ezmedicationinput 0.1.44 → 0.1.45

package/dist/segment.js

@@ -1,203 +0,0 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.splitSigSegments = splitSigSegments;
-const COMMA_SEGMENT_STARTERS = new Set([
-    "apply",
-    "take",
-    "instill",
-    "inject",
-    "spray",
-    "use",
-    "od",
-    "os",
-    "ou",
-    "re",
-    "le",
-    "be",
-    "right",
-    "left",
-    "both",
-    "each"
-]);
-const DIRECTIONAL_SEGMENT_STARTERS = new Set(["right", "left", "both", "each"]);
-const SEPARATOR_RULES = [
-    {
-        name: "double-slash",
-        match: (input, index) => (input.startsWith("//", index) ? 2 : 0)
-    },
-    {
-        name: "line-break",
-        match: (input, index) => {
-            const ch = input[index];
-            if (ch === "\r" && input[index + 1] === "\n") {
-                return 2;
-            }
-            return ch === "\n" || ch === "\r" ? 1 : 0;
-        }
-    },
-    {
-        name: "pipe",
-        match: (input, index) => input[index] === "|" && hasNonWhitespaceAround(input, index) ? consumePipeRun(input, index) : 0
-    },
-    {
-        name: "plus",
-        match: (input, index) => input[index] === "+" && hasNonWhitespaceAround(input, index) ? 1 : 0
-    },
-    {
-        name: "slash-divider",
-        match: (input, index) => input[index] === "/" && isDividerSlash(input, index) ? 1 : 0
-    },
-    {
-        name: "comma-clause",
-        match: (input, index, currentStart) => input[index] === "," && shouldSplitComma(input, index, currentStart) ? 1 : 0
-    }
-];
-function splitSigSegments(input) {
-    const segments = [];
-    let currentStart = 0;
-    let depth = 0;
-    const pushSegment = (rawStart, rawEnd) => {
-        let start = rawStart;
-        let end = rawEnd;
-        while (start < end && /\s/.test(input[start])) {
-            start += 1;
-        }
-        while (end > start && /\s/.test(input[end - 1])) {
-            end -= 1;
-        }
-        if (end <= start) {
-            return;
-        }
-        segments.push({
-            text: input.slice(start, end),
-            start,
-            end
-        });
-    };
-    for (let index = 0; index < input.length; index += 1) {
-        const ch = input[index];
-        if (ch === "(" || ch === "[" || ch === "{") {
-            depth += 1;
-            continue;
-        }
-        if ((ch === ")" || ch === "]" || ch === "}") && depth > 0) {
-            depth -= 1;
-            continue;
-        }
-        if (depth > 0) {
-            continue;
-        }
-        for (const rule of SEPARATOR_RULES) {
-            const length = rule.match(input, index, currentStart);
-            if (!length) {
-                continue;
-            }
-            pushSegment(currentStart, index);
-            currentStart = index + length;
-            index = currentStart - 1;
-            break;
-        }
-    }
-    pushSegment(currentStart, input.length);
-    if (segments.length > 0) {
-        return segments;
-    }
-    const fallback = input.trim();
-    if (!fallback) {
-        return [];
-    }
-    const start = input.indexOf(fallback);
-    return [{ text: fallback, start, end: start + fallback.length }];
-}
-function hasNonWhitespaceAround(input, index) {
-    const left = previousNonWhitespace(input, index - 1);
-    const right = nextNonWhitespace(input, index + 1);
-    return left !== undefined && right !== undefined;
-}
-function isDividerSlash(input, index) {
-    var _a, _b;
-    if (input[index - 1] === "/" || input[index + 1] === "/") {
-        return false;
-    }
-    const previous = previousNonWhitespace(input, index - 1);
-    const next = nextNonWhitespace(input, index + 1);
-    if (previous === undefined || next === undefined) {
-        return false;
-    }
-    if (/\d/.test(previous) && /\d/.test(next)) {
-        return false;
-    }
-    return /\s/.test((_a = input[index - 1]) !== null && _a !== void 0 ? _a : "") || /\s/.test((_b = input[index + 1]) !== null && _b !== void 0 ? _b : "");
-}
