buildKeywordsFromDictionary function presentation exported ✓ 100.0%

Last updated: 2026-02-24T21:07:57.551Z

Location

Metrics

LOC: 21 Complexity: 4 Params: 0 Coverage: 100.0% (9/9 lines, 4x executed)

Signature

buildKeywordsFromDictionary(): : Record<FieldType, string[]>

Summary

Generates a Record<FieldType, string[]> keyword map directly from the field dictionary. This can replace / supplement the static FIELD_TYPE_KEYWORDS in tensorflow-generator.ts.

Source Code

export function buildKeywordsFromDictionary(): Record<FieldType, string[]> {
  const result = {} as Record<FieldType, string[]>;

  for (const entry of FIELD_DICTIONARY) {
    const keywords: string[] = [];

    // Derive keywords from tags (tags are the classification hints)
    if (entry.tags.length > 0) {
      keywords.push(...entry.tags);
    }

    // Derive keywords from entry type itself (e.g. "cpf" → ["cpf"])
    if (!keywords.includes(entry.type)) {
      keywords.push(entry.type);
    }

    result[entry.type] = [...new Set(keywords)]; // dedupe
  }

  return result;
}

Dependencies (Outgoing)

graph LR buildKeywordsFromDictionary["buildKeywordsFromDictionary"] FieldType["FieldType"] FIELD_DICTIONARY["FIELD_DICTIONARY"] flattenStructuredSignals["flattenStructuredSignals"] getTrainingDistribution["getTrainingDistribution"] TrainingSample["TrainingSample"] augmentShuffle["augmentShuffle"] augmentDrop["augmentDrop"] augmentTypo["augmentTypo"] checkDatasetHealth["checkDatasetHealth"] AccuracyThresholds["AccuracyThresholds"] DatasetHealthReport["DatasetHealthReport"] classifyField["classifyField"] FormField["FormField"] fromFlatSignals["fromFlatSignals"] buildKeywordsFromDictionary -->|uses| FieldType buildKeywordsFromDictionary -->|uses| FIELD_DICTIONARY buildKeywordsFromDictionary -->|uses| flattenStructuredSignals buildKeywordsFromDictionary -->|uses| getTrainingDistribution buildKeywordsFromDictionary -->|uses| TrainingSample buildKeywordsFromDictionary -->|uses| augmentShuffle buildKeywordsFromDictionary -->|uses| augmentDrop buildKeywordsFromDictionary -->|uses| augmentTypo buildKeywordsFromDictionary -->|uses| checkDatasetHealth buildKeywordsFromDictionary -->|uses| AccuracyThresholds buildKeywordsFromDictionary -->|uses| DatasetHealthReport buildKeywordsFromDictionary -->|uses| classifyField buildKeywordsFromDictionary -->|uses| FormField buildKeywordsFromDictionary -->|uses| fromFlatSignals style buildKeywordsFromDictionary fill:#dbeafe,stroke:#2563eb,stroke-width:2px click buildKeywordsFromDictionary "dfb7d9bfef2aba38.html" click FieldType "e49ed8a601306673.html" click FIELD_DICTIONARY "9c3a352ea31b8001.html" click flattenStructuredSignals "a5d367e2c926f76e.html" click getTrainingDistribution "a91cb4e44bdcf500.html" click TrainingSample "01996f51712dacc4.html" click augmentShuffle "e927efe70011297a.html" click augmentDrop "fc8dabee39176edb.html" click augmentTypo "d345acd14e868d5a.html" click checkDatasetHealth "3be8e84862588bff.html" click AccuracyThresholds "6615b48ce086e8dd.html" click DatasetHealthReport "31043e5ebc950c0b.html" click classifyField "aa03a8b1140f5f42.html" click FormField "85cdbded927c91b0.html" click fromFlatSignals "de70361bba9181a6.html"

Target	Type
FieldType	uses
FIELD_DICTIONARY	uses
flattenStructuredSignals	uses
getTrainingDistribution	uses
TrainingSample	uses
augmentShuffle	uses
augmentDrop	uses
augmentTypo	uses
checkDatasetHealth	uses
AccuracyThresholds	uses
DatasetHealthReport	uses
classifyField	uses
FormField	uses
fromFlatSignals	uses

Impact (Incoming)

graph LR buildKeywordsFromDictionary["buildKeywordsFromDictionary"] EvalMisclassified["EvalMisclassified"] FIELD_DICTIONARY["FIELD_DICTIONARY"] EvalMisclassified -->|uses| buildKeywordsFromDictionary FIELD_DICTIONARY -->|uses| buildKeywordsFromDictionary style buildKeywordsFromDictionary fill:#dbeafe,stroke:#2563eb,stroke-width:2px click buildKeywordsFromDictionary "dfb7d9bfef2aba38.html" click EvalMisclassified "4e0d00ef54656ad2.html" click FIELD_DICTIONARY "9c3a352ea31b8001.html"

Source	Type
EvalMisclassified	uses
FIELD_DICTIONARY	uses