| | <<<Back 1 day (to 2014/01/10) | 2014/01/11 |
| Noldorin | can postscript do basic string parsing? | 01:35.11 |
| Robin_Watts | Noldorin: postscript can do most things - it's just a question of how hard you want to work. | 01:42.51 |
| Noldorin | Robin_Watts. yeah. i'm just reading through the cookbook at the moment, and pondering how hard it might be, that's all :) | 01:43.19 |
| | aha | 01:54.23 |
| | the array ops also work on strings | 01:54.29 |
| | very good | 01:54.30 |
| | Robin_Watts, any idea how to push a literal onto the stack? | 02:16.16 |
| | i mean | 02:16.22 |
| | a literal constructed from some string, say | 02:16.31 |
| | bah | 02:23.54 |
| | no-one here now :( | 02:23.58 |
| mobjax | Hi | 13:22.02 |
| ghostbot | que tal, mobjax | 13:22.02 |
| mobjax | good Morning | 13:22.09 |
| | I succesfully installed MuPDF for one of my Android Apps | 13:22.23 |
| | But the problems is, | 13:22.35 |
| | I cannot fit the PDF to the screen | 13:22.53 |
| | can anybody suggest me how to achieve this | 13:23.05 |
| jo0nas | mobjax: Don't contact me discretely, that's bad behaviour on IRC! | 13:33.01 |
| | mobjax: do *NOT*^contact me directly!!! | 13:37.51 |
| mobjax | ok | 13:38.05 |
| | then how can I contact you | 13:38.08 |
| jo0nas | I have no interest in talking with you now | 13:38.26 |
| mobjax | ok | 13:38.36 |
| | Thanks | 13:38.39 |
| jo0nas | good luck - be nice and there's a higher chance folks want to help you | 13:39.17 |
| mobjax | Thanks Again! | 13:39.47 |
| Robin_Watts | mobjax: Hi. irc has time lags; be prepared to wait around for a while for answers. | 14:03.32 |
| | In particular, over weekends here, it's slower. | 14:03.43 |
| | So, you have build MuPDF into your android app. Is this a commercial app, or a GPL one? | 14:03.59 |
| mobjax | it is a GPL one | 14:04.34 |
| | thanks that somebody answered at last | 14:05.09 |
| Robin_Watts | Ok. So are you just taking our example android viewer and hacking on that? | 14:05.13 |
| | Or are you writing at the C API level ? | 14:05.26 |
| mobjax | i built the entire library using Cigwin and all | 14:05.38 |
| | and am able to render the PDF | 14:05.43 |
| | on Android | 14:05.48 |
| | but it gets displayed in the middle of the screen | 14:05.59 |
| Robin_Watts | But are you coding in C or in java? | 14:06.06 |
| mobjax | in Java | 14:06.18 |
| | What i want is, i have a layout and I want to show the PDF inside it | 14:06.39 |
| | and it should fill the layout | 14:06.47 |
| Robin_Watts | Ok. The published API for the MuPDF is the C level one. | 14:06.47 |
| mobjax | that is correct | 14:06.57 |
| | but by using CigWyn and Android NDK | 14:07.09 |
| | i converted the native one to a library and using it inside my Android project | 14:07.27 |
| Robin_Watts | In order to make the C level API accessible from java, we have had to wrap it in some JNI code and expose that through various java classes. | 14:07.59 |
| mobjax | yes | 14:08.07 |
| | that is what i did | 14:08.11 |
| | and I am able to do that | 14:08.22 |
| Robin_Watts | When I first did the Android build, I didn't expose all the entrypoints. I wrapped up just enough for what I needed. | 14:08.33 |
| mobjax | but i am stuck on, scaling the PDF to the screen | 14:08.35 |
| | ohh | 14:08.47 |
| Robin_Watts | And paulgardiner has now taken over from me on the android app. He's exposed some more stuff, but it's not a general purpose solution. | 14:09.16 |
| | Consequently, the android app as it is now has been built specifically to cope with displaying things in the way we need for our app; pages are spread out horizontally, and you can flip through them. | 14:09.56 |
| | If you're trying to make MuPDF layout it's rendered pages differently that will require some significant changes to the java. | 14:10.23 |
| mobjax | yes | 14:11.10 |
| | rendering I am able to achieve, horizontal thought | 14:11.23 |
| | but i want the pages to be fit the screen resolution | 14:11.35 |
| Robin_Watts | I have been working on exposing the MuPDF C API through java in a more natural way - a way which will allow java programmers to open files/load pages/draw pages with full control of size/position/scale etc. | 14:11.40 |
| mobjax | ohh | 14:11.53 |
| | can you share some thoughts | 14:12.00 |
| | so that we also can do the same here | 14:12.06 |
| Robin_Watts | While this code is close to the point where someone could try it out, it will most probably not work first time, and will need some tweaking as we get it working. | 14:12.19 |
| | The latest version of this work is here: http://git.ghostscript.com/?p=user/robin/mupdf.git;a=shortlog;h=refs/heads/jni | 14:12.55 |
