2012-03-17T00:01:21 rebased a bit to get rid of a couple of trivial commits, and withdrew the pref_lib ordering, but retained the "set search library to pref library at login" behaviour 2012-03-17T00:01:24 user/dbs/unapi_improve_limits_and_tpac_display 2012-03-17T00:04:37 *** Harrison has quit IRC 2012-03-17T00:56:30 *** busycoder has joined #evergreen 2012-03-17T00:56:43 hi there 2012-03-17T00:57:12 I wanna join gsoc 2012 and work for evergreen 2012-03-17T00:57:23 can anyone guide me through 2012-03-17T02:05:09 *** busycoder has quit IRC 2012-03-17T02:41:00 *** rohit12 has joined #evergreen 2012-03-17T02:52:48 *** Rohit_ has joined #evergreen 2012-03-17T02:53:14 *** Rohit_ is now known as Guest19610 2012-03-17T02:55:59 *** rohit12 has quit IRC 2012-03-17T03:10:15 *** Guest19610 is now known as rohit12 2012-03-17T03:10:48 *** rohit12 has joined #evergreen 2012-03-17T03:29:26 *** rohit12 has quit IRC 2012-03-17T03:53:15 *** avinash has joined #evergreen 2012-03-17T03:54:19 *** avinash has quit IRC 2012-03-17T04:11:03 *** umashankar has joined #evergreen 2012-03-17T05:09:19 *** umashankar has quit IRC 2012-03-17T05:27:22 *** maheshwar has joined #evergreen 2012-03-17T05:27:31 *** akdroid has joined #evergreen 2012-03-17T05:27:48 helo 2012-03-17T05:28:21 i am applying for this year's gsoc and want to know what exactly is evergreen 2012-03-17T05:28:23 ?? 2012-03-17T05:30:13 *** akdroid has quit IRC 2012-03-17T05:30:37 *** akdroid has joined #evergreen 2012-03-17T05:39:58 *** akdroid has quit IRC 2012-03-17T05:51:38 *** kmlussier has joined #evergreen 2012-03-17T05:56:32 *** kmlussier_ has joined #evergreen 2012-03-17T05:56:32 *** kmlussier has quit IRC 2012-03-17T06:00:47 *** kmlussier_ has quit IRC 2012-03-17T06:22:09 *** umashankar has joined #evergreen 2012-03-17T06:45:02 *** umashankar_ has joined #evergreen 2012-03-17T06:45:31 *** umashankar has quit IRC 2012-03-17T06:46:48 *** umashankar_ has left #evergreen 2012-03-17T08:38:11 *** Mirko_ has joined #evergreen 2012-03-17T08:39:12 *** tater-laptop has quit IRC 2012-03-17T08:45:37 sorry I could tell where I can find a mentor Dan Wells? I would like to participate in the project application from Android you proposal. 2012-03-17T08:48:46 *** Mirko_ has left #evergreen 2012-03-17T08:51:25 *** tater-laptop has joined #evergreen 2012-03-17T09:01:53 @later tell jamesrf re 905$u, that evergreen specific. all of the 905-based ingest-time stuff is evergreen-specific, it's not about using another ILS' magic fields if that's what you meant 2012-03-17T09:01:53 eeevil: The operation succeeded. 2012-03-17T09:28:07 *** Kevin__H has joined #evergreen 2012-03-17T09:53:06 *** manasCh has joined #evergreen 2012-03-17T09:58:06 *** granitize has joined #evergreen 2012-03-17T09:58:13 *** granitize has left #evergreen 2012-03-17T10:01:10 *** akdroid has joined #evergreen 2012-03-17T10:30:50 *** Rohit_ has joined #evergreen 2012-03-17T10:31:04 *** Rohit_ is now known as rohit12 2012-03-17T10:40:51 anyone here to discuss ideas on android clients for evergreen 2012-03-17T10:41:01 ? 2012-03-17T10:41:28 *** maheshwar has quit IRC 2012-03-17T10:43:29 I can discuss ideas, but I am not on the mentor list for that particular area in regards to GSoC. 2012-03-17T11:01:57 rohit12: i am also interested for the same 2012-03-17T11:02:36 but i dont know anything about evergreen 2012-03-17T11:03:16 Evergreen is an integrated library system (ILS) used to manage library holdings and circulations. 2012-03-17T11:10:44 tsbere: wwho are the users like ?? 2012-03-17T11:11:17 Libraries and library consortia plus their patrons. 