Throughout the French LibreOffice UI we can find "English quotation marks". Instead, « French guillemets » should always be used at the first level, according to French language typographical rules that read as follow: * « French guillemets » should always be used at the first level (main quotation) * « Within a quotation (using French guillemets), a second quotation will be surrounded by ‘‘double English quotation marks’’ » * « Should we have three levels of quotation, French guillemets will first be used, ‘‘then double English quotation marks, then ‘single English quotation marks will be used’, if need be’’ ». A non-breaking space will be used after the opening French guillemets and before the closing French guilemets.
On which env are you (Windows, Linux, MacOs)? What LO version do you use? Do you reproduce this on a brand new document?
Created attachment 154052 [details] English quotation marks in the French UI
I’m on Linux. This is not document related, since this English quotation marks appear in the French UI. A general check should be done in the whole French UI.
French.l10n: thank you for your feedback, I understood with the screenshot. Sophie/Jean-Baptiste: On Pootle, I made the changes on 6.2 and 6.3UI for: - Afficher l'alerte "L'aide locale n'est pas installée" - Impression : spécification du statut "Document modifié" Is it ok for you? Any idea when they'll be available in translations repository?
Thank you Julien Nabet for your fast answer and reaction. These were only two examples, but not the only occurrences. Doesn’t Pootle offer a search feature to globally chase all " in the French translation and replace them with « with the proper spacing?
Indeed https://translations.documentfoundation.org/fr/libo_ui/translate/#search=%22&sfields=target,notes :-( (208 items only for UI 6.3)
Good hunting and replace :-)
(In reply to French.l10n from comment #0) > Throughout the French LibreOffice UI we can find "English quotation marks". > Instead, « French guillemets » should always be used at the first level, > according to French language typographical rules that read as follow: > > * « French guillemets » should always be used at the first level (main > quotation) > * « Within a quotation (using French guillemets), a second quotation will be > surrounded by ‘‘double English quotation marks’’ » > * « Should we have three levels of quotation, French guillemets will first > be used, ‘‘then double English quotation marks, then ‘single English > quotation marks will be used’, if need be’’ ». > > A non-breaking space will be used after the opening French guillemets and > before the closing French guilemets. What is your reference ? I am not sure if these typographical rules are shared by each country where French is spoken. Best regards. JBF
@JBF I do not know what typographical rules are used in every French speaking countries in the world, but I’m quite sure that « French guillemets » should always be used at the first level (main quotation). Best,
Well many character string with double quotes instead of French guillemets are in the Tips of the day. So it is possible to make the change offline in the cui/message.po file. But there is a difficulty: I need to differentiate opening \" from closing \" (" are escaped in po files because each string is enclosed between a double quote pair. Do you know a regular expression to replace \"abc\" by «abc» were abc is an arbitrary character string? For remaining single quotes, I do not know how to find them because they are identical to apostrophes. Best regards. JBF
(In reply to Jean-Baptiste Faure from comment #10) > ... > Do you know a regular expression to replace \"abc\" by «abc» were abc is an > arbitrary character string? Something like: sed -i 's/"\([^"]*\)"/«\1»/g' <input file> ?
(In reply to Julien Nabet from comment #11) > (In reply to Jean-Baptiste Faure from comment #10) > > ... > > Do you know a regular expression to replace \"abc\" by «abc» were abc is an > > arbitrary character string? > Something like: > sed -i 's/"\([^"]*\)"/«\1»/g' <input file> > ? Thank you Julien, but this command replace all " even the ones that are not escaped. Best regards. JBF
(In reply to Jean-Baptiste Faure from comment #12) > ... > Thank you Julien, but this command replace all " even the ones that are not > escaped. Sorry, I had forgotten the escaped quote constraint: sed -i 's/\\"\([^"]*\)\\"/«\1»/g' <input file> ?
Thank you very much ! That works and leaves only 14 occurrences where opening and closing \" are on different lines. Best regards. JBF
There is 145 double quotes remaining to be converted and several of them must be done by hand because the English string contains " too. Work is on the way. Best regards. JBF
Dear Jean-Baptiste Faure, This bug has been in ASSIGNED status for more than 3 months without any activity. Resetting it to NEW. Please assign it back to yourself if you're still working on this.
Fixed now in weblate. I changed all relevant pairs "..." and '...' I found in « ... ». Will be available in the next release 6.4.3. Please reopen if you find unchanged pair. Best regards. JBF