THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1616746649
Thanks, fixed.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To uns
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1546911147
> Hi, @THausherr ! These changes are also available for line 159
(PDSeparation.toRGBImage).
![grafik](https://github.com/apache/pdfbox/assets/6665575/655276b8-a278-4175-b461-aa20b9a
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1529374910
Thanks, all fixed. Amusingly, I had a personal comment in the
PDSpecialColorSpace class ("//TODO why is this needed? Super method is
identical") and the file is from 2016. I'm keeping the
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1465128114
Good find, fixed. Not a bug, operators really aren't written in object
streams. Probably copied from other code.
--
This is an automated message from the Apache Git Service.
To respond t
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1464999224
typos fixed, thanks. I think it's just debug output. Re my misunderstanding
about overloading: I understand now that I have to see this "at compile time",
it's not dynamic. Maybe I had ano
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1460102180
Re the latest ContentStreamWriter.java change, I was wondering if we could
overload all methods but when I put a log here
```
else if( o instanceof Operator)
{
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1449477685
@lehmi Indeed, thanks! Done.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific co
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1449313244
> Hi, @THausherr ! Several possible performance improvements. Inside
GlyphSubstitutionTable.readScriptList method we create a scriptTables variable.
Then we fill it in loop. We then use it
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1257194406
I don't really understand the last comment, but this may be related to
changes in PDFBOX-5489.
--
This is an automated message from the Apache Git Service.
To respond to the message, ple
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1229395121
ttfFont is still null if parse() fails. One could argue that
`ff2Stream.getCOSObject().createView()` should be closed if parse() fails, but
that wasn't your argument.
--
This is an auto
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1229364861
Re 1: this is a known problem, a comment mentions PDFBOX-4963
Re 2: TTFParser does not have a close method
--
This is an automated message from the Apache Git Service.
To respond to th
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1207141365
> Hi, @THausherr ! The testPDFBox5484. Don't you need to free the doc
variable?
Thanks, will fix this later.
--
This is an automated message from the Apache Git Service.
To respon
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1186577537
> As I understood, my commit has no bug and it is simpler then old version.
Yes + yes and I've of course committed it. Sorry again for my mistake.
--
This is an automated message f
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1186576212
Sorry, I completely missed that the original returns had been removed. I
don't know how this happened because it's obvious.
--
This is an automated message from the Apache Git Service.
T
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1186512434
re
[85368a5](https://github.com/apache/pdfbox/pull/107/commits/85368a5312e69b3a863e9163f19d63c8945b6c3f)
- this is good but shouldn't the returns be changed as well?
re those 2 LGTM
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1186450214
> PDDocument.saveIncrementalForExternalSigning.
> If the getSignatureDictionaries method returns an empty collection
(string...), you throw an IllegalStateException. You break the loop i
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1186433593
> Hi, @THausherr! The few my observations:
> 1 AnnotationBorder.underline - unused field.
> 2 CCITTFactory.readshort - useless if-statement (wrong if body?).
> 3 PDCaretAppearanceHa
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1179741986
> > existing code is better IMHO. And whoever wrote this, intended to use an
array instead of a list.
>
> You are converting an array to a list in the getXRefRanges method. You use
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1155945323
> Hi, @THausherr ! You did commit
[53bd9bb](https://github.com/apache/pdfbox/commit/53bd9bb6cdaf01300aad13fa5632d830444a).
How about line 249?
fixed, thanks
--
This is an aut
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1121395542
> > > Hi, @THausherr ! 1 The method PDFPrintable.print has affine transform
clonning (260). Is this operation really needed or is it old code? 2 the
classes from org.apache.pdfbox.pdmodel.
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1114219783
> generateAppearanceStreams
> Proposition: move the method generateAppearanceStreams to
PDAbstractAppearanceHandler
Thanks, done.
--
This is an automated message fr
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1114028675
> Proposition: use java.util.concurrent.ConcurrentHashMap for CMapManager
.cMapCache
thanks, done and found and removed unneeded related code
--
This is an automated message from
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1113574945
> Okay. Third is warning. Just a strange code. How about my propositions?
