Potential bug - NPE in COSArray.add with big HTML document in PDF A_1_A mode and Arial font

Hello,

First of all, thanks for the work performed on this lib, it's really appreciated.

I work on a project using openhtmltopdf to generate PDF A_1_A compliant from quite big HTML files.
We ran into the following stacktrace in some cases:
`java.lang.NullPointerException: Cannot invoke "org.apache.pdfbox.pdmodel.common.COSObjectable.getCOSObject()" because "object" is null

at org.apache.pdfbox.cos.COSArray.add(COSArray.java:62)
at com.openhtmltopdf.pdfboxout.PdfBoxAccessibilityHelper.finishNumberTree(PdfBoxAccessibilityHelper.java:858)
at com.openhtmltopdf.pdfboxout.PdfBoxFastOutputDevice.finish(PdfBoxFastOutputDevice.java:908)
at com.openhtmltopdf.pdfboxout.PdfBoxRenderer.writePDFFast(PdfBoxRenderer.java:674)
at com.openhtmltopdf.pdfboxout.PdfBoxRenderer.createPdfFast(PdfBoxRenderer.java:564)
at com.openhtmltopdf.pdfboxout.PdfBoxRenderer.createPDF(PdfBoxRenderer.java:490)
at com.openhtmltopdf.pdfboxout.PdfBoxRenderer.createPDF(PdfBoxRenderer.java:427)
at com.openhtmltopdf.pdfboxout.PdfBoxRenderer.createPDF(PdfBoxRenderer.java:409)
at com.openhtmltopdf.pdfboxout.PdfRendererBuilder.run(PdfRendererBuilder.java:46)
at com.openhtmltopdf.pdfa.testing.PdfATester.run(PdfATester.java:95)
at com.openhtmltopdf.pdfa.testing.PdfATester.cannotRenderInArial(PdfATester.java:153)
at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:77)
at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.base/java.lang.reflect.Method.invoke(Method.java:568)
at org.junit.runners.model.FrameworkMethod$1.runReflectiveCall(FrameworkMethod.java:59)
at org.junit.internal.runners.model.ReflectiveCallable.run(ReflectiveCallable.java:12)
at org.junit.runners.model.FrameworkMethod.invokeExplosively(FrameworkMethod.java:56)
at org.junit.internal.runners.statements.InvokeMethod.evaluate(InvokeMethod.java:17)
at org.junit.runners.ParentRunner$3.evaluate(ParentRunner.java:306)
at org.junit.runners.BlockJUnit4ClassRunner$1.evaluate(BlockJUnit4ClassRunner.java:100)
at org.junit.runners.ParentRunner.runLeaf(ParentRunner.java:366)
at org.junit.runners.BlockJUnit4ClassRunner.runChild(BlockJUnit4ClassRunner.java:103)
at org.junit.runners.BlockJUnit4ClassRunner.runChild(BlockJUnit4ClassRunner.java:63)
at org.junit.runners.ParentRunner$4.run(ParentRunner.java:331)
at org.junit.runners.ParentRunner$1.schedule(ParentRunner.java:79)
at org.junit.runners.ParentRunner.runChildren(ParentRunner.java:329)
at org.junit.runners.ParentRunner.access$100(ParentRunner.java:66)
at org.junit.runners.ParentRunner$2.evaluate(ParentRunner.java:293)
at org.junit.internal.runners.statements.RunBefores.evaluate(RunBefores.java:26)
at org.junit.runners.ParentRunner$3.evaluate(ParentRunner.java:306)
at org.junit.runners.ParentRunner.run(ParentRunner.java:413)
at org.junit.runner.JUnitCore.run(JUnitCore.java:137)
at com.intellij.junit4.JUnit4IdeaTestRunner.startRunnerWithArgs(JUnit4IdeaTestRunner.java:69)
at com.intellij.rt.junit.IdeaTestRunner$Repeater$1.execute(IdeaTestRunner.java:38)
at com.intellij.rt.execution.junit.TestsRepeater.repeat(TestsRepeater.java:11)
at com.intellij.rt.junit.IdeaTestRunner$Repeater.startRunnerWithArgs(IdeaTestRunner.java:35)
at com.intellij.rt.junit.JUnitStarter.prepareStreamsAndStart(JUnitStarter.java:235)
at com.intellij.rt.junit.JUnitStarter.main(JUnitStarter.java:54)
`
It happens only when:
- We use A_1_A mode;
- The HTML file reached a certain size;
- The used font is Arial.

If I disable the A_1_A mode, it will work with Arial.
If I use Karla instead of Arial, it will work with the A_1_A mode.

I tried to debug a bit to understand what leads to this NPE but without reaching an understanding so far.

I forked your repo from the 1.0.10 release and created a branch to reproduce the issue.
You can find it here: https://github.com/CedricMtta/openhtmltopdf/blob/cedricmtta-error-with-big-html-and-some-fonts-reproduction/openhtmltopdf-pdfa-testing/src/test/java/com/openhtmltopdf/pdfa/testing/PdfATester.java

You can run PdfATester#canRenderInArial to see a quite big html document being rendered in Arial and A_1_A compliance successfully.
You can run PdfATester#cannotRenderInArial to see the failure described above.

I don't understand how it could work in Karla but not Arial. The main difference I see between those two fonts is that Arial supports nearly any language, while Karla supports only western language.

I wonder whether I'm doing something wrong in the config or if the source HTML is broken, or if it's an issue in openhtmltopdf.

I'd appreciate any help, thanks a lot in advance :)

EDIT: It works with the Open Sans regular font, that can be found here https://www.fontsquirrel.com/fonts/open-sans
This work is compatible with non-western characters, like Arial.
I still cannot understand why the rendering with Arial throws an NPE :(

Start with PdfATester#cannotRenderInArial in openhtmltopdf-pdfa-testing/src/test/java/com/openhtmltopdf/pdfa/testing/PdfATester.java and reproduce the failure using the reported Arial, large HTML, and A_1_A conditions. Trace the stack from PdfBoxRenderer through PdfBoxFastOutputDevice.finish and PdfBoxAccessibilityHelper.finishNumberTree; done means the reproduction completes without the COSArray NPE and produces a compliant PDF.

Written by the indexing model from the issue text.