Age | Commit message (Collapse) | Author | |
---|---|---|---|
2019-06-21 | Remove integer type task from TODO | Jay Berkenbilt | |
2019-06-21 | Update casting policy in the documentation | Jay Berkenbilt | |
2019-06-21 | Enable int warnings by default | Jay Berkenbilt | |
Now that there aren't any more... | |||
2019-06-21 | Remove implicit conversion from int/pointer to bool | Jay Berkenbilt | |
This fixes cases of warning C4800 from msvc | |||
2019-06-21 | Fix sign and conversion warnings (major) | Jay Berkenbilt | |
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. | |||
2019-06-21 | Pl_Flate: constructor's out_bufsize is now unsigned int | Jay Berkenbilt | |
This is the type we need for the underlying zlib implementation. | |||
2019-06-21 | Change QPDFObjectHandle::pipeStreamData's encode_flags type | Jay Berkenbilt | |
Change from unsigned long to int since we pass enumerated type values to this field. | |||
2019-06-21 | Add new integer accessors to QPDFObjectHandle | Jay Berkenbilt | |
2019-06-21 | QUtil: add unsigned int/string functions | Jay Berkenbilt | |
2019-06-21 | New safe type converters in QIntC | Jay Berkenbilt | |
2019-06-21 | Add 32-bit Linux build to CI | Jay Berkenbilt | |
2019-06-21 | configure: determine wordsize automatically | Jay Berkenbilt | |
Based on sizeof(size_t). Assumes 64 if not 32. | |||
2019-06-21 | configure: add --enable-int-warnings | Jay Berkenbilt | |
2019-06-21 | Appearance streams: some floating point values were truncated | Jay Berkenbilt | |
Bounding box X coordinates could be truncated, causing them to be off by a fraction of a point. This was most likely not visible, but it was still wrong. | |||
2019-06-18 | Fix typo in doc for configure | Jay Berkenbilt | |
2019-06-18 | TODO | Jay Berkenbilt | |
2019-06-18 | Remove qpdf_read_memory_fuzzer | Jay Berkenbilt | |
2019-06-15 | Run tests with sanitizers in CI | Jay Berkenbilt | |
2019-06-15 | As a test suite, run stand-alone fuzzer on seed corpus | Jay Berkenbilt | |
Temporarily skip fuzz tests on Windows. There are Windows-specific failures to address later. | |||
2019-06-15 | Populate seed corpus | Jay Berkenbilt | |
2019-06-15 | Add corpus from previous fuzzer | Jay Berkenbilt | |
This is the set of files from the latest corpus of running the older fuzzer (qpdf_read_memory_fuzzer) at the time of adding the new fuzzer (qpdf_fuzzer) after running a merge operation to minimize the corpus. | |||
2019-06-15 | Add test files from oss-fuzz bugs (fixes #335) | Jay Berkenbilt | |
2019-06-15 | Fix problems found in fuzz corpus | Jay Berkenbilt | |
2019-06-15 | Implement fuzzer with good coverage | Jay Berkenbilt | |
2019-06-15 | Give up reading objects with too many consecutive errors | Jay Berkenbilt | |
2019-06-15 | Eliminate one potential integer overflow | Jay Berkenbilt | |
There are more to handle, but this resolves an issue already caught by oss-fuzz. | |||
2019-06-14 | oss-fuzz-build: support branch switching | Jay Berkenbilt | |
2019-06-13 | oss-fuzz initial integration | Jay Berkenbilt | |
2019-06-12 | TODO | Jay Berkenbilt | |
2019-06-12 | Fix typo (#334) | Marco Scarpetta | |
2019-06-10 | Commit forgotten test files | Jay Berkenbilt | |
2019-06-09 | Handle empty DecodeParams array for (fixes #331) | Jay Berkenbilt | |
On read, ignore /DecodeParms when empty list; on write, delete it. Some files have been found that include an empty list for /DecodeParms, but this is not technically compliant with the spec, and the only sensible interpretation is to treat it as if there are no decode parameters. | |||
2019-05-18 | Regenerate autofilesrelease-qpdf-8.4.2 | Jay Berkenbilt | |
2019-05-18 | Prepare 8.4.2 release | Jay Berkenbilt | |
2019-05-18 | Provide link to public key file | Jay Berkenbilt | |
2019-05-16 | TODO | Jay Berkenbilt | |
2019-05-16 | Fix Windows memory error (fixes #330) | Jay Berkenbilt | |
2019-04-28 | Prepare 8.4.1 releaserelease-qpdf-8.4.1 | Jay Berkenbilt | |
2019-04-28 | Improve Unicode filename testing | Jay Berkenbilt | |
Remove dependency on the behavior of perl for reliable creation of Unicode file names on Windows. | |||
2019-04-28 | Remove stray comment | Jay Berkenbilt | |
2019-04-22 | Fix typo | Jay Berkenbilt | |
2019-04-21 | Spell check | Jay Berkenbilt | |
2019-04-21 | Compare versions between CLI and library | Jay Berkenbilt | |
2019-04-21 | Add --remove-page-labels option (fixes #317) | Jay Berkenbilt | |
2019-04-21 | Add --keep-files-open-threshold (fixes #288) | Jay Berkenbilt | |
2019-04-21 | Support Unicode in filenames (fixes #298) | Jay Berkenbilt | |
2019-04-21 | Tighten isPageObject (fixes #310) | Jay Berkenbilt | |
2019-04-21 | Document that linearize disables qdf (fixes #312) | Jay Berkenbilt | |
2019-04-21 | Revert preservations of outlines with --split-pages | Jay Berkenbilt | |
The preservation of outlines didn't provide very useful behavior anyway as it copied all outlines but most didn't work. This implementation also caused a very significant performance hit and so is being reverted until a proper solution can be coded. The eventual solution will not be compatible with the reverted solution anyway, so it's best not to leave this in. | |||
2019-04-21 | TODO: ABI note | Jay Berkenbilt | |