2012-03-17T11:11:17 akdroid: you might want to start by reading the Evergreen blog post about GSoC: http://evergreen-ils.org/blog/?p=750 2012-03-17T11:13:39 * denials notes that http://www.evergreen-ils.org/dokuwiki/doku.php?id=faqs:evergreen_faq_1#what_new_functions_will_evergreen_offer_in_the_future discusses features that will be added in 2010 2012-03-17T11:13:57 also, our roadmap is woefully out of date: http://www.evergreen-ils.org/dokuwiki/doku.php?id=faqs:evergreen_roadmap 2012-03-17T11:15:53 *** akdroid has quit IRC 2012-03-17T11:17:46 * tsbere wonders why we keep some of that around. <_< 2012-03-17T11:20:57 *** akdroid has joined #evergreen 2012-03-17T11:21:37 *** rohit12 has quit IRC 2012-03-17T11:27:06 *** manasCh has quit IRC 2012-03-17T11:29:12 *** jenny has joined #evergreen 2012-03-17T11:51:07 *** tater-laptop has quit IRC 2012-03-17T11:54:12 *** akdroid has left #evergreen 2012-03-17T12:09:29 *** sourabh0612__ has joined #evergreen 2012-03-17T12:32:13 *** Rohit has joined #evergreen 2012-03-17T12:32:43 *** Rohit is now known as Guest63979 2012-03-17T12:33:29 *** Guest63979 is now known as rohit12 2012-03-17T12:35:50 *** Conlin has joined #evergreen 2012-03-17T12:36:58 Hello 2012-03-17T12:38:24 Conlin: hi 2012-03-17T12:39:51 Is there any typ eof evergreen community? Outside of the IRC and Mailing-List 2012-03-17T12:46:24 *** rohit12 has quit IRC 2012-03-17T12:48:36 *** Conlin has quit IRC 2012-03-17T12:50:53 @later tell Conlin Yes, there's an annual conference - the next conference is in April 2012 (you can see details from the Evergreen home page) 2012-03-17T12:50:53 denials: The operation succeeded. 2012-03-17T12:55:56 *** Rohit_ has joined #evergreen 2012-03-17T12:56:23 *** Rohit_ is now known as Guest3817 2012-03-17T12:58:46 *** sourabh0612__ has quit IRC 2012-03-17T13:02:51 *** Guest3817 is now known as rohit12 2012-03-17T13:07:33 *** Rohit__ has joined #evergreen 2012-03-17T13:08:54 *** rohit12 has quit IRC 2012-03-17T13:09:21 *** jenny has quit IRC 2012-03-17T13:14:08 *** umashankar_ has joined #evergreen 2012-03-17T13:15:20 *** imharpreet has joined #evergreen 2012-03-17T13:17:34 Hi, i am new here. I was interested in contributing to the Evergreen project via GSoc. 2012-03-17T13:17:38 *** Rohit__ is now known as rohit12 2012-03-17T13:21:57 * denials circles in on pref_lib showing copy counts outside of pref_lib scope, other goodness 2012-03-17T13:22:44 hi imharpreet! No interest in contributing to Evergreen outside of GSoC? 2012-03-17T13:23:01 *** Ashen_ has joined #evergreen 2012-03-17T13:24:50 denials: Certainly would not mind that too. 2012-03-17T13:26:18 *** maheshwar has joined #evergreen 2012-03-17T13:26:31 *** umashankar_ has left #evergreen 2012-03-17T13:30:31 Great! 2012-03-17T13:32:57 *** Umashankar has joined #evergreen 2012-03-17T13:34:28 I am new here where to start leaning of org 2012-03-17T13:36:38 Please reply 2012-03-17T13:38:48 No reply yet 2012-03-17T13:39:37 Umashankar: you can check this http://evergreen-ils.org/ 2012-03-17T13:40:32 Please reply 2012-03-17T13:42:59 *** Umashankar has quit IRC 2012-03-17T13:47:40 rohit12++ 2012-03-17T13:48:27 what happened? 2012-03-17T13:49:34 ? 2012-03-17T13:59:46 rohit12: giving you positive karma for helping someone else 2012-03-17T13:59:51 @karma rohit12 2012-03-17T13:59:51 denials: Karma for "rohit12" has been increased 1 time and decreased 0 times for a total karma of 1. 2012-03-17T14:05:04 *** tsbere_ has joined #evergreen 2012-03-17T14:05:18 *** tsbere has quit IRC 2012-03-17T14:05:22 *** tsbere_ is now known as tsbere 2012-03-17T14:05:24 rohit12: Please don't ask for things like DCC chats. Not all of us are at our machines all the time. It is better to ask your questions in the channel, or in the event of no activity in the channel you may be better off on our mailing list(s) (time zone differences and all). 2012-03-17T14:06:38 *** imharpreet has left #evergreen 2012-03-17T14:07:23 tsbere: sorry sir, it won't happen again. 