-function shouldSplitComma(input, index, currentStart) {
-    var _a, _b;
-    if (!hasNonWhitespaceAround(input, index)) {
-        return false;
-    }
-    const left = input.slice(currentStart, index).trim();
-    const right = input.slice(index + 1).trim();
-    if (!left || !right) {
-        return false;
-    }
-    if (startsWithTimeExpression(right)) {
-        return false;
-    }
-    const rightToken = (_b = (_a = right.match(/^([a-z]+|\d+(?:\.\d+)?)/i)) === null || _a === void 0 ? void 0 : _a[1]) === null || _b === void 0 ? void 0 : _b.toLowerCase();
-    if (!rightToken) {
-        return false;
-    }
-    if (/^\d/.test(rightToken)) {
-        return true;
-    }
-    if (COMMA_SEGMENT_STARTERS.has(rightToken)) {
-        if (DIRECTIONAL_SEGMENT_STARTERS.has(rightToken)) {
-            return looksLikeDirectionalClause(right);
-        }
-        return true;
-    }
-    return false;
-}
-function looksLikeDirectionalClause(text) {
-    const normalized = text.trim().toLowerCase();
-    if (!normalized) {
-        return false;
-    }
-    if (/\b\d+(?:\.\d+)?\b/.test(normalized)) {
-        return true;
-    }
-    return /\b(once|twice|thrice|daily|bid|tid|qid|q\d+[a-z0-9/-]*|every|prn|hs|morning|lunch|dinner|noon|night|weekly|monthly)\b/.test(normalized);
-}
-function startsWithTimeExpression(text) {
-    const trimmed = text.replace(/^\s+/, "");
-    if (!trimmed) {
-        return false;
-    }
-    return (/^@\s*\d{1,2}([:.]\d{2})?\s*(am|pm)?\b/i.test(trimmed) ||
-        /^\d{1,2}[:.]\d{2}\s*(am|pm)?\b/i.test(trimmed) ||
-        /^\d{1,2}\s*(am|pm)\b/i.test(trimmed));
-}
-function previousNonWhitespace(input, index) {
-    for (let cursor = index; cursor >= 0; cursor -= 1) {
-        const ch = input[cursor];
-        if (!/\s/.test(ch)) {
-            return ch;
-        }
-    }
-    return undefined;
-}
-function nextNonWhitespace(input, index) {
-    for (let cursor = index; cursor < input.length; cursor += 1) {
-        const ch = input[cursor];
-        if (!/\s/.test(ch)) {
-            return ch;
-        }
-    }
-    return undefined;
-}
-function consumePipeRun(input, index) {
-    let cursor = index;
-    while (cursor < input.length && input[cursor] === "|") {
-        cursor += 1;
-    }
-    return cursor - index;
-}

package/dist/site-phrases.js

@@ -1,344 +0,0 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.isTimingOnlySitePhrase = isTimingOnlySitePhrase;
-exports.hasExternalSurfaceModifier = hasExternalSurfaceModifier;
-exports.extractExplicitSiteCandidate = extractExplicitSiteCandidate;
-exports.selectBestResidualSiteCandidate = selectBestResidualSiteCandidate;
-exports.inferRouteHintFromSitePhrase = inferRouteHintFromSitePhrase;
-const maps_1 = require("./maps");
-const meaning_1 = require("./lexer/meaning");
-const types_1 = require("./types");
-function getInferableDefaultRouteHint(canonical, definition) {
-    const routeHint = definition === null || definition === void 0 ? void 0 : definition.routeHint;
-    if (!routeHint) {
-        return undefined;
-    }
-    switch (routeHint) {
-        case types_1.RouteCode["Ophthalmic route"]:
-            return canonical.includes("eye") || canonical.includes("eyelid")
-                ? routeHint
-                : undefined;
-        case types_1.RouteCode["Otic route"]:
-            return canonical.includes("ear") ? routeHint : undefined;
-        case types_1.RouteCode["Nasal route"]:
-            return canonical.includes("nostril") ||
-                canonical.includes("naris") ||
-                canonical.includes("nares") ||
-                canonical === "nose"
-                ? routeHint
-                : undefined;
-        case types_1.RouteCode["Oral route"]:
-            return canonical === "mouth" ? routeHint : undefined;
-        case types_1.RouteCode["Sublingual route"]:
-            return canonical.includes("tongue") ? routeHint : undefined;
-        case types_1.RouteCode["Buccal route"]:
-            return canonical.includes("cheek") ? routeHint : undefined;
-        case types_1.RouteCode["Intravenous route"]:
