I can't get compound word errors to be split properly

[wnm3]

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

[wnm3]

I realized after creating this issue that I wasn't loading any dictionaries. It would help if the README.md showed examples for loading the two dictionaries you have so people can see the spell checker working. Also, I was calling lookup rather than lookupCompound. After looking at the test cases, I figured out what I was missing.

One thing that was also confusing to me is how Verbosity.ALL is used in the SymSpellC?heck class. While the lookup method takes Verbosity as a parameter, the other methods ignore what was passed and provide a hardcoded Verbosity.TOP value. I'd tried referencing what was passed but that broke a bunch of test cases so I backed out that approach.

I'll close this hoping you might update the README to include references to loading the dictionaries to help others wanting to try your libraries.

[MighTguY]

HI @wnm3 , for bringing out this one, I am reopening the same. and will close once I update the ReadMe. :)
Thanks again 🙌