||||
stack.pop_back();
|
||||
frame = &stack.back();
|
||||
} else {
|
||||
// QTC::TC("qpdf", "QPDFParser bad dictionary close");
|
||||
warn("unexpected dictionary close token");
|
||||
if (tooManyBadTokens()) {
|
||||
return {QPDF_Null::create()};
|
||||
}
|
||||
is_null = true;
|
||||
}
|
||||
break;
|
||||
|
||||
case QPDFTokenizer::tt_array_open:
|
||||
case QPDFTokenizer::tt_dict_open:
|
||||
if (stack.size() > 500) {
|
||||
// QTC::TC("qpdf", "QPDFParser too deep");
|
||||
warn("ignoring excessively deeply nested data structure");
|
||||
object = QPDF_Operator::create(value);
|
||||
} else if (value == "endobj") {
|
||||
// We just saw endobj without having read anything. Treat this as a null and do
|
||||
// not move the input source's offset.
|
||||
input->seek(input->getLastOffset(), SEEK_SET);
|
||||
empty = true;
|
||||
return {QPDF_Null::create()};
|
||||
} else {
|
||||
b_contents = false;
|
||||
stack.emplace_back(
|
||||
input,
|
||||
(tokenizer.getType() == QPDFTokenizer::tt_array_open) ? st_array
|
||||
: st_dictionary);
|
||||
frame = &stack.back();
|
||||
return parseRemainder(content_stream); // NEW
|
||||
continue;
|
||||
QTC::TC("qpdf", "QPDFParser treat word as string");
|
||||
warn("unknown token while reading object; treating as string");
|
||||
object = QPDF_String::create(value);
|
||||
}
|
||||
|
||||
case QPDFTokenizer::tt_bool:
|
||||
object = QPDF_Bool::create((tokenizer.getValue() == "true"));
|
||||
break;
|
||||
|
||||
case QPDFTokenizer::tt_null:
|
||||
is_null = true;
|
||||
++frame->null_count;
|
||||
|
||||
break;
|
||||
|
||||
case QPDFTokenizer::tt_integer:
|
||||
object = QPDF_Integer::create(QUtil::string_to_ll(tokenizer.getValue().c_str()));
|
||||
break;
|
||||
|
||||
case QPDFTokenizer::tt_real:
|
||||
object = QPDF_Real::create(tokenizer.getValue());
|
||||
break;
|
||||
|
||||
case QPDFTokenizer::tt_name:
|
||||
{
|
||||
auto const& name = tokenizer.getValue();
|
||||
object = QPDF_Name::create(name);
|
||||
|
||||
if (name == "/Contents") {
|
||||
b_contents = true;
|
||||
} else {
|
||||
b_contents = false;
|
||||
}
|
||||
}
|
||||
break;
|
||||
|
||||
case QPDFTokenizer::tt_word:
|
||||
{
|
||||
auto const& value = tokenizer.getValue();
|
||||
auto size = frame->olist.size();
|
||||
if (content_stream) {
|
||||
object = QPDF_Operator::create(value);
|
||||
} else if (
|
||||
value == "R" && frame->state != st_top && size >= 2 && frame->olist.back() &&
|
||||
frame->olist.back()->getTypeCode() == ::ot_integer &&
|
||||
!frame->olist.back()->getObjGen().isIndirect() && frame->olist.at(size - 2) &&
|
||||
frame->olist.at(size - 2)->getTypeCode() == ::ot_integer &&
|
||||
!frame->olist.at(size - 2)->getObjGen().isIndirect()) {
|
||||
if (context == nullptr) {
|
||||
// QTC::TC("qpdf", "QPDFParser indirect without context");
|
||||
throw std::logic_error("QPDFObjectHandle::parse called without context on "
|
||||
"an object with indirect references");
|
||||
}
|
||||
auto ref_og = QPDFObjGen(
|
||||
QPDFObjectHandle(frame->olist.at(size - 2)).getIntValueAsInt(),
|
||||
QPDFObjectHandle(frame->olist.back()).getIntValueAsInt());
|
||||
if (ref_og.isIndirect()) {
|
||||
// This action has the desirable side effect of causing dangling references
|
||||
// (references to indirect objects that don't appear in the PDF) in any
|
||||
// parsed object to appear in the object cache.