I assume you mean the other comments. I'll read them when I have the time. I
have about 50 github mails from
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1108901829
> Hi, @THausherr ! 1 The method PDFPrintable.print has affine transform
clonning (260). Is this operation really needed or is it old code? 2 the
classes from org.apache.pdfbox.pdmodel.font
THausherr commented on PR #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1101283438
> > This pull request **introduces 1 alert** when merging
[d3c1021](https://github.com/apache/pdfbox/commit/d3c10211f5202cb1baa98cb4b365ff6b6a5334b0)
into
[9db582c](https://github.com/apa
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1077993440
> If PDDocument.saveIncrementalForExternalSigning will be called several
times then signingSupport instance will be createad again but released only
once (PDDocument.close())
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1077945750
> Hi, @THausherr! I found a comment of SequenceRandomAccessRead.createView
method was removed.
Which one / when?
--
This is an automated message from the Apache Git Se
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-1045995533
> Hi, @THausherr! Method PDTerminalField.importFDF (103). Can the
getWidgets() method return an emply list for this case (probably an
IndexOutOfBoundsException)?
Thanks,
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-997327723
I agree that the method is also flawed because the offset isn't set to the
starting point - 1 after a fail, but I doubt troubles would happen as long as
we only search for %%EOF
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-997278739
Good find. I think this can be fixed by changing the code to be like this:
```
else if (counter > 0)
{
counter = 0;
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-997237347
My guess is that the comment is misleading. Should probably rather be a text
like "should not happen, likely broken PDF". The code is related to
https://issues.apache.org/jira/
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-969855014
>
>
> Hi, @THausherrI found one strange method. COSInteger.equals. Why it has no
isValid method checks?
Yeah maybe it should be done. But isValid is about big value
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-969853797
>
>
> Hi, @THausherr ! The classes JBIG2Filter (decode method) and COSWriter
(getDataToSig method) uses a SequenceInputStream class with constructor with
Vector creation.
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-968209106
Thanks, fixed. I missed that one but did it correctly in the 2.0 branch.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-964417948
>
>
> Hi @THausherr! Please, check the COSDictionary.getDictionaryString method.
It looks like there are error on line 1372 ("if (base instanceof COSStream)" in
the wrong
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-960429567
>
>
> Hi, @THausherr! You have changed the PDColorSpace class. It looks like the
ColorConvertOp field might be static? Why do you need many instances of
ColorConvertOp fo
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-947908126
Thanks, fixed in PDFBOX-5298. It's the first time I hear about this class.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-938320827
>
>
> PDSignature.getByteRange able to return empty array but getContents(...)
methods have no checks. A comment in getContents() method says "@throws
IOException if the
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-906629051
>
>
> PDColor.PDColor(COSArray, PDColorSpace). Line 58. Index is zero based (was
-1 lost?).
I don't see the problem - it retrieves the last element, which is a name
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-905787547
>
>
> StandardSecurityHandler.prepareForDecryption. Line 153:
encryption.get__String__FilterName()?
good observation! fixed.
--
This is an automated message from
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-904865473
> WinAnsiEncoding
>
>
> WinAnsiEncoding constructor. Check comment (276) and creation of index
(277).
So what's the problem? I don't see a contradict
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-904772908
>
>
> FDFJavaScript.getDoc. This method returns null or empty map but not map
with keys and values.
Thanks, will be fixed in https://issues.apache.org/jira/browse/P
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-904772530
>
>
> Hi, @THausherr! MapBackedScriptFeature.equals. Line 105. This is right
fields comparing? The maps compares keys in the equals method and you compares
keys with type
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-899034394
> COSParser.parseTrailer. If a nextLine variable will start with a space
character then a nextLine.startsWith("trailer") return false. Π‘an the nextLine
start with a space?
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-894651433
Fixed in https://issues.apache.org/jira/browse/PDFBOX-5253 . However it
turns out that this is almost impossible to happen (see PDPage.getMediaBox()),
unless a user has a page wi
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-890442972
I'm aware that a developer would be able to change it. These types of fields
have been flagged by Sonar. They shouldn't change the page size after creation,
and instead set the r
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-890394495
re 1, 5, 6, 7, 8, 9: fixed, thanks
re 3: π for now π
re 4: will wait until it happens. Not sure if this can ever happen.
--
This is an automated message from the Apach
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-890381198
> PDAnnotation.getBorder. If border.size() >= 3 don't need to "create a copy
to avoid changing the PDF"?
No. The "create copy" is a workaround to be able to work with a bro
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-882731846
Rather not, this would mean even more work, for tiny advantages.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-882731846
Rather not, this would mean even more work, for tiny advantages.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-882731846
Rather not, this would mean even more work, for tiny advantages.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-869173102
filter() only returns a new image if dest is null ("dest color converted
from src or a new, converted BufferedImage if dest is null") so it doesn't
matter.
Btw please don'
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-868305576
Isn't `buffer` closed in the `finally` segment? Same for `filteredBuffer`.
Or did I misunderstand this?
--
This is an automated message from the Apache Git Service.
To respond
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-861149415
Thanks, fixed.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific commen
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-856120740
Oops, you're right. Gotta think about it.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above t
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-856111219
Your code removes all after 0 and including this 0, but that is different to
what the current code does. We don't need this at this time.
--
This is an automated message from t
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-855626092
Re PDPageLabels: my understanding of the code is that it removes the 0 bytes
at the end, and only these, see
https://issues.apache.org/jira/browse/PDFBOX-1047 . The problem strin
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-854882614
Why make it protected? They're both private and very different. To make this
less confusing, I'm renaming them to something very different.
--
This is an automated message from
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-850540365
Re ASCII85Filter / DCT filter - the problem is that I would have to close
the top level object and I no longer have it. No harm is done currently, it's
just "not a good style".
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-850088834
1) maybe. I'll do something at a late time
2) no, bytes being not null means that this has worked in the past.
3) yes, but this would mean a broken project.
--
This is an
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-848083747
Please discuss this in those issues. They are both self-assigned to Andreas
so maybe he has already done something.
--
This is an automated message from the Apache Git Service.
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-847489084
> BaseParser.readStringNumber. A buffer.length() cannot be bigger than
Integer.MAX_VALUE (max length of any java array).
The name of that constant is a bit confusing, MAX_L
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-847230592
> ASCII85Filter.decode. Should the InputStream encoded be closed inside? The
same for DCTFilter, ASCII85Filter.
No, this is done by the caller, see earlier text here
--
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-847183618
I committed two changes re: FileSystemFontProvider; one of those you
mentioned I didn't because it's already been taken care of, see at the bottom
of addTrueTypeFontImpl.
--
T
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-847131678
PDFPrintable suggestion done in
https://issues.apache.org/jira/browse/PDFBOX-5200 , thanks
--
This is an automated message from the Apache Git Service.
To respond to the messag
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-846879740
> DecodeResult.DEFAULT uses COSDictionary field. Why DEFAULT field uses
modifiable COSDictionary and colorSpace? The COSDictionary can be changed via
getParameters method.
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-846866094
Thanks, I removed the unneeded code with the HashMap search.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-846614634
> private constructor of PDStream. If input will be null then NPE will be
thrown inside a row "IOUtils.copy(input, output);" (137). And redurant closing
of stream. IMHO, the inpu
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-846419514
> CosStream.createView: if filterList.size() will be zero then the input
will not be closed.
In that case that part of the code isn't used, look for `if
(filterList.isEmpt
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-846406659
Re filter: My understanding is that the caller should close streams unless
the javadoc tell something else. So I did search a bit, and found places in
COSInputStream, COSStream a
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-840411665
> ASCII85InputStream.read(byte[] data, int offset, int len). Row 182. Maybe
logical OR.
No, test fails.
--
This is an automated message from the Apache Git Service.
To r
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-837160661
> PDResources.getExtGState. row 249. The indirect variable cannot be null
here or it is acceptable value? The same for getShading, getPattern,
getProperties, getXObject. Actually
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-837142071
> @THausherr , will you fix this cosname-test? 'PDFBOX-5190: revert changes
as those were too strict', but not test.
Andreas will likely do this.
--
This is an automated
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-837047675
> FDFAnnotationCaret.setSymbol. Maybe it is not a bug but looks weird.
Yeah, but I prefer not to change it. I don't know what the intention was and
it isn't used much.
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-835846473
I realize that we don't have to clone the points. These aren't changed in
PDFBox, and maybe people outside shouldn't change them. I'm too influenceable
in my effort to produce th
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-835785826
Oops, you're of course right. I reviewed the rest to find the creation of
these arrays and these are new Point2D values each time, but who knows what a
future developer will do,
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-835768285
> COSDictionary.getObjectFromPath method. If the first object found is
sufficient, you need to add two break operators to the loop.
