summaryrefslogtreecommitdiff
path: root/core/fpdfapi
AgeCommit message (Collapse)Author
2016-11-14Make CPDF_PageContentGenerator methods take object numberstsepez
This patch fixes a possibility that an owned CPDF_Stream is handed to the indirect object holder inside RealizeResource(). Its arguments are changed to take an object number, as is done elsewhere in the code, to suggest that only indirect objects are acceptable. BUG=660756 Review-Url: https://codereview.chromium.org/2489423002
2016-11-14Properly release caches in CPDF_DocRenderDatanpm
Currently, only RemoveRef() is being called when releasing. The problem is that when the object is no longer being used, the map is not updated accordingly. In the pdf in the bug below, the pointer of a font used in the second page coincides with the pointer of a font in the first page (that was released). So the cached values for the font in the first page are used, resulting in incorrect rendering. BUG=pdfium:629 Review-Url: https://codereview.chromium.org/2501053002
2016-11-11Relax colorspace checks in CPDF_DIBSource::CreateDecoder().chromium/2917thestig
Previously the log just made sure the colorspace and the image have exact matches for the number of colorspace components. Now, for some colorspace types, check the type and make sure the number of components meets or exceeds what is required by the spec. Also do some refactoring. BUG=chromium:650230 Review-Url: https://codereview.chromium.org/2486123002
2016-11-11Fix sevg above CPDF_PageContentGenerator::ProcessImage()tsepez
Adds a test to process an image for which a stream hasn't been set. BUG=660756 Review-Url: https://codereview.chromium.org/2494683007
2016-11-09Do not load main cross ref table for first page in linearized pdf.art-snake
For the first page in linearized PDFs, it is not necessary to load the main xref table, because it have separate special xref table in the file header. This decreasing the time to show first page in PDF plugin. Review-Url: https://codereview.chromium.org/2483633002
2016-11-09Fix receiving page, if it have not obj num.art-snake
In some PDF's the page may not have the obj num. For example: testing\corpus\fx\other\jetman_std.pdf in pdfium repository. And CPDF_Document::GetPage failed on second call for this page. Restart the traversing of pages, to fix this Also added test. Review-Url: https://codereview.chromium.org/2491583002
2016-11-09Return unique_ptr from CPDF_Object::Clone().tsepez
Because that's what clone does. There are numerous release() calls that will go away as more code is converted. Review-Url: https://codereview.chromium.org/2484033002
2016-11-08Cleanup CPDF_DeviceCS and friends.chromium/2914thestig
Review-Url: https://codereview.chromium.org/2482273002
2016-11-07Force compiler to deduce src type for checked_cast<dst, src>.tsepez
Otherwise, we might be silently doing an unsafe cast before applying the check if the actual argument doesn't match the exact src type. Review-Url: https://codereview.chromium.org/2484953003
2016-11-07Hold trailers via unique_ptrs.tsepez
Follow-up from ParseIndirectObject() CL. This will get cleaner once CPDF_Object::Clone itself returns unique_ptrs. Pack some bools together while we're at it. Review-Url: https://codereview.chromium.org/2479353002
2016-11-07Use unique_ptr return from CPDF_Parser::ParseIndirectObject()tsepez
In turn, propgate to callers. This introduces a few release() calls that will go away as more code is converted. It also removes a couple of WrapUnique calls that are no longer needed as ownership of the object flows along. Review-Url: https://codereview.chromium.org/2479303002
2016-11-07Rename CPDF_Linearized to CPDF_LinearizedHeadertsepez
My OCD insists that classes be named after nouns, and "linearized" feels like an adjective. Remove a redundant "if" while at it. Review-Url: https://codereview.chromium.org/2482973002
2016-11-07Clean up fpdf_page_parsersnpm
- The code in fpdf_page_parser is only called by CPDF_StreamContentParser, so moved there. - Split fpdf_page_parser_old into its two classes - Renamed the corresponding unittests accordingly. - Moved PDF_ReplaceAbbr to namespace - Fixed few nits - Added TODO because CPDF_StreamParser has a lot of code similar to CPDF_SyntaxParser Review-Url: https://codereview.chromium.org/2474303003
2016-11-07Reland of Unify some codeart-snake
