Mirror/MarginaliaSearch
1
0
Fork 0
mirror of https://github.com/MarginaliaSearch/MarginaliaSearch.git synced 2025-02-23 21:18:58 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

(coded-sequence) Handle weird legacy HTML that puts everything in a heading

Browse Source
This commit is contained in:
Viktor Lofgren 2024-08-26 12:49:15 +02:00
parent 7d471ec30d
commit 67a98fb0b0
1 changed files with 31 additions and 27 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

58
code/index/java/nu/marginalia/index/results/IndexResultScoreCalculator.java
View File

@ -251,36 +251,40 @@ public class IndexResultScoreCalculator {
int firstPosition = 1;
for (int i = 0; i < weightedCounts.length; i++) {
if (positions[i] != null && ctx.regularMask.get(i)) {
searchableKeywordsCount ++;
int[] posArray = positions[i].toIntArray();
if (positions[i] == null || !ctx.regularMask.get(i))
continue;
for (int idx = 0; idx < positions[i].size(); idx++) {
int pos = positions[i].getInt(idx);
firstPosition = Math.max(firstPosition, pos);
}
searchableKeywordsCount ++;
int[] posArray = positions[i].toIntArray();
int cnt;
if ((cnt = spans.title.countIntersections(posArray)) != 0) {
unorderedMatchInTitleCount++;
weightedCounts[i] += 2.5f * cnt;
}
if ((cnt = spans.heading.countIntersections(posArray)) != 0) {
for (int idx = 0; idx < positions[i].size(); idx++) {
int pos = positions[i].getInt(idx);
firstPosition = Math.max(firstPosition, pos);
}
int cnt;
if ((cnt = spans.title.countIntersections(posArray)) != 0) {
unorderedMatchInTitleCount++;
weightedCounts[i] += 2.5f * cnt;
}
if ((cnt = spans.heading.countIntersections(posArray)) != 0) {
if (spans.heading.size() < 64) {
// Correct for the case where there's a lot of headings everywhere, or the entire document is a heading
unorderedMatchInHeadingCount++;
weightedCounts[i] += 2.5f * cnt;
}
if ((cnt = spans.code.countIntersections(posArray)) != 0) {
weightedCounts[i] += 0.25f * cnt;
}
if ((cnt = spans.anchor.countIntersections(posArray)) != 0) {
weightedCounts[i] += 0.2f * cnt;
}
if ((cnt = spans.nav.countIntersections(posArray)) != 0) {
weightedCounts[i] += 0.1f * cnt;
}
if ((cnt = spans.body.countIntersections(posArray)) != 0) {
weightedCounts[i] += 1.0f * cnt;
}
weightedCounts[i] += 2.5f * cnt;
}
if ((cnt = spans.code.countIntersections(posArray)) != 0) {
weightedCounts[i] += 0.25f * cnt;
}
if ((cnt = spans.anchor.countIntersections(posArray)) != 0) {
weightedCounts[i] += 0.2f * cnt;
}
if ((cnt = spans.nav.countIntersections(posArray)) != 0) {
weightedCounts[i] += 0.1f * cnt;
}
if ((cnt = spans.body.countIntersections(posArray)) != 0) {
weightedCounts[i] += 1.0f * cnt;
}
}
@ -290,7 +294,7 @@ public class IndexResultScoreCalculator {
}
if (!verbatimMatches.get(HtmlTag.HEADING) && unorderedMatchInHeadingCount == searchableKeywordsCount) {
verbatimMatchScore += 2.0f * unorderedMatchInHeadingCount;
verbatimMatchScore += 1.0f * unorderedMatchInHeadingCount;
}
double overallPart = averageSentenceLengthPenalty