2018-04-24 GnuCash IRC logs
00:08:13 *** ptoal1 has quit IRC
00:10:43 *** MPL has joined #gnucash
00:11:14 *** ptoal1 has joined #gnucash
00:22:21 *** ptoal1 has quit IRC
00:32:18 *** ptoal1 has joined #gnucash
00:38:15 *** MPL has quit IRC
00:41:24 *** ptoal1 has quit IRC
00:46:59 *** ptoal1 has joined #gnucash
01:05:18 *** MPL has joined #gnucash
01:11:05 *** marusich has quit IRC
01:21:34 *** harshitaneja has joined #gnucash
01:23:10 *** Mechtilde has joined #gnucash
01:28:08 *** fell has quit IRC
01:29:16 *** fell has joined #gnucash
01:35:23 *** harshitaneja has quit IRC
01:36:42 *** harshitaneja has joined #gnucash
01:39:58 *** karelk has joined #gnucash
01:42:32 <karelk> my scheduled transactions are not executed while gnucash is running
01:43:00 <karelk> I have to close Gnucash
01:43:25 <karelk> and when I start Gnucash, then my scheduled transactions are executed
01:43:30 <karelk> is it expected behavior ?
01:47:28 *** MPL has quit IRC
01:51:57 *** MPL has joined #gnucash
01:56:19 *** MPL has quit IRC
02:02:13 <warlord> karelk: Actions -> Scheduled Transactions -> Since Last Run
02:02:41 <warlord> That only gets executed when gnucash starts; otherwise you need to manually run it.
02:03:11 *** gour has joined #gnucash
02:03:59 <karelk> OK, is this a bug or a feature ?
02:04:09 <karelk> or is this a design limitation ?
02:06:04 <warlord> Well, it is behaving as designed, so technically not a bug. But it's a reasonable feature request to ask gnucash to recheck periodically.
02:06:59 *** MPL has joined #gnucash
02:18:38 *** Mechtilde has quit IRC
02:42:35 *** gjanssens has joined #gnucash
02:42:35 *** ChanServ sets mode: +o gjanssens
02:42:40 <gjanssens> .
02:44:16 <warlord> .
02:45:11 <warlord> kkknewton: You can operate it at different times, so long as you share the data file between users and they completely exit the program between them. GnuCash does not support simultaneous operation.
02:48:44 *** fekepp has quit IRC
02:53:58 *** fabior has joined #gnucash
02:57:28 <gjanssens> .
02:58:28 <warlord> hiya gjanssens
03:02:31 <gjanssens> warlord: hi
03:02:37 <gjanssens> It seems you're in Europe now ?
03:02:53 <warlord> Yes. Currently in Sophia Antipolis, FR
03:04:13 <warlord> Speaking of which -- gotta run now. BBL.
03:04:30 *** warlord has quit IRC
03:05:41 *** Jacques has joined #gnucash
03:10:54 *** Jacques has quit IRC
03:28:11 <kkknewton> Warlord: can you use 2 separate files in separate locations? For Example: we have volunteer to clean up a nonprofit company"s account books that go back three years. So we thought we could divide the expenses into 3 piles representing 1 of 3 years. We then use 3 bookkeepers to do each year's posting.
03:31:55 *** warlord has joined #gnucash
03:31:56 <kkknewton> Can we import these files into 1 file.? Or are we asking too much?
03:33:36 <kkknewton> By the way your help is greatly appreciate thank you
03:39:05 <warlord> kkknewton: it is generally hard to merge two gnucash data files together.
