- py-configparser-devel@lists.common-lisp.net

Moving the repository to Git+Gitlab?
by Erik Huelsmann 03 Nov '18

03 Nov '18

Hi, How about moving our repository to Git+Gitlab? The result will be 2 repositories: 1 for code, the other for the site. Regards, -- Bye, Erik. http://efficito.com -- Hosted accounting and ERP. Robust and Flexible. No vendor lock-in.

1 0

py-configparser returning (vector character) but not string
by Philipp Marek 06 Mar '18

06 Mar '18

Hi everybody, I'm using py-configparser in one of my projects. After quite some debugging I found out that it returns (VECTOR CHARACTER <length>) instead of plain strings; and in some cases this makes a difference, for example SB-BSD-SOCKETS:GET-HOST-BY-NAME throws an CASE-FAILURE. Perhaps the results should become SIMPLE-STRINGS -- although I'm not sure how widespread such problems/ checks are. Regards, Phil

1 0

[py-configparser-devel] Minor bugs
by Станислав Кондратьев 08 Apr '15

08 Apr '15

parser.lisp (defun is-whitespace (c) (or (eq c #\Space) (eq c #\Tab) (eq c #\Return))) (defun is-comment-char (c) (or (eq c #\;) (eq c #\#))) eq should be changed into char=, since "Common Lisp makes no guarantee that eq is true even when both its arguments are ``the same thing'' if that thing is a character or number." Best regards, S. Kondratyev.

1 0

[py-configparser-devel] get-option: inheritance from `default' broken.
by Станислав Кондратьев 08 Apr '15

08 Apr '15

Hello, In Python's configparser the default section is supposed to supply missing options for the rest sections. In py-configparser this doesn't work. I suggest fixing %get-option as follows: (defun %get-option (config section-name option-name if-does-not-exist) (let* ((section (%get-section config section-name)) (norm-option (norm-option-name config option-name)) (option (or (assoc norm-option (section-options section) :test #'string=) (assoc norm-option (section-options (config-defaults config)) :test #'string=)))) (if (null option) (if (eq if-does-not-exist :error) (error 'no-option-error) ;; no such option error (values (car (push (list (%validate-option-name norm-option)) (section-options section))) section)) (values option section)))) Best regards, Stanislav Kondratyev.

1 1

[py-configparser-devel] get-option: handling the case
by Станислав Кондратьев 08 Apr '15

08 Apr '15

Hello, I've stumbled upon an unexpected behaviour of get-option and related functions. By default the options are converted to lower case. However, if a config file is read by read-files, the case is not converted and the options become inaccessible. The bug must be due to the code of %get-option, where %validate-option-name receives option-name rather than norm-option. Best regards, S. Kondratyev.

1 0

[py-configparser-devel] CLOS, structs and ABCL?
by Abhishek Reddy 07 Aug '10

07 Aug '10

Hi, A comment in config.lisp suggests that structs are used instead of CLOS due to an issue with ABCL as of 2008. Has the situation with ABCL improved, given its substantial development since then? If so, would it make sense to abandon structs for CLOS in py-configparser? -- Abhishek Reddy http://abhishek.geek.nz

2 2

[py-configparser-devel] Incorrect keyword exported
by Abhishek Reddy 09 Mar '09

09 Mar '09

Hi, There is a typo in package.lisp, exporting #:no-section-erorr instead of #:no-section-error. Below is a specific patch. Cheers, Abhishek Index: package.lisp =================================================================== --- package.lisp (revision 46) +++ package.lisp (working copy) @@ -23,7 +23,7 @@ ;; Configuration management ;; Error classes - #:no-section-erorr + #:no-section-error #:duplicate-section-error #:no-option-error #:interpolation-error

2 1

Re: [py-configparser-devel] missing files treatment
by Erik Huelsmann 20 Jun '08

20 Jun '08

On Sat, Jun 21, 2008 at 12:13 AM, Erik Huelsmann <ehuels(a)gmail.com> wrote: >> Well, I committed the change to trunk, so unless you have additional >> changes (or anyone else for that matter), all that remains is creating >> a new release. And the new release is named 1.0.3. Bye, Erik.

