| | <<<Back 1 day (to 2017/11/14) | 20171115 |
| f-a | Hello, I am ps2pdf a quite large (~300 pages, abc-notation typeset music) ps file | 16:37.54 |
| | I get an /stackoverflow in --%ztokenexec_continue-- | 16:37.57 |
| kens | Sounds like your PostScript is broken then | 16:38.16 |
| | Its leaving stuff on the operand stack, and eventually it fills up, leading to an error | 16:38.34 |
| f-a | thanks kens , I thught maybe making the stack bigger would have make it go away | 16:39.02 |
| kens | Well, it might work around the problem, but it wouldn't make the problem go away | 16:39.21 |
| | A larger file would presumably still trigger the problem | 16:39.34 |
| | Of course, this is all guesswork when we can't see the file | 16:39.44 |
| chrisl | Also, the the more items on the stack, likely the slower it will get | 16:40.27 |
| f-a | I can of course show, it's just quite big, (80k lines). Let me upload it | 16:40.30 |
| | http://www.ariis.it/link/t/Out.ps | 16:41.12 |
| kens | According to the PLRM typical implementation limits for a PostScript interpreter have the operand stack at 500 elements | 16:41.20 |
| | Well I get a stackoverflow only a few dozen pages in | 16:42.13 |
| | Assuming that the program is leaving stuff on the stack, and its a per-page problem, then you won't get to the end of the job unless you make the aoperand stack *really* big | 16:42.50 |
| f-a | mhhh, that's interesting. I"ll drop a message to the maintainer | 16:43.31 |
| kens | I haven't really started looking yet, give me a minute | 16:43.49 |
| | I need to reduce the file | 16:43.55 |
| f-a | ok | 16:44.05 |
| kens | Yep, each page leaves 2 strings with the content (eintro) in it on the operand stack. | 16:45.02 |
| f-a | thanks for diagnosing! | 16:45.19 |
| kens | NP, it does seem to be a problem with the program | 16:45.38 |
| f-a | yeah I'll write abcm2ps author a line now | 16:45.56 |
| kens | I notice that the end of each page has: | 16:46.01 |
| | (sintro)349.9 -65.7 st | 16:46.01 |
| | (eintro)682.1 -65.7 nd | 16:46.01 |
| | which is probably where the problem arises. | 16:46.09 |
| | Yes, the /st and /nd functions don't take any operands, which means they leave the string argument untouched on the stack | 16:46.44 |
| | That seems too be the source of the error | 16:47.01 |
| | Oh, tghe program defines those functions on each page :-( | 16:47.57 |
| f-a | yeah I am checking it | 16:48.08 |
| | awful | 16:48.11 |
| kens | If I fix them by adding a 'pop' at the end of the function I can sort each one, but it needs to be sorted properly in the producer, sorry | 16:48.29 |
| f-a | ok, I guess the tunes were standalone files. thanks for your extremely quick support kens & chrisl | 16:51.56 |
| kens | NP good lcuk with getting it fixed. | 16:52.09 |
| | If you can't, then a global search and replace on /st and /nd would work | 16:52.29 |
| | Just put a pop at the end of each function | 16:52.43 |
| f-a | well just so you know (maybe someone else will ask): | 16:53.13 |
| | abcm2ps is a tool that converts text files to music | 16:53.29 |
| | and the directive which is most likely causing this has a description that starts with: | 16:53.51 |
| | This command assumes a good knowledge of the abcm2ps/abc2svg internals. [...] | 16:53.57 |
| | so the author of abcm2ps is likely not at fault, but the author of the tunes | 16:54.20 |
| kens | O.O assuming a good knowledge of the internals seems like a big assumption :-) | 16:54.29 |
| f-a | yyyyyyyyyyyyyep ;P | 16:54.55 |
| kens | I don't think its the tunes, it looks to me like the abcm2ps code that has created the functiona, and used it | 16:54.57 |
| | I would imagine for small files this is not a problem | 16:55.18 |
| | smal as in not many pages | 16:55.25 |
| | Forward 1 day (to 2017/11/16)>>> | |
Log of #ghostscript at irc.freenode.net.