|
||||
object = context->getObject(ref_og).obj;
|
||||
indirect_ref = true;
|
||||
} else {
|
||||
// QTC::TC("qpdf", "QPDFParser indirect with 0 objid");
|
||||
is_null = true;
|
||||
}
|
||||
frame->olist.pop_back();
|
||||
frame->olist.pop_back();
|
||||
} else if ((value == "endobj") && (frame->state == st_top)) {
|
||||
// We just saw endobj without having read anything. Treat this as a null and do
|
||||
// not move the input source's offset.
|
||||
is_null = true;
|
||||
input->seek(input->getLastOffset(), SEEK_SET);
|
||||
empty = true;
|
||||
} else {
|
||||
// QTC::TC("qpdf", "QPDFParser treat word as string");
|
||||
warn("unknown token while reading object; treating as string");
|
||||
if (tooManyBadTokens()) {
|
||||
return {QPDF_Null::create()};
|
||||
}
|
||||
object = QPDF_String::create(value);
|
||||
}
|
||||
}
|
||||
break;
|
||||
|
||||
case QPDFTokenizer::tt_string:
|
||||
{
|
||||
auto const& val = tokenizer.getValue();
|
||||
if (decrypter) {
|
||||
if (b_contents) {
|
||||
frame->contents_string = val;
|
||||
frame->contents_offset = input->getLastOffset();
|
||||
b_contents = false;
|
||||
}
|
||||
std::string s{val};
|
||||
decrypter->decryptString(s);
|
||||
object = QPDF_String::create(s);
|
||||
} else {
|
||||
object = QPDF_String::create(val);
|
||||
}
|
||||
}
|
||||
break;
|
||||
|
||||
default:
|
||||
warn("treating unknown token type as null while reading object");
|
||||
if (tooManyBadTokens()) {
|
||||
return {QPDF_Null::create()};
|
||||
}
|
||||
is_null = true;
|
||||
break;
|
||||
}
|
||||
break;
|
||||
|
||||
if (object == nullptr && !is_null) {
|
||||
throw std::logic_error("QPDFParser:parseInternal: unexpected uninitialized object");
|
||||
case QPDFTokenizer::tt_string:
|
||||
if (decrypter) {
|
||||
std::string s{tokenizer.getValue()};
|
||||
decrypter->decryptString(s);
|
||||
object = QPDF_String::create(s);
|
||||
} else {
|
||||
object = QPDF_String::create(tokenizer.getValue());
|
||||
}
|
||||
break;
|
||||
|
||||
switch (frame->state) {
|
||||
case st_dictionary:
|
||||
case st_array:
|
||||
if (is_null) {
|
||||
object = null_oh;
|
||||
// No need to set description for direct nulls - they probably will become implicit.
|
||||
} else if (!indirect_ref && !set_offset) {
|
||||
setDescription(object, input->getLastOffset());
|
||||
}
|
||||
set_offset = true;
|
||||
frame->olist.push_back(object);
|
||||
break;
|
||||
|
||||
case st_top:
|
||||
done = true;
|
||||
break;
|
||||
}
|
||||
default:
|
||||
warn("treating unknown token type as null while reading object");
|
||||
return {QPDF_Null::create()};
|
||||
}
|
||||
|
||||
if (is_null) {
|
||||
object = QPDF_Null::create();
|
||||
}
|
||||
if (!set_offset) {
|
||||
setDescription(object, frame->offset);
|
||||
}
|
||||
setDescription(object, frame->offset);
|
||||
return object;
|
||||
}
|
||||
|
||||
@ -363,18 +145,15 @@ QPDFParser::parseRemainder(bool content_stream)
|
||||
// logic error to be thrown from QPDF::inParse().
