| <<<Back 1 day (to 2016/10/05) | 20161006 |
norbertj | kens: chrisl: so remove the -dNOPAUSE in ps_impl_allocate_interp_instance() and adding XCFLAGS=-DPSI_INCLUDED (to use ps_end_page_top (which handles the keypress) i.s.o. gs_output_page. This works for me. Thanks. | 05:51.47 |
| HenryStiles: it's just for me to have an easy all-in-one pdl on my desktop on which I can drop a ps/pdf/xps/pcl testfile to have a quick look. Before this I always had to tweak the makefiles (i.e. adding the xps to the ls-build) This is less work for me ;) And I use this as a base to determine what I need (removing devs from device_devs/feature_devs that I don't need) to build the pdl-core for our controller. | 05:58.18 |
kens | norbertj *either* of those approaches will 'work' but using XCFLAGS should result in ps_end_page_top() being called which is probably the more useful method. | 07:42.40 |
| You shouldn't need to do both | 07:43.30 |
deekej | hello guys! :) Is there any significant API/ABI change between ghostscript versions 9.16 -> 9.20? | 10:48.53 |
kens | No | 10:50.44 |
deekej | thank :) | 10:51.38 |
| *thansk | 10:51.43 |
kens | The Ghostscritp API has been essentially unchanged for years, possibly decades. That will be changing soon though | 10:51.48 |
deekej | you plan to do it for 9.21? | 10:52.09 |
kens | at some point soon | 10:52.19 |
deekej | ok, thanks for the info :) | 10:52.27 |
kens | And when we do it won't be a minor point change | 10:52.28 |
deekej | good to know | 10:52.46 |
sirowain | hi everyone | 13:28.32 |
| I need help with mupdf, I basically need to add an image to an existing pdf and save it, no need to render it | 13:30.27 |
kens | sirowain : looks like you're going to have to wait a little for help, the relevant people are at lunch or something. | 13:36.32 |
sirowain | kens : thanks | 13:44.49 |
deekej | do you plan to release new version of ghostscript-9.20, after yesterday's fixes? :) | 13:44.58 |
chrisl_x250 | deekej: No | 13:45.30 |
deekej | thanks :) | 13:45.49 |
chrisl_x250 | Besides, how could we do a "new" version of 9.20? It wouldn't be 9.20, then..... | 13:46.17 |
tor8 | sirowain: do you need to add the image on top of an existing page? | 13:47.16 |
deekej | well, I was thinking something like 9.20.1 or something like that :) | 13:47.48 |
sirowain | yes, i need to 'cover' an portion of page with an image | 13:47.50 |
kens | deekej : we only do releases at 6 month intervals, barring disasters | 13:49.07 |
chrisl_x250 | deekej: No, if we did that, it would be 9.21. But the "exploits" in question are very low likelihood, and have been there for better than 15 years without causing a problem - not worth rushing out a new version for them | 13:49.24 |
deekej | ah, ok :) | 13:49.43 |
| every upstream has it different, so I'm just trying to remember all the differences between their release processes :) | 13:50.33 |
chrisl_x250 | deekej: the thing is, most distros grab patches from our repo as and when it suits them, so....... | 13:51.35 |
deekej | yes, that's exactly what I will do :) | 13:51.51 |
tor8 | sirowain: you can use 'mutool run' to edit a pdf | 13:52.24 |
| including adding images | 13:52.26 |
| gimme a minute and I'll cook up an example script | 13:52.34 |
sirowain | tor8: thanks but I've not been very clear, I'm doing this in a c program | 13:53.54 |
| I can't use mutool | 13:54.32 |
tor8 | then you can use the same approach | 13:54.57 |
| load the image using fz_new_image_from_file | 13:56.52 |
| create a PDF object from the image using pdf_add_image | 13:57.01 |
| add the PDF object as a named resource in the page's resource dictionary (pdf_dict_put in the page /Resources /XObject dictionary) | 13:57.41 |
| add a drawing command using the named resource to the content stream | 13:57.51 |
| using pdf_load_stream(), fz_buffer_printf(), pdf_update_stream() | 13:59.08 |
| and finally save the file using pdf_save_document() | 13:59.22 |
sirowain | tor8: thanks, I'll give it a try, I missed something | 14:01.22 |
tor8 | sirowain: http://pastebin.com/raw/k3gieyAT has a mutool run script you can look at for reference | 14:02.21 |
| it should not be difficult to translate that to the equivalent C | 14:02.40 |
sirowain | tor8: again, thank you very much | 14:03.09 |
Robin_Watts | tor8: Is js_error tagged as no return? | 14:38.13 |
| looks to be. | 14:38.49 |
tor8 | it should be | 14:38.59 |
Robin_Watts | http://marcos.ghostscript.com:8080/artifex/mupdf/6870dfecd94c1dec0e7970464b11d3b9d5d1250a/mupdf-cppcheck.txt throws up stuff in murun.c | 14:39.04 |
tor8 | yes, they're all tagged as noreturn | 14:39.22 |
| does cppcheck not set __GNUC__? | 14:40.35 |
Robin_Watts | not a clue, sorry. | 14:40.49 |
tor8 | that's the only reason I can think of why it wouldn't pick up noreturn | 14:41.03 |
| but then it ought to be confused by fz_throw as well | 14:41.29 |
sirowain | tor8: I'm trying to use pdf_add_image but I get an error (EXC_BAD_ACCESS on iOS) | 15:10.24 |
| here's my code: http://pastebin.com/eS8gezLv | 15:10.38 |
| the fz_image is not null | 15:11.02 |
tor8 | sirowain: I don't see any obvious errors | 15:39.07 |
| sirowain: what type is the 'page' variable? | 15:39.41 |
| if it's a fz_page/pdf_page then that could be where your error is. it needs to be the pdf_page->obj object | 15:40.09 |
| pdf_lookup_page_obj if you have the page number | 15:40.32 |
| sirowain: you can use the pdf_page_resources() and pdf_page_contents() functions | 15:41.51 |
| and pdf_update_stream needs the pdf_page_contents object, not the page object | 15:42.26 |
mvrhel_laptop | tor8: since you are nont feeling well I will not bother you with the pdf create questions now | 15:42.27 |
| s/nont/not/ | 15:42.33 |
tor8 | mvrhel_laptop: hopefully I will be up to the task by monday, if it can wait that long? | 15:42.45 |
mvrhel_laptop | tor8: easily | 15:42.58 |
| it is not pressing | 15:43.02 |
tor8 | mvrhel_laptop: I may not give coherent answers if you ask me today :) | 15:43.06 |
mvrhel_laptop | Yes. Get some rest etc. | 15:43.27 |
sirowain | tor8: thanks, I'm figuring out why I don't have pdf_page_resources() and pdf_page_contents() functions | 15:56.28 |
tor8 | sirowain: you might want to upgrade to the latest git version | 16:07.11 |
| the pdf_page_contents function was added in july | 16:07.33 |
| you should also check that the XObject dictionary exists, and create it if it doesn't, before adding to it | 16:08.16 |
sirowain | tor8: I found that pdf_add_image goes wrong because of the call pdf_insert_resource(ctx, doc->resources->image, digest, imref), in fact my doc->resources is null | 16:17.25 |
tor8 | sirowain: hm, that bug *should* be fixed. I remember approving the bug fix for it... | 17:09.09 |
| Forward 1 day (to 2016/10/07)>>> | |