2012-03-17T14:08:59 rohit12: It is more of a speeding your responses thing. If you are talking to just one of us then you have to wait for the one person to respond. If you ask in the channel anyone who is around at the time may be able to answer instead. 2012-03-17T14:10:01 For example, your helping of Umashankar earlier. Had he poked me instead of the channel at large you couldn't have pointed him at the evergreen-ils.org website. He would have been stuck waiting for me instead. 2012-03-17T14:11:11 * denials is just popping into the channel every 15 minutes or so while sorting out the final bits (really!) of the preferred library fixes 2012-03-17T14:12:02 * tsbere actually went on a walk (!) and grabbed lunch while he was out. Yes, he could have popped in via his phone but walking and monitoring IRC at the same time sounds dangerous when the entire route you are walking is a construction area. 2012-03-17T14:12:45 i've understood sir and agree fully with you. 2012-03-17T14:14:25 is there any problem in adding the patron registration feature in the android client? 2012-03-17T14:16:04 rohit12: Well, I have some issues with the staging area (mainly required database fields, easily changed). Beyond that the response wouldn't be immediate, as for various reasons most libraries would want a staff member to approve the registrations. But there is a staging are you could provide an interface for populating. 2012-03-17T14:16:20 s/are/area/ near the end there. >_> 2012-03-17T14:22:38 *** Reganeet has joined #evergreen 2012-03-17T14:33:39 * denials hastily force-pushes a tiny fix onto bug 907056 2012-03-17T14:33:39 Launchpad bug 907056 in Evergreen "In-db unAPI has little control over subobject inclusion" (affected: 1, heat: 10) [Undecided,New] https://launchpad.net/bugs/907056 2012-03-17T14:38:10 *** mayu has joined #evergreen 2012-03-17T14:40:34 * denials finds it kind of funny that we don't have any projects around tpac in the ideas list 2012-03-17T14:43:06 *** mayu has left #evergreen 2012-03-17T14:47:34 * denials is pretty sure that user/dbs/unapi_improve_limits_and_tpac_display in working is "the one" 2012-03-17T14:49:30 denials: I dunno if it is safe to put tpac anything on the ideas list. We are too busy changing tpac for someone to catch up in time to propose anything ;) 2012-03-17T14:50:59 true! but surely in a month or two we'll settle down :) 2012-03-17T14:59:00 *** avinash has joined #evergreen 2012-03-17T16:13:02 *** maheshwar has quit IRC 2012-03-17T16:13:53 I'm trying to exceute some subroutines within specific perl modules and I am getting the following error: No boostrap configs exists, Have you bootstrapped? .. Is there any way I can get around this.. I've tried a few different modules from different directories ... so subroutines work, others just produce this message 2012-03-17T16:20:14 *** avinash has quit IRC 2012-03-17T16:21:03 Kevin__H: You writing your own perl program? Because if so you need to, well, "bootstrap" the system (to tell it how to do things like connect to XMPP and such) 2012-03-17T16:21:45 Are there instructions available on how to do that? 2012-03-17T16:22:25 You could look at existing programs. Say hold_targeter.pl as a small example? 2012-03-17T16:23:07 Under support_scripts? 2012-03-17T16:24:02 yea 2012-03-17T16:33:31 I'm sorry, one more question, I am trying to do what hold_targeter.pl does ... I simply added these: 2012-03-17T16:33:55 my $config = shift || die "bootstrap config required\n" 2012-03-17T16:34:20 OpenSRF::System->bootstrap_client( config_file => $config ); 2012-03-17T16:34:49 now I understand that upon running I have to append a 2012-03-17T16:34:49 I dunno what you are asking 2012-03-17T16:36:19 It says usage: hold_targeter.pl 2012-03-17T16:36:49 I assume I would have to adopt that as well into what I'm doing in order to be able to use those two lines of code to bootstrap the system? 2012-03-17T16:39:32 Well, you can usually set a default as well. There are examples of that elsewhere, say in clark_kent.pl I believe? 2012-03-17T16:43:03 $config = 'SYSCONFDIR/opensrf_core.xml' 2012-03-17T16:43:08 That looks pretty promising 2012-03-17T16:43:23 Is that what you were refering to? 2012-03-17T16:44:04 That kind of thing. Granted, you may need more absolute paths (SYSCONFDIR is replaced when the file is installed) but that is a way to set defaults. 2012-03-17T16:46:37 k 2012-03-17T16:46:42 thanks, I'll give it a try 2012-03-17T16:48:39 Kevin__H: a slightly more modern way is to rely on OpenILS::Utils::Cronscript 2012-03-17T16:49:56 "perldoc OpenILS::Utils::Cronscript" might even be accurate :) 2012-03-17T16:53:30 *** rohit12 has quit IRC 2012-03-17T17:36:57 I'm trying to figure out OpenILS::Utils::Cronscript 2012-03-17T17:38:08 and I am getting this error: www.pastebin.ca/2129371 2012-03-17T18:19:25 Kevin__H: Umm. why "perl -T"? 2012-03-17T18:19:39 * denials heads off to dinner 2012-03-17T18:19:59 It's for my test file trying to access a perl module 2012-03-17T18:20:32 So you're checking for taints? 2012-03-17T18:20:46 * denials doesn't get it, but really needs to go eat 2012-03-17T18:57:08 :) 2012-03-17T18:58:22 *** yohan has joined #evergreen 2012-03-17T19:03:57 *** yohan has quit IRC 2012-03-17T19:52:47 denials: I think I may be approaching testing in a confused manner, can you enlighten me? 2012-03-17T20:01:16 Kevin__H: if you're running perl with the -T option, you shouldn't be =) 2012-03-17T20:02:28 I think it complains if I run a .t file without it 2012-03-17T20:03:09 it shouldn't, unless your test has turned on taint checking 2012-03-17T20:03:15 which is what the -t/-T options do 2012-03-17T20:03:31 Hmm 2012-03-17T20:04:31 You mean the first line is : #! perl -T 2012-03-17T20:05:15 I did that by simply following the convention found under Evergreen/Open-ILS/src/perlmods/t 2012-03-17T20:07:19 *** ronb54 has joined #evergreen 2012-03-17T20:10:02 Kevin__H: are you trying to run the tests with prove, or with perl? 2012-03-17T20:10:18 personally I'd recommend writing your tests without the -T in that line 2012-03-17T20:11:00 I am adding to the directory t, and to keep things cleaner when I test to make sure my tests work, I have been shortcutting and doing perl -T testcase.t 2012-03-17T20:11:32 yeah, I would take out the -T from tests you write 2012-03-17T20:11:32 I was doing prove, but I just found it was cluttering for how often I was executing my tests (to see the results of my newly written test cases) 2012-03-17T20:11:51 *nodnod* that should be fine 2012-03-17T20:11:55 Okay.. I will do that .. I was just following the convention found in the other testcases.t found in that directory 2012-03-17T20:12:00 but taint mode is probably more than you want to fight with 2012-03-17T20:12:19 do 'perldoc perlsec' and scroll down to "taint mode" to read about the uphill battle you'd be up against =) 2012-03-17T20:12:42 Will do 2012-03-17T20:12:52 Could that be the reason why I was having issues with bootstrapping? 2012-03-17T20:14:02 hmm, dunno. bootstrapping is never something I've run into an issue with 2012-03-17T20:16:44 I was finding when calling subroutines within perl modules I was getting this error: No boostrap confi exists. Have you bootstrapped? 2012-03-17T20:17:26 www.pastebin.ca/2129394 2012-03-17T20:17:43 Sorry if is a little hard to read 2012-03-17T20:17:52 The commented out stuff at the bottom is stuff I tried, but didn't work 2012-03-17T20:28:19 Kevin__H: that error message is coming out of ./opensrf-2.0.1/src/perl/lib/OpenSRF/Utils/SettingsClient.pm 2012-03-17T20:28:34 can you use srfsh to do the 'math add 2 2' thing? 