1 0

Re: [py-configparser-devel] missing files treatment
by Ury Marshak 14 Jun '08

14 Jun '08

Erik Huelsmann wrote: > I'd say this is indeed the intended behaviour. The code which is > currently there just seems broken to me. I'll commit your change later > tonight and release a new version. (Unless you have more > bugs/improvements to report, in which case it's probably better to > release all of it at once...) > > If I may be so bold: what do you (plan to) use py-configparser for? > > Well, I didn't use it for much (yet), so that's the only thing that I've seen for now. My needs are pretty basic: read some values from a file, which is in a format familiar to many people (especially Windows administrators kind of people). So I can have a small tool that hooks up with an existing database and I can just tell the admin to "edit the .INI" and there's a good chance they'll feel right at home. Easy syntax, basically just a=b, not a lot of opportunities to screw up. The other options are to use the Lisp reader (problems with security, hard to get syntax right for non-lispers) and to invent some other format (what's the point when this one works well), so py-configparser seems perfect for me. I actually _do_ have one suggestion, although it's not directly related to the code and also may seem a little drastic :) , and that is to rename the library (maybe to something like just "configparser"). TBH, I don't see the point of "py-" part of the name. Of course, it's interface is (somewhat, I guess) modeled after python's module, but that warrants only a honorable mention in the docs, not the name. As it is, "py" in the name suggests for me some kind of "runtime" connection to python, which is misleading and actually unfair for a nice pure-lisp general-purpose library. There's also a chance that with the interface can be changed slightly (no, nothing specific, just that some patterns may emerge with usage that will be expressed easily in a more lispy way) and it's good not to be limited by the name. Of course that's not a serious issue, but if we are talking suggestions, I couldn't but mention it. And yes, that can happen in practice - I was actually mislead by the name and read it as "a library to parse (some sort of) configuration of python itself" at first, before having a look at it. Certainly, it could've been just me being dense :) Thanks, Ury

1 0

Re: [py-configparser-devel] missing files treatment
by Erik Huelsmann 14 Jun '08

14 Jun '08

On Fri, Jun 13, 2008 at 6:24 PM, Erik Huelsmann <ehuels(a)gmail.com> wrote: > On 6/13/08, Ury Marshak <urym(a)two-bytes.com> wrote: >> Hi, >> >> It seems at the moment (version 1.0.2) if I specify some files that are >> not present I get an error in (read-files). Python's docs at >> http://docs.python.org/lib/RawConfigParser-objects.html state that "If a >> file named in filenames cannot be opened, that file will be ignored", >> which actually is very useful behavior. So, maybe I'll be possible to >> make a change in this, something like: >> >> diff -u -r py-configparser-1.0.2-orig/parser.lisp >> py-configparser-1.0.2/parser.lisp >> --- py-configparser-1.0.2-orig/parser.lisp 2008-03-16 >> 01:05:40.000000000 +0200 >> +++ py-configparser-1.0.2/parser.lisp 2008-06-03 15:53:01.406250000 +0300 >> @@ -190,7 +190,7 @@ >> >> Returns as values the configuration and the list of files actually read." >> (let (files-read) >> - (dolist (filename (mapcar #'probe-file filenames) >> + (dolist (filename (remove-if-not #'probe-file filenames) >> (values config files-read)) >> (with-open-file (s filename >> :direction :input > > > Hi! > > I'd say this is indeed the intended behaviour. The code which is > currently there just seems broken to me. I'll commit your change later > tonight and release a new version. (Unless you have more > bugs/improvements to report, in which case it's probably better to > release all of it at once...) > > If I may be so bold: what do you (plan to) use py-configparser for? > > Regards, > > Erik. >

1 1