03:40:07 <warlord> if it's a FORK (started with one file, then added transactions to multiple files, somehow), you can replay the log files -- provided it's only regular transactions. If you added accounts or perfromed any business features, then the logs wont replay that and you'll have to redo it by hand
04:18:12 *** Cuare has quit IRC
04:20:50 *** Cuare has joined #gnucash
04:29:14 *** tuxd00d has quit IRC
04:29:28 *** fabior has quit IRC
04:37:25 *** pilotauto_ has joined #gnucash
04:38:44 *** pilotauto has quit IRC
04:47:22 *** gour_ has joined #gnucash
04:49:45 *** gour has quit IRC
05:09:38 *** ncv__ has joined #gnucash
05:10:37 *** harshitaneja has quit IRC
05:11:37 *** harshitaneja has joined #gnucash
05:31:09 *** gour_ is now known as gour
05:33:20 *** fekepp has joined #gnucash
05:45:58 *** fekepp has quit IRC
05:48:53 *** fekepp has joined #gnucash
06:22:33 *** pilotauto_ has quit IRC
06:44:04 *** Cuare has quit IRC
06:49:55 *** fekepp has quit IRC
06:52:37 *** fekepp has joined #gnucash
06:56:06 *** Jimraehl1 has joined #gnucash
06:57:03 *** Jimraehl1 has left #gnucash
07:19:40 *** ggherdov_ has quit IRC
07:19:54 *** ggherdov_ has joined #gnucash
07:20:14 *** ggherdov_ has joined #gnucash
07:20:33 *** ggherdov_ has joined #gnucash
07:20:53 *** ggherdov_ has joined #gnucash
07:33:21 *** oozer has joined #gnucash
07:41:01 <warlord> gjanssens: when you get the chance, could you point me to a bug that has an attachment? Any valid bug will do.
07:45:30 <gjanssens> warlord: sure - https://bugzilla.gnome.org/show_bug.cgi?id=792833
07:46:05 <warlord> gjanssens: great. thanks!
07:48:28 *** fekepp has quit IRC
07:48:38 *** fekepp has joined #gnucash
07:49:22 <warlord> Hmm, looks like I lose attachment IDs
07:50:12 <warlord> Oh, nevermind -- it's in there.
07:50:27 <warlord> Thanks, gjanssens
07:51:42 *** fekepp has quit IRC
07:53:50 *** fekepp has joined #gnucash
07:58:28 *** fekepp has quit IRC
08:03:48 *** fekepp has joined #gnucash
08:04:45 *** boldstripe has joined #gnucash
08:06:22 *** fekepp1 has joined #gnucash
08:06:53 *** fekepp has quit IRC
08:08:37 <gjanssens> You're welcome
08:09:22 *** fekepp1 has quit IRC
08:12:04 *** fekepp has joined #gnucash
08:25:15 <boldstripe> Is there an option to disable file locking? Opening an existing file prompts the 'Overide Lock?' dialog even if there is no LCK file there. Overiding the lock works, but making changes, or making a new file, fails on save with another message about not being able to get the lock.
08:25:47 <boldstripe> Crucially, the file is on a FUSE mounted volume that does not implement locks (yet).
08:26:24 <boldstripe> I already looked in the Preferences of the app as well as Dconf settings, but could not see anything.
08:27:04 <gjanssens> boldstripe: I don't know of a way to disable locking, though I believe it only happens on xml based books.
08:27:25 <gjanssens> Perhaps you can try with an sqlite3 based book (which has its own caveats of course)
08:27:46 *** fekepp has quit IRC
08:38:01 <warlord> boldstripe: what is the underlying FS? The lock file doesn't really use flock() per se, but it does require the ability to create the lock file... If it's failing for you then more likely it's a permission issue, or your fuse filesystem is not POSIX-compliant enough.
08:41:29 <boldstripe> warlord: I am not sure. It is an experimental version of Cryptomator that mounts a FUSE volume from encrypted files. https://github.com/cryptomator/cryptomator/issues/669
08:43:21 <boldstripe> The volume can save files from an application like LibreOffice without issue.
08:49:49 *** fekepp has joined #gnucash
08:52:32 *** ptoal1 has quit IRC
08:58:50 *** fekepp has quit IRC
08:59:29 <boldstripe> Is there anything I can do to help in providing the Cryptomator developer with more information about what the lock file is and what permissions it needs?
09:02:55 *** kael has joined #gnucash
09:22:04 <gjanssens> boldstripe: the lockfile itself i being created here https://github.com/Gnucash/gnucash/blob/3.0/libgnucash/backend/xml/gnc-xml-backend.cpp#L637
09:22:40 <gjanssens> A little further down in the code it will also try to create a hard link to this lock file.
09:23:13 <gjanssens> If your filesystem reports it doesn't support hard links gnucash will just continue. If there's another error, it will warn.
