forked from kit-sdq/artemis4j
-
Notifications
You must be signed in to change notification settings - Fork 0
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
implement annotation merging kit-sdq#41
- Loading branch information
Showing
9 changed files
with
467 additions
and
11 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
186 changes: 186 additions & 0 deletions
186
src/main/java/edu/kit/kastel/sdq/artemis4j/grading/AnnotationMerger.java
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,186 @@ | ||
| /* Licensed under EPL-2.0 2024. */ | ||
| package edu.kit.kastel.sdq.artemis4j.grading; | ||
|
|
||
| import java.text.MessageFormat; | ||
| import java.util.ArrayList; | ||
| import java.util.Collection; | ||
| import java.util.LinkedHashMap; | ||
| import java.util.List; | ||
| import java.util.Locale; | ||
| import java.util.Map; | ||
| import java.util.StringJoiner; | ||
| import java.util.stream.Collectors; | ||
|
|
||
| import edu.kit.kastel.sdq.artemis4j.i18n.FormatString; | ||
|
|
||
| /** | ||
| * Merges annotations based on their classifiers. | ||
| */ | ||
| final class AnnotationMerger { | ||
| // TODO: where should the translations be stored? | ||
| private static final FormatString MERGED_ANNOTATIONS_FORMAT = new FormatString( | ||
| // TODO: should default locale be English or German? | ||
| new MessageFormat("{0}Other problems in {1}.", Locale.ENGLISH), | ||
| Map.of(Locale.GERMAN, new MessageFormat("{0}Weitere Probleme in {1}.", Locale.GERMAN))); | ||
|
|
||
| private AnnotationMerger() {} | ||
|
|
||
| /** | ||
| * Merges annotations based on their classifiers or not if they have none. | ||
| * <p> | ||
| * This method assumes that the annotation uuids are unique. | ||
| * | ||
| * @param unreducedAnnotations the list of annotations to merge | ||
| * @param upperAnnotationLimit the maximum number of annotations for the first classifier | ||
| * @param locale the locale to use for the format string | ||
| * @return the merged list of annotations | ||
| */ | ||
| static List<Annotation> mergeAnnotations( | ||
| Collection<Annotation> unreducedAnnotations, int upperAnnotationLimit, Locale locale) { | ||
| // -1 means no limit (useful for unit tests, where one wants to see all annotations) | ||
| if (upperAnnotationLimit == -1) { | ||
| return new ArrayList<>(unreducedAnnotations); | ||
| } | ||
|
|
||
| // first group all problems by the first classifier: | ||
| Map<String, List<Annotation>> groupedAnnotations = unreducedAnnotations.stream() | ||
| .collect(Collectors.groupingBy( | ||
| annotation -> { | ||
| List<String> classifiers = annotation.getClassifiers(); | ||
| if (classifiers.isEmpty()) { | ||
| // do not merge annotations without a classifier: | ||
| return annotation.getUUID(); | ||
| } else { | ||
| return classifiers.get(0); | ||
| } | ||
| }, | ||
| LinkedHashMap::new, | ||
| Collectors.toList())); | ||
|
|
||
| List<Annotation> result = new ArrayList<>(); | ||
| for (List<Annotation> annotationsForClassifier : groupedAnnotations.values()) { | ||
| // if the annotation limit is set, use it (if it does not exceed the upper limit), | ||
| // otherwise use the upper limit | ||
| int targetNumberOfAnnotations = Math.min( | ||
| upperAnnotationLimit, | ||
| annotationsForClassifier.get(0).getAnnotationLimit().orElse(upperAnnotationLimit)); | ||
|
|
||
| if (annotationsForClassifier.size() <= targetNumberOfAnnotations) { | ||
| result.addAll(annotationsForClassifier); | ||
| continue; | ||
| } | ||
|
|
||
| // Further partition the annotations by their remaining classifiers: | ||
| annotationsForClassifier.stream() | ||
| .collect(Collectors.groupingBy( | ||
| annotation -> { | ||
| List<String> classifiers = annotation.getClassifiers(); | ||
| if (classifiers.size() <= 1) { | ||
| return annotation.getUUID(); | ||
| } else { | ||
| // to simplify the grouping code, we merge the remaining classifiers | ||
| // into a single string | ||
| return String.join(" ", classifiers.subList(1, classifiers.size())); | ||