|
||||
|
||||
const static std::shared_ptr<QPDFObject> null_oh = QPDF_Null::create();
|
||||
// QPDF::ParseGuard pg(context);
|
||||
|
||||
// empty = false;
|
||||
|
||||
std::shared_ptr<QPDFObject> object;
|
||||
bool set_offset = false;
|
||||
|
||||
// std::vector<StackFrame> stack{{input, st_top},};
|
||||
bool done = false;
|
||||
bool b_contents = false;
|
||||
bool is_null = false;
|
||||
frame = &stack.back(); // CHANGED
|
||||
bad_count = 0;
|
||||
|
||||
while (!done) {
|
||||
bool indirect_ref = false;
|
||||
@ -389,19 +168,17 @@ QPDFParser::parseRemainder(bool content_stream)
|
||||
|
||||
switch (tokenizer.getType()) {
|
||||
case QPDFTokenizer::tt_eof:
|
||||
if (stack.size() > 1) {
|
||||
warn("parse error while reading object");
|
||||
}
|
||||
warn("parse error while reading object");
|
||||
if (content_stream) {
|
||||
// In content stream mode, leave object uninitialized to indicate EOF
|
||||
return {};
|
||||
}
|
||||
QTC::TC("qpdf", "QPDFParser eof in parse");
|
||||
QTC::TC("qpdf", "QPDFParser eof in parseRemainder");
|
||||
warn("unexpected EOF");
|
||||
return {QPDF_Null::create()};
|
||||
|
||||
case QPDFTokenizer::tt_bad:
|
||||
QTC::TC("qpdf", "QPDFParser bad token in parse");
|
||||
QTC::TC("qpdf", "QPDFParser bad token in parseRemainder");
|
||||
if (tooManyBadTokens()) {
|
||||
return {QPDF_Null::create()};
|
||||
}
|
||||
@ -410,7 +187,7 @@ QPDFParser::parseRemainder(bool content_stream)
|
||||
|
||||
case QPDFTokenizer::tt_brace_open:
|
||||
case QPDFTokenizer::tt_brace_close:
|
||||
QTC::TC("qpdf", "QPDFParser bad brace");
|
||||
QTC::TC("qpdf", "QPDFParser bad brace in parseRemainder");
|
||||
warn("treating unexpected brace token as null");
|
||||
if (tooManyBadTokens()) {
|
||||
return {QPDF_Null::create()};
|
||||
@ -434,7 +211,7 @@ QPDFParser::parseRemainder(bool content_stream)
|
||||
stack.pop_back();
|
||||
frame = &stack.back();
|
||||
} else {
|
||||
QTC::TC("qpdf", "QPDFParser bad array close");
|
||||
QTC::TC("qpdf", "QPDFParser bad array close in parseRemainder");
|
||||
warn("treating unexpected array close token as null");
|
||||
if (tooManyBadTokens()) {
|
||||
return {QPDF_Null::create()};
|
||||
@ -519,7 +296,7 @@ QPDFParser::parseRemainder(bool content_stream)
|
||||
stack.pop_back();
|
||||
frame = &stack.back();
|
||||
} else {
|
||||
QTC::TC("qpdf", "QPDFParser bad dictionary close");
|
||||
QTC::TC("qpdf", "QPDFParser bad dictionary close in parseRemainder");
|
||||
warn("unexpected dictionary close token");
|
||||
if (tooManyBadTokens()) {
|
||||
return {QPDF_Null::create()};
|
||||
@ -582,7 +359,7 @@ QPDFParser::parseRemainder(bool content_stream)
|
||||
if (content_stream) {
|
||||
object = QPDF_Operator::create(value);
|
||||
} else if (
|
||||
value == "R" && frame->state != st_top && size >= 2 && frame->olist.back() &&
|
||||
value == "R" && size >= 2 && frame->olist.back() &&
|
||||
frame->olist.back()->getTypeCode() == ::ot_integer &&
|
||||
!frame->olist.back()->getObjGen().isIndirect() && frame->olist.at(size - 2) &&
|
||||
frame->olist.at(size - 2)->getTypeCode() == ::ot_integer &&
|
||||
@ -607,14 +384,8 @@ QPDFParser::parseRemainder(bool content_stream)
|
||||
}
|
||||
frame->olist.pop_back();
|
||||
frame->olist.pop_back();
|
||||
} else if ((value == "endobj") && (frame->state == st_top)) {
|
||||
// We just saw endobj without having read anything. Treat this as a null and do
|
||||
// not move the input source's offset.
