lingucomponent/source/spellcheck/languagetool/languagetoolimp.cxx | 18
+++++-----
1 file changed, 9 insertions(+), 9 deletions(-)
New commits:
commit a57c8fdeb21f4ee77763833a3c92ea337a649cda
Author: Mike Kaganski <mike.kagan...@collabora.com>
AuthorDate: Thu Nov 23 13:21:02 2023 +0100
Commit: Xisco Fauli <xiscofa...@libreoffice.org>
CommitDate: Fri Nov 24 16:29:48 2023 +0100
tdf#158117: encode the username sent to LanguageTool
Change-Id: I762ebb7487b3bcb80cd88b265b4e75c0d8b7c639
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/159764
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kagan...@collabora.com>
Signed-off-by: Xisco Fauli <xiscofa...@libreoffice.org>
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/159910
diff --git a/lingucomponent/source/spellcheck/languagetool/languagetoolimp.cxx
b/lingucomponent/source/spellcheck/languagetool/languagetoolimp.cxx
index 7e61e7a8cca9..63c8307aadf9 100644
--- a/lingucomponent/source/spellcheck/languagetool/languagetoolimp.cxx
+++ b/lingucomponent/source/spellcheck/languagetool/languagetoolimp.cxx
@@ -84,15 +84,16 @@ PropertyValue lcl_GetLineColorPropertyFromErrorId(const
std::string& rErrorId)
return comphelper::makePropertyValue("LineColor", aColor);
}
-OString encodeTextForLanguageTool(const OUString& text)
+OString encodeTextForLT(const OUString& text)
{
// Let's be a bit conservative. I don't find a good description what needs
encoding (and in
// which way) at https://languagetool.org/http-api/; the "Try it out!"
function shows that
// different cases are handled differently by the demo; some
percent-encode the UTF-8
// representation, like %D0%90 (for cyrillic А); some turn into entities
like ! (for
- // exclamation mark !); some other to things like \u0027 (for apostrophe
').
+ // exclamation mark !); some other to things like \u0027 (for apostrophe
'). So only keep
+ // RFC 3986's "Unreserved Characters" set unencoded, use UTF-8
percent-encoding for the rest.
static constexpr auto myCharClass = rtl::createUriCharClass(
- u8"0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz");
+
u8"-._~0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz");
return OUStringToOString(
rtl::Uri::encode(text, myCharClass.data(), rtl_UriEncodeStrict,
RTL_TEXTENCODING_UTF8),
RTL_TEXTENCODING_ASCII_US);
@@ -202,10 +203,9 @@ std::string makeHttpRequest(std::u16string_view aURL,
HTTP_METHOD method, const
{
OString apiKey
= OUStringToOString(LanguageToolCfg::ApiKey::get().value_or(""),
RTL_TEXTENCODING_UTF8);
- OString username =
OUStringToOString(LanguageToolCfg::Username::get().value_or(""),
- RTL_TEXTENCODING_UTF8);
+ OUString username = LanguageToolCfg::Username::get().value_or("");
if (!apiKey.isEmpty() && !username.isEmpty())
- realPostData += "&username=" + username + "&apiKey=" + apiKey;
+ realPostData += "&username=" + encodeTextForLT(username) +
"&apiKey=" + apiKey;
}
return makeHttpRequest_impl(aURL, method, realPostData, nullptr,
nStatusCode);
@@ -460,7 +460,7 @@ ProofreadingResult SAL_CALL
LanguageToolGrammarChecker::doProofreading(
}
else
{
- postData = "text=" + encodeTextForLanguageTool(aText) + "&language=" +
langTag;
+ postData = "text=" + encodeTextForLT(aText) + "&language=" + langTag;
}
if (auto cachedResult = mCachedResults.find(postData); cachedResult !=
mCachedResults.end())
commit 6b4bcad9da516ee11f52f96452ba83ba5234b1b1
Author: Gökay Şatır <gokaysa...@collabora.com>
AuthorDate: Mon Oct 16 12:26:10 2023 +0300
Commit: Xisco Fauli <xiscofa...@libreoffice.org>
CommitDate: Fri Nov 24 16:29:42 2023 +0100
Don't encode the text when sending it to Duden corrector.
Signed-off-by: Gökay Şatır <gokaysa...@collabora.com>
Change-Id: I9e1a2adede04858e5c43b878786bbcc28922aa5f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/158023
Tested-by: Jenkins CollaboraOffice <jenkinscollaboraoff...@gmail.com>
(cherry picked from commit 8989cba47fce3763229005b1ed2fec74da7cfb72)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/158333
Tested-by: Jenkins
Reviewed-by: Caolán McNamara <caolan.mcnam...@collabora.com>
Signed-off-by: Xisco Fauli <xiscofa...@libreoffice.org>
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/159889
diff --git a/lingucomponent/source/spellcheck/languagetool/languagetoolimp.cxx
b/lingucomponent/source/spellcheck/languagetool/languagetoolimp.cxx
index c5062b84fbe7..7e61e7a8cca9 100644
--- a/lingucomponent/source/spellcheck/languagetool/languagetoolimp.cxx
+++ b/lingucomponent/source/spellcheck/languagetool/languagetoolimp.cxx
@@ -444,14 +444,14 @@ ProofreadingResult SAL_CALL
LanguageToolGrammarChecker::doProofreading(
= std::min(xRes.nStartOfNextSentencePosition, aText.getLength());
OString langTag(LanguageTag::convertToBcp47(aLocale, false).toUtf8());
- OString postData = encodeTextForLanguageTool(aText);
+ OString postData;
const bool bDudenProtocol =
LanguageToolCfg::RestProtocol::get().value_or("") == "duden";
if (bDudenProtocol)
{
std::stringstream aStream;
boost::property_tree::ptree aTree;
aTree.put("text-language", langTag.getStr());
- aTree.put("text", postData.getStr());
+ aTree.put("text", aText.toUtf8()); // We don't encode the text in
Duden Corrector tool case.
aTree.put("hyphenation", false);
aTree.put("spellchecking-level", 3);
aTree.put("correction-proposals", true);
@@ -460,7 +460,7 @@ ProofreadingResult SAL_CALL
LanguageToolGrammarChecker::doProofreading(
}
else
{
- postData = "text=" + postData + "&language=" + langTag;
+ postData = "text=" + encodeTextForLanguageTool(aText) + "&language=" +
langTag;
}
if (auto cachedResult = mCachedResults.find(postData); cachedResult !=
mCachedResults.end())
