 I think you should now see the slides that I continue to confirm. Okay, I am Tomás Vajngel and I'm about to talk about the improvements that I did for PDF support in LibreOffice and Collabor Online, actually. So PDF support is very important because it's used a lot for sharing of written documents and forms and whatever on the internet and there's a big demand to have good support of course in LibreOffice and Collabor Online. So first I want to talk about a little bit what PDF support is currently in LibreOffice and we have this PDF import that is based on this popular library and what it does is that it converts these PDF elements into LibreOffice shapes and objects. The problem with this is that it's not exact as all PDF positioning we do, we try to do good job we did but it's not always possible so it's not always presented exact as it is in the PDF document. It's not possible because we don't support all the primitives that PDF supports and there is also font issues and there is PDF that has font subsets and these are not supported so well when importing it like in draw into the LibreOffice. So fidelity is compromised it's not the same as it would be rendered like in Adobe PDF reader or any more and better PDF reader like the ones that are usually in operating systems. So but what you get is we can edit it, we can edit the PDF, we can edit all the text and a lot of this is good and later we can export it and we have a new PDF document edited but it doesn't look the same so this is our problem. So we have another alternative inside LibreOffice and this is to import the PDF with PDFium which we then import as a graphic object and because we are using PDFium to render this as a graphic object the fidelity is better almost like we expect from a PDF viewer and we can import this like in graphic normally in LibreOffice but what we also can do is we can import the whole PDF document into the draw where each PDF page will be then inserted into draw as page and each draw page will have this PDF page graphic inserted for the whole draw page so it will look like draw is actually like a PDF reader. The problem with this is that of course because there are inserted as graphic we cannot edit them and this is let's say not ideal but it's a compromise and another thing is like this how we import is only this importer with PDFium is only available if you are unable to the environment variable so now so we go to collaborate online we had like some request from the customers they want to use instead of another PDF viewer they wanted to use collaborate online as the PDF viewer and of course we implemented this just for viewing of PDFs and use PDFium import as described before to do this in collaborate online we get this by default so if you open the PDF document it will be opened in draw like each page will go have a PDF page as a graphic but there are some missing features here which PDF viewers have for mostly the search is missing because we have this like PDF graphics we cannot search inside the graphics so and another I saw this was requested by OWN cloud they wanted to search another thing was that they are also in PDF like PDF viewers usually support annotations and these are comments and this was requested by Micro Focus so we first try to then create PDF search so how to do this the PDF pages are graphics so how can we search this is a problem but each graphic is still kept in SAP PDF in memory so it can be re-rendered so the idea is just that we can implement like we can use the existing search in draw what we do in this case is then that when we search draw we go through all objects and see okay this object is text and then we search inside text to find a match but when we have like this PDF graphics then we also go with search we go to PDF graphics but normally we just ignore it because we cannot do anything with it but we can extend now the search and say okay if we discover a PDF graphics then we can just search inside it we can use PDF and we can ask PDF to perform the search of the graphics to perform the search of the PDF inside the graphic and return result does there exist a string that we are searching or not inside PDF page so another issue here is also we want to show where in the graphics is the string that we are searching but we cannot select just some graphics so we have to extend like this to add new way how to select rectangles or to draw rectangles inside a graphic object and again we can ask PDF just can you give us whatever you found the text can you give us where this rectangle relative to the pages and then we can just draw the graphic like this graphic selection subselection how to say it okay and after some time we managed to do it so we so I have some examples some movies some videos how this search is working so now this is in like on our own cloud instance with collaboration line and here first this is open ODP documents so presentation as it's like draw or presentation or impress it's the same thing but the what is important for search PDF search is that we are searching for if there is a PDF graphic inside so it doesn't matter if it's a draw or impress so I first made this example for searching inside impress and I'm searching like this ABC now in this case you see like this title ABC this is like normal title page title object just text object but this ABC 123 this other this big ABC and other are actually PDFs and maybe this is not too clear from from this example so this is now we are searching ABC and also automatically goes to the next page and it also finds search results there maybe a little bit more clearly will be this example when we are like we load open cloud user manual and search inside this one so on cloud user on cloud manually is a PDF so now this is all these pages are actually are actually loaded inside draw as graphical pages and now we will try to search inside the search for the string on cloud and just go through now this is the result is our search inside PDF file actually and we can also go back one big problem here is that or like not say problem missing feature actually that we can search but we cannot copy this to clipboard if you find it find the text but we cannot do anything like copy to clipboard and paste it somewhere else for now this is something it's not yet been implemented so the next thing I want to talk about is the PDF annotation so so normally we have PDF can have many kind of annotations but inside draw we we have only okay I may be sure explain a little bit you have you can have inside PDF annotations like text or not annotation you have you can have like free text is just text like this in line node on the picture on the right side if you see this is free text it's more like a node but it's not pop especially in a pop up but in line inside of the document and then you have all like this