|
||||
is_null = true;
|
||||
input->seek(input->getLastOffset(), SEEK_SET);
|
||||
// empty = true;
|
||||
} else {
|
||||
QTC::TC("qpdf", "QPDFParser treat word as string");
|
||||
QTC::TC("qpdf", "QPDFParser treat word as string in parseRemainder");
|
||||
warn("unknown token while reading object; treating as string");
|
||||
if (tooManyBadTokens()) {
|
||||
return {QPDF_Null::create()};
|
||||
|
||||
@ -57,11 +57,14 @@ QPDF trailer lacks size 0
|
||||
QPDF trailer size not integer 0
|
||||
QPDF trailer prev not integer 0
|
||||
QPDFParser bad brace 0
|
||||
QPDFParser bad brace in parseRemainder 0
|
||||
QPDFParser bad array close 0
|
||||
QPDFParser bad array close in parseRemainder 0
|
||||
QPDF stream without length 0
|
||||
QPDF stream length not integer 0
|
||||
QPDF missing endstream 0
|
||||
QPDFParser bad dictionary close 0
|
||||
QPDFParser bad dictionary close in parseRemainder 0
|
||||
QPDF can't find xref 0
|
||||
QPDFTokenizer bad ) 0
|
||||
QPDFTokenizer bad > 0
|
||||
@ -258,6 +261,7 @@ QPDFParser indirect with 0 objid 0
|
||||
QPDF object id 0 0
|
||||
QPDF recursion loop in resolve 0
|
||||
QPDFParser treat word as string 0
|
||||
QPDFParser treat word as string in parseRemainder 0
|
||||
QPDFParser found fake 1
|
||||
QPDFParser no val for last key 0
|
||||
QPDF resolve failure to null 0
|
||||
@ -289,7 +293,9 @@ QPDFObjectHandle coalesce called on stream 0
|
||||
QPDFObjectHandle coalesce provide stream data 0
|
||||
QPDF_Stream bad token at end during normalize 0
|
||||
QPDFParser bad token in parse 0
|
||||
QPDFParser bad token in parseRemainder 0
|
||||
QPDFParser eof in parse 0
|
||||
QPDFParser eof in parseRemainder 0
|
||||
QPDFObjectHandle array bounds 0
|
||||
QPDFObjectHandle boolean returning false 0
|
||||
QPDFObjectHandle integer returning 0 0
|
||||
|
||||
@ -17,7 +17,7 @@ my $td = new TestDriver('parsing');
|
||||
my $n_tests = 17;
|
||||
|
||||
$td->runtest("parse objects from string",
|
||||
{$td->COMMAND => "test_driver 31 good1.qdf"},
|
||||
{$td->COMMAND => "test_driver 31 bad39.qdf"},
|
||||
{$td->FILE => "parse-object.out", $td->EXIT_STATUS => 0},
|
||||
$td->NORMALIZE_NEWLINES);
|
||||
$td->runtest("EOF terminating literal tokens",
|
||||
|
||||
102
qpdf/qtest/qpdf/bad39.qdf
Normal file
102
qpdf/qtest/qpdf/bad39.qdf
Normal file
@ -0,0 +1,102 @@
|
||||
%PDF-1.3
|
||||
%¿÷¢þ
|
||||
%QDF-1.0
|
||||
|
||||
%% Original object ID: 1 0
|
||||
1 0 obj
|
||||
<<
|
||||
/Pages 2 0 R
|
||||
/Type /Catalog
|
||||
>>
|
||||
endobj
|
||||
|
||||
%% Original object ID: 2 0
|
||||
2 0 obj
|
||||
<<
|
||||
/Count 1
|
||||
/Kids [
|
||||
3 0 R
|
||||
]
|
||||
/Type /Pages
|
||||
>>
|
||||
endobj
|
||||
|
||||
%% Page 1
|
||||
%% Original object ID: 3 0
|
||||
3 0 obj
|
||||
<<
|
||||
/Contents 4 0 R
|
||||
/MediaBox [
|
||||
0
|
||||
0
|
||||
612
|
||||
