Skip to content

Commit

Permalink
Remove a redundant text from the translation (#38)
Browse files Browse the repository at this point in the history
  • Loading branch information
jazzman committed Apr 16, 2023
2 parents d4def6c + a2cd9b1 commit af8db85
Show file tree
Hide file tree
Showing 3 changed files with 28 additions and 11 deletions.
13 changes: 5 additions & 8 deletions src/main/java/pro/jazzman/odmiana/services/wikislownik/Html.java
Original file line number Diff line number Diff line change
Expand Up @@ -43,21 +43,18 @@ public String translation(String language) {
return null;
}

String text = element.text();

return Arrays.stream(
text.substring(text.indexOf(":") + 2) // remove language
.replaceAll("\\(\\d+\\.\\d+\\) ", "") // remove enumeration
element
.text()
.substring(element.text().indexOf(":") + 2)
.split(";")
)
.map(e -> e.split(","))
.flatMap(Stream::of)
.map(String::trim)
.map(e -> e.replaceAll("\\s[żmwn]$", "")) // remove kind
.map(e -> e.replaceAll("\\s*\\(\\S+\\)\\s*", ""))
.map(e -> e.replaceAll("\\s[żmwn]$", ""))
.map(String::trim)
.distinct()
.collect(Collectors.joining(", "));
}


}
Original file line number Diff line number Diff line change
@@ -1,7 +1,6 @@
package pro.jazzman.odmiana.services.wikislownik;

import org.jsoup.nodes.Document;
import org.junit.jupiter.api.BeforeEach;
import org.junit.jupiter.api.DisplayName;
import org.junit.jupiter.api.Test;
import org.junit.jupiter.api.extension.ExtendWith;
Expand All @@ -11,7 +10,6 @@
import pro.jazzman.odmiana.parsers.Parser;

import static org.mockito.ArgumentMatchers.any;
import static org.mockito.Mockito.verify;
import static org.assertj.core.api.Assertions.*;
import static org.mockito.Mockito.when;

Expand All @@ -30,4 +28,26 @@ void parseWordNotFoundThrowsException() throws Exception {
() -> new Html("<html></html>", parser).parse(LANG)
).isExactlyInstanceOf(ApplicationRuntimeException.class);
}

@Test
void translation() {
String markup = """
<ul>
<li>ukraiński:
<span>(1.1)</span>
<a href="/">test</a>
</li>
<li>
angielski:
<span>(1.1)</span> <a href="/">foo</a> <link rel="" href=""><span><a href="/"><span><span>m</span></span></a></span> (qux);
<span>(1.2)</span> <a href="/">bar</a> <link rel="" href=""><span><a href="/"><span><span>ż</span></span></a></span> (quux),
<span>(1.3)</span> <a href="/">baz</a> <link rel="" href=""><span><a href="/"><span><span>n</span></span></a></span> (quuux);
<span>(1.4)</span> <a href="/">foo</a> <link rel="" href=""><span><a href="/"><span><span>n</span></span></a></span> (quuuux),
<span>(1.5)</span> <a href="/">navy blue</a> <link rel="" href=""><span><a href="/"><span><span>n</span></span></a></span> (blue)
</li>
</ul>
""";

assertThat(new Html(markup, parser).translation("en")).isEqualTo("foo, bar, baz, navy blue");
}
}
2 changes: 1 addition & 1 deletion src/test/resources/telegram/responses/verb.success.txt
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
*kochać* - love, be in love (with), make love
*kochać* - love, be in love, make love

*Liczba Pojedyncza* | *Mnoga*

Expand Down

0 comments on commit af8db85

Please sign in to comment.