Next ==== * Find PDF header in the first 1024 bytes of the file. Treat the location of the PDF header as offset 0 for purposes of resolving explicit file locations as this is what other implementations appear to do. General ======= * See if I can support the encryption format used with /R 5 /V 5 (AESV3), even though a qpdf-announce subscriber with an adobe.com email address mentioned that this is deprecated. There is also a new encryption format coming in a future release (PDF 2.0), which may be better to support. As of the qpdf 3.0 release, the specification was not publicly available yet. AESV3 encryption is supported with PDF 1.7 extension level 3 and is being deprecated, but there are plenty of files out there. The encryption format is decribed in adobe_supplement_iso32000.pdf. Such a file must specify that it uses these extensions in its document catalog: << /Type /Catalog /Extensions << /ADBE << /BaseVersion /1.7 /ExtensionLevel 3 >> >> >> Possible sha256 implementations: http://sol-biotech.com/code/sha2/, http://hashlib2plus.sourceforge.net/ * Consider the possibility of doing something locale-aware to support non-ASCII passwords. Update documentation if this is done. * Look for %PDF header somewhere within the first 1024 bytes of the file. Also accept headers of the form "%!PS−Adobe−N.n PDF−M.m". See Implementation notes 13 and 14 in appendix H of the PDF 1.7 specification. This is bug 3267974. * Consider impact of article threads on page splitting/merging. Subramanyam provided a test file; see ../misc/article-threads.pdf. Email Q-Count: 431864 from 2009-11-03. Other things to consider: outlines, page labels, thumbnails, zones. There are probably others. * See whether it's possible to remove the call to flattenScalarReferences. I can't easily figure out why I do it, but removing it causes strange test failures in linearization. I would have to study the optimization and linearization code to figure out why I added this to begin with and what in the code assumes it's the case. For enqueueObject and unparseChild in QPDFWriter, simply removing the checks for indirect scalars seems sufficient. Looking back at the branch in the apex epub repository, before flattening scalar references, there was special case code in QPDFWriter to avoid writing out indirect nulls. It's still not obvious to me why I did it though. To pursue this, remove the call to flattenScalarReferences in QPDFWriter.cc and disable the logic_error exceptions for indirect scalars. Just search for flattenScalarReferences in QPDFWriter.cc since the logic errors have comments that mention flattenScalarReferences. Then run the test suite. Several files that explicitly test flattening of scalar references fail, but the indirect scalars are properly preserved and written. But then there are some linearized files that have a bunch of unreferenced objects that contain scalars. Need to figure out what these are and why they're there. Maybe they're objects that used to be stream lengths. Probably we just need to make sure don't traverse through a stream's /Length stream when enqueueing stream dictionaries. This could potentially happen with any object that QPDFWriter replaces when writing out files. Such objects would be orphaned in the newly written file. This could be fixed, but it may not be worth fixing. If flattenScalarReferences is removed, a new method will be needed for checking PDF files. * See if we can avoid preserving unreferenced objects in object streams even when preserving the object streams. * For debugging linearization bugs, consider adding an option to save pass 1 of linearization. This code is sufficient. Change the interface to allow specification of a pass1 file, which would change the behavior as in this patch. ------------------------------ Index: QPDFWriter.cc =================================================================== --- QPDFWriter.cc (revision 932) +++ QPDFWriter.cc (working copy) @@ -1965,11 +1965,15 @@ // Write file in two passes. Part numbers refer to PDF spec 1.4. + FILE* XXX = 0; for (int pass = 1; pass <= 2; ++pass) { if (pass == 1) { - pushDiscardFilter(); +// pushDiscardFilter(); + XXX = fopen("/tmp/pass1.pdf", "w"); + pushPipeline(new Pl_StdioFile("pass1", XXX)); + activatePipelineStack(); } // Part 1: header @@ -2204,6 +2208,8 @@ // Restore hint offset this->xref[hint_id] = QPDFXRefEntry(1, hint_offset, 0); + fclose(XXX); + XXX = 0; } } } ------------------------------ * Handle embedded files. PDF Reference 1.7 section 3.10, "File Specifications", discusses this. Once we can definitely recognize all embedded files in a document, we can update the encryption code to handle it properly. In QPDF_encryption.cc, search for cf_file. Remove exception thrown if cf_file is different from cf_stream, and write code in the stream decryption section to use cf_file instead of cf_stream. In general, add interfaces to get the list of embedded files and to extract them. To handle general embedded files associated with the whole document, follow root -> /Names -> /EmbeddedFiles -> /Names to get to the file specification dictionaries. Then, in each file specification dictionary, follow /EF -> /F to the actual stream. There may be other places file specification dictionaries may appear, and there are also /RF keys with related files, so reread section 3.10 carefully. * The description of Crypt filters is unclear with respect to how to use them to override /StmF for specific streams. I'm not sure whether qpdf will do the right thing for any specific individual streams that might have crypt filters. The specification seems to imply that only embedded file streams and metadata streams can have crypt filters, and there are already special cases in the code to handle those. Most likely, it won't be a problem, but someday someone may find a file that qpdf doesn't work on because of crypt filters. There is an example in the spec of using a crypt filter on a metadata stream. For now, we notice /Crypt filters and decode parameters consistent with the example in the PDF specification, and the right thing happens for metadata filters that happen to be uncompressed or otherwise compressed in a way we can filter. This should handle all normal cases, but it's more or less just a guess since I don't have any test files that actually use stream-specific crypt filters in them. * The second xref stream for linearized files has to be padded only because we need file_size as computed in pass 1 to be accurate. If we were not allowing writing to a pipe, we could seek back to the beginning and fill in the value of /L in the linearization dictionary as an optimization to alleviate the need for this padding. Doing so would require us to pad the /L value individually and also to save the file descriptor and determine whether it's seekable. This is probably not worth bothering with. * The whole xref handling code in the QPDF object allows the same object with more than one generation to coexist, but a lot of logic assumes this isn't the case. Anything that creates mappings only with the object number and not the generation is this way, including most of the interaction between QPDFWriter and QPDF. If we wanted to allow the same object with more than one generation to coexist, which I'm not sure is allowed, we could fix this by changing xref_table. Alternatively, we could detect and disallow that case. In fact, it appears that Adobe reader and other PDF viewing software silently ignores objects of this type, so this is probably not a big deal. * Pl_PNGFilter is only partially implemented. If we ever decoded images, we'd have to finish implementing it along with the other filter decode parameters and types. For just handling xref streams, there's really no need as it wouldn't make sense to use any kind of predictor other than 12 (PNG UP filter). * If we ever want to have check mode check the integrity of the free list, this can be done by looking at the code from prior to the object stream support of 4/5/2008. It's in an if (0) block and there's a comment about it. There's also something about it in qpdf.test -- search for "free table". On the other hand, the value of doing this seems very low since no viewer seems to care, so it's probably not worth it. * QPDFObjectHandle::getPageImages() doesn't notice images in inherited resource dictionaries. See comments in that function. * Based on an idea suggested by user "Atom Smasher", consider providing some mechanism to recover earlier versions of a file embedded prior to appended sections. * From a suggestion in bug 3152169, consider having an option to re-encode inline images with an ASCII encoding. * From github issue 2, provide more in-depth output for examining hint stream contents.