From mboxrd@z Thu Jan 1 00:00:00 1970 From: Rasmus Subject: Re: [bug?, ox-odt] Format DATE Date: Mon, 10 Nov 2014 11:06:49 +0100 Message-ID: <874mu774jq.fsf@gmx.us> References: <87oasky213.fsf@gmx.us> <87a940o40u.fsf@gmx.us> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Return-path: Received: from eggs.gnu.org ([2001:4830:134:3::10]:39058) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1XnlsG-00067x-NT for emacs-orgmode@gnu.org; Mon, 10 Nov 2014 05:07:11 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1XnlsA-0002s8-Db for emacs-orgmode@gnu.org; Mon, 10 Nov 2014 05:07:04 -0500 Received: from mout.gmx.net ([212.227.15.18]:54082) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1XnlsA-0002rs-4W for emacs-orgmode@gnu.org; Mon, 10 Nov 2014 05:06:58 -0500 In-Reply-To: (Christian Moe's message of "Mon, 10 Nov 2014 09:53:20 +0100") List-Id: "General discussions about Org-mode." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: emacs-orgmode-bounces+geo-emacs-orgmode=m.gmane.org@gnu.org Sender: emacs-orgmode-bounces+geo-emacs-orgmode=m.gmane.org@gnu.org To: mail@christianmoe.com Cc: emacs-orgmode@gnu.org Christian Moe writes: >>> Going by the documentation of org-odt-use-date-fields, the data styles >>> "OrgDate1" and "OrgDate2" are supposed to be mapped from >>> org-time-stamp-custom-formats, rather than >>> org-export-date-timestamp-format.=20 >> >> Yeah, I saw that, but the description of >> `org-time-stamp-custom-formats' looks a bit opaque. I wondered if it >> would affect the recognition of timestamps in buffer (since the >> variable is not an org-export-=C2=B7 variable). If it's truly *only* >> affecting exports it should be renamed. > > No, judging by the manual entry, it was intended for customizing the > appearance of timestamps in the buffer, which it does, see below. To > accommodate people who think ISO time format is just too darned > /sensible/... O-o-okay. > But it has the nice (and little-mentioned) side effect that you can use > it to change the appearance of timestamps in exports, where > non-geeky-looking everyday-language dates are often comme il faut. Works > nicely in HTML, for instance, except that you still get geeky-looking > brackets around the dates (may need a filter to remove those). While I really want this functionality, I think that this way of using it is a bug. It should depend on a separate org-export-=C2=B7 variable. .. > It's nice, for instance, if you have a document with a lot of dates in a > European format and need to switch to an American format; you can do it > by changing a setting. Wait, so does this variable change how dates are interpreted or not? I.e. can I "choose" whether <2010-01-10> is January 10th or October first? Or is an Org timestamp always in sensible ISO? > And the ODT exporter builds on this to give date fields that can be > further changed in LibreOffice. Right. >>> So this will apply to the output from the DATE keyword too. >>> >>> To make this happen, org-display-custom-times must be non-nil.=20 >>> This affects not only the date in the heading from the DATE keyword, >>> but also all other timestamps in the document. >>> >>> Having org-display-custom-times turned on all the time also puts >>> overlays on the timestamps in your buffer, but if this annoys you you >>> can bind it to be set during export only. >> >> OK. I don't know this functionality. It sounds less bad that what I >> feared, but still the org-export-=C2=B7 variables should probably be >> sufficient. >> >> Is `org-time-stamp-custom-formats' the recommend way of formatting >> regular time stamps for export? > > I don't know about "recommended". It doesn't seem to be documented for > export at all. And as for ODT, the code comments say the feature > translating from custom time stamp formats to ODT date styles is > "experimental". Seems to work OK, though. I like the possibility, but I think export and buffer display should be separated. >>> But doing it this way still ignores >>> org-export-date-timestamp-format, and any solution based on copying that >>> variable into org-time-stamp-custom-formats makes unsafe assumptions ab= out >>> user preferences. >> >> Yes. >> >>> It seems to me that the export of the DATE keyword ought to honor a >>> non-nil org-export-date-timestamp-format, whether or not the user is >>> applying custom formatting to other timestamps.=20 >>> But that would take some >>> changes to several parts of ox-odt.el, I think. >> >> Yeah. It's can be made even more complicated than that, since the >> *document language* of the output also affects how dates are >> formatted. . . > > In Org, or in LibreOffice once you set the language there?=20 In LO some of the date formatting is a function of the document language. The document language (or maybe region language, I don't really know...) is seen at the bottom. Date has a gray background indicating it's special... With sufficient time and creativity you can get to a menu for formatting the date. Some of these formats (especially the "human-friendly" ones) will depend on the language. On the other hand, "Table of Contents" seems completely independent of the document language. . . > If the #+LANGUAGE keyword is supposed to affect date formatting in Org > output, I must be missing a trick. I think #+DATE might not be truly affected by #+LANGUAGE, after all, but rather be affected of different system languages on my two home-PC and work-PC (I'm unable to get this computer=E2=80=94which has English as a sys= tem language=E2=80=94to export dates in other languages via org-export-date-timestamp-format).=20=20 This is probably a separate bug. >> So a two step method would be: (0) make sure that the document >> language is set correctly (similar to how the right babel language is >> selected in LaTeX), and (1) be able to change the format of the date. >> >> Or we could lose the date-stamp-feature and insert the date as >> plaintext. This is probably simpler, but I don't know if this is the >> "correct" way. >>> Rasmus, will you be pursuing this? >> If you are thinking about fixing this, I won't stop you! I dread >> =C2=B7xml. This weekend sort of disappeared in (other kinds of) wasted >> efforts, so I haven't progressed on this. > Not anytime soon, I'm afraid; pressed for time. That's fine. I will try to look into this; hopefully soon. =E2=80=94Rasmus --=20 What will be next?