@cyanheads/openalex-mcp-server 0.7.15 → 0.7.16

package/dist/services/openalex/provider-text.js

@@ -0,0 +1,276 @@
+/**
+ * @fileoverview Normalization of provider text — the titles, abstracts, names, and affiliation
+ * strings OpenAlex passes through from its sources unsanitized. Decodes HTML character
+ * references against the full WHATWG table in one pass, then removes HTML comments and unwraps
+ * a small allowlist of HTML/JATS/MathML elements. Everything else stays literal, so text such as
+ * `A < B`, `Fish <Actinopterygii>`, or a SICI DOI survives untouched. Output is plain text;
+ * Markdown escaping for `content[]` happens at the render boundary, never here.
+ * @module services/openalex/provider-text
+ */
+import entityTable from './html-entities.json' with { type: 'json' };
+/**
+ * Named character references, keyed without the `&` and `;`. The data is the WHATWG table
+ * (https://html.spec.whatwg.org/entities.json), which the HTML standard freezes — it is never
+ * extended — so a vendored copy cannot drift. A `Map` rather than an object literal, because an
+ * object answers `constructor` and every other `Object.prototype` member as if it were a name.
+ */
+const NAMED_REFERENCES = new Map(Object.entries(entityTable.named));
+/** The 106 legacy names the standard also accepts without a trailing `;` (`&amp`, `&copy`, …). */
+const LEGACY_REFERENCES = new Set(entityTable.legacy);
+/**
+ * Common references OpenAlex also carries with their case changed, because a source upper- or
+ * title-cased the whole title entities and all (`BLEEDING IN&NBSP;CHILDREN`, `Nations:&Nbsp;`).
+ * A spelling the table does not hold decodes as the lowercase name; one it does hold keeps its
+ * own meaning, so `&Lt;` stays ≪.
+ */
+const CASE_FOLDED_REFERENCES = new Set([
+    'amp',
+    'apos',
+    'gt',
+    'lt',
+    'nbsp',
+    'quot',
+]);
+function namedReference(name) {
+    const exact = NAMED_REFERENCES.get(name);
+    if (exact !== undefined)
+        return exact;
+    const folded = name.toLowerCase();
+    return CASE_FOLDED_REFERENCES.has(folded) ? NAMED_REFERENCES.get(folded) : undefined;
+}
+const MAX_UNICODE_CODE_POINT = 0x10ffff;
+/**
+ * One character reference: decimal, hex, or a maximal alphanumeric name, each with an optional
+ * `;`. The name run is maximal, so the character after an unterminated name is never an
+ * alphanumeric — the only lookahead the legacy rule still needs is `=`.
+ */
+const CHARACTER_REFERENCE = /&(?:#(\d+)|#[xX]([0-9a-fA-F]+)|([A-Za-z][A-Za-z0-9]*))(;?)/g;
+/** Whether `&name;` is a named character reference, for callers that must not let one decode. */
+export function isNamedCharacterReference(name) {
+    return NAMED_REFERENCES.has(name);
+}
+function codePointToString(code, fallback) {
+    return Number.isInteger(code) && code >= 0 && code <= MAX_UNICODE_CODE_POINT
+        ? String.fromCodePoint(code)
+        : fallback;
+}
+/**
+ * Decode HTML character references in a single pass, so `&amp;lt;` becomes `&lt;` and never `<`.
+ *
+ * - Numeric references (`&#38;`, `&#x27E9;`) decode with or without the `;` — OpenAlex sends
+ *   `&#38 Hepatology` — and a code point past U+10FFFF stays literal rather than corrupting text.
+ * - Named references are case-sensitive (`&Alpha;` is Α, `&alpha;` is α) and need the `;`,
+ *   except for the legacy names, which follow the standard's attribute-value rule: decoded
+ *   without the `;` unless an `=` follows, so a URL's `&copy=2` query parameter stays intact.
+ *   A recased common reference with its `;` (`&NBSP;`, `&Quot;`) decodes as the lowercase name.
+ * - Anything else — `&constructor;`, `AT&T`, `&madeupentity;` — stays literal.
+ */
+function decodeCharacterReferences(text) {
+    if (!text.includes('&'))
+        return text;
+    return text.replace(CHARACTER_REFERENCE, (match, decimal, hex, name, semicolon, offset) => {
+        if (decimal !== undefined)
+            return codePointToString(Number.parseInt(decimal, 10), match);
+        if (hex !== undefined)
+            return codePointToString(Number.parseInt(hex, 16), match);
+        const reference = name;
+        if (semicolon)
+            return namedReference(reference) ?? match;
+        const next = text[offset + match.length];
+        return LEGACY_REFERENCES.has(reference) && next !== '='
+            ? (NAMED_REFERENCES.get(reference) ?? match)
+            : match;
+    });
+}
+/** Inline elements whose tags are dropped and whose text is kept. */
+const INLINE_ELEMENTS = new Set([
+    'b',
+    'em',
+    'i',
+    'inline-formula',
+    'jats:bold',
+    'jats:italic',
+    'sc',
+    'scp',
+    'span',
+    'strong',
+    'tex-math',
+    // MathML presentation markup, bare; `mml:`-prefixed names are matched by prefix.