09:26:34 <boldstripe> gjanssens: thanks, I will pass that on
09:32:08 *** ptoal1 has joined #gnucash
09:34:09 *** oozer has quit IRC
09:34:59 *** fabior has joined #gnucash
09:45:15 *** oozer has joined #gnucash
10:00:48 *** fabior has quit IRC
10:02:05 *** fabior has joined #gnucash
10:04:12 *** warlord has quit IRC
10:09:51 *** Mechtilde has joined #gnucash
10:10:53 *** kkknewton has quit IRC
10:14:27 *** Mechtilde has quit IRC
10:19:11 <chris> ANZAC Day tomorrow, respect o/
10:19:27 <chris> it always surprises me how much everyone is keen to push boundaries of gnucash
10:19:36 <chris> present company included :)
10:21:13 *** fekepp has joined #gnucash
10:26:56 *** warlord has joined #gnucash
10:42:14 *** ArtGravity has joined #gnucash
10:45:49 *** Mechtilde has joined #gnucash
10:49:29 *** gour_ has joined #gnucash
10:51:38 *** gour has quit IRC
10:57:31 *** Mechtilde has quit IRC
11:11:27 *** ptoal1 has quit IRC
11:38:27 *** fabior has quit IRC
11:43:04 *** O01eg has quit IRC
11:43:15 *** O01eg has joined #gnucash
11:46:58 *** MPL has quit IRC
11:54:28 *** karelk has quit IRC
11:55:24 <gjanssens> jralls: I'm slowly making progress on https://bugzilla.gnome.org/show_bug.cgi?id=795031 and https://bugzilla.gnome.org/show_bug.cgi?id=795040
11:56:06 <gjanssens> That's the good news. The bad news is that the core issue turns out to be the partial conversion from Timespec to time64 :(
11:56:44 <gjanssens> On linux this doesn't appear to be an issue (or perhaps on 64-bit ?), but on windows it can lead to a crash
11:59:50 <gjanssens> The issue is that all engine objects define qofaccess functions for their properties
12:00:22 <gjanssens> The qof access function for date properties expects a Timespec, but not all functions passed in return one.
12:01:36 <gjanssens> For example the gncInvoiceGetDate{Opened,Posted,Due} functions return a time64. This works fine on my Fedora system (64bit), but not so on Windows.
12:03:18 <gjanssens> There are other offenders which are probably even worse if really used. For example xaccSchedXactionGetStartDate returns a pointer to a GDate, which I doubt is anywhere near compatible with a Timespec.
12:04:10 <gjanssens> We don't use that access function directly as far as I can see, which circumvents "bad things from happening"
12:07:48 *** kael has quit IRC
12:08:29 *** MPL has joined #gnucash
12:11:17 <gjanssens> So my next steps will be to clean this up.
12:11:37 <gjanssens> Unfortunately I probably won't have time until the end of the week again.
12:11:42 *** bits has joined #gnucash
12:15:09 *** ptoal1 has joined #gnucash
12:22:08 *** fabior has joined #gnucash
12:26:23 *** ptoal1 has quit IRC
12:26:56 *** kael has joined #gnucash
12:29:11 *** ptoal1 has joined #gnucash
12:30:11 *** gour_ is now known as gour
12:30:46 *** KaiForce has joined #gnucash
12:39:13 *** ptoal1 has quit IRC
12:42:01 *** ptoal1 has joined #gnucash
12:44:40 <bits> Hi, aspiring GnuCash user here, but I'm having difficulty getting started. I can't make sense of how to import QIF files reflecting accounts in several currencies.
12:44:48 <bits> I've exported each Quicken account into its own QIF. Starting with a new, unconfigured GnuCash 3.0 book that I'd like to default to the CAD currency, I import all of the CAD-denominated QIF files in a single import. The first sign of difficulty is that transfers that have a USD account on the other side have been created as CAD on both ends. Then, when I import all the USD-denominated QIF files, the
12:44:54 <bits> importer creates duplicate asset+liability+income+expense accounts in USD and does not match any of the cross-currency transfers. I end up with, for example, a "Rent" account in CAD and a "Rent 2" acccount in USD. How can I migrate from Quicken so that cross-currency transfers are matched and recorded as a single transaction in each of two accounts of differing currency? Secondly, is having a "Rent CAD"
12:45:00 <bits> and another "Rent USD" the "normal" way to record this, or is it sensible to have a single expense account, e.g. "Rent", in CAD that has transactions whose other end might be in another currency?