Unify some code Move parsing of linearized header into separate CPDF_Linearized class. Original review: https://codereview.chromium.org/2466023002/ Revert review: https://codereview.chromium.org/2474283005/ Revert reason was: Breaking the chrome roll. See https://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_rel_ng/builds/331856 ___ Added Fix for fuzzers. Review-Url: https://codereview.chromium.org/2477213003
2016-11-04Revert of Unify some code (patchset #14 id:260001 of ↵chromium/2912chromium/2911dsinclair
https://codereview.chromium.org/2466023002/ ) Reason for revert: Breaking the chrome roll. See https://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_rel_ng/builds/331856 Original issue's description: > Unify some code > > Move parsing of linearized header into separate CPDF_Linearized class. > > Committed: https://pdfium.googlesource.com/pdfium/+/71333dc57ac7e4cf7963c83333730b3882ab371f TBR=thestig@chromium.org,brucedawson@chromium.org,art-snake@yandex-team.ru # Skipping CQ checks because original CL landed less than 1 days ago. NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true Review-Url: https://codereview.chromium.org/2474283005
2016-11-04Unify some codeart-snake
Move parsing of linearized header into separate CPDF_Linearized class. Review-Url: https://codereview.chromium.org/2466023002
2016-11-04Pass object to ReplaceIndirectObjectIfHigherGeneration() by unique_ptr.tsepez
Review-Url: https://codereview.chromium.org/2477003002
2016-11-04CHECK for re-addition of indirect objectstsepez
Review-Url: https://codereview.chromium.org/2471263005
2016-11-04Traverse PDF page tree only once in CPDF_Document Try 3npm
Now, we do not start traversal from where we were at, but from the top. This makes the code less prone to bugs, as now there is no need to call methods to recursively fix things. This will save a lot of time when the trees are rather flat, as in the PDF file in the bug. It can still be slow, for instance if we have a chain of page nodes, and the last in the chain contains all of the pages (this is artificial). Try 2 at https://codereview.chromium.org/2442403002/ Also added test where Try 2 would have failed. Tested the pdf from the bug on my Mac: With this CL: load in 21 seconds Without this CL: did not load in 4 minutes, got tired of waiting BUG=chromium:638513 Review-Url: https://codereview.chromium.org/2470803003
2016-11-04Reland "Remove CPDF_Object::Release() in favor of direct delete"tsepez
This reverts commit f0d5b6c35fa343108a3ab7a25bc2cc2b3cf105b3. Review-Url: https://codereview.chromium.org/2478303002
2016-11-04Assert objnum non-zero in ReplaceIndirectObjectIfHigherGeneration()tsepez
Removes a path where it might not have taken ownership of an object, at the expense of making callers check first. This eases making the second argument an unique_ptr someday. Review-Url: https://codereview.chromium.org/2475823002
2016-11-04Revert of Remove CPDF_Object::Release() in favor of direct delete (patchset ↵dsinclair
#11 id:200001 of https://codereview.chromium.org/2384883003/ ) Reason for revert: Looks like it's blocking the roll. https://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_compile_dbg_ng/builds/186619 Original issue's description: > Remove CPDF_Object::Release() in favor of direct delete > > Follow-on once we prove Release always deletes in previous CL. > > Committed: https://pdfium.googlesource.com/pdfium/+/4de3d095c9d9e961f93750cf1ebd489fd515be12 TBR=thestig@chromium.org,tsepez@chromium.org # Skipping CQ checks because original CL landed less than 1 days ago. NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true Review-Url: https://codereview.chromium.org/2478253002
2016-11-03Remove CPDF_Object::Release() in favor of direct deletetsepez
Follow-on once we prove Release always deletes in previous CL. Review-Url: https://codereview.chromium.org/2384883003
2016-11-03Clean up CPDF_SyntaxParser a little bitnpm
- Added a private method to read a character. - Added enum for parsing status. - Deleted unused method. Review-Url: https://codereview.chromium.org/2469833002
2016-11-03Move CPDF_Document insert methods from namespacenpm
Making the insert methods private allows us to use private members, as I will need on https://codereview.chromium.org/2470803003/ Review-Url: https://codereview.chromium.org/2472473005
2016-11-02Remove FX_BOOL from coretsepez
Review-Url: https://codereview.chromium.org/2477443002