792
|
||||
]
|
||||
/Parent 2 0 R
|
||||
/Resources <<
|
||||
/Font <<
|
||||
/F1 6 0 R
|
||||
>>
|
||||
/ProcSet 7 0 R
|
||||
>>
|
||||
/Type /Page
|
||||
>>
|
||||
endobj
|
||||
|
||||
%% Contents for page 1
|
||||
%% Original object ID: 4 0
|
||||
4 0 obj
|
||||
<<
|
||||
/Length 5 0 R
|
||||
>>
|
||||
stream
|
||||
BT
|
||||
/F1 24 Tf
|
||||
72 720 Td
|
||||
(Potato) Tj
|
||||
ET
|
||||
endstream
|
||||
endobj
|
||||
|
||||
5 0 obj
|
||||
44
|
||||
endobj
|
||||
|
||||
%% Original object ID: 6 0
|
||||
6 0 obj
|
||||
<<
|
||||
/BaseFont /Helvetica
|
||||
/Encoding /WinAnsiEncoding
|
||||
/Name /F1
|
||||
/Subtype /Type1
|
||||
/Type /Font
|
||||
>>
|
||||
endobj
|
||||
|
||||
%% Original object ID: 5 0
|
||||
7 0 obj
|
||||
[
|
||||
/PDF
|
||||
/Text
|
||||
]
|
||||
endobj
|
||||
|
||||
xref
|
||||
0 8
|
||||
0000000000 65535 f
|
||||
0000000052 00000 n
|
||||
0000000133 00000 n
|
||||
0000000242 00000 n
|
||||
0000000484 00000 n
|
||||
0000000583 00000 n
|
||||
0000000629 00000 n
|
||||
0000001113 00000 n
|
||||
trailer <<
|
||||
/Root 1 0 R
|
||||
/Size 8
|
||||
/ID [<31415926535897932384626433832795><31415926535897932384626433832795>]
|
||||
>>
|
||||
startxref
|
||||
809
|
||||
%%EOF
|
||||
7 0 obj
|
||||
@ -2,4 +2,10 @@
|
||||
logic error parsing indirect: QPDFObjectHandle::parse called without context on an object with indirect references
|
||||
trailing data: parsed object (trailing test): trailing data found parsing object from string
|
||||
WARNING: parsed object (offset 9): unknown token while reading object; treating as string
|
||||
WARNING: parsed object: treating unexpected brace token as null
|
||||
WARNING: parsed object: treating unexpected brace token as null
|
||||
WARNING: parsed object: unexpected dictionary close token
|
||||
WARNING: bad39.qdf (object 7 0, offset 1121): unexpected EOF
|
||||
WARNING: bad39.qdf (object 7 0, offset 1121): expected endobj
|
||||
WARNING: bad39.qdf (object 7 0, offset 1121): EOF after endobj
|
||||
test 31 done
|
||||
|
||||
@ -1195,6 +1195,13 @@ test_31(QPDF& pdf, char const* arg2)
|
||||
// mistakenly parsed as an indirect object.
|
||||
assert(QPDFObjectHandle::parse(&pdf, "[5 0 R 0 R /X]").unparse() == "[ 5 0 R 0 (R) /X ]");
|
||||
assert(QPDFObjectHandle::parse(&pdf, "[1 0 R]", "indirect test").unparse() == "[ 1 0 R ]");
|
||||
// TC:QPDFParser bad brace
|
||||
assert(QPDFObjectHandle::parse(&pdf, "}").unparse() == "null");
|
||||
assert(QPDFObjectHandle::parse(&pdf, "{").unparse() == "null");
|
||||
// TC:QPDFParser bad dictionary close
|
||||
assert(QPDFObjectHandle::parse(&pdf, ">>").unparse() == "null");
|
||||
// TC:QPDFParser eof in parse
|
||||
assert(QPDFObjectHandle::parse(&pdf, "[7 0 R]").getArrayItem(0).isNull());
|
||||
}
|
||||
|
||||
static void
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user