graphical you cannot have line and polygon and like ink square and circles and polylines and then you have you can also also highlighted the text with highlight or you can underline it or with squiggly line or strike out and a lot of different you can also add a graphical stamp and and there are many many types of annotation we we we don't even support so many so and this is a problem and because we have PDF annotations are many varieties and we want to allow a lot of them so but in draw we have only comments and comments are just like note annotations texts where we go and we have a pop up window where the annotation is annotation text is presented so first idea is just to implement this text note annotation this is most common one and most easy to implement and how to do this is on import we can just read the PDF annotations that are again we just use PDF and see what annotations we have and then and and we read the annotations and forward it to the import where we create all this PDF graphics and at that time we we can say okay this PDF annotation we can support and add PDF of the annotation as a pro comment okay so here here is for example on the left side there is like this note or pop up annotation how it's presented inside PDF viewer and this is how we managed to do it as a comment in LibreOffice in this case so of course when we import the PDF annotations we have like these draw comments and we want to change them and but what to do next we also want to preserve and want to export them again so we can just use the existing PDF exports that this already supports that already supports the comments and comments writes like PDF annotations again and what I had to do for to get these is that most of these was already quite a good implemented and but there were some properties missing like pop-up has like modification date we just not exported and to make sure that this is working in the future I implemented a test that imports exports and imports so that this will keep working in the future there is one like discrepancy here is that in comments that we don't support colors but cores are supported by PDF notes and we didn't I didn't implement this yet in draw the comments are always hard coded whatever the I think it depends on the user username so that each user has always the same color but if there is another user it gets another color that is then I don't know in a list so so this is not ideal we cannot choose our own colors so another problem then when we get this implemented in LibreOffice and it works like so we can work with them maybe I need to implement also this in collaborate online and we had to extend LibreOffice kit another problem is that like drawing collaborate online is quite basic it's not it currently doesn't have any support you cannot use it except for PDF showing the PDF documents so because but we import we support impress and we need to show impress documents impress comments so we had some support for for that so what I needed to add is to show the comments on a draw page and the problem is like when you have like these comments you have to send okay I have on this page I have this and that comments and this needs to be then transported via adjacent to to the collaborate online instance and this I had to implement and of course when you are changing on call online you are changing or adding or removing comments you have to report this to the core then then there was no icons yet because when you had like this comments was only present on the right side but not no icon was shown inside the documents where the comment was added or you can you could not move it around this was to edit later on so what about other notations yes this is now work in progress I want later I'll show you what's the current state but what is the idea here is that we can support a lot of these graphic are like vector graphic annotation pdf annotations just by changing the default default marker to to draw a vector graphic instead depend on which type of pdf annotation is like we get polygon ink lines square or circle we can just draw this and get this graphic like graphic information from the pdf annotation and convert this to our own another type is then the text highlight annotations this is a little bit different because we are annotating the text itself and so this is a little bit bound to the text ideally like if we annotate text it would be highlight would be behind the text but we have a graphic the pdf as a graphic is not possible here so we have to compromise this and just have a like transparent rectangle over it but underlying a strike I should meet problem steering there is a problem later on when we try to annotate it ourselves when we try to add new annotations so this is now the current state I have already shown the left side which is a pf you are on the right side on the right side you can see then in the imported pdf inside LibreOffice maybe I can show what is the latest state of this so this is current state we already import also the lines but I have also other examples here okay here this is has more types of annotations like you can have field polygon like field elites rectangle okay this is freehand and this is the highlight sign so they are also working okay thanks for listening any questions so if there's no no questions we could probably move on to the next talk which is going to be from Angkor I have a question already asked in the chat is it I don't know how it works but when you scan the pdf and you have images and then with OCR you have the text behind this image so the quality of the pdf is always the same because you see the image but you can search because the text is somewhere in the background I don't know how it works but because it's the same issue you had you can't search in the pdf and yes would it be useful to use something like that for the searching or is already you fix the issue well we searched I like that we have a pdf graphic and then we asked the pdfium library to search it but I think that how OCR this is then a feature of a feature of pdf that you can have graphic and then like text added to the graphic so that you can search inside so this is a feature of the pdf itself I think but I'm not really familiar with this but I think this is something different what the problem here is because sometimes when I want to for example edit and pdf with draw it's very cool because everything was imported as text but in the end sometimes I only want to have the 100% same pdf and for example remove some text and add some new text so sometimes it's frustrated because the quality of the edited document is not the same gentlemen could you could you maybe take this this conversation off to the one of the other rooms that we have like the chat room that would be very helpful for us to keep everything going thank you thank you