2012-03-17T20:28:39 yes 2012-03-17T20:33:50 hmmm 2012-03-17T20:34:12 what it's telling you (I think) is that it can't find your opensrf client config file 2012-03-17T20:34:25 if you do 'perldoc opensrf-2.0.1/src/perl/lib/OpenSRF/Utils/Config.pm' that might shed some light 2012-03-17T20:34:32 what user are you running as when you're trying to run the tests? 2012-03-17T20:35:40 hmm.. good question -- I think I've done it with both opensrf and root 2012-03-17T20:35:43 let me check again 2012-03-17T20:36:43 same result for both 2012-03-17T20:37:00 you've been able to do the srfsh thing since upgrading though? 2012-03-17T20:37:25 yes, I just did it again to test 2012-03-17T20:37:37 hmmm 2012-03-17T20:37:47 sorry, I've got nothing more for you at the moment =/ 2012-03-17T20:38:16 I'll try looking at that config file you suggested 2012-03-17T21:57:30 Kevin__H: took a look at your pastebin; 1) use strict; use warnings; 2) replace "perl -T" with "/usr/bin/perl"; 3) you're missing a semicolon at the end of line 11 2012-03-17T21:59:07 4) quote the names of the @subs 2012-03-17T22:00:50 thank you 2012-03-17T22:01:12 also, not sure if you know this, but you can specify the test you want to run with prove, like "prove -l t/29.foo.t" 2012-03-17T22:01:58 I didn't, thank you 2012-03-17T22:02:13 No problem. Great to see somebody else tackling some tests! 2012-03-17T22:02:20 any suggestions on my bootsrapping problem? I am still getting that error 2012-03-17T22:03:23 Strange, works for me 2012-03-17T22:03:59 You've configured Evergreen with ./configure? 2012-03-17T22:04:13 should be working 2012-03-17T22:04:36 I have OpenSRF / Evergreen fully installed and I can do the test : request opensrf.math add 2 2 2012-03-17T22:04:42 and I can login evergreen evergreen 2012-03-17T22:06:34 You see that bootstrapping error with the exact script that you pastebinned? 2012-03-17T22:06:54 sorry 2012-03-17T22:07:04 I typed: cronsript, not cronscript 2012-03-17T22:07:26 it works! 2012-03-17T22:07:37 string / warnings help with typos a lot :) 2012-03-17T22:07:45 \o/ 2012-03-17T22:28:18 any suggestions for writing tests for subroutines which take parameters? 2012-03-17T22:30:09 specifically look at Application.pm, register_method subroutine 2012-03-17T22:33:19 That is one of the subroutines I would personally be least concerned about. But that is me. <_< 2012-03-17T22:33:50 If someone breaks that then large portions of the system will stop working. 2012-03-17T22:34:16 In fact, if you run tests on an actual running system then you are indirectly testing that one 2012-03-17T22:34:45 I see 2012-03-17T22:35:25 I don't suppose you can suggest, some areas that may be better to work on? 2012-03-17T22:37:33 * tsbere isn't 100% sure what Kevin__H is working on 2012-03-17T22:37:53 You writing OpenSRF level tests or Evergreen tests? (I am assuming you are writing tests in general, anyway) 2012-03-17T22:38:34 In summary: white box tests for the whole suite 2012-03-17T22:38:56 and then my intent is to use Devel::Cover to show coverage of the tests 2012-03-17T22:39:07 ok. Have you figured out how to grab specific applications and make requests against them? 2012-03-17T22:39:47 no 2012-03-17T22:40:57 Most of the system is expecting to be talked to over XMPP in some fashion, so you may want to start there, then 2012-03-17T22:41:15 There are areas you could test without XMPP running, but not many on the Evergreen side that I know of. 2012-03-17T22:41:56 Okay 2012-03-17T22:42:31 Is that.... ejabberd? 2012-03-17T22:42:31 Although I suspect that you can do things I am not aware of. <_< 2012-03-17T22:43:38 ejabberd is the XMPP service recommended for use. Let me dig up some "call things on services from perl" examples quick, though. 