Describe the bug
I have a sentence like:

String resp = "Yes because it's: $~num~ for ~misc~ payment contract $~num~ for planTOTAL $~num~";

but I can't configure the parameters correctly to get "plan total". I have tried the default settings and various combinations of parameters like:

         SpellCheckSettings spellCheckSettings = SpellCheckSettings.builder()
//            .countThreshold(1).deletionWeight(1).insertionWeight(1)
//            .replaceWeight(1).maxEditDistance(2).transpositionWeight(1).topK(5)
//            .prefixLength(10).verbosity(Verbosity.ALL)
            .build();

         DataHolder dataHolder = new InMemoryDataHolder(spellCheckSettings,
            new Murmur3HashFunction());

         StringDistance weightedDamerauLevenshteinDistance = new WeightedDamerauLevenshteinDistance(
            spellCheckSettings.getDeletionWeight(),
            spellCheckSettings.getInsertionWeight(),
            spellCheckSettings.getReplaceWeight(),
            spellCheckSettings.getTranspositionWeight(), new QwertyDistance());

         SymSpellCheck checker = new SymSpellCheck(dataHolder,
            weightedDamerauLevenshteinDistance, spellCheckSettings);

         List<SuggestionItem> suggestions = checker.lookupCompound(resp, 1.0d, true );

To Reproduce
Steps to reproduce the behavior:

1. Use code above and examine the suggestions.get(0).getTerm() to see that there is no split.

Expected behavior
The term should show "plan total" (or better, preserving case... (is there an option for this?) "plan TOTAL") and not "plantotal"

Desktop (please complete the following information):

• OS: MacOS 10.15.7
• openjdk version "1.8.0_252"

Additional context
Providing more details on the impact of settings would be helpful for people not familiar with the art.

This may be related to #53

Upgrade the current Symspell algorithm from 6.6 to 6.7.1

SpellHelper.earlyExit(suggestionItems, phrase, maxEditDistance);
Adding items inside this method seems to be wrong

Precondition:
SpellCheckSettings is initiated with maxEditDistance > 0.

I want to separately cover corner case with missed space, but only between correct words (maxEditdistance=0 for each word separately).
But it's impossible to do with the same SymSpell if it was created with SpellCheckSettings with maxEditDistance > 0.

To cover the case with missed space, lookupCompound() has method lookupSplitWords().
Inside it split a word into part1 and part2.
For each lookup() is called. It has the following code:

    if (maxEditDistance <= 0) {
      maxEditDistance = spellCheckSettings.getMaxEditDistance();
    }

Now the scenario:
query: {applewatch}

Scenario:
I want to lookup for missed space between only correctly spelled words, which means maxEditDistance = 1 (missed space).

With the current implementation, SymSpell will look for extra space between 2 words with additional edit distance by 1. And there is no way to prevent this.
Total maxEditDistance:
lookup(part1, maxEditDistance) = 1
lookup(part2, maxEditDistance) = 1
lookupCompaund(part1+part1) = 1
Total = 3

Depending on the dictionary following results are possible:
{apple watch}, editDistance = 1
{apple patch}, editDistance = 2 (if watch is not present in the dictionary)
{apply patch}, editDistance = 3 (if both apple and watch is not present in the dictionary)

So, i am trying to use symspell to do post ocr correction in financial domain.

So for a sample sentence like Pro f i t in the year 2020 symspell is giving suggestions as prof it in the year 20 20

Could anyone suggest in

1. correct pro f i t as profit
2. for numbers or for some regex don't apply any segmentation

Hello, I have found this case that seems strange for the input string "I am the begt spell cherken!":

int maxEd = 2;
suggestionItems = symSpellCheck.lookupCompound("I am the begt spell cherken!", maxEd);
    for (SuggestionItem elem : suggestionItems) {
      System.out.println("compound : " + elem.getTerm().trim());
    }

I'm getting compound : a am the best spell cher ken

My setup is the default one:

    SpellCheckSettings spellCheckSettings = SpellCheckSettings.builder().countThreshold(1).deletionWeight(1f)
        .insertionWeight(1f).replaceWeight(1f).maxEditDistance(2).transpositionWeight(1f).topK(5).prefixLength(10)
        .verbosity(Verbosity.ALL).build();

    dataHolder = new InMemoryDataHolder(spellCheckSettings, new Murmur3HashFunction());