12:45:04 *** boldstripe has quit IRC
12:46:11 *** boldstripe has joined #gnucash
12:47:26 <gjanssens> bits: I don't have experience with Qif import, so I can't tell how well it deals with multi-currency imports.
12:48:16 <gjanssens> If you don't get a better answer here, I'd suggest to ask your question on the gnucash-user mailing list.
12:48:36 <gjanssens> There are several users who do use qif import and will probably better placed to guide you
12:49:59 *** fabior has quit IRC
12:51:48 <bits> gjanssens: Thanks. Is QIF import the only way, or is there another way to migrate data?
12:52:23 <gjanssens> Gnucash can also import from csv (which got a serious overhaul in 3.0) and ofx
12:53:27 <gjanssens> I know csv does handle multi-currency imports but with limitations. It needs a value and conversion rate, you can't at present import with a value in CAD to a value in USD or the other way around
12:54:00 <gjanssens> I don't know if Quicken can export into a suitable csv format though.
12:54:14 <gjanssens> (I never used Quicken)
12:54:49 <gjanssens> Note also you may want to read up on double-entry accounting as gnucash does. IIUC Quicken works with categories (or is that QuickBooks ?)
12:55:05 <warlord> bits: qif really sucks for cross-currency
12:55:23 <gjanssens> The concept is slightly different and it will help you al lot if you fully understand this.
12:55:34 <gjanssens> *a lot
12:56:02 <gjanssens> You can read the Tutorial and Concepts guide on the gnucash website. It has a good primer on this topic.
12:56:46 <bits> yes, Quicken calls income/expense accounts categories. I'm new to double-entry but I understand it, if lacking in the correct terminology
12:59:56 <bits> I've read the guide, manual and searched extensively, including through chat logs, but I haven't found any reasonable treatment for migrating multi-currency. If you know of anything covering the subject, please point me at it
13:01:18 *** kael has quit IRC
13:01:19 *** kael has joined #gnucash
13:02:28 <warlord> bits: Why not just start over?
13:02:36 *** boldstripe has quit IRC
13:03:23 <bits> glanssens: Could you expand on "needs a value and conversion rate"? The CSV has a transaction amount and a conversion rate to be applied to the amount on the other end? If not, could you please expand your explanation?
13:05:29 <bits> warlord: I have over 20 years of data in Quicken. It's quite nice to be able to look back to reference an old transaction, and to see multi-year trends in charts.
13:05:42 *** boldstripe has joined #gnucash
13:07:22 *** kael has quit IRC
13:09:52 <linuturk> how stable are the SQL based backends?
13:10:02 <linuturk> is it easy to import my existing data?
13:14:13 *** harshitaneja has quit IRC
13:14:24 *** ptoal1 has quit IRC
13:14:52 *** harshitaneja has joined #gnucash
13:16:30 *** Mechtilde has joined #gnucash
13:16:39 *** boldstripe has quit IRC
13:17:37 *** boldstripe has joined #gnucash
13:27:27 <jralls> linuturk: A lot more stable in 3.0 than before, but there have been a couple of issues so you might want to wait for 3.1. We'll release that in about 3 weeks.
13:27:56 *** ncv__ has quit IRC
13:28:12 <jralls> linuturk: As for importing your existing data, that depends on what format you can export it to.
13:28:56 *** kael has joined #gnucash
13:29:16 <jralls> gjanssens: Quicken is single-entry and QuickBooks is double-entry.
13:30:50 <gjanssens> Ok so I remembered correctly
13:34:38 <jralls> gjanssens: Did you see that Chris found where the strings are getting converted for javascript yesterday?
13:35:55 <jralls> Do you have any thoughts about how to resolve that?
13:35:55 <gjanssens> I did
13:36:27 <gjanssens> Not yet
13:36:48 <gjanssens> I only followed it superficially as I had other priorities today
13:37:15 <gjanssens> Let me dig a bit deeper...
13:38:32 <jralls> Here's a thought on qofaccess functions: Turn it into a template. Write two specializations, one for Timespec, the other for time64.
13:39:39 <jralls> The better alternative, of course, is to change all of the functions that need to be exported via qof access functions to time64, but that will take a lot of work.
13:42:02 <linuturk> jralls: thanks for the answer! It's currently the standard .gnucash file (xml I think?)
