| <<<Back 1 day (to 2013/07/30) | 2013/07/31 |
chrisl | kens: is there anything on your desk that might influence the 9.08 release? | 07:18.35 |
kens | chrisl I go the majority of the colour stuff in, but we won't be turning it on because it has some funny colour conversions, and we casn't do guraranteed V2 ICC profiles yet. SO I guess the answer is no, but I haven't had time to go through all my emial yet, or look at he 4 new bugs that have materialised :-( | 07:19.42 |
chrisl | kens: there didn't *seem* any serious problems for you over the last few days. I'll do the release branch, and associated stuff, so marcosw1 can get rolling on the testing - if you do stumble on something, we can call a halt..... | 07:21.34 |
kens | O | 07:21.46 |
veli_ | hey guys how are you | 07:30.45 |
ghostbot | veli_: just great | 07:30.45 |
veli_ | i have an issue bout mupdf | 07:30.57 |
| installing mupdf on ios | 07:31.38 |
chrisl | veli_: the mupdf devs probably won't be around for another hour or two | 07:36.51 |
veli_ | ok then thanks | 07:44.50 |
| any mupdf developer there? | 07:56.04 |
tor7 | veli_: morning. | 07:56.52 |
veli_ | good morning | 07:59.48 |
| tor7: i have an issue about building mupdf on ios, can you help? | 08:01.34 |
tor7 | veli_: maybe, you'll have to tell me what issues you have before I can answer that question. | 08:02.08 |
veli_ | ok then, first i installed and used mupdf library on my android device and working just fine | 08:03.16 |
| but i dont have any clue how to build in ios environment | 08:03.39 |
| i build mupdf source 1.2 with my android ndk and cygwin but, dont know how to build for ios | 08:04.36 |
tor7 | do you intend to run the ios code on an actual device, or just the simulator? | 08:07.39 |
| to build the iOS version of mupdf you have to use Xcode and the Xcode project in "platform/ios/" | 08:08.19 |
| in order to run the code on a device, you need to pay the apple developer tax first. | 08:09.58 |
veli_ | i will use it on simulator, and want to see if runs without problem | 08:10.23 |
| i downloaded mupdf 1.2 source, in ios folder opened project and run, but gave me error and cannot find solution | 08:12.15 |
| like in android version, do i need apps like cygwin or android-ndk | 08:13.17 |
chrisl | Off for the rest of the day - back tomorrow morning...... | 08:25.55 |
tor7 | veli_: well, if you don't tell us about the error specifics, we can't help you. | 08:26.20 |
| a common error when building iOS is not having installed the "command line tools" SDK pack or what they call it | 08:27.08 |
| it's hidden in a dialog in Xcode itself | 08:27.15 |
veli_ | the error is | 08:28.05 |
| </bin/sh: ./build/debug-ios-i386/cmapdump: No such file or directory< | 08:28.15 |
tor7 | and the lines above it? don't they say something about running "make generate" in the mupdf directory? | 08:28.47 |
veli_ | Run the 'make generate' command manually from now on! If you see an error while running GEN, you've forgotten. | 08:29.37 |
| it sayz | 08:29.53 |
tor7 | open the Terminal, cd to where you have the mupdf directory, and run "make generate" | 08:30.26 |
veli_ | ok trying | 08:30.34 |
veli1 | tor7, thanks a lot it is working now | 08:50.43 |
tor7 | veli1: fab! | 09:44.49 |
veli1 | tor7: i have a question about moving builded file | 10:04.34 |
| when i moved compiled mupdf folder, it says library missing | 10:05.17 |
tor7 | with the iOS simulator? | 10:05.57 |
veli1 | yep, and i cannot rename or modify project name | 10:07.14 |
vtorri | tor7, hey | 10:13.09 |
tor7 | veli1: then I'm sorry, I have no idea what that problem could be. | 10:13.42 |
| vtorri: hi. | 10:13.48 |
vtorri | tor7, i'm asking again a question i asked some time ago | 10:14.07 |
| when i try to open a png with mupdf 1.2 (the lib), mupdf still tries to open it | 10:14.36 |
