From kiuma72 at gmail.com Tue Mar 4 16:27:41 2008 From: kiuma72 at gmail.com (Andrea Chiumenti) Date: Tue, 4 Mar 2008 17:27:41 +0100 Subject: [cl-l10n-devel] Some modification requested Message-ID: <4d3bc9370803040827n6d333f39i7e5237ddd9f93dcd@mail.gmail.com> Hello, I'm new here nice library I'd like to point out some possible implementations: the possibility to not specify the decimal digits: (defun format-number (stream arg no-dp no-ts &optional (locale (current-locale))) (let ((locale (locale locale)) (float-part (float-part (coerce (abs arg) 'double-float)))) (cl:format stream (getf (printers locale) (if no-ts :number-no-ts :number-ts)) (get-sign arg locale) (truncate (abs arg)) (unless (and (string= "" float-part) no-dp) (list (locale-decimal-point locale) (if *float-digits* (fix-float-string float-part *float-digits*) float-part)))) (values))) this will permit the following: CL-L10N> (let ((cl-l10n:*locale* "it_CH")(cl-l10n:*float-digits* nil)) (format nil "~:/cl-l10n:format-number/" (coerce 1/100000000000000000000000000000000 'double-float))) "0,00000000000000000000000000000001" Secondly it_IT is missing: a solution should be if "it_XX" is not found use the locale "it" and if "xx_XX" and "xx" doesn't exist use the "en" locale. Let me know, cheers kiuma -------------- next part -------------- An HTML attachment was scrubbed... URL: From rosssd at gmail.com Wed Mar 5 10:30:08 2008 From: rosssd at gmail.com (Sean Ross) Date: Wed, 5 Mar 2008 10:30:08 +0000 Subject: [cl-l10n-devel] Some modification requested In-Reply-To: <4d3bc9370803040827n6d333f39i7e5237ddd9f93dcd@mail.gmail.com> References: <4d3bc9370803040827n6d333f39i7e5237ddd9f93dcd@mail.gmail.com> Message-ID: <5bef28df0803050230g561394bava52e58b13cbcd49@mail.gmail.com> On 3/4/08, Andrea Chiumenti wrote: > Hello, > I'm new here nice library > > I'd like to point out some possible implementations: Hi there, thanks for the suggestions and i will look over them (hopefull tomorrow) when i am slightly less stressed. cheers, sean. From attila.lendvai at gmail.com Wed Mar 5 11:05:51 2008 From: attila.lendvai at gmail.com (Attila Lendvai) Date: Wed, 5 Mar 2008 12:05:51 +0100 Subject: [cl-l10n-devel] Some modification requested In-Reply-To: <4d3bc9370803040827n6d333f39i7e5237ddd9f93dcd@mail.gmail.com> References: <4d3bc9370803040827n6d333f39i7e5237ddd9f93dcd@mail.gmail.com> Message-ID: > CL-L10N> (let ((cl-l10n:*locale* "it_CH")(cl-l10n:*float-digits* nil)) > (format nil "~:/cl-l10n:format-number/" (coerce > 1/100000000000000000000000000000000 'double-float))) > "0,00000000000000000000000000000001" thanks, pushed. > Secondly it_IT is missing: > a solution should be if "it_XX" is not found use the locale "it" it's already done: CL-L10N> (locale "it_IT") # but i've pushed a little more dwim-ish code that works in your case, too. CL-L10N> (locale 'it_IT) # -- attila From kiuma72 at gmail.com Wed Mar 5 15:29:06 2008 From: kiuma72 at gmail.com (Andrea Chiumenti) Date: Wed, 5 Mar 2008 16:29:06 +0100 Subject: [cl-l10n-devel] problem in compiling hungarian with sbcl 1.0.15 Message-ID: <4d3bc9370803050729m5ee76c63w9cac9f7f1073c419@mail.gmail.com> When doing for the first time (having no .fasl) (asdf:oos 'asdf:load-op :cl-l10n) I've this error on swank ; /home/kiuma/.fasls/sbcl-1.0.15-gentoo-linux-x86-64/home/kiuma/lisp/cl-l10n/languages/hungarian.fasl written ; compilation finished in 0:00:01 ; ; compilation unit finished ; printed 26 notes ") ;; Event history end. ;; Backtrace: ;; Connection to Emacs lost. [ ;; condition: Unable to display error condition: # is closed. ;; type: SB-INT:STREAM-ENCODING-ERROR ;; encoding: NIL style: :SPAWN