Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

expose document level term positions #1337

Merged
merged 6 commits into from
Aug 4, 2020
Merged

expose document level term positions #1337

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
3f7892f
add get_document_posting
nsndimt Jun 21, 2020
6b2bd2c
improve variable and function naming
nsndimt Jul 27, 2020
2a78ca9
Merge branch 'master' into add_getPostionList
nsndimt Aug 3, 2020
36561cd
use camel case naming conversion; add testcase
nsndimt Aug 4, 2020
bcc5bbe
use int instead of long
nsndimt Aug 4, 2020
48ed904
use primitive type to avoid autoboxing
nsndimt Aug 4, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
42 changes: 42 additions & 0 deletions src/main/java/io/anserini/index/IndexReaderUtils.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -383,6 +383,48 @@ public static Map<String, Long> getDocumentVector(IndexReader reader, String doc
return docVector;
}

/**
* Returns the term position mapping for a particular document. Note that this method explicitly returns
* {@code null} if the document does not exist (as opposed to an empty map), so that the caller is explicitly forced
* to handle this case.
*
* @param reader index reader
* @param docid collection docid
* @return term position mapping for a particular document or {@code null} if document does not exist.
* @throws IOException if error encountered during query
* @throws NotStoredException if the term vector is not stored
*/
public static Map<String, List<Integer>> getTermPositions(IndexReader reader, String docid) throws IOException, NotStoredException {
int ldocid = convertDocidToLuceneDocid(reader, docid);
if (ldocid == -1) {
return null;
}
Terms terms = reader.getTermVector(ldocid, IndexArgs.CONTENTS);
if (terms == null) {
throw new NotStoredException("Document vector not stored!");
}
TermsEnum termIter = terms.iterator();
if (termIter == null) {
throw new NotStoredException("Document vector not stored!");
}

Map<String, List<Integer>> termPosition = new HashMap<>();
PostingsEnum positionIter = null;

while ((termIter.next()) != null) {
List<Integer> positions = new ArrayList<>();
long termFreq = termIter.totalTermFreq();
positionIter = termIter.postings(positionIter, PostingsEnum.POSITIONS);
positionIter.nextDoc();
for ( int i = 0; i < termFreq; i++ ) {
positions.add(positionIter.nextPosition());
}
termPosition.put(termIter.term().utf8ToString(), positions);
}

return termPosition;
}

/**
* Returns the Lucene {@link Document} based on a collection docid. The method is named to be consistent with Lucene's
* {@link IndexReader#document(int)}, contra Java's standard method naming conventions.
Expand Down
32 changes: 32 additions & 0 deletions src/test/java/io/anserini/index/IndexReaderUtilsTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -381,6 +381,38 @@ public void testDocumentVector() throws Exception {
dir.close();
}

@Test
public void testTermPositions() throws Exception {
Directory dir = FSDirectory.open(tempDir1);
IndexReader reader = DirectoryReader.open(dir);

Map<String, List<Integer>> termPositions;

termPositions = IndexReaderUtils.getTermPositions(reader, "doc1");
assertEquals(Integer.valueOf(0), termPositions.get("here").get(0));
assertEquals(Integer.valueOf(4), termPositions.get("here").get(1));
assertEquals(Integer.valueOf(2), termPositions.get("some").get(0));
assertEquals(Integer.valueOf(6), termPositions.get("some").get(1));
assertEquals(Integer.valueOf(3), termPositions.get("text").get(0));
assertEquals(Integer.valueOf(8), termPositions.get("text").get(1));
assertEquals(Integer.valueOf(7), termPositions.get("more").get(0));
assertEquals(Integer.valueOf(9), termPositions.get("citi").get(0));

termPositions = IndexReaderUtils.getTermPositions(reader, "doc2");
assertEquals(Integer.valueOf(0), termPositions.get("more").get(0));
assertEquals(Integer.valueOf(1), termPositions.get("text").get(0));

termPositions = IndexReaderUtils.getTermPositions(reader, "doc3");
assertEquals(Integer.valueOf(0), termPositions.get("here").get(0));
assertEquals(Integer.valueOf(3), termPositions.get("test").get(0));

// Invalid docid.
assertTrue(IndexReaderUtils.getDocumentVector(reader, "foo") == null);

reader.close();
dir.close();
}

@Test
public void testGetDocumentRaw() throws Exception {
Directory dir = FSDirectory.open(tempDir1);
Expand Down