OpenSystemsDevelopment / qpdf

* Make it a runtime error, not a logic error
* Include additional information
* Capture it properly in checkLinearization

This also reverts the addition of a new checkLinearization that
distinguishes errors from warnings. There's no practical distinction
between what was considered an error and what was considered a
warning.

* Several assertions in linearization were not always true; change
  them to run time errors
* Handle a few cases of uninitialized objects
* Handle pages with no contents when doing form operations
* Handle invalid page tree nodes when traversing pages

This makes all integer type conversions that have potential data loss
explicit with calls that do range checks and raise an exception. After
this commit, qpdf builds with no warnings when -Wsign-conversion
-Wconversion is used with gcc or clang or when -W3 -Wd4800 is used
with MSVC. This significantly reduces the likelihood of potential
crashes from bogus integer values.

There are some parts of the code that take int when they should take
size_t or an offset. Such places would make qpdf not support files
with more than 2^31 of something that usually wouldn't be so large. In
the event that such a file shows up and is valid, at least qpdf would
raise an error in the right spot so the issue could be legitimately
addressed rather than failing in some weird way because of a silent
overflow condition.

Instead of calling assert for problems found during checking
linearization data, throw an exception which is later caught and
issued as an error. Ideally we would handle errors more robustly, but
this is still a significant improvement.

Give objects descriptions and context so it is possible to issue
warnings instead of fatal errors for attempts to access objects of the
wrong type.

Pushing member variables into a nested class enables addition of new
member variables without breaking binary compatibility.

This commit adds several API methods that enable control over which
types of filters QPDF will attempt to decode. It also adds support for
/RunLengthDecode and /DCTDecode filters for both encoding and
decoding.

Preparing to refactor some pattern searching code to use these instead
of their own memchr loops. This should simplify the code that replaces
PCRE.

For std::string and std::vector, replace operator[] with at.  This was
done using an automated process.  See README.hardening for details.

Ideally, the library should never call assert outside of test code,
but it does in several places.  For some cases where the assertion
might conceivably fail because of a problem with the input data,
replace assertions with exceptions so that they can be trapped by the
calling application.  This commit surely misses some cases and
replaced some cases unnecessarily, but it should still be an
improvement.

In places where std::vector<T>(size_t) was used, either validate that
the size parameter is sane or refactor code to avoid the need to
pre-allocate the vector.

The faulty code was only used during explicit checks of linearization
data.  Those checks are not part of normal reading or writing of PDF
files.

Even though this case is not valid according to the spec, it has been
seen, and caused an internal error.

In internal code and examples, replace calls to getObjectID() and
getGeneration() with calls to getObjGen() where possible.

This makes it possible to use two different writers to write
linearized files from the same QPDF object.

off_t is used internally only when needed to talk to standard
libraries.  This requires that the "long long" type be supported by
the compiler.

Significantly improve the code's use of off_t for file offsets, size_t
for memory sizes, and integer types in cases where there has to be
compatibility with external interfaces.  Rework sections of the code
that would have prevented qpdf from working on files larger than 2 (or
maybe 4) GB in size.

… of delete, thus making it useful to run valgrind over qpdf during its test suite

git-svn-id: svn+q:///qpdf/trunk@1056 71b93d88-0707-0410-a8cf-f5a4172ac649

git-svn-id: svn+q:///qpdf/trunk@1049 71b93d88-0707-0410-a8cf-f5a4172ac649

git-svn-id: svn+q:///qpdf/trunk@1035 71b93d88-0707-0410-a8cf-f5a4172ac649

…uffer; also fix a bug in BufferInputSource::seek

git-svn-id: svn+q:///qpdf/trunk@1030 71b93d88-0707-0410-a8cf-f5a4172ac649

git-svn-id: svn+q:///qpdf/trunk@829 71b93d88-0707-0410-a8cf-f5a4172ac649

git-svn-id: svn+q:///qpdf/trunk@796 71b93d88-0707-0410-a8cf-f5a4172ac649

git-svn-id: svn+q:///qpdf/trunk@709 71b93d88-0707-0410-a8cf-f5a4172ac649

git-svn-id: svn+q:///qpdf/trunk@690 71b93d88-0707-0410-a8cf-f5a4172ac649

git-svn-id: svn+q:///qpdf/trunk@688 71b93d88-0707-0410-a8cf-f5a4172ac649

Fix libtests test suites to pass on Windows, mostly by dealing with
ascii vs. binary and NL vs. CRNL change ($td->NORMALIZE_NEWLINES).
Convert some test suites to use fread instead of read.
PCRE.hh: define PCRE_STATIC if on Windows.
Provide cross-platform function for getting current time instead of
using time(0).


git-svn-id: svn+q:///qpdf/trunk@678 71b93d88-0707-0410-a8cf-f5a4172ac649

git-svn-id: svn+q:///qpdf/trunk@651 71b93d88-0707-0410-a8cf-f5a4172ac649