2016-10-31Fix a memory leak in CPDF_SecurityHandler.thestig
AES256_CheckPassword() leaks memory on failure. Clean up some nits. BUG=chromium:659468 Review-Url: https://codereview.chromium.org/2463183002
2016-10-31Fix loading page using hint tables.art-snake
When linearized document have hint table, The FPDFAvail_IsPageAvail return true, but FPDF_LoadPage return nullptr, for non first pages. This happens, bacause document not use hint tables, to load page. To fix this, I force save the page's ObjNum in document. This is restoring of original fix: https://codereview.chromium.org/2437773003/ Review-Url: https://codereview.chromium.org/2444903002
2016-10-28Revert of Traverse PDF page tree only once in CPDF_Document Try 2 (patchset ↵npm
#3 id:40001 of https://codereview.chromium.org/2442403002/ ) Reason for revert: Not quite right yet. Original issue's description: > Traverse PDF page tree only once in CPDF_Document > > Try 2: main fix was recursively popping elements from the stack. Since > the Traverse method can be called on non-root nodes from GetPage(), we > have to make sure to properly update the parents. > > Try 1 at https://codereview.chromium.org/2414423002/ > > In our current implementation of CPDF_Document::GetPage, we traverse > the PDF page tree until we find the index we are looking for. This is > slow when we do calls GetPage(0), GetPage(1), ... since in this case > the page tree will be traversed n times if there are n pages. This CL > makes sure the page tree is only traversed once. > > Time to load the PDF from the bug below in chrome official build: > Before this CL: around 1 minute 25 seconds > After this CL: around 4 seconds > > BUG=chromium:638513 > > Committed: https://pdfium.googlesource.com/pdfium/+/d3a2009d75eac3cda442f545ef0865afae7b35cf TBR=tsepez@chromium.org,weili@chromium.org,thestig@chromium.org # Not skipping CQ checks because original CL landed more than 1 days ago. BUG=chromium:638513 Review-Url: https://codereview.chromium.org/2461063003
2016-10-28Use CompositeMask instead of TransferBitmap when drawing type 3 textnpm
TransferBitmap seems to work improperly when the glyph boxes overlap. In particular, after drawing a glyph, the next glyph's blanks will override what the previous glyph drew, and this is not the correct behavior. While on it, use CheckedNumeric to do operations safely. For reference of somewhere where something similar is done, see: https://cs.chromium.org/chromium/src/third_party/pdfium/core/fxge/ge/cfx_renderdevice.cpp?sq=package:chromium&rcl=1477581616&l=988 BUG=513954 Review-Url: https://codereview.chromium.org/2461743002
2016-10-26Fix some FX_BOOL / int noise in fxcrt.chromium/2902tsepez
Review-Url: https://codereview.chromium.org/2450183003
2016-10-26Take advantage of implicit std::unique_ptr<>(nulltpr_t) ctor.tsepez
Review-Url: https://codereview.chromium.org/2453163002
2016-10-26Traverse PDF page tree only once in CPDF_Documentnpm
Try 2: main fix was recursively popping elements from the stack. Since the Traverse method can be called on non-root nodes from GetPage(), we have to make sure to properly update the parents. Try 1 at https://codereview.chromium.org/2414423002/ In our current implementation of CPDF_Document::GetPage, we traverse the PDF page tree until we find the index we are looking for. This is slow when we do calls GetPage(0), GetPage(1), ... since in this case the page tree will be traversed n times if there are n pages. This CL makes sure the page tree is only traversed once. Time to load the PDF from the bug below in chrome official build: Before this CL: around 1 minute 25 seconds After this CL: around 4 seconds BUG=chromium:638513 Review-Url: https://codereview.chromium.org/2442403002
2016-10-26Fix some bool/int mismatches.tsepez
Found by winxfa bot when fx_bool defined to bool. Review-Url: https://codereview.chromium.org/2449293002
2016-10-24Rename IFX_ stream nameschromium/2900tsepez
It's been troubling for some time that an IFX_FileStream might actually be an in-memory buffer with no backing file. Review-Url: https://codereview.chromium.org/2443723002
2016-10-24Fix root dictionary leak in cpdf_document_unittestNicolas Pena
The CPDF_Document does not own its root dictionary, so add ownership in CPDF_TestDocumentForPages, using ReleaseDeleter because the dictionary cannot be deleted. R=tsepez@chromium.org Review URL: https://codereview.chromium.org/2445753002 .