| error: cannot recognize version marker | 10:14.55 |
| warning: trying to repair broken xref | 10:14.55 |
| etc... | 10:14.57 |
| but i would like to have an error instead of mupdf trying to open it | 10:15.14 |
| is it possible ? | 10:15.22 |
veli1 | it does not try to open it, because in ChoosePDFActivity.java, there is a property named accept, and in it it controls whether file is in corret format | 10:20.09 |
vtorri | veli1, it's for me ? | 10:20.38 |
veli1 | yep | 10:20.59 |
vtorri | well, as i don't use java... | 10:21.08 |
veli1 | what is platform? | 10:21.18 |
vtorri | it's not a platform prblem | 10:21.30 |
| it's a mupdf (the lib) feature | 10:21.40 |
tor7 | vtorri: I get no such error... is this something that has been fixed since 1.2? | 10:38.12 |
vtorri | i've not tried mupdf > 1.2 yet | 10:39.00 |
tor7 | right, so 1.2 from git seems to have that problem | 10:40.14 |
| vtorri: commit 09bb1dc adds support for opening .png files directly | 10:42.38 |
| any files we don't recognise the file type by filename extension, we try to open as PDF. that's intended behaviour. | 10:43.14 |
vtorri | tor7, ok, but maybe the same problem will be here for another image file... | 10:43.15 |
| but the problem is that i get an awful output | 10:43.49 |
tor7 | that commit is from May 3, so after release 1.2 | 10:44.05 |
vtorri | with no error | 10:44.07 |
tor7 | vtorri: mupdf: error: cannot open document | 10:44.28 |
| you get an error. | 10:44.31 |
vtorri | imho, you should provide a way to return an error for a wrong format | 10:44.36 |
| tor7, can I redirect mupdf output ? | 10:45.08 |
tor7 | well, "wrong format" is not so easy to detect. you have no idea how many broken PDF files that show up that we still have to make a best-attempt to handle. | 10:45.47 |
vtorri | the output is actually a problem (terrible 'biiip' from the computer) | 10:45.50 |
tor7 | files with missing headers, etc | 10:45.54 |
vtorri | i can understand that | 10:46.01 |
tor7 | vtorri: the console beep is probably from the warning: ignoring invalid character in hex string: '%c' errors. we've trimmed that warning down so it shouldn't beep anymore. | 10:46.38 |
vtorri | indeed, that's the error in the console | 10:47.04 |
| but anyway, can i redirect the output ? | 10:47.20 |
tor7 | I thought you were going to write a patch to redirect errors using the API? | 10:47.23 |
vtorri | well | 10:47.31 |
| Robin tried, iirc | 10:47.36 |
tor7 | mupdf foo.pdf >log.txt 2>&1 | 10:47.41 |
vtorri | haha | 10:47.46 |
| i use the library :) | 10:47.53 |
| nice try :) | 10:47.59 |
| maybe i should update to mupdf 1.3 | 10:48.37 |
tor7 | vtorri: you can get the error message from a try/catch using the fz_caught_message function | 10:48.39 |
vtorri | ok | 10:48.46 |
| i'll try that | 10:48.55 |
| thank you | 10:48.57 |
tor7 | but do upgrade (and give the release candidate a shake) | 10:49.18 |
vtorri | :) | 10:51.02 |
| ho, it's just an RC | 10:52.08 |
| ok | 10:52.09 |
MrSMAK | Hello Every one | 11:07.07 |
| Guys, has anyone implemented MuPDF on Android ? | 11:09.52 |
| Hello ! | 11:15.21 |
| any one here ? | 11:15.23 |
tor7 | MrSMAK: yes. | 11:15.54 |
paulgardiner | MrSMAK: Either tor7 or I may be able to help. | 11:33.17 |
ManDay | I've been sent a PDF which was "annotated" (whatever that means). Oepning it in mupdf I see little yellow, icon-ish boxes with speechbubbles - do you have any idea how to see the annotations? | 15:06.37 |
tor7 | ManDay: I believe that's one for paulgardiner. the speechbubbles are probably the kind of annotation that you need to click to toggle. | 15:09.35 |
paulgardiner | ManDay: We don't support display of those yet, but it is on our list. | 15:10.36 |
ManDay | paulgardiner: any idea how I can figure out the content? | 15:12.12 |
