Re: package.el strings

[Jean-Christophe Helary]

It took me a lot longer than I thought but here is a diff for your comments. 
There is a dozen modifications in the file and a few comments.

There is 1 modification that is not related to (future) l10n, it was discussed 
in a different thread (using format to create https vs http). The rest is about 
native strings.

I have checked:

* DONE %
* DONE (... :tag "..."
* DONE (cl-defstruct
* DONE (completing-read
* DONE (completing-read-multiple
* DONE (concat -> problems
* DONE (defconst
* DONE (defcustom
* DONE (defface
* DONE (define-error
* DONE (defun
* DONE (defvar
* DONE (easy-menu-define
* DONE (error
* DONE (format
* DONE (format-message
* DONE (insert -> problems
* DONE (interactive -> prompt should be defined separately
* DONE (interactive-only
* DONE (mapconcat
* DONE (message
* DONE (prin1
* DONE (princ
* DONE (read-string
* DONE (setq
* DONE (user-error
* DONE (with-demoted-errors

There are string issues that I could not solve with some concat and some insert 
and as I wrote in May (describe-package-1) seems to use a lot of magic to 
create strings. I've fixed the prin1/princ block but there are other places 
where I'm not sure how to proceed.

Here is the "git diff master" file.

Jean-Christophe

l10n_package.diff
Description: Binary data

> On May 24, 2017, at 14:08, Jean-Christophe Helary <address@hidden> wrote:
> 
> I've tried to straighten up the strings so that they stand on their own and 
> don't rely of complex concatenations and variable substitutions. That makes 
> for much less "smart" code but the resulting code/strings are more readable.
> 
> There is one big chunk left though (I put 2 TODOs there). It is the 
> (describe-package-1) function. The function itself is about 200 lines long 
> and there are 2 places I'm not sure yet what to do with, one uses prin1/princ 
> and the other uses insert to generate strings. I'll check that part later.
> 
> Let me know what you think.
> 
> Jean-Christophe
> 
> <package.el_0524.diff>