dedicated: NIL] The problem doesn't appear anymode on successive load-op Cheers, kiuma -------------- next part -------------- An HTML attachment was scrubbed... URL: From rosssd at gmail.com Wed Mar 5 15:46:53 2008 From: rosssd at gmail.com (Sean Ross) Date: Wed, 5 Mar 2008 15:46:53 +0000 Subject: [cl-l10n-devel] problem in compiling hungarian with sbcl 1.0.15 In-Reply-To: <4d3bc9370803050729m5ee76c63w9cac9f7f1073c419@mail.gmail.com> References: <4d3bc9370803050729m5ee76c63w9cac9f7f1073c419@mail.gmail.com> Message-ID: <5bef28df0803050746p31fbacc2m69fe63665c39ec1@mail.gmail.com> On 3/5/08, Andrea Chiumenti wrote: > When doing for the first time (having no .fasl) > (asdf:oos 'asdf:load-op :cl-l10n) > > I've this error on swank Does this error occur when compiling with sbcl in a terminal? sean. From kiuma72 at gmail.com Wed Mar 5 16:13:28 2008 From: kiuma72 at gmail.com (Andrea Chiumenti) Date: Wed, 5 Mar 2008 17:13:28 +0100 Subject: [cl-l10n-devel] Problem with it_IT locale Message-ID: <4d3bc9370803050813h6437c09dg7027929acf3e8655@mail.gmail.com> Am I doing something wrong or there is a small bug into locale definition? (let ((cl-l10n:*locale* "it_IT")(cl-l10n:*float-digits* nil)) (format nil "~:/cl-l10n:format-number/" (coerce 1/100000000000000000000000000000000 'double-float))) arithmetic error DIVISION-BY-ZERO signalled Operation was SB-KERNEL::DIVISION, operands (0 0). [Condition of type DIVISION-BY-ZERO] Restarts: 0: [ABORT] Return to SLIME's top level. 1: [TERMINATE-THREAD] Terminate this thread (#) Backtrace: 0: (TRUNCATE 0 0) 1: (SB-FORMAT::FORMAT-ADD-COMMAS "0" #\, 0) 2: (SB-FORMAT::FORMAT-PRINT-INTEGER # 0 T NIL 10 0 #\ #\, 0) 3: (SB-FORMAT::D-FORMAT-DIRECTIVE-INTERPRETER # #<~,,',,0:D> (#<~{> #<~A> #<~}>) # #) 4: (SB-FORMAT::INTERPRET-DIRECTIVE-LIST # (#<~,,',,0:D> #<~{> #<~A> #<~}>) ("" 0 ("," "00000000000000000000000000000001")) (0 ("," "00000000000000000000000000000001"))) 5: (SB-FORMAT::%FORMAT # "~A~,,',,0:D~{~A~}" ("" 0 ("," "00000000000000000000000000000001")) ("" 0 ("," "00000000000000000000000000000001"))) 6: (COMMON-LISP:FORMAT # "~A~,,',,0:D~{~A~}")[:EXTERNAL] 7: (FORMAT-NUMBER # 1.d-32 T NIL "it_IT") 8: (SB-FORMAT::/-FORMAT-DIRECTIVE-INTERPRETER # #<~:/cl-l10n:format-number/> NIL # #) 9: (SB-FORMAT::INTERPRET-DIRECTIVE-LIST # (#<~:/cl-l10n:format-number/>) (1.d-32) (1.d-32)) 10: (SB-FORMAT::%FORMAT # "~:/cl-l10n:format-number/" (1.d-32) (1.d-32)) 11: (COMMON-LISP:FORMAT NIL "~:/cl-l10n:format-number/")[:EXTERNAL] 12: ((LAMBDA ())) 13: (SB-INT:SIMPLE-EVAL-IN-LEXENV (LET ((*LOCALE* "it_IT") (*FLOAT-DIGITS* NIL)) (FORMAT NIL "~:/cl-l10n:format-number/" (COERCE 1/100000000000000000000000000000000 'DOUBLE-FLOAT))) #) 14: (SWANK::EVAL-REGION " (let ((cl-l10n:*locale* \"it_IT\")(cl-l10n:*float-digits* nil)) (format nil \"~:/cl-l10n:format-number/\" (coerce 1/100000000000000000000000000000000 'double-float))) ") 15: ((LAMBDA ())) 16: (SWANK::TRACK-PACKAGE #) 17: ((LAMBDA (SWANK-BACKEND::FN)) #) 18: (SWANK::CALL-WITH-BUFFER-SYNTAX #) 19: (SWANK::REPL-EVAL " (let ((cl-l10n:*locale* \"it_IT\")(cl-l10n:*float-digits* nil)) (format nil \"~:/cl-l10n:format-number/\" (coerce 1/100000000000000000000000000000000 'double-float))) ") 20: (SB-INT:SIMPLE-EVAL-IN-LEXENV (SWANK:LISTENER-EVAL " (let ((cl-l10n:*locale* \"it_IT\")(cl-l10n:*float-digits* nil)) (format nil \"~:/cl-l10n:format-number/\" (coerce 1/100000000000000000000000000000000 'double-float))) ") #) 21: ((LAMBDA ())) 22: ((LAMBDA (SWANK-BACKEND::HOOK SWANK-BACKEND::FUN)) # #) 23: ((LAMBDA ())) 24: ((LAMBDA (SWANK-BACKEND::HOOK SWANK-BACKEND::FUN)) # #) 25: (SWANK::CALL-WITH-REDIRECTED-IO # #) 26: (SWANK::CALL-WITH-CONNECTION # #) 27: (SWANK::HANDLE-REQUEST #) 28: (SWANK::REPL-LOOP #) 29: (SWANK::REPL-LOOP #)[:EXTERNAL] 30: (SWANK::CALL-WITH-BINDINGS NIL #) 31: ((FLET SB-THREAD::WITH-MUTEX-THUNK)) 32: ((FLET #:WITHOUT-INTERRUPTS-BODY-[CALL-WITH-MUTEX]438)) 33: (SB-THREAD::CALL-WITH-MUTEX # #S(SB-THREAD:MUTEX :NAME "thread result lock" :%OWNER # :STATE 1) # T) 34: ((LAMBDA ())) 35: ("foreign function: #x41D832") 36: ("foreign function: #x415644") -------------- next part -------------- An HTML attachment was scrubbed... URL: From attila.lendvai at gmail.com Wed Mar 5 16:44:51 2008 From: attila.lendvai at gmail.com (Attila Lendvai) Date: Wed, 5 Mar 2008 17:44:51 +0100 Subject: [cl-l10n-devel] Problem with it_IT locale In-Reply-To: <4d3bc9370803050813h6437c09dg7027929acf3e8655@mail.gmail.com> References: <4d3bc9370803050813h6437c09dg7027929acf3e8655@mail.gmail.com> Message-ID: > (let ((cl-l10n:*locale* "it_IT")(cl-l10n:*float-digits* nil)) (format nil > "~:/cl-l10n:format-number/" (coerce 1/100000000000000000000000000000000 > 'double-float))) use with-locale, it's not that simple as binding a string... -- attila From attila.lendvai at gmail.com Wed Mar 5 16:52:47 2008 From: attila.lendvai at gmail.com (Attila Lendvai) Date: Wed, 5 Mar 2008 17:52:47 +0100 Subject: [cl-l10n-devel] problem in compiling hungarian with sbcl 1.0.15 In-Reply-To: <4d3bc9370803050729m5ee76c63w9cac9f7f1073c419@mail.gmail.com> References: <4d3bc9370803050729m5ee76c63w9cac9f7f1073c419@mail.gmail.com> Message-ID: > When doing for the first time (having no .fasl) > (asdf:oos 'asdf:load-op :cl-l10n) i've pushed a patch that will complain if the .lisp files are not processed in utf-8 encoding. for now this is #+(and sbcl sb-unicode) and it's only in hungarian.lisp, but it may be crucial later for the entire project when/if other languages get some utilities/resources that need unicode... -- attila From rosssd at gmail.com Wed Mar 5 17:00:56 2008 From: rosssd at gmail.com (Sean Ross) Date: Wed, 5 Mar 2008 17:00:56 +0000 Subject: [cl-l10n-devel] Problem with it_IT locale In-Reply-To: <4d3bc9370803050813h6437c09dg7027929acf3e8655@mail.gmail.com> References: <4d3bc9370803050813h6437c09dg7027929acf3e8655@mail.gmail.com> Message-ID: <5bef28df0803050900x2e894661v68c837e4ef174bf5@mail.gmail.com> On 3/5/08, Andrea Chiumenti wrote: > Am I doing something wrong or there is a small bug into locale definition? > > (let ((cl-l10n:*locale* "it_IT")(cl-l10n:*float-digits* nil)) (format nil > "~:/cl-l10n:format-number/" (coerce > 1/100000000000000000000000000000000 'double-float))) hmm, it appears that the grouping in LC_NUMERIC for it_IT is 0;0 (as opposed to it_CH's 3;3) which is causing this particular error. grepping through the various locales seems to indicate that this is not an anomaly but suggests that cl-l10n does not handle grouping correctly when printing numeric values (although monetary printing is fine which is rather odd). I'm unsure what 0;0 means in this context and at first glance google hasn't turned up anything particularly useful. any ideas? From rosssd at gmail.com Wed Mar 5 17:01:05 2008 From: rosssd at gmail.com (Sean Ross) Date: Wed, 5 Mar 2008 17:01:05 +0000 Subject: [cl-l10n-devel] Problem with it_IT locale In-Reply-To: References: <4d3bc9370803050813h6437c09dg7027929acf3e8655@mail.gmail.com> Message-ID: <5bef28df0803050901y2ce73acfvd8c43da504ef282e@mail.gmail.com> On 3/5/08, Attila Lendvai wrote: > > (let ((cl-l10n:*locale* "it_IT")(cl-l10n:*float-digits* nil)) (format nil > > "~:/cl-l10n:format-number/" (coerce 1/100000000000000000000000000000000 > > 'double-float))) > > > use with-locale, it's not that simple as binding a