+    'math',
+    'menclose',
+    'mfenced',
+    'mfrac',
+    'mi',
+    'mmultiscripts',
+    'mn',
+    'mo',
+    'mover',
+    'mpadded',
+    'mphantom',
+    'mroot',
+    'mrow',
+    'mspace',
+    'msqrt',
+    'mstyle',
+    'msub',
+    'msubsup',
+    'msup',
+    'mtable',
+    'mtd',
+    'mtext',
+    'mtr',
+    'munder',
+    'munderover',
+    'semantics',
+]);
+/** Block elements, each replaced by a paragraph break. */
+const BLOCK_ELEMENTS = new Set([
+    'br',
+    'jats:p',
+    'jats:sec',
+    'jats:title',
+    'li',
+    'p',
+]);
+const PARAGRAPH_BREAK = '\n\n';
+/**
+ * Sub/superscripts open as TeX-style `_{`/`^{` and close with `}`. Unicode has no
+ * sub/superscript for most letters, unwrapping would read `10<sup>-3</sup>` as `10-3`, and
+ * OpenAlex's TeX-bearing abstracts already use this notation.
+ */
+const SCRIPT_OPENERS = new Map([
+    ['sub', '_{'],
+    ['jats:sub', '_{'],
+    ['sup', '^{'],
+    ['jats:sup', '^{'],
+]);
+/** A complete tag, checked against one `<…>` candidate that holds no other `<` or `>`. */
+const TAG = /^<(\/?)([A-Za-z][\w:.-]*)(?:\s[^<>]*)?\/?>$/;
+/**
+ * What an allowlisted tag becomes, or `undefined` for any other `<…>` span, which stays literal.
+ * Matching a whole tag name — never a prefix — is what keeps `i1<i2<` and `<Actinopterygii>`.
+ */
+function replaceTag(candidate) {
+    const tag = TAG.exec(candidate);
+    if (!tag)
+        return;
+    const closing = tag[1] === '/';
+    const name = tag[2].toLowerCase();
+    if (INLINE_ELEMENTS.has(name) || name.startsWith('mml:'))
+        return '';
+    if (BLOCK_ELEMENTS.has(name))
+        return PARAGRAPH_BREAK;
+    const opener = SCRIPT_OPENERS.get(name);
+    if (opener === undefined)
+        return;
+    if (candidate.endsWith('/>'))
+        return '';
+    return closing ? '}' : opener;
+}
+const COMMENT_OPEN = '<!--';
+/** `<!---->` — the shortest span that closes a comment it opened. */
+const MIN_COMMENT_LENGTH = 7;
+function endsWith(out, suffix) {
+    if (out.length < suffix.length)
+        return false;
+    for (let i = 0; i < suffix.length; i++) {
+        if (out[out.length - suffix.length + i] !== suffix[i])
+            return false;
+    }
+    return true;
+}
+/**
+ * Remove HTML comments and rewrite allowlisted tags in one left-to-right pass, reaching the
+ * result that repeating a regex replacement until nothing changes would reach — no allowlisted
+ * tag can re-form in the output — in linear time. A replace-until-stable loop costs one full
+ * pass per nesting level, which is quadratic on nested input such as `<<<i>i>i>`.
+ *
+ * Characters are copied to `out`. Removing a tag or comment can join the text on either side
+ * into a new one (`<<i>i>`, `<!-<!---->- x -->`), so every removal truncates `out` and scanning
+ * simply continues: `opens` holds the positions of each `<` that could still begin a tag, and
+ * `commentStart` the earliest unclosed `<!--`. A tag ends at the first `>` after its `<`, so a
+ * `>` settles the innermost open `<`: kept as text, it also blocks every `<` below it, except
+ * those before an unclosed comment, which come back into play if that comment closes. Each
+ * character is examined a bounded number of times.
+ */
+function stripMarkup(text) {
+    if (!text.includes('<'))
+        return text;
+    const out = [];
+    const opens = [];
+    let commentStart = -1;
+    for (const ch of text) {
+        out.push(ch);
+        if (ch === '<') {
+            opens.push(out.length - 1);
+            continue;
+        }
+        if (ch === '-') {
+            if (commentStart < 0 && endsWith(out, COMMENT_OPEN))
+                commentStart = out.length - 4;
+            continue;
+        }
+        if (ch !== '>')
+            continue;
+        if (commentStart >= 0 &&
+            out.length - commentStart >= MIN_COMMENT_LENGTH &&
+            endsWith(out, '-->')) {
+            out.length = commentStart;
+            commentStart = -1;
+            while (opens.length > 0 && opens.at(-1) >= out.length)
+                opens.pop();
+            continue;
+        }
+        // Inside an unclosed comment, only a `<` opened after it can form a tag.
+        const top = opens.at(-1);
+        if (top === undefined || top <= commentStart)
+            continue;
+        opens.pop();
+        const replacement = replaceTag(out.slice(top).join(''));
+        if (replacement === undefined) {
+            while (opens.length > 0 && opens.at(-1) > commentStart)
+                opens.pop();
+            continue;
+        }
+        out.length = top;
+        if (replacement)
+            out.push(replacement);
+    }
+    // A break element can only come from a block tag (input is copied one character at a time),
+    // and one inside a comment was truncated with it.
+    const stripped = out.join('');
+    if (!out.includes(PