French: false positive report of missing non-breaking space

DidierB · March 24, 2016, 10:36am

Hello,

I use LanguageTool 3.2 embedded in OmegaT development version.

Whenever I enter a “double” punctuation sign, I have the warning “Point d’interrogation (or whatever sign it is) est précédé d’une espace fine insécable” (Question mark should have a leading small non-breaking space) even if I actually use a non-breaking space.

Previously, OmegaT used LanguageTool 2.2 and it works fine.

If I try the online test page I have the same false positive, so it seems unlikely the error comes from OmegaT embedding.

I thought perhaps the rule had become more strict, really requiring a small (“fine”) non-breaking space, but that doesn’t seem the case according to the source code:

github.com

languagetool-org/languagetool/blob/master/languagetool-language-modules/fr/src/main/java/org/languagetool/rules/fr/QuestionWhitespaceRule.java

/* LanguageTool, a natural language style checker
 * Copyright (C) 2005 Daniel Naber (http://www.danielnaber.de)
 *
 * This library is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * This library is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with this library; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301
 * USA
 */
package org.languagetool.rules.fr;

This file has been truncated. show original

// Strictly speaking, the character before ?!;: should be an
// “espace fine insécable” (U+202f). In practise, an
// “espace insécable” (U+00a0) is also often used. Let’s accept both.

Does someone else has the issue?

Thank you for your attention,

Didier

amake · March 24, 2016, 1:00pm

This appears to be either a bug with the rule or bad guidance in the documentation regarding the required Unicode normalization of the input text:

github.com/languagetool-org/languagetool

Inconsistent requirements for Unicode normalization and French spaces

opened 12:58PM - 24 Mar 16 UTC

amake

The French [`QuestionWhitespaceRule`](https://github.com/languagetool-org/langua…getool/blob/6f6c8dd6dbabc92787a0950e51294a64a8c46317/languagetool-language-modules/fr/src/main/java/org/languagetool/rules/fr/QuestionWhitespaceRule.java#L108) requires `U+202f` or `U+00a0` before `[?!:;]`. However, the documentation for `JLanguageTool` states that it [expects input text to be NFKC-normalized](https://github.com/languagetool-org/languagetool/blob/6f6c8dd6dbabc92787a0950e51294a64a8c46317/languagetool-core/src/main/java/org/languagetool/JLanguageTool.java#L58). [Both `U+202f` and `U+00a0` normalize to `U+0020 SPACE` under NFKC](http://www.unicode.org/charts/normalization/chart_Separator-Space.html), so this rule will always be violated if the caller is properly using NFKC. So which is correct?