string... Although it is undocumented and rather unused, i would expect it to work, as the cl-l10n operators do accept locale designators rather than locale objects. From rosssd at gmail.com Wed Mar 5 17:01:05 2008 From: rosssd at gmail.com (Sean Ross) Date: Wed, 5 Mar 2008 17:01:05 +0000 Subject: [cl-l10n-devel] Problem with it_IT locale In-Reply-To: References: <4d3bc9370803050813h6437c09dg7027929acf3e8655@mail.gmail.com> Message-ID: <5bef28df0803050901y2ce73acfvd8c43da504ef282e@mail.gmail.com> On 3/5/08, Attila Lendvai wrote: > > (let ((cl-l10n:*locale* "it_IT")(cl-l10n:*float-digits* nil)) (format nil > > "~:/cl-l10n:format-number/" (coerce 1/100000000000000000000000000000000 > > 'double-float))) > > > use with-locale, it's not that simple as binding a string... Although it is undocumented and rather unused, i would expect it to work, as the cl-l10n operators do accept locale designators rather than locale objects. From rosssd at gmail.com Wed Mar 5 17:15:52 2008 From: rosssd at gmail.com (Sean Ross) Date: Wed, 5 Mar 2008 17:15:52 +0000 Subject: [cl-l10n-devel] Problem with it_IT locale In-Reply-To: <5bef28df0803050900x2e894661v68c837e4ef174bf5@mail.gmail.com> References: <4d3bc9370803050813h6437c09dg7027929acf3e8655@mail.gmail.com> <5bef28df0803050900x2e894661v68c837e4ef174bf5@mail.gmail.com> Message-ID: <5bef28df0803050915w7c091918t210e6689533044ca@mail.gmail.com> On 3/5/08, Sean Ross wrote: > any ideas? I do believe i'm being a little slow. I'm now under the impression that the 0;0 indicates that NO grouping is to be performed (which is supported by a blank thousands-seperator). I've pushed a quick patch to allow numeric printing for locales with blank seperators and will look to getting a more complete patch (which handles the #\; seperated values more appropriately) in the future. cheers, sean. From attila.lendvai at gmail.com Wed Mar 5 18:01:31 2008 From: attila.lendvai at gmail.com (Attila Lendvai) Date: Wed, 5 Mar 2008 19:01:31 +0100 Subject: [cl-l10n-devel] Problem with it_IT locale In-Reply-To: <5bef28df0803050901y2ce73acfvd8c43da504ef282e@mail.gmail.com> References: <4d3bc9370803050813h6437c09dg7027929acf3e8655@mail.gmail.com> <5bef28df0803050901y2ce73acfvd8c43da504ef282e@mail.gmail.com> Message-ID: > > use with-locale, it's not that simple as binding a string... > > Although it is undocumented and rather unused, i would expect it to > work, as the cl-l10n operators do accept locale designators rather > than locale objects. it's arguable and i'm not strongly against it. the rational behind the current setup is that if *locale* is always a normalized list of locale instances (precedence list) then all code dealing with *locale* can assume this and avoid the headache (and some performance loss if each function needs to look up the locales for itself). while using with-locale, this normalization is only at one place. i'd instead argue for not exporting *locale* and advertising current-locale and with-locale. -- attila From kiuma72 at gmail.com Wed Mar 5 18:46:31 2008 From: kiuma72 at gmail.com (Andrea Chiumenti) Date: Wed, 5 Mar 2008 19:46:31 +0100 Subject: [cl-l10n-devel] Problem with it_IT locale In-Reply-To: <4d3bc9370803051045t862a56ka281f0976168fed0@mail.gmail.com> References: <4d3bc9370803050813h6437c09dg7027929acf3e8655@mail.gmail.com> <5bef28df0803050901y2ce73acfvd8c43da504ef282e@mail.gmail.com> <4d3bc9370803051045t862a56ka281f0976168fed0@mail.gmail.com> Message-ID: <4d3bc9370803051046l3e1c087an240845988ed0619d@mail.gmail.com> Sorry! On Wed, Mar 5, 2008 at 7:45 PM, Andrea Chiumenti wrote: > > (with-locale (locale 'it_IT) (let ((cl-l10n:*float-digits* nil)) (format > nil "~:/cl-l10n:format-number/" 1000.04)) ) > => "1,0,0,0,0399780273438" > should be 1'000,0399780273438 > > (with-locale (locale 'it_IT) (format-number *standard-output* 99.04 t > nil)) > => 9,9,04; > should be > 99,44 > > > > On Wed, Mar 5, 2008 at 7:01 PM, Attila Lendvai > wrote: > > > > > use with-locale, it's not that simple as binding a string... > > > > > > Although it is undocumented and rather unused, i would expect it to > > > work, as the cl-l10n operators do accept locale designators rather > > > than locale objects. > > > > it's arguable and i'm not strongly against it. the rational behind the > > current setup is that if *locale* is always a normalized list of > > locale instances (precedence list) then all code dealing with *locale* > > can assume this and avoid the headache (and some performance loss if > > each function needs to look up the locales for itself). > > > > while using with-locale, this normalization is only at one place. > > > > i'd instead argue for not exporting *locale* and advertising > > current-locale and with-locale. > > > > -- > > attila > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From attila.lendvai at gmail.com Wed Mar 5 18:55:16 2008 From: attila.lendvai at gmail.com (Attila Lendvai) Date: Wed, 5 Mar 2008 19:55:16 +0100 Subject: [cl-l10n-devel] Problem with it_IT locale In-Reply-To: <5bef28df0803050900x2e894661v68c837e4ef174bf5@mail.gmail.com> References: <4d3bc9370803050813h6437c09dg7027929acf3e8655@mail.gmail.com> <5bef28df0803050900x2e894661v68c837e4ef174bf5@mail.gmail.com> Message-ID: > any ideas? i tried to get a repository where the latest locales are and i've found it in the libc cvs in libc/localedata (http://www.gnu.org/software/libc/resources.html). i updated cl-l10n based on this and i was happy for a while but then i realized that the hungarian locale has many wrong entries and generally it's not what my gnome programs are working by, because some of the wrongness was clearly good in the gui programs. so i'm not sure where to get the bunch of good locale files and what is used by gnome if not these locales. then i became desperate and fixed a few things in the hu_HU locale in the cl-l10n repo, but this is clearly a bad idea in the long run... any ideas? -- attila From kiuma72 at gmail.com Wed Mar 5 19:24:40 2008 From: kiuma72 at gmail.com (Andrea Chiumenti) Date: Wed, 5 Mar 2008 20:24:40 +0100 Subject: [cl-l10n-devel] Problem with it_IT locale In-Reply-To: References: <4d3bc9370803050813h6437c09dg7027929acf3e8655@mail.gmail.com> <5bef28df0803050900x2e894661v68c837e4ef174bf5@mail.gmail.com> Message-ID: <4d3bc9370803051124w57274f17hc5972410480cbc1d@mail.gmail.com> Yes! found problem with it_IT the patch applied helped me to find the problem! LC_NUMERIC was LC_NUMERIC decimal_point "" thousands_sep "" grouping 0;0 END LC_NUMERIC but should be LC_NUMERIC decimal_point "" thousands_sep "" grouping 3;3 END LC_NUMERIC On Wed, Mar 5, 2008 at 7:55 PM, Attila Lendvai wrote: > > any ideas? > > i tried to get a repository where the latest locales are and i've > found it in the libc cvs in libc/localedata > (http://www.gnu.org/software/libc/resources.html). > > i updated cl-l10n based on this and i was happy for a while but then i > realized that the hungarian locale has many wrong entries and > generally it's not what my gnome programs are working by, because some > of the wrongness was clearly good in the gui programs. so i'm not sure > where to get the bunch of good locale files and what is used by gnome > if not these locales. > > then i became desperate and fixed a few things in the hu_HU locale in > the cl-l10n repo, but this is clearly a bad idea in the long run... > > any ideas? > > -- > attila > -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: it_IT Type: application/octet-stream Size: 5012 bytes Desc: not available URL: From rosssd at gmail.com Thu Mar 6 11:02:31 2008 From: rosssd at gmail.com (Sean Ross) Date: Thu, 6 Mar 2008 11:02:31 +0000 Subject: [cl-l10n-devel] Problem with it_IT locale In-Reply-To: <4d3bc9370803051124w57274f17hc5972410480cbc1d@mail.gmail.com> References: <4d3bc9370803050813h6437c09dg7027929acf3e8655@mail.gmail.com> <5bef28df0803050900x2e894661v68c837e4ef174bf5@mail.gmail.com> <4d3bc9370803051124w57274f17hc5972410480cbc1d@mail.gmail.com> Message-ID: <5bef28df0803060302j2aa26c8bwbd33d628fcf2dba7@mail.gmail.com> On 3/5/08, Andrea Chiumenti wrote: > Yes! > found problem with it_IT the patch applied helped me to find the problem! > > LC_NUMERIC was > > LC_NUMERIC > decimal_point "" > thousands_sep "" > grouping 0;0 > END LC_NUMERIC > > but should be > > LC_NUMERIC > decimal_point "" > thousands_sep "" > grouping 3;3 > END LC_NUMERIC As always I'm a little nervous about editing the locale files directly as this sort of thing is easy to revert when changing to newer locale files although that doesn't mean that having this in the interem is a bad thing. Out of curiosity is it the norm in italy to use seperators and group numbers into thousands? sean. From attila.lendvai at gmail.com Sat Mar 8 21:11:20 2008 From: attila.lendvai at gmail.com (Attila Lendvai) Date: Sat, 8 Mar 2008 22:11:20 +0100 Subject: [cl-l10n-devel] cldr refactoring/experiment Message-ID: dear list, i've created a branch at darcs get http://common-lisp.net/project/cl-l10n/darcs/cl-l10n.cldr/ it is a first scratch at how it would look like if we parsed the cldr xmls instead of the libc locale files which are kind of obsolete now. i think it's much more promising, but there's plenty of work left to do... :| also note that the code is really in a half-done stage, but the stuff in cldr-parsing.lisp is ok, just needs much more work. you need to run bin/update-cldr.sh and then (asdf:oos 'asdf:test-op :cl-l10n) should work with a few tests that test the locale inheritance. i've done some controversial changes like using stefil for the test suite or depending on alexandria, but i hope the usefulness of the quickly delivered results will out weight this impoliteness eventually... -- attila From rosssd at gmail.com Mon Mar 10 11:57:20 2008 From: rosssd at gmail.com (Sean Ross) Date: Mon, 10 Mar 2008 11:57:20 +0000 Subject: [cl-l10n-devel] cldr refactoring/experiment In-Reply-To: References: Message-ID: <5bef28df0803100457x3fadc76cp4f660087e725970f@mail.gmail.com> On 3/8/08, Attila Lendvai wrote: > dear list, > > i've created a branch at > > darcs get http://common-lisp.net/project/cl-l10n/darcs/cl-l10n.cldr/ > > it is a first scratch at how it would look like if we parsed the cldr > xmls instead of the libc locale files which are kind of obsolete now. > i think it's much more promising, but there's plenty of work left to > do... :| > > also note that the code is really in a half-done stage, but the stuff > in cldr-parsing.lisp is ok, just needs much more work. > > you need to run bin/update-cldr.sh and then (asdf:oos 'asdf:test-op > :cl-l10n) should work with a few tests that test the locale > inheritance. > > i've done some controversial changes like using stefil for the test > suite or depending on alexandria, but i hope the usefulness of the > quickly delivered results will out weight this impoliteness > eventually... > thanks for all that, hopefully (work permitting) I'll be able to take a look at this during this week. sean.