// weighted Damerau-Levenshtein
    weightedDamerauLevenshteinDistance = new WeightedDamerauLevenshteinDistance(spellCheckSettings.getDeletionWeight(),
        spellCheckSettings.getInsertionWeight(), spellCheckSettings.getReplaceWeight(),
        spellCheckSettings.getTranspositionWeight(), null);

    symSpellCheck = new SymSpellCheck(dataHolder, weightedDamerauLevenshteinDistance, spellCheckSettings);

First I just want to say thank you for the awesome project!

I ported it to Kotlin Multiplatform and had a few questions, I know it's been a long time, so I understand if you don't remember.

Some of my tests are failing, very slight errors in word distance and such which I'm investigating, but one in particular puzzled me:
testWordBreak()

Passes in:
"itwasabrightcolddayinaprilandtheclockswerestrikingthirteen"

And expects back:
"it was bright cold day in april and the clock were striking thirteen"

Which is missing the "a" between "was" and "bright", as well as the "s" at the end of "clocks".

My Kotlin version returns what looks more correct to me:
"it was a bright cold day in april and the clocks were striking thirteen"

Now my question is this: Did I inadvertently fix a bug? Or is the output you are expecting in your test actually correct, and an expected side effect of the algorithm?

Plugin for Solr, to use symspell lib

Describe the bug
The build is failing for long, the deployment step is failing with unauthorised error

Hello.
now, I am testing customized-symspell at solr.
I wanted to see the situation when I put slives query, I get slices, not olives.
But I faced a situation where I couldn't see the result that I want.

Before the result,
first, I indexed document that include slices and olives keyword.
Second, I put jar files on specific core.

[jar files]

• commons-collections4-4.4.jar
• murmur-1.0.0.jar
• solr-commons-1.2.34.jar
• symspell-lib-6.6.154.jar
• symspell-service-6.6.154.jar
• symspell-solr-6.6.154.jar

Third, updated solrconfig.xml file like the image below. (lib path, component, requesthandler)

[add the component and requesthandler]

Is there anything else that I have to do?
please help me ! 😢

private static Double getEdistance(double maxEditDistance, int length) {
double factor = 0.3;
double computedEd = Math.round(factor * length);
if (Math.min(maxEditDistance, computedEd) == maxEditDistance) {
return maxEditDistance;
}
return computedEd;
}

This factor should be configurable in Spellchecksettings.

SymSpellCheck.class
lookupCompound()
Line 113
``` joinedTerm = joinedTerm.concat(si.getTerm()).concat(" ");````
results that the joinedTerm end with space.

For example, the next line that calculates editDistance trim() the joinedTerm, while it should trim() the stored result.

phrase: "ipd"
extected joinedTerm: "ipad"
actual joinedTerm: "ipad "

Same for the multiword phase.

From docs to Verbosity.ALL:

All suggestions within maxEditDistance, suggestions ordered by edit distance, 
then by term frequency (slower, no early termination).

Several times faced cases when suggestions are not sorted correctly.
Neither by editDistance neither by count.
Settings mostly default. Please see them on the screenshots.

Hello, thanks a lot for this porting! I was using another SymSpell Java port, but out of date now.
I have successfully compiled the project. Do you have a basic execution example?
Thanks a lot.

Could you please create maven repository so that we can use this as library?

Hi, for e it looks like the method isTransposition is bugged:

private boolean isTransposition(int i, int j, String a, String b) {
return i > 2
&& j > 2
&& a.charAt(i - 2) == a.charAt(i - 1)
&& b.charAt(j - 2) == b.charAt(j - 1);
}

The letters of the strings a and b are not compared with each other. You check whether a and b have consecutive characters being identical. Shouldn't it be the following?

private boolean isTransposition(int i, int j, String a, String b) {
return i > 2
&& j > 2
&& a.charAt(i - 2) == b.charAt(j - 1)
&& b.charAt(j - 2) == a.charAt(i - 1);
}