Console View
|
|
|
|||
| 022c41206fa0... |
Adam Sampson
ats@offog.org |
|
|
|
glib: Correct assert in convert_color This didn't match the equivalent code in StructElement.cc's isRGBColor; the length test was inverted. Use isArrayOfLength instead. |
|||
| 3398be722736... |
Adam Sampson
ats@offog.org |
|
|
|
Object: Add isArrayOfLengthAtLeast 25 places in the Poppler code wanted to check whether an object was an array with at least N elements. Add a member function for this, along the lines of isArrayOfLength. There were two cases of a check for > 10000 where the intent is to catch an absurdly large array rather than to require a specific limit; in these cases it's now checking for >= 10000. |
|||
| 5c8c83e7780a... |
Adam Sampson
ats@offog.org |
|
|
|
Object: Add isArrayOfLength 42 places in the Poppler code wanted to check whether an object was an array with a specific length, e.g.: if (obj.isArray() && obj.arrayGetLength() == 4) ... Add a member function to simplify this. This also avoids doing the typecheck twice, in the unlikely case that the compiler hasn't already optimized it away. |
|||
| 515217456664... |
Juraj Šarinay
juraj@sarinay.com |
|
|
| Replace the constant maxSupportedSignatureSize by a generous estimate based on the sizes of the certificates involved. | |||
| e1f3ecdbfa35... |
Sune Vuorela
sune@vuorela.dk |
|
|
| GPGME: Fix marking of qualified keys | |||
| 3aa4e28f3427... |
Albert Astals Cid
aacid@kde.org |
|
|
| Update (C) | |||
| 6173e3db241d... |
Albert Astals Cid
aacid@kde.org |
|
|
| Make a few Dict * be const & | |||
| 7b641b6bc271... |
Sune Vuorela
sune@vuorela.dk |
|
|
| Document array-ownership before handed to object | |||
| fdc6a40da7bd... |
Sune Vuorela
sune@vuorela.dk |
|
|
| Document dict-ownership before handed to object | |||
| 3608d840588f... |
Albert Astals Cid
aacid@kde.org |
|
|
|
Remove USE_FLOAT cmake option We are in 2026, hoping no one needs float vs double for speed |
|||
| d9d0b54f7fdb... |
Albert Astals Cid
aacid@kde.org |
|
|
| Increase version so people that track master can add ifdefs on version | |||
| 30927d02c1fd... |
Albert Astals Cid
aacid@kde.org |
|
|
| poppler 26.02.0 | |||
| f40c4f3a4017... |
Sune Vuorela
sune@vuorela.dk |
|
|
| Build clang-17 in debian-trixie (current stable) image | |||
| 09c4c1ba2039... |
Albert Astals Cid
aacid@kde.org |
|
|
| Update (C) | |||
| a83aa3971607... |
Albert Astals Cid
aacid@kde.org |
|
|
|
SplashFont: Make the clip const & Makes it clear we don't take ownership of the pointer |
|||
| 64fc550a8769... |
Stefan Brüns
stefan.bruens@rwth-aachen.de |
|
|
|
TextOutputDev: Fix non-transitive comparision for text fragment sorting Comparision functions used for std::ranges::sort must be transitive, i.e. (a > b) && (b > c) -> (a > c). Using a fuzzy comparision for the y coordinate breaks this proberty (assuming rot==0, likewise for others): - [0 0] < [2 0.008] && [2 -0.008] < [4 -0.016] (y ~ 0 -> x is compared) - [0 0] > [4 -0.016] (y is compared) As the comparision is only used in a single place, and the y sorted fragments are then split into ranges with similar coordinate (less than maxIntraLineDelta) which are sorted by x coordinate, the overall result is very likeyly identical. |
|||
| 19c291eb62ee... |
Stefan Brüns
stefan.bruens@rwth-aachen.de |
|
|
|
TextOutputDev: Fix incorrect flag for column recalculation In case an area is passed in, some blocks/fragments may be discarded, and the columns should be recalculated for the remaining fragments. The remaining fragments may still have varying rotations, and it is necessary to use the correct algorithm. Otherwise, the sorting may use a comparision function which is not transitive for varying orientations, and may even crash due to violated requirements. |