13:42:46 <linuturk> assuming the current .gnucash xml files can be imported?
13:43:09 <jralls> linuturk: Oh, in that case you can just do a "save as" and pick the SQL backend you want... which is probably SQLite3 unless you have some compelling reason and the expertise to use a server.
13:43:37 <jralls> GnuCash files don't need to be imported. They're already GnuCash files.
13:44:05 <linuturk> +1 do you have details on the issues found in 3.0?
13:44:36 <jralls> Of course. More than can be relayed here. If you want to catch up, read the bugs.
13:45:57 <bits> I found some old references to a plugin that can aid in migrating a multi-currency Quicken into GnuCash. Where would one look for such a thing? Is there a plugin/macro/scripting repo or aggregator?
13:48:44 <bits> Alternatively, are there techniques or scripts to bulk edit transactions to deduplicate or merge?
13:49:37 <gjanssens> jralls: I'll get back to this later. Got to go for some time...
13:49:39 *** gjanssens is now known as gjanssens_afk
13:49:59 <jralls> bits: None of either that I know about.
13:51:22 <bits> How about just bulk edit?
13:52:55 <jralls> bits: Already written tools? Again, nothing that I know about.
14:34:14 *** frakturfreak has joined #gnucash
14:39:34 *** bertbob has quit IRC
14:40:54 *** ptoal1 has joined #gnucash
14:42:36 *** Agfarmer18 has joined #gnucash
14:50:47 *** Agfarmer18 has quit IRC
15:04:33 *** MPL has quit IRC
15:20:11 *** O01eg has quit IRC
15:23:47 *** gour has quit IRC
15:35:04 *** O01eg has joined #gnucash
15:40:07 *** ncv__ has joined #gnucash
15:41:58 *** ptoal1 has quit IRC
15:44:42 *** MPL has joined #gnucash
15:52:57 *** MPL has quit IRC
16:02:28 *** ncv__ has quit IRC
16:02:38 *** ncv__ has joined #gnucash
16:03:34 *** KaiForce has quit IRC
16:04:17 *** MPL has joined #gnucash
16:09:49 *** ncv_ has joined #gnucash
16:11:01 *** ncv__ has quit IRC
16:14:23 *** ncv_ has quit IRC
16:14:41 *** gjanssens_afk is now known as gjanssens
16:17:32 *** ncv has joined #gnucash
16:27:28 *** frakturfreak has quit IRC
16:28:15 *** Mechtilde has quit IRC
16:28:48 *** ptoal1 has joined #gnucash
16:28:55 *** ncv has quit IRC
16:30:47 *** ncv has joined #gnucash
16:31:41 <gjanssens> jralls: I've looked at the report issue some more
16:31:56 <gjanssens> I don't agree we can completely drop all the escapes.
16:32:23 <gjanssens> Most of them are unneeded. But we do need to escape a double quote.
16:33:41 <gjanssens> Unfortunately the way we currently do it is clearly not working either on Windows
16:34:04 <jralls> OK. I tried some ways to get the regex to not whack unicode and didn't find one. Any attempt to set LC_CTYPE crashes guile, so ISTM we need to find a way different than guile regexes.
16:35:12 <jralls> One way would be to have our own regex wrapper that uses utf8 explicitly.
16:37:13 <jralls> Meanwhile, in the course of testing that I discovered that undoing the scheme define of scm_(to|from)_utf8_string re-breaks utf-8 handling in saved-reports-2.4.
16:38:26 <jralls> Though now it's crashing in a new place, gnc-gettext-helper, which seems to be calling scm_from_string.
16:40:41 <gjanssens> Have you brought this regex issue up on #guile already ?
16:41:48 <jralls> No. That hasn't been terribly productive with any of the other transcoding questions, and in this case it's documented behavior.
16:44:13 <gjanssens> Ok. I wasn't sure how to interpret the documentatoin exactly. But your tests are clear enough.
16:44:50 <gjanssens> So the other option is as you say to rewrite jqplot-escape-string in C(++) and wrap it.
16:45:21 <gjanssens> It's silly because at first sight we really only have to escape a double quote.
16:46:07 <gjanssens> And perhaps a backslash if users really want to have that in their account names (which is possible)
16:46:09 <jralls> Surely there's a way to do that in scheme without resorting to regexes.