-            return canonical.includes("vein") ? routeHint : undefined;
-        case types_1.RouteCode["Per vagina"]:
-            return canonical.includes("vagina") ? routeHint : undefined;
-        case types_1.RouteCode["Per rectum"]:
-            return canonical.includes("rectum") || canonical === "anus"
-                ? routeHint
-                : undefined;
-        case types_1.RouteCode["Topical route"]:
-            return canonical.startsWith("affected ") ? routeHint : undefined;
-        default:
-            return undefined;
-    }
-}
-function mergeBodySiteRouteHint(canonical, customDefinition, defaultDefinition) {
-    var _a;
-    const definition = customDefinition !== null && customDefinition !== void 0 ? customDefinition : defaultDefinition;
-    if (!definition) {
-        return undefined;
-    }
-    const routeHint = (_a = customDefinition === null || customDefinition === void 0 ? void 0 : customDefinition.routeHint) !== null && _a !== void 0 ? _a : getInferableDefaultRouteHint(canonical, defaultDefinition);
-    if (routeHint === definition.routeHint) {
-        return definition;
-    }
-    return Object.assign(Object.assign({}, definition), { routeHint });
-}
-function isExplicitSiteBoundaryToken(lower, options, services) {
-    if (!lower) {
-        return true;
-    }
-    if ((0, meaning_1.isWorkflowInstructionWord)(lower)) {
-        return true;
-    }
-    if (maps_1.EVENT_TIMING_TOKENS[lower] ||
-        maps_1.TIMING_ABBREVIATIONS[lower] ||
-        maps_1.WORD_FREQUENCIES[lower] ||
-        services.hasFrequencyLikeWord(lower) ||
-        (0, meaning_1.isCountKeywordWord)(lower) ||
-        services.mapFrequencyAdverb(lower) ||
-        services.mapIntervalUnit(lower) ||
-        maps_1.MEAL_KEYWORDS[lower]) {
-        return true;
-    }
-    if (services.hasRouteLikeWord(lower, options) || services.normalizeUnit(lower, options)) {
-        return true;
-    }
-    if ((0, meaning_1.isApplicationVerbWord)(lower)) {
-        return true;
-    }
-    if (lower === "prn" || lower === "as" || lower === "needed" || lower === "for") {
-        return true;
-    }
-    if (services.isNumericToken(lower) && !services.isOrdinalToken(lower)) {
-        return true;
-    }
-    return false;
-}
-function isTimingOnlySitePhrase(words) {
-    if (!words.length) {
-        return true;
-    }
-    let phrase = "";
-    let allTimingWords = true;
-    for (let index = 0; index < words.length; index += 1) {
-        const word = words[index];
-        if (index > 0) {
-            phrase += " ";
-        }
-        phrase += word;
-        if (!maps_1.EVENT_TIMING_TOKENS[word] &&
-            !maps_1.TIMING_ABBREVIATIONS[word] &&
-            !maps_1.MEAL_KEYWORDS[word]) {
-            allTimingWords = false;
-        }
-    }
-    if (maps_1.EVENT_TIMING_TOKENS[phrase] ||
-        maps_1.TIMING_ABBREVIATIONS[phrase] ||
-        maps_1.MEAL_KEYWORDS[phrase]) {
-        return true;
-    }
-    return allTimingWords;
-}
-function hasExternalSurfaceModifier(siteText) {
-    const normalized = (0, maps_1.normalizeBodySiteKey)(siteText);
-    if (!normalized) {
-        return false;
-    }
-    const words = normalized.split(/\s+/);
-    for (const word of words) {
-        if (word.length > 0 && (0, meaning_1.isSiteSurfaceModifierWord)(word)) {
-            return true;
-        }
-    }
-    return false;
-}
-function extractExplicitSiteCandidate(tokens, consumed, startIndex, options, services) {
-    var _a;
-    const anchor = tokens[startIndex];
-    if (!anchor || consumed.has(anchor.index)) {
-        return undefined;
-    }
-    const anchorLower = services.normalizeTokenLower(anchor);
-    if (!(0, meaning_1.isSiteAnchorWord)(anchorLower)) {
-        return undefined;
-    }
-    if (!services.hasExplicitSiteIntroduction(startIndex)) {
-        return undefined;
-    }
-    const collected = [anchor.index];
-    const contentWords = [];
-    const candidateTextParts = [anchor.original];
-    let hasSiteHint = false;
