tedkulp/Readarr
1
0
Fork 0
mirror of https://github.com/Readarr/Readarr.git synced 2026-04-23 22:25:09 -04:00
Code Issues Projects Releases Wiki Activity

Fixed: Improve fuzzy matching algorithm to match around word boundaries

Browse Source 
Fixes READARR-C1
This commit is contained in:
ta264
2022-01-24 20:58:00 +00:00
parent ecf1e1a130
commit c0e193dd1f
5 changed files with 149 additions and 84 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/NzbDrone.Core/Parser/Parser.cs
+6 -56
View File
@@ -203,6 +203,7 @@ namespace NzbDrone.Core.Parser
private static readonly Regex YearInTitleRegex = new Regex(@"^(?<title>.+?)(?:\W|_)?(?<year>\d{4})",
RegexOptions.IgnoreCase | RegexOptions.Compiled);
private static readonly HashSet<char> WordDelimiters = new HashSet<char>(" .,_-=()[]|\"`'’");
private static readonly Regex WordDelimiterRegex = new Regex(@"(\s|\.|,|_|-|=|\(|\)|\[|\]|\|)+", RegexOptions.Compiled);
private static readonly Regex PunctuationRegex = new Regex(@"[^\w\s]", RegexOptions.Compiled);
private static readonly Regex CommonWordRegex = new Regex(@"\b(a|an|the|and|or|of)\b\s?", RegexOptions.IgnoreCase | RegexOptions.Compiled);
@@ -352,7 +353,7 @@ namespace NzbDrone.Core.Parser
simpleTitle = CleanTorrentSuffixRegex.Replace(simpleTitle);
var bestBook = books
.OrderByDescending(x => simpleTitle.FuzzyContains(x.Editions.Value.Single(x => x.Monitored).Title))
.OrderByDescending(x => simpleTitle.FuzzyMatch(x.Editions.Value.Single(x => x.Monitored).Title, wordDelimiters: WordDelimiters))
.First()
.Editions.Value
.Single(x => x.Monitored);
@@ -419,69 +420,18 @@ namespace NzbDrone.Core.Parser
Logger.Trace($"Finding '{name}' in '{report}'");
var (locStart, score) = report.ToLowerInvariant().FuzzyMatch(name.ToLowerInvariant(), 0.6);
var (locStart, matchLength, score) = report.ToLowerInvariant().FuzzyMatch(name.ToLowerInvariant(), 0.6, WordDelimiters);
if (locStart == -1)
{
return null;
}
var diff = (int)Math.Round((1.0 - score) * name.Length, 0);
var length = Math.Min(name.Length + diff, report.Length - locStart);
var found = report.Substring(locStart, matchLength);
var reportReversed = new string(report.Substring(locStart, length).ToLowerInvariant().Reverse().ToArray());
var nameReversed = new string(name.ToLowerInvariant().Reverse().ToArray());
var locEnd = locStart + reportReversed.Length - reportReversed.FuzzyFind(nameReversed, 0.6);
var boundaries = WordDelimiterRegex.Matches(report);
if (boundaries.Count == 0)
if (score >= 0.8)
{
return null;
}
var starts = new List<int>();
var finishes = new List<int>();
if (boundaries[0].Index == 0)
{
starts.Add(boundaries[0].Length);
}
else
{
starts.Add(0);
}
foreach (Match match in boundaries)
{
var start = match.Index + match.Length;
if (start < report.Length)
{
starts.Add(start);
}
var finish = match.Index - 1;
if (finish >= 0)
{
finishes.Add(finish);
}
}
var lastMatch = boundaries[boundaries.Count - 1];
if (lastMatch.Index + lastMatch.Length < report.Length)
{
finishes.Add(report.Length - 1);
}
var wordStart = starts.OrderBy(x => Math.Abs(x - locStart)).First();
var wordEnd = finishes.OrderBy(x => Math.Abs(x - locEnd)).First();
var found = report.Substring(wordStart, wordEnd - wordStart + 1);
if (found.ToLowerInvariant().FuzzyMatch(name.ToLowerInvariant()) >= 0.8)
{
remainder = report.Remove(wordStart, wordEnd - wordStart + 1);
remainder = report.Remove(locStart, matchLength);
return found.Replace('.', ' ').Replace('_', ' ');
}