2012-03-17T22:44:49 There we are. I suppose a good example is in hold_targeter.pl - It creates a standard open-ils.storage and a multisession open-ils.storage call in the same short file. 2012-03-17T22:45:40 The OpenSRF::AppSession and related calls in particular 2012-03-17T22:45:49 okay 2012-03-17T22:46:56 Note that OpenSRF defines several things you can use as really simple tests. 2012-03-17T22:47:47 Can you elaborate? 2012-03-17T22:48:05 Is this within AppSession.pm? 2012-03-17T22:48:07 For example, I think everything defines, due to the OpenSRF API layer, the "opensrf.system.echo" function. So once you ->create your appsession to whatever service you are testing you can request against that methd. It should return what you pass in for an argument. 2012-03-17T22:48:31 Doesn't test much more, but it is a good starting point for making sure your code is working ;) 2012-03-17T22:49:19 So basically.. I could make a test call OpenSRF -- then OpenSRF calls, let's say OpenILS/Application/Claims.pm? 2012-03-17T22:49:44 and then that's indirectly testing Claims.pm? 2012-03-17T22:49:48 oops 2012-03-17T22:49:55 Application/Acq/Claims.pm 2012-03-17T22:50:15 Somewhat more directly, really. Most of the code can't really be tested *off* of a running system given interactions between services and such. 2012-03-17T22:51:21 I *think* you can do things like load a service directly in your script and call the functions it defines....it would still need to be able to call outside of itself, mind you, so you would still need a running system. I just don't know as much about doing that. 2012-03-17T22:52:14 So you don't believe writing individual test cases for individual perl modules is a reasonable approach? 2012-03-17T22:52:38 Looks like Open-ILS/src/extras/import/direct_loader.pl might do the "load things directly in your script" bit. Not 100% sure, though. :/ 2012-03-17T22:54:06 Kevin__H: What I am saying is that when you test, say, OpenILS::Application::Circ it is going to want to talk to other services as it does things, like OpenILS::Application::Storage. So yes, you can run tests on individual services, but those tests will indirectly test other areas at the same time. 2012-03-17T22:55:02 I suppose that's why when I looked at my coverage report, I was seeing more than expected 2012-03-17T22:55:44 Running tests directly against services tells us if those services are working, though. So testing Storage there and finding it works but Circ doesn't for some reason would, if all of the appropriate Storage calls are passing, point the finger at Circ itself. If you aren't testing Storage directly then it could be Circ or Storage that is broken. 2012-03-17T22:56:39 thus a test would still fail and we'd have a starting point at locating where the failure is... so it would still be ... okay? 2012-03-17T22:57:28 I am a fan of "even if something is testing it indirectly, running a battery of tests against it directly is good too". Your direct tests may cover things that indirect tests don't do in most situations, for example. 2012-03-17T22:58:35 But if you want to skip direct tests for things that are being tested indirectly according to your coverage reports then I am not going to tell you to go to the extra effort. ;) 2012-03-17T22:59:07 good idea 2012-03-17T22:59:15 BTW, have you looked at the srfsh "introspect" command? 2012-03-17T22:59:37 nope... the only thing I've done at the srfsh is login and add 2012-03-17T22:59:56 Well, try this: introspect opensrf.math 2012-03-17T23:00:00 Nice short one ;) 2012-03-17T23:00:31 General idea is it hands you every method call that service is exporting. Sometimes with useful things like what the commands expect for input, sometimes not. 2012-03-17T23:01:44 should I login with my admin ..? 