-    for (let cursor = startIndex + 1; cursor < tokens.length; cursor += 1) {
-        const candidate = tokens[cursor];
-        if (!candidate || consumed.has(candidate.index)) {
-            break;
-        }
-        const lower = services.normalizeTokenLower(candidate);
-        if (/^[;:(),]+$/.test(lower)) {
-            break;
-        }
-        if (services.siteFillerWords.has(lower) && contentWords.length === 0) {
-            collected.push(candidate.index);
-            candidateTextParts.push(candidate.original);
-            continue;
-        }
-        if ((0, meaning_1.isSiteListConnectorWord)(lower) || lower === ",") {
-            let hasFollowingContent = false;
-            for (let lookahead = cursor + 1; lookahead < tokens.length; lookahead += 1) {
-                const lookaheadToken = tokens[lookahead];
-                if (!lookaheadToken || consumed.has(lookaheadToken.index)) {
-                    break;
-                }
-                const lookaheadLower = services.normalizeTokenLower(lookaheadToken);
-                if (services.siteFillerWords.has(lookaheadLower)) {
-                    continue;
-                }
-                hasFollowingContent = !isExplicitSiteBoundaryToken(lookaheadLower, options, services);
-                break;
-            }
-            if (!hasFollowingContent) {
-                break;
-            }
-            collected.push(candidate.index);
-            candidateTextParts.push(candidate.original);
-            continue;
-        }
-        if (isExplicitSiteBoundaryToken(lower, options, services)) {
-            if (lower === "top") {
-                const nextToken = services.getNextActiveToken(cursor);
-                if (nextToken && services.normalizeTokenLower(nextToken) === "of") {
-                    collected.push(candidate.index);
-                    candidateTextParts.push(candidate.original);
-                    contentWords.push(lower);
-                    hasSiteHint = true;
-                    continue;
-                }
-            }
-            break;
-        }
-        collected.push(candidate.index);
-        candidateTextParts.push(candidate.original);
-        if (!services.siteConnectors.has(lower) && !services.siteFillerWords.has(lower)) {
-            contentWords.push(lower);
-            if (services.isBodySiteHint(lower, services.customSiteHints) ||
-                (0, meaning_1.isSiteSurfaceModifierWord)(lower) ||
-                services.isOrdinalToken(lower)) {
-                hasSiteHint = true;
-            }
-        }
-    }
-    if (!contentWords.length || isTimingOnlySitePhrase(contentWords)) {
-        return undefined;
-    }
-    const candidateText = candidateTextParts.join(" ").replace(/\s+/g, " ").trim();
-    if (candidateText && ((_a = services.isInstructionLikeText) === null || _a === void 0 ? void 0 : _a.call(services, candidateText))) {
-        return undefined;
-    }
-    if (!hasSiteHint &&
-        contentWords.length === 1 &&
-        contentWords[0] !== "area" &&
-        anchorLower === "in") {
-        return undefined;
-    }
-    return {
-        tokenIndices: collected,
-        source: "explicit"
-    };
-}
-function selectBestResidualSiteCandidate(groups, prnSiteSuffixIndices, services) {
-    var _a;
-    let bestTokenIndices;
-    let bestScore = Number.NEGATIVE_INFINITY;
-    for (const group of groups) {
-        const filteredTokens = [];
-        const contentWords = [];
-        const candidateTextParts = [];
-        let hasWorkflowWord = false;
-        let knownWordCount = 0;
-        let modifierCount = 0;
-        for (const token of group.tokens) {
-            if (prnSiteSuffixIndices.has(token.index)) {
-                continue;
-            }
-            filteredTokens.push(token);
-            const lower = services.normalizeTokenLower(token);
-            if (lower.length === 0 ||
-                services.siteConnectors.has(lower) ||
-                services.siteFillerWords.has(lower) ||
-                (0, meaning_1.isSiteListConnectorWord)(lower) ||