2016-10-21Add CPDF_Document::GetPage() unittestschromium/2899npm
Added a nontrivial page tree and a test that pages are being fetched properly, both when requested in order and in reverse order. This will help prevent introducing bugs while changing the way the page tree is processed. BUG=chromium:638513 Review-Url: https://chromiumcodereview.appspot.com/2435783006
2016-10-21Re-enable CHECK() than only 0-numbered objects are released.tsepez
Review-Url: https://chromiumcodereview.appspot.com/2424933002
2016-10-21Revert of Fix loading page using hint tables. (patchset #5 id:80001 of ↵npm
https://codereview.chromium.org/2437773003/ ) Reason for revert: CPDF_DataAvail::IsPageAvail is causing crashes. BUG=chromium:658168, chromium:658170 Original issue's description: > Fix loading page using hint tables. > > When linearized document have hint table, > The FPDFAvail_IsPageAvail return true, but > FPDF_LoadPage return nullptr, for non first pages. > > This happens, bacause document not use hint tables, to load page. > > To fix this, I force save the page's ObjNum in document. > > R=npm, dsinclair > > Committed: https://pdfium.googlesource.com/pdfium/+/ef38283688c1ee7c08bcf4204cfb78e09c039782 TBR=dsinclair@chromium.org,tsepez@chromium.org,thestig@chromium.org,art-snake@yandex-team.ru # Skipping CQ checks because original CL landed less than 1 days ago. NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true Review-Url: https://chromiumcodereview.appspot.com/2442663005
2016-10-20Fix loading page using hint tables.chromium/2897art-snake
When linearized document have hint table, The FPDFAvail_IsPageAvail return true, but FPDF_LoadPage return nullptr, for non first pages. This happens, bacause document not use hint tables, to load page. To fix this, I force save the page's ObjNum in document. R=npm, dsinclair Review-Url: https://chromiumcodereview.appspot.com/2437773003
2016-10-20Revert of Traverse PDF page tree only once in CPDF_Document (patchset #4 ↵dsinclair
id:60001 of https://codereview.chromium.org/2414423002/ ) Reason for revert: Possible cause of crbug.com/657897 reverting to find out. BUG=657897 Original issue's description: > Traverse PDF page tree only once in CPDF_Document > > In our current implementation of CPDF_Document::GetPage, we traverse > the PDF page tree until we find the index we are looking for. This is > slow when we do calls GetPage(0), GetPage(1), ... since in this case > the page tree will be traversed n times if there are n pages. This CL > makes sure the page tree is only traversed once. > > Time to load the PDF from the bug below in chrome official build: > Before this CL: 1 minute 40 seconds > After this CL: 5 seconds > > BUG=chromium:638513 > > Committed: https://pdfium.googlesource.com/pdfium/+/7c29e27dae139a205755c1a29b7f3ac8b36ec0da TBR=thestig@chromium.org,tsepez@chromium.org,npm@chromium.org # Not skipping CQ checks because original CL landed more than 1 days ago. BUG=chromium:638513 Review-Url: https://chromiumcodereview.appspot.com/2430313006
2016-10-20Merge the CFWL_*Imp classes into the IFWL_* classes.dsinclair
This Cl merges the implementation classes up into the IFWL classes as the intermediate classes were just acting as proxies. Review-Url: https://chromiumcodereview.appspot.com/2432423002
2016-10-18Traverse PDF page tree only once in CPDF_Documentchromium/2895npm
In our current implementation of CPDF_Document::GetPage, we traverse the PDF page tree until we find the index we are looking for. This is slow when we do calls GetPage(0), GetPage(1), ... since in this case the page tree will be traversed n times if there are n pages. This CL makes sure the page tree is only traversed once. Time to load the PDF from the bug below in chrome official build: Before this CL: 1 minute 40 seconds After this CL: 5 seconds BUG=chromium:638513 Review-Url: https://codereview.chromium.org/2414423002
2016-10-17Fix segv in AddImage()tsepez
The assumption that only indirect objects would be passed is wrong. Restore old behaviour despite muddying ownership constraints. R=thestig@chromium.org BUG=656145 Review-Url: https://codereview.chromium.org/2425843002
2016-10-17Rename CFX_CountRef to CFX_SharedCopyOnWritetsepez
Avoid confusing this class with other ref-counted objects. Review-Url: https://codereview.chromium.org/2426673002
2016-10-17Revert "Make CPDF_Object containers hold objects via unique pointers."tsepez
This reverts commit 1d023881cd53485303c0fcc0b5878e700dc470fd. Reason for revert -- fuzzers hit issues. TBR=thestig@chromium.org Review-Url: https://codereview.chromium.org/2425783002
2016-10-14Revert "Update CPDF_IndirectObjectHolder APIs for unique objects."Tom Sepez
This reverts commit 3ba098595ae56b64eacc0c25ab76b89a4d78d920. TBR=thestig@chromium.org,weili@chromium.org Review URL: https://codereview.chromium.org/2424533003 .
2016-10-14Update CPDF_IndirectObjectHolder APIs for unique objects.tsepez
Doing so highlights a few places where ownership is dubious. Add convenience functions to return an unowned reference to a newly-created indirect object. Review-Url: https://codereview.chromium.org/2419173002
2016-10-13Make CPDF_Object containers hold objects via unique pointers.tsepez
This tweaks the implementation while leaving the API the same. The API change is more disruptive, so break this part off first. Review-Url: https://codereview.chromium.org/2385293002
2016-10-13Optimize roll operator in CPDF_PSEngine some more.thestig
Also move code that belongs in an anonymous namespace into one. Review-Url: https://codereview.chromium.org/2416703002
generated by cgit v1.2.3 (git 2.25.1) at 2024-12-24 16:23:47 +0000