Francesco Pretto

A quick search for LibSourcey.cmake shows no results, so it seems something specific to your system. Instead, it happens that the dependency on [LibWebRTC](https://github.com/aisouard/libwebrtc) is not mentioned at all.

I discovered what was the problem. Google servers weren't the problem: in some circumstances two STUN servers binded to different IPs and TCP ports are needed for NAT traversal[1]. I...

The actual extracted text is not read by those fragments you highlighted but by the `Tj` PDF operators, and the text to be rendered is encoded in a different way....

To give you the sense of what to do, I am not in a rush in implementing the system font lookup fallback for text extraction/rendering (which was identified quite some...

I made all the relevant analyses, which are based solely on the [adobe_creat2.pdf](https://github.com/podofo/podofo/files/13765156/adobe_creat2.pdf) document, where there are multiple fonts, one is SimSun which is used for the headers text. I...

Text extraction ignores that `/Artifact` (which appear to be metadata) by design, because text needs an exact position which is determined by other operators. As a proof of this, you...

Then I guess you can do `if (CMAKE_CXX_COMPILER_ID STREQUAL "GNU" AND MINGW)`, still within the else() branch.

Hello. No, it doesn't. I also noticed that Adobe and other libraries (eg. Pdf.js) can do some repair of the cross-reference table, so I previously added an entry in the...

Some triage, after a while: it makes sense. Something should be done so newlines are processed this way (and `\r` ignored, as we don't support MacOS classic, version

Decisamente quei metodi vengono da una versione forkata di PoDoFo. Con la vecchia API 0.9.x la firma era supportata solo come operazione di basso livello, e proprio per questo motivo...