2012-03-17T23:02:10 You don't need a login for that 2012-03-17T23:02:30 In fact, many parts of the system don't require an authtoken. Those that do usually require it to be the first parameter. 2012-03-17T23:03:08 Anyway, for opensrf.math, one of the api names is "add". That is what happens when you do "request opensrf.math add 2 2" and similar. It is like, in perl, doing an OpenSRF::AppSession->create('opensrf.math')->request('add', 2, 2); I believe. 2012-03-17T23:04:50 Oh 2012-03-17T23:05:24 The authtoken is an evergreen-specific thing. OpenSRF itself doesn't understand those ;) 2012-03-17T23:05:41 ah because evergreen is ontop of opensrf 2012-03-17T23:06:45 * jeff nods 2012-03-17T23:07:20 Yes. Although srfsh does have at least one evergreen specific command. The "login" command in srfsh specifically calls an evergreen service: open-ils.auth - It actually runs two different method calls there. An authentication init call followed by a complete call. 2012-03-17T23:07:58 If you need an authtoken for your tests you will need to run the same dance 2012-03-17T23:08:15 Me? 2012-03-17T23:08:41 Your tests, to get an authtoken for those services that need one, will need to do the authentication init and complete calls to get one. 2012-03-17T23:09:08 I believe cronscript has a helper function for that, though. I could be wrong. 2012-03-17T23:11:09 Okay, I ran that introspect opensrf.math (it worked this time -- I had to restart my VM) 2012-03-17T23:12:52 So what I'm getting about this authtoken is : 2012-03-17T23:12:56 The cronscript file does indeed have the helper function - authenticate. It takes a single argument of a hash of information: username, password, and if you want something other than staff logins a login type. 2012-03-17T23:13:14 If I come to a point where I'm writing a test that uses it... I will 2012-03-17T23:13:45 nvm that last message 2012-03-17T23:14:41 So I'm understanding that if I come to a point where I'm testing and I need an authtoken... I have to use the cronscript::authenticate function.. and for that I can use my evergreen admin as parameters 2012-03-17T23:17:06 Yea. Although it dawns on me that for test purposes being able to bypass "give me a username and password" might be a good idea. 2012-03-17T23:19:40 If you are willing to change what cronscript is doing to allow you to hand it a MD5 instead of an actual password you can probably log in without asking for username/password information 2012-03-17T23:20:07 Will I need to be doing this for a lot of modules? 2012-03-17T23:20:58 a request to open-ils.cstore for the open-ils.cstore.direct.actor.user.search method with a parameter of {"super_user":"t","deleted":"f"} will get you a list of all superuser accounts. Grab the password hash and the username and do the auth dance with them.... 2012-03-17T23:22:21 request open-ils.cstore.direct.actor.user.search("super_user" : "t", "deleted": "f") 2012-03-17T23:22:23 ? 2012-03-17T23:22:31 Kevin__H: A good portion of Evergreen requires authtokens. In opensrf_core.xml you can find a list of services on the "public" router near the top of the file. Anything published from those services will either be considered "safe" or will require an authtoken. 2012-03-17T23:25:01 Kevin__H: For srfsh the command is 'request open-ils.cstore open-ils.cstore.direct.actor.user.search {"super_user":"t","deleted":"f"}' - For perl it differs slightly. The parameter could be formatted in perl as, say, {super_user => 't', deleted => 'f'} (including the { and }) 2012-03-17T23:25:31 Kevin__H: May also be a good idea to test with those that aren't super-users, but for an initial test suite I wouldn't worry too much about that. 2012-03-17T23:25:48 yes 2012-03-17T23:25:59 sounds good to me! 2012-03-17T23:30:03 it didn't work 2012-03-17T23:30:10 What didn't work? 2012-03-17T23:30:16 the request on srfsh 2012-03-17T23:30:37 Make sure on the srfsh one you aren't including the single quotes? 