| | but I can't offer you a trivial "if you change this bit here, you can set the scale for the pages" answer. | 14:13.48 |
| mobjax | ohh ok | 14:14.50 |
| | u want me to try this now? | 14:14.57 |
| Robin_Watts | If you come back on monday morning, you may be able to speak to paulgardiner, who is more up on the java side of things than me. | 14:15.24 |
| | Have you read platform/android/ClassStructure.txt ? | 14:15.56 |
| mobjax | ok | 14:16.17 |
| | I will try with the link that you provided | 14:16.26 |
| | if I am unable to crack it, I will speak to Paul on Monday as you suggested, | 14:16.49 |
| | yes, | 14:16.53 |
| | after reading it only, I understood that I need to change some values in MuPDFCore.Java | 14:17.17 |
| | to achive the full sized rendering of the PDF | 14:17.27 |
| Robin_Watts | ok. Making use of the code at the link I provided will require you to rework a significant amount of work. | 14:17.34 |
| | mobjax: Are you calling MuPDFCore yourself ? | 14:17.44 |
| mobjax | coz it is in native | 14:17.50 |
| | no, | 14:17.52 |
| Robin_Watts | or are you using the existing class structure? | 14:17.58 |
| mobjax | it is done automatically whenever I select any PDF | 14:18.06 |
| Robin_Watts | So you're still using MuPDFReaderView etc ? | 14:18.29 |
| mobjax | yes | 14:19.05 |
| Robin_Watts | so you don't want to fiddle with MuPDFCore itself. You want to fiddle with how the classes built upon it call it. | 14:22.35 |
| mobjax | yes | 14:32.49 |
| | that would be more easy for me | 14:32.57 |
| Robin_Watts | Right, so it's paul you need to talk to. | 14:33.08 |
| mobjax | ohh ok | 14:33.14 |
| | then I will wait till monday and talk to him | 14:33.22 |
| | thanks for your time and effort | 14:33.33 |
| | btw, where are u from > | 14:33.38 |
| Robin_Watts | Both Paul and I are in the UK. | 14:33.47 |
| mobjax | ohh | 14:34.15 |
| | I am from India | 14:34.18 |
| ray_laptop | Noldorin: I'm here if you still have PS questions | 17:44.14 |
| Noldorin | ray_laptop, thank you. as it happens i've solved it now though, and taught myself the rudiments of postscript in the process. :) | 17:44.52 |
| | i feel comfortable around PS now at least | 17:44.57 |
| ray_laptop | Noldorin: great. Always nice to have more PS hackers in the mix | 17:45.33 |
| Noldorin | :) | 17:45.45 |
| ray_laptop | I saw you were looking for string parsing. search anchorsearch and token are the handy ones. Ghostscript also has concatstrings that was missing from PS | 17:47.21 |
| | Also, for debugging, 'pstack' is your friend | 17:48.22 |
| Noldorin | ray_laptop, ah interesting. so token is like an 'eval' for PS? | 18:10.29 |
| ray_laptop | Noldorin: token just gets the next delimitted element. It doesn't 'exec' it | 18:11.51 |
| Noldorin | ray_laptop, element? doesn't it get all the tokens in a given string? | 18:12.13 |
| ray_laptop | Noldorin: and it isn't flexible as far as token delmitters -- it does match PS rules: whitespace, of course, but also ( and { and other things are self delimiting tokens). | 18:13.30 |
| | Noldorin: token only returns the next token (actually it returns a boolean to tell you whether or not it got a token, then the token, then the rest of the string) | 18:15.13 |
| Noldorin | ray_laptop, ah k. got it | 18:15.26 |
| ray_laptop | so only one token at a time | 18:15.27 |
| Noldorin | ray_laptop, and you execute it just by popping it off the stack? | 18:16.08 |
| ray_laptop | Noldorin: no, if you pop it, you discard it. To execute it, you use 'exec' | 18:16.43 |
| Noldorin | ray_laptop, oh, so it's a procedure huh? | 18:16.55 |
| ray_laptop | Noldorin: no, you aren't guaranteed that it's executable. Most often it's a literal, and exec of a literal just pushes itself on the stack | 18:18.28 |
| Noldorin | ray_laptop, but what if it's a normal token...? | 18:19.43 |
| ray_laptop | Noldorin: don't know what you mean by 'normal token' | 18:20.08 |
| Noldorin | well, a normal token | 18:20.15 |
| | like exch | 18:20.18 |
| | pop | 18:20.19 |
| | roll | 18:20.21 |
| | etc | 18:20.23 |
| | not a literal or number :P | 18:20.26 |
| | or procedure | 18:20.30 |
| ray_laptop | Noldorin: those are operators, and no different to procedure names | 18:21.09 |
| Noldorin | k | 18:21.15 |
| ray_laptop | if it isn't a literal, it is looks it up in the dictstack and if it finds it, it will exec it. | 18:22.02 |
| | otherwise 'undefined' error | 18:22.42 |
| | Noldorin: sorry -- phone call... bbiab | 18:23.32 |
| | Forward 1 day (to 2014/01/12)>>> | |
IRC Logs