|||
| ad1645e782cb... |
Stefan Brüns
stefan.bruens@rwth-aachen.de |
|
|
|
qt: Pass default text area as nullopt, avoid incorrect warning In case a default constructed rect is passed to Page::text(), it should be passed down as is instead of using the cropbox. Using the cropbox is incorrect, as the passed down rect refers to the area of the logical page, while the cropbox refers to the physical media. E.g. when page 1 is the right-hand A5 area of a A4+ paper, the cropbox rect would have an xMin of >= 148mm, and would not overlap with the logical page [0 0 148 210]. This also avoids the warning for ReadingOrder text layout which is triggered when a rect is passed down. |
|||
| f145613c3734... |
Adam Sampson
ats@offog.org |
|
|
|
cpp: tests: Use range-for loops in poppler-dump The "Enable modernize-use-auto" changes revealed a few more cases where range-for loops can be used here. |
|||
| 68a9a2f46146... |
Albert Astals Cid
aacid@kde.org |
|
|
| JBIG2Stream: Add another getReadPastEndOfStream check | |||
| 62f8b62514f7... |
Azhar Momin
azhar-momin@outlook.com |
|
|
| Move ossfuzz build scripts to poppler repository | |||
| 9eea384d06c4... |
Albert Astals Cid
aacid@kde.org |
|
|
|
XRef::removeIndirectObject: Don't manually call the destructor The other 2 places on the file that do it is "ok" because we are manually holding that memory in a freed/malloced array and we never call that object again, but here we can use it again and that's not allowed |
|||
| 414fd172aec3... |
Albert Astals Cid
aacid@kde.org |
|
|
|
JBIG2Stream: Check we discarded the correct amount of bytes If not, we reached EOF, so report it to the caller |
|||
| 16ccb3c65171... |
Albert Astals Cid
aacid@kde.org |
|
|
| Mark discardChars as nodiscard | |||
| b6bf563dae16... |
Albert Astals Cid
aacid@kde.org |
|
|
| Update (C) | |||
| eee94a13b930... |
Ojas Maheshwari
workonlyojas@gmail.com |
|
|
| pdftotext: Fix page level bounds calculation in tsv mode | |||
| 618da9e31f4e... |
Albert Astals Cid
aacid@kde.org |
|
|
| JBIG2Stream::readGenericBitmap: Sprinkle some more EOF checks | |||
| 640373f32e1d... |
Albert Astals Cid
aacid@kde.org |
|
|
| JBIG2Stream::readSymbolDictSeg: Remove gotos | |||
| dfbd101c24cd... |
Albert Astals Cid
aacid@kde.org |
|
|
| JBIG2Stream::readTextRegion: Take const & vector | |||
| ad16a22d1ccf... |
Albert Astals Cid
aacid@kde.org |
|
|
|
JBIG2Stream::readSymbolDictSeg: Make bitmaps be a vector And collect the ones we have to delete in another vector |
|||
| 0ef782a0c23e... |
Albert Astals Cid
aacid@kde.org |
|
|
| Some initial tweaking for JBIG2Stream::readSymbolDictSeg | |||
| 86c30db439a2... |
Albert Astals Cid
aacid@kde.org |
|
|
| Add a bit of constness around JArithmeticDecoderStats | |||
| a61aec2d0e21... |
Albert Astals Cid
aacid@kde.org |
|
|
| JBIG2Bitmap::combine: take bitmap by const & | |||
| 0e1de7f2f782... |
Albert Astals Cid
aacid@kde.org |
|
|
| Tweak JBIG2Stream::readGenericRefinementRegionSeg | |||
| 0a7d51df66e9... |
Albert Astals Cid
aacid@kde.org |
|
|
| JBIG2Stream: make pageBitmap a unique_ptr | |||
| 0e63e0c8229e... |
Albert Astals Cid
aacid@kde.org |
|
|
| JBIG2SymbolDict: I forgot the resize ^_^ | |||
| 72e5d850ebc6... |
Albert Astals Cid
aacid@kde.org |
|
|
|
JBIG2Stream: Move vector max sizecheck to a templated inline This way we don't need a pragma for clang anymore |
|||
| ed4718e9389b... |
Albert Astals Cid
aacid@kde.org |
|
|
| JArithmeticDecoderStats: Rename reset to resetContext | |||
|
|||
| d588c22ef7dc... |
Albert Astals Cid
aacid@kde.org |
|
|
| JArithmeticDecoderStats: copy now returns a unique_ptr | |||
|
|||
| 8cb72f4dfd81... |
Albert Astals Cid
aacid@kde.org |
|
|
| JBIG2SymbolDict: Make bitmaps a vector | |||
|
|||