paulgardiner | Each of those speechbubbles will correspond to a text annotation in the document. The annotation will have a "Contents" field containing text that should be displayed when the speechbubble is clicked on | 15:12.24 |
ManDay | is the bubble-thing part of the pdf or is it rendered by mupdf? | 15:12.41 |
paulgardiner | It's part of the PDF. It's the appearance stream associated with the annotation. | 15:13.14 |
ManDay | Okay. Is there any way I can find out what the annotation is? Some regex in the pdf's file, perhaps? | 15:14.50 |
tor7 | ManDay: if you just want the "Contents" field, you could get with with "mutool show" and some digging | 15:17.28 |
| "mutool show p" to list the object numbers of each page | 15:17.49 |
| then you can start following object numbers until you get to the annotation contents | 15:18.07 |
paulgardiner | ManDay: Do you need to be able to do this systematically and repeatedly, or just this once for this particular file? | 15:19.15 |
ManDay | paulgardiner: just once, I guess | 15:19.26 |
| I never needed annotations but this file I've been sent I have to examine | 15:19.38 |
| tor7: cool, I even have mutool, must be part of mupdf - I'll try that | 15:19.57 |
paulgardiner | Then mutool clean -d might be another option. That will give a file that you an load into a text editor then search for /Contents. You'd need to then guess which lumps of text went with which visible speechbubble. | 15:21.02 |
ManDay | When I search for /someVisibleString in the out.pdf of `mutool clean -d` I get no hits | 15:30.25 |
| In fact, the out.pdf looks anything but clean. | 15:31.03 |
| tor7: Could you please give me an example of how to use mutool show | 15:33.04 |
paulgardiner | Can you find occurrences of the string "Contents"? ... although tor7's approach may work better for you. | 15:33.29 |
ManDay | paulgardiner: yes I can | 15:38.23 |
paulgardiner | Do you see something like /Contents 343 0 obj ? | 15:38.58 |
| where 343 will be some other number probably | 15:39.10 |
| ManDay: If so then "mutool show file.pdf 343" should display the text you are looking for (where obviously you replace 343 by the actual number you find in the doc). | 15:42.33 |
ManDay | paulgardiner: things like /Contents 128 0 R | 15:43.14 |
sebras | ManDay: yes. | 15:43.25 |
| paul made a typo, it should be R not obj. | 15:43.42 |
paulgardiner | Sorry, yes that's right. Not exactly what I said | 15:43.44 |
ManDay | paulgardiner: There are thousands of /Contents .... entries. Do you suggest I try them all out? | 15:44.05 |
| My bad, not thousands | 15:45.21 |
paulgardiner | ManDay: that's not so good if there are thousands. This could be difficult to do by had then. There is a way to determine which ones, but it could be fiddly | 15:45.35 |
ManDay | You couldnt, by any chance, elaborate on what tor7 suggested? | 15:46.05 |
| (If that's likely to help me) | 15:46.16 |
paulgardiner | ManDay: I'm not exactly sure what tor7 was suggesting, but if it is what I suspect then it wont be any quicker. | 15:47.18 |
| Each of the /Content 128 0 R type things will apear in brackets << >> | 15:47.48 |
| The ones you want are those that also have /Subtype /Text in them | 15:48.06 |
| And /Type /Annot | 15:48.16 |
| sebras: please correct me if I've got any more of this wrong. | 15:48.37 |
ManDay | I tried a few "mutool show file.pdf <number>" but they all give me cryptic output tailed by "endstream endobj" | 15:48.52 |
| there is no text whatsoever | 15:48.58 |
| Well, I find "/Type /Annot" pdf's source, though | 15:50.28 |
paulgardiner | And /Subtype /Text between the same << >> brackets? | 15:51.38 |
ManDay | Yeah, I think I found some content there! | 15:51.50 |
| Indeed | 15:52.19 |
| thank you paulgardiner! | 15:52.22 |