16:46:34 <gjanssens> Perhaps with substrings
16:48:52 <jralls> How about (string-replace) https://www.gnu.org/software/guile/manual/html_node/Miscellaneous-String-Operations.html#Miscellaneous-String-Operations ?
16:49:22 *** harshitaneja has quit IRC
16:49:38 <gjanssens> Yeah, let me try...
16:53:13 <jralls> Back in an hour...
16:53:27 <gjanssens> Hmm, string-replace will not do it in itself but we may write a loop to look for the characters to replace
16:53:34 <gjanssens> I'll be in bed by then :)
16:53:45 <gjanssens> See you later.
17:00:27 *** ArtGravity has quit IRC
17:01:16 *** bertbob has joined #gnucash
17:11:09 *** gjanssens has quit IRC
17:17:30 *** fekepp has quit IRC
17:17:43 *** fekepp has joined #gnucash
17:30:20 *** ArtGravity has joined #gnucash
17:31:02 *** ncv has quit IRC
17:48:29 *** kael has quit IRC
17:51:59 *** fekepp has quit IRC
17:52:30 *** fekepp has joined #gnucash
18:07:07 *** MPL has quit IRC
18:12:28 *** bertbob has quit IRC
18:14:08 *** bertbob has joined #gnucash
18:14:49 *** Robert847 has joined #gnucash
18:19:05 <chris> this is how (json builder) will do string substitution https://github.com/aconchillo/guile-json/blob/master/json/builder.scm#L126
18:23:24 <Robert847> Hi, I have been using backported release 2.6.11 in Debian Jessie without problems for quite a while but I am experiencing a few issues recently with seg faults while importing OFX files and other LibOFX errors. I am not sure if a recent Synaptic update stomped on LibOFX or if there is some other issue. I do not want to spend much time on this if it will automatically go away when I finally update to Stretch, except that may be about
18:23:25 <Robert847> a month from now at the earliest.
18:26:57 <jralls> chris: I think that's OK as long as the port has been configured correctly to transcode SCM to utf-8.
18:27:29 <Robert847> I forgot t ask a qestion. Will LibOFX change in release 3.0.1?
18:28:07 *** MPL has joined #gnucash
18:29:20 <jralls> Robert847: If the question is "does GnuCash 3.0 use LibOFX?" the answer is yes. If it's "is the version of LibOFX different?" that's up to the Debian LibOFX packager. Note that at present GnuCash 3.0 is built only for Debian Sid (i.e. unstable).
18:33:37 <chris> jralls I spoke too soon; the current maint on ubuntu can process emoji http://i.imgur.com/yTJJkEf.png but my chartjs-on-guile-json can't https://imgur.com/a/5H9ajLS
18:33:38 *** kus has joined #gnucash
18:35:20 <jralls> chris: I think that's because it uses (display x). The port that display writes to must be configured to emit UTF-8. The default is for it to use the current locale.
18:37:24 <Robert847> It sounds like I need to do more learning and testing in VMs before charting my migration path. Thanks
18:40:08 <chris> hmm here's a fun (current) bug - include html entity eg © in account-name. the html will be interpreted in any report (including text reports). this means ALL strings in reports should really be escaped. Bah.
18:41:20 <jralls> chris: On Unix (including MacOS) it's mostly not a problem because the native locales are mostly UTF-8 anyway so guile does the right thing when left to itself. On Windows it seems to be reading the codepage and unless forced to do something else it converts strings to that codepage, which is never what we want.
18:53:51 <chris> jralls - no, I meant: in current maint, in gnc register, enter arbitrary html plaintext in Description or Account-names - they are sent to html raw. this is not good. see https://imgur.com/a/krVR4hJ
18:55:40 <jralls> Isn't that because you removed the call to jqplot-escape-string?
18:56:07 <chris> this is because jqplot-escape-string is only called in chart-type reports.
18:56:56 <chris> I think html-document.scm will need to escape strings
18:57:38 *** bertbob has quit IRC
18:58:39 <jralls> Ah, of course. Yes. I raised that issue with gjanssens and warlord the other day: The escape function only does backslashes. gjanssens was of the opinion that & and < didn't need escaping. Clearly not true, at least for &.
18:58:59 <jralls> Though perhaps in his test he didn't try using an html entity.