2012-03-17T23:30:45 yes 2012-03-17T23:30:50 the single quotes around the whole thing? 2012-03-17T23:31:02 I copied and pasted out of srfsh after using it to make sure it did work. 2012-03-17T23:31:04 Hmmm 2012-03-17T23:31:21 The single quotes around the whole thing were to show you where it started/ended. You don't want them when you copy/paste it ;) 2012-03-17T23:31:34 request open-ils.cstore open-ils.cstore.direct.actor.user.search{"super_user":"t":"deleted":"f"} 2012-03-17T23:31:43 You need a space before the { 2012-03-17T23:31:46 I'm working on a VM - I can't copy/past :) 2012-03-17T23:32:04 and before deleted you put a : instead of a , 2012-03-17T23:32:31 and I copy/paste in and out of VMs all the time. I just SSH into them ;) 2012-03-17T23:33:09 Is that a quick process? 2012-03-17T23:33:14 This has been painful 2012-03-17T23:34:17 If the SSH service is running and you can load the catalog in your browser then I don't think it would be bad. Just put the hostname or IP into an SSH client. 2012-03-17T23:34:33 If you can't see the VM from outside of the VM then....that is more complicated. 2012-03-17T23:35:41 and beyond my comfort zone to the point where I can't speculate much. Most VM software includes some form of "the host machine can see the IP the VM is using" though, so if you can grab that IP (say, from inside the VM with ifconfig) you may be able to connect. Even if it is only from the host machine. 2012-03-17T23:37:13 I have the IP address that my virtual machine is using 2012-03-17T23:38:01 You also need SSH software, obviously. I use putty on windows. Others have their own choices. 2012-03-17T23:38:23 putty I have 2012-03-17T23:41:51 Well, if you can get putty to connect to the VM you should have an easier time. At least you can then copy/paste in normal putty fashion. Makes working with things a lot easier. 2012-03-17T23:42:20 I'll take a look at it 2012-03-17T23:42:27 let me try that srfsh command again 2012-03-17T23:42:34 Well, unless you are more used to GUI than command line. But I suggest getting used to the command line in that case ;) 2012-03-17T23:43:02 I have 2012-03-17T23:43:35 request open-ils.cstore open-ils.cstore.direct.actor.user.search {"super_user":"t","deleted":"f"} 2012-03-17T23:44:16 That looks right. Works for me, even. 2012-03-17T23:44:39 Uh oh 2012-03-17T23:45:16 * tsbere dislikes Uh oh followed by CTCP hits on IRC 2012-03-17T23:45:33 sorry, curious 2012-03-17T23:46:06 anything you can suggest? 2012-03-17T23:46:22 I'm not keeping you up am I? 2012-03-17T23:46:46 I am not afraid to say "I am going to bed now, good luck!" and similar ;) 2012-03-17T23:48:13 Thanks for all your help 2012-03-17T23:50:20 Did you get the cstore request there working? 2012-03-17T23:51:27 no... 2012-03-17T23:51:46 what did it tell you? 2012-03-17T23:52:06 Received no data from server 2012-03-17T23:54:04 You able to run the login command? 2012-03-17T23:54:15 login evergreen evergreen 2012-03-17T23:54:16 works 2012-03-17T23:54:22 ok 2012-03-17T23:54:35 Not sure why the cstore one isn't then 2012-03-17T23:54:59 Hmm. Is your srfsh configured for the private router, or the public one? 2012-03-17T23:55:26 Not sure... I followed the steps on evergreen-ils.org 2012-03-17T23:55:37 That's about all I did 2012-03-17T23:55:43 ok 2012-03-17T23:55:53 I set passwords for public/private for both router and opensrf 2012-03-17T23:57:38 Hmm. Could be you don't have any super users. 2012-03-17T23:58:05 Don't believe I do, I only configured the evergreen//evergreen admin 2012-03-17T23:58:22 Try this: request open-ils.cstore open-ils.cstore.direct.actor.user.retrieve 1