paulgardiner | It may be just of the form /Contents (The actual text) | 15:52.31 |
ManDay | It is | 15:52.36 |
paulgardiner | Brilliant. | 15:52.41 |
ManDay | Now I just need to figure out which annotation is where :] | 15:52.45 |
paulgardiner | :-) Yeah quite | 15:52.53 |
tor7 | ManDay: if you run "mutool show file.pdf p" you'll get a list of pages and the object number for each page | 15:53.24 |
paulgardiner | We'll probably add proper support for this soon, so it should be easier next time | 15:53.33 |
tor7 | then run "mutool show file.pdf <number>" to display the page object for a page | 15:53.44 |
| in there you'll see some entries like /AcroForm X 0 R | 15:53.54 |
| mutuul show file.pdf X | 15:54.02 |
| etc | 15:54.05 |
| until you find the /Contents | 15:54.09 |
ManDay | tor7: Ah, "p" wasn't documented , it appears | 15:54.15 |
tor7 | you can give multiple numbers on a command line | 15:54.21 |
| "p" short for "pages" | 15:54.25 |
ManDay | oh, ok | 15:54.36 |
tor7 | if you give no arguments, I think it just prints the "xref" table which lists all objects and their location in the file | 15:54.57 |
paulgardiner | tor7: these may possibly not appear in the /AcroForm hierarchy, not being form widgets. | 15:55.04 |
tor7 | paulgardiner: right. could be some other key, like /Annots? | 15:55.32 |
paulgardiner | Yeah, I reckon | 15:55.43 |
sebras | tor7: mutuul?! your new keyboard isn't perchance kens' old one? ;) | 16:24.36 |
kens | LOL | 16:24.46 |
sebras | tor7: is /Annots for the old type of annotations? | 16:24.50 |
| well, forms really. | 16:25.04 |
tor7 | sebras: I was just pulling /Keys out of a hat there | 16:25.24 |
kens | OK goodnight all | 16:47.45 |
dogisfat_ | What is the preferred way to allocate a 2d array in GS? e.g. int **array_2d. | 16:47.46 |
dogisfat | Or maybe the better question is how should I allocate the array I am going to use to store image information for libpng? | 16:50.51 |
ray_laptop | Question for anyone: When NumCopies != 1 and we are saving pages for rendering later, should we ignore the NumCopies on individual pages, and just use this as the copy count for the subsequent rendering/printing of the job ? | 17:44.48 |
| or should we use the copy count for each page within the jiob (unless .IgnoreNumCopies=true), then also allow the job "copies" to produce multiple collated copies of the job, potentially including multiple copies of some pages | 17:46.36 |
| which is most likely to produce the least 'surprise' or confusion. | 17:47.24 |
| ? | 17:50.31 |
marcosw1 | henrys: I've assigned a P1 bug to you since Robin is on holiday: <http://bugs.ghostscript.com/show_bug.cgi?id=694451>. | 18:49.27 |
vtorri | ray_laptop, hey | 19:18.10 |
| ray_laptop, do you know if the build system of mupdf 1.3 will change before the release ? | 19:18.35 |
| 1.3 RC | 19:18.48 |
ray_laptop | vtorri: Sorry. I was away. I don't know -- you probably need to ask tor when he shows up | 19:46.53 |
rodfantana | i am trying to convert a PDF to a multipage TIFF using gs9.0.7... PDF contains pages of different sizes/resolution. When GS creates the tiff file, it keeps resolution of all pages constant. Is there any setting that will keep the orignal resolution of the PDF page? | 21:27.01 |
| anyone? | 21:35.30 |
ray_laptop | rodfantana: PDF doesn't have "different resolutions" PDF coordinates are initially at 72 units per inch, per the PDF spec. If what you want is the images from a PDF to be extracted and converted to TIFF, then gs doesn't do that (it can, but it involves writing PS to do it) | 21:56.51 |
| you may want to try 'mutool extract'. I don't use it, so I don't know offhand what format images are written into. | 22:02.21 |
sebras | vtorri: what part of the buildsysten are you depending on? | 22:56.28 |
| Forward 1 day (to 2013/08/01)>>> | |