19:00:09 *** bertbob has joined #gnucash
19:03:04 *** kus has quit IRC
19:07:49 <chris> Ok I did it - need to change html-style-info.scm Line 259 to escape strings in the strings renderer. See output https://imgur.com/LAGFuKu
19:08:36 <jralls> chris: We don't want to use jqplot-escape-string, it eats unicode.
19:08:47 <chris> oh yes
19:09:14 <chris> we still want to protect reports from arbitrary html tho
19:09:17 <jralls> Or maybe I should say we need to rewrite jqplot-escape-string so that it doesn't eat unicode.
19:09:55 <jralls> Of course. There's also another escape function, give me a moment...
19:13:09 *** kus has joined #gnucash
19:13:31 <jralls> Ah, it's "catenate-escaped-strings", appearing separately in several files. It also uses guile regexes so another thing to fix.
19:14:40 *** kus has quit IRC
19:14:46 <jralls> Hmm, probably the best place to do the escaping is in gnucash/html. That's all in C, so we could convert a file to C++ and use boost::regex.
19:14:50 *** kus has joined #gnucash
19:17:09 <chris> so, conclusion, we need a generic "sanitize-string-for-html" function str->str ?
19:19:31 *** ArtGravity has quit IRC
19:20:23 *** chris is now known as chris_nap
19:20:38 <jralls> Yes, and if it needs to go to C it should use scm_to_utf8_string() going out and scm_from_utf8_string() coming back.
19:22:28 <jralls> Beware of Guile functions, like regex, that process in C, they do the wrong thing. Maybe they've fixed it in 2.2, I'll wander over there and ask.
19:24:10 *** MPL_ has joined #gnucash
19:24:27 *** MPL has quit IRC
19:25:07 *** yuriks has quit IRC
19:26:01 *** bertbob has quit IRC
19:27:37 *** yuriks has joined #gnucash
19:28:04 *** kus has quit IRC
19:33:23 <chris_nap> i'll try make a good one, and i'll avoid regex
19:51:22 <jralls> chris_nap: Coordinate with gjanssens, he's thinking about the same problem.
20:03:59 *** bertbob has joined #gnucash
20:07:15 *** Robert847 has left #gnucash
20:19:13 *** MPL_ has quit IRC
20:19:41 *** MPL_ has joined #gnucash
20:24:55 *** MPL_ has quit IRC
20:30:54 *** MPL_ has joined #gnucash
20:34:03 *** MPL_ has quit IRC
20:51:35 *** Aussie_matt has joined #gnucash
20:53:47 <bits> ll
20:54:36 *** bits has quit IRC
21:10:08 *** MPL_ has joined #gnucash
21:10:56 <jralls> gjanssens, fell: My last push didn't go through to github so you'll need to pull from code before you can push.
21:11:08 *** meb has joined #gnucash
21:11:39 <jralls> @tell gjanssens My last push didn't go through to github so you'll need to pull from code before you can push.
21:11:39 <gncbot> jralls: The operation succeeded.
21:11:45 *** jralls has quit IRC
21:12:00 <fell> NP
21:12:29 *** gncbot sets mode: +o fell
21:13:18 *** MPL_ has quit IRC
21:17:20 *** meb has quit IRC
21:22:24 *** ptoal1 has quit IRC
21:23:11 *** MPL_ has joined #gnucash
21:23:41 *** ptoal1 has joined #gnucash
21:29:18 *** meb has joined #gnucash
21:42:07 *** meb has quit IRC
21:43:43 *** meb has joined #gnucash
21:44:00 *** oozer has quit IRC
21:45:12 *** MPL_ has quit IRC
22:05:03 *** MPL_ has joined #gnucash
22:09:55 *** ArtGravity has joined #gnucash
22:24:45 *** boldstripe has quit IRC
22:27:10 *** meb has quit IRC
22:31:26 *** boldstripe has joined #gnucash
22:34:26 *** kus has joined #gnucash
22:38:55 *** boldstripe_ has joined #gnucash
22:46:34 *** boldstripe_ has quit IRC
23:00:54 *** Agfarmer18 has joined #gnucash
23:03:53 *** Agfarmer18 has quit IRC
23:05:42 <chris_nap> I'll do a loop-based function and plenty of TDD tests
23:05:57 *** chris_nap is now known as chris_away
23:28:29 *** kus has quit IRC