-                lower === ",") {
-                continue;
-            }
-            candidateTextParts.push(token.original);
-            contentWords.push(lower);
-            if ((0, meaning_1.isWorkflowInstructionWord)(lower)) {
-                hasWorkflowWord = true;
-            }
-            if (services.isBodySiteHint(lower, services.customSiteHints)) {
-                knownWordCount += 1;
-            }
-            if ((0, meaning_1.isSiteSurfaceModifierWord)(lower) || services.isOrdinalToken(lower)) {
-                modifierCount += 1;
-            }
-        }
-        if (!filteredTokens.length) {
-            continue;
-        }
-        if (!contentWords.length || isTimingOnlySitePhrase(contentWords)) {
-            continue;
-        }
-        if (candidateTextParts.length) {
-            const candidateText = candidateTextParts.join(" ").replace(/\s+/g, " ").trim();
-            if (candidateText && ((_a = services.isInstructionLikeText) === null || _a === void 0 ? void 0 : _a.call(services, candidateText))) {
-                continue;
-            }
-        }
-        if (hasWorkflowWord) {
-            continue;
-        }
-        const startsWithApplicationContext = services.hasApplicationVerbBefore(filteredTokens[0].index);
-        const previous = services.getPreviousActiveToken(filteredTokens[0].index);
-        const anchoredBySiteConnector = previous
-            ? (0, meaning_1.isSiteAnchorWord)(services.normalizeTokenLower(previous))
-            : false;
-        if (knownWordCount === 0 &&
-            modifierCount === 0 &&
-            !anchoredBySiteConnector) {
-            continue;
-        }
-        const score = knownWordCount * 5 +
-            modifierCount * 2 +
-            (startsWithApplicationContext ? 2 : 0) +
-            (anchoredBySiteConnector ? 2 : 0) -
-            filteredTokens.length * 0.2 +
-            filteredTokens[0].index * 0.001;
-        if (score > bestScore) {
-            bestScore = score;
-            bestTokenIndices = [];
-            for (const token of filteredTokens) {
-                bestTokenIndices.push(token.index);
-            }
-        }
-    }
-    if (!bestTokenIndices || !bestTokenIndices.length) {
-        return undefined;
-    }
-    return {
-        tokenIndices: bestTokenIndices,
-        source: "residual"
-    };
-}
-function inferRouteHintFromSitePhrase(siteText, options, services) {
-    const canonical = (0, maps_1.normalizeBodySiteKey)(siteText);
-    if (!canonical) {
-        return undefined;
-    }
-    const customExact = services.lookupBodySiteDefinition(options === null || options === void 0 ? void 0 : options.siteCodeMap, canonical);
-    const exact = mergeBodySiteRouteHint(canonical, customExact, maps_1.DEFAULT_BODY_SITE_SNOMED[canonical]);
-    if (exact === null || exact === void 0 ? void 0 : exact.routeHint) {
-        return exact.routeHint;
-    }
-    if (hasExternalSurfaceModifier(siteText)) {
-        return undefined;
-    }
-    const words = [];
-    for (const word of canonical.split(/\s+/)) {
-        if (word.length > 0) {
-            words.push(word);
-        }
-    }
-    for (let length = words.length; length >= 1; length -= 1) {
-        for (let start = 0; start + length <= words.length; start += 1) {
-            let candidate = "";
-            for (let offset = 0; offset < length; offset += 1) {
-                if (offset > 0) {
-                    candidate += " ";
-                }
-                candidate += words[start + offset];
-            }
-            const customDefinition = services.lookupBodySiteDefinition(options === null || options === void 0 ? void 0 : options.siteCodeMap, candidate);
-            const definition = mergeBodySiteRouteHint(candidate, customDefinition, maps_1.DEFAULT_BODY_SITE_SNOMED[candidate]);
-            if (definition === null || definition === void 0 ? void 0 : definition.routeHint) {
-                return definition.routeHint;
-            }
-        }
-    }
-    return undefined;
-}