No, this is a bit like XPath. The real p
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-835723488
I fixed the checkbox thing, thanks. I don't understand what you mean with
the Point2d thing. What I can tell:
- it works fine (there are almost no public tests due to file copy
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-835472443
> Type4ShadingContext. Unnecessary bitsPerFlag field.
There's a TODO about it.
--
This is an automated message from the Apache Git Service.
To respond to the message, ple
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-835224869
ShadingContext classes now refactored
https://issues.apache.org/jira/browse/PDFBOX-5189
--
This is an automated message from the Apache Git Service.
To respond to the message
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-835132621
> Note: Children of ShadingContext uses an interface PaintContext with has a
dispose method too.
Yes it is kindof weird that ShadingContext isn't an implementation of
Pain
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-835124674
> Maybe it is not bug but a little strange. You changed the
PDCIEDictionaryBasedColorSpace.setWhitePoint and setBlackPoint. Do getters of
these whitepoint and blackpoint good? Yo
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-835123235
> COSOutputStream.flush. @THausherr, I think, you need use a super if buffer
is null. To revert null check.
Thanks. The risk of early morning work.
--
This is an automat
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-835121361
> TilingPaint.getAnchorRect. Can the pattern.getBBox method return null in
this case (using in the method getAnchorRect)?
Yeah it could be. I've fixed this and this brought
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-835042564
Thanks, fixed in https://issues.apache.org/jira/browse/PDFBOX-5188
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHu
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-834681664
> CosInputStream.flush. Why not if buffer is null then super.flush()?
Please clarify
--
This is an automated message from the Apache Git Service.
To respond to the messag
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-834680450
> JPEGFactory.retrieveDimensions(ByteArrayInputStream). Same as
Filter.findImageReader
thanks, fixed, but in 3.0 only.
--
This is an automated message from the Apache Gi
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-834665252
Re Patch: thanks, fixed. Re "ShadingContext is public why dispose package
private"? Don't know. It's possible that somebody wanted to reference the
object without needing to disp
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-833746897
Thanks, the javadoc was wrong, only whitepoint is required. I've updated the
code accordingly.
--
This is an automated message from the Apache Git Service.
To respond to the me
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-833261642
Yeah the two things in the images are bugs, I'll fix these, thanks.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitH
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-833245397
No you didn't make a bug. What I mean is that you were optimizing an error
handling. This is a bad PDF but we don't abort because people would complain
"but it renders with Adobe
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-831358700
Thanks, fixed
https://issues.apache.org/jira/browse/PDFBOX-5183
and also inspired me for
https://issues.apache.org/jira/browse/PDFBOX-5184
--
This is an automated messa
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-830849092
> SmallMap.putAll - if NullPointerException will be thrown the mapArr field
will not be cleared. It can produce new bugs.
Doesn't matter, if an NPE happens this is a progra
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-830848548
> There is strange behavior. The method PdCalGray.toBGB returns clon in
first case and returns not clon in second case.
It's always a new object. In the first case a clone
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-827769059
I don't know and I prefer not to test this, because in the worst case we
would either not clone enough, or have doubles. This is a very sensitive part
of the code, I had some har
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-819692863
No, although there is a cloning class, but I prefer not to use it because it
could create an incredible bloat. Not so much because these two dictionaries,
but because of the rest
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-818922235
I don't see how this would make it faster, a type is probably also some sort
of number in a table. And the way you're setting currentReturnType is a side
effect. The fonts aren't
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-818918758
Re PDViewerPreferences: The same problem also applies to the document
information. And probably to other parts of the document, e.g. fonts or images.
Yes all of this carries risk
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-813324291
Yes it is a tricky code segment. I prefer to wait until somebody brings up
such a file. IIRC one of the problems is that we can't throw an exception.
Another problem is that we c
THausherr commented on pull request #107:
URL: https://github.com/apache/pdfbox/pull/107#issuecomment-813274532
> Do you want to say there is difference in memory usage for a field and
local variable? But it is only one additional parameter.
Yes I assume that. It's pushed on the stac
1 - 100 of 126 matches
Mail list logo