[en] Case conversion does not give the correct result for UPPER CASE text

[MikeUnwalla]

"Note: LT automatically adjusts the case of suggestions if they are added as plain text." (https://dev.languagetool.org/tips-and-tricks#changing-the-case-of-matched-word)

        <rule id="CASE_CONVERSION_TEST" name="Case conversion test">
            <pattern>
                <token>cat</token>
            </pattern>
            <message>Use '<suggestion>dog</suggestion>'.</message>
            <example correction="dog">My <marker>cat</marker> is happy.</example>
            <example correction="Dog">My <marker>Cat</marker> is fat.</example>
            <example correction="DOG">MY <marker>CAT</marker> SAT ON THE MAT.</example>
        </rule>

Testrules gives this message (LanguageTool-20220919-snapshot.zip):

Exception in thread "main" org.languagetool.rules.patterns.PatternRuleTest$PatternRuleTestFailure: Test failure for rule CASE_CONVERSION_TEST[1] in file /org/languagetool/rules/en/grammar.xml: Incorrect suggestions: Expected 'DOG', got: 'Dog' on input: 'MY CAT SAT ON THE MAT.'

[jaumeortola]

You are right. The default behaviour of the suggestions should be similar to case_converson="preserve" for all languages. I will take a look. Probably, we will need to adjust some tests in many languages.

[jaumeortola]

All-uppercase suggestions have been implemented.

Once we preserve the case for all-uppercase words in the original sentence automatically, rule creators will need a way to make exceptions to all-uppercase suggestions. The way to do that, is to use a <match> with a case_conversion attribute:

<suggestion><match no="2" regexp_match="^.*$" regexp_replace="Liebe Grüße" case_conversion="startupper"/></suggestion>

The <match> has to be at the start of the first suggestion. This won't work:

<suggestion>L<match no="2" regexp_match="^.*$" regexp_replace="iebe" case_conversion="alllower/></suggestion>

[MikeUnwalla]

@jaumeortola , thank you.