| } | ||
| }, | ||
| LinkedHashMap::new, | ||
| Collectors.toList())) | ||
| .values() | ||
| .stream() | ||
| .flatMap(list -> merge(list, targetNumberOfAnnotations, locale).stream()) | ||
| .forEach(result::add); | ||
| } | ||
|
|
||
| return result; | ||
| } | ||
|
|
||
| private static List<Annotation> merge(List<Annotation> annotations, int limit, Locale locale) { | ||
| // use a dumb algorithm: keep the first limit - 1 annotations, and merge the remainder into a single annotation | ||
| if (annotations.size() <= limit) { | ||
| return annotations; | ||
| } | ||
|
|
||
| List<Annotation> result = new ArrayList<>(annotations.subList(0, limit - 1)); | ||
| List<Annotation> toMerge = annotations.subList(limit - 1, annotations.size()); | ||
|
|
||
| // first we try to find an annotation with a custom message, this is the main one that will be shown to the user | ||
| int firstIndexWithCustomMessage = 0; | ||
| for (int i = 0; i < toMerge.size(); i++) { | ||
| if (toMerge.get(i).getCustomMessage().isPresent()) { | ||
| firstIndexWithCustomMessage = i; | ||
| break; | ||
| } | ||
| } | ||
|
|
||
| // the first annotation with the custom message is removed, so it doesn't appear twice | ||
| Annotation firstAnnotation = toMerge.remove(firstIndexWithCustomMessage); | ||
| // if there are no annotations with a custom message, the firstIndexWithCustomMessage will be 0, | ||
| // and because it doesn't have a custom message it will be null | ||
| String message = firstAnnotation.getCustomMessage().orElse(null); | ||
| if (message == null) { | ||
| message = ""; | ||
| } else { | ||
| // some messages might not end with a period, which would look weird with the above format string, | ||
| // so this adds one if necessary | ||
| if (!message.endsWith(".")) { | ||
| message += "."; | ||
| } | ||
|
|
||
| message += " "; | ||
| } | ||
|
|
||
| String customMessage = MERGED_ANNOTATIONS_FORMAT | ||
| .format(message, displayLocations(firstAnnotation, toMerge)) | ||
| .translateTo(locale); | ||
|
|
||
| result.add(new Annotation( | ||
| firstAnnotation.getMistakeType(), | ||
| firstAnnotation.getFilePath(), | ||
| firstAnnotation.getStartLine(), | ||
| firstAnnotation.getEndLine(), | ||
| customMessage, | ||
| firstAnnotation.getCustomScore().orElse(null), | ||
| firstAnnotation.getSource())); | ||
|
|
||
| return result; | ||
| } | ||
|
|
||
| private static String displayLocations(Annotation first, Collection<Annotation> others) { | ||
| Map<String, List<Annotation>> positionsByFile = others.stream() | ||
| .collect(Collectors.groupingBy(Annotation::getFilePath, LinkedHashMap::new, Collectors.toList())); | ||
|
|
||
| // if all annotations are in the same file, we don't need to display the filename | ||
| boolean withoutFilename = positionsByFile.size() == 1 && positionsByFile.containsKey(first.getFilePath()); | ||
|
|
||
| StringJoiner joiner = new StringJoiner(", "); | ||
| // Format should look like this: File:(L1, L2, L3), File2:(L4, L5), File3:L5 | ||
| for (Map.Entry<String, List<Annotation>> entry : positionsByFile.entrySet()) { | ||
| String path = entry.getKey(); | ||
| List<Annotation> filePositions = entry.getValue(); | ||
|
|
||
| String lines = filePositions.stream() | ||
| .map(position -> "L%d".formatted(position.getStartLine())) | ||
| .collect(Collectors.joining(", ")); | ||
|
|
||
| if (filePositions.size() > 1 && !withoutFilename) { | ||
| lines = "(%s)".formatted(lines); | ||
| } | ||
|
|
||
| if (withoutFilename) { | ||
| joiner.add(lines); | ||
| continue; | ||
| } | ||
|
|
||
| joiner.add("%s:%s".formatted(getFilenameWithoutExtension(path), lines)); | ||
| } | ||
|
|
||
| return joiner.toString(); | ||
| } | ||
|
|
||
| private static String getFilenameWithoutExtension(String path) { | ||
| String[] parts = path.split("[\\\\\\/]"); | ||
| String file = parts[parts.length - 1]; | ||
|
|
||
| return file.split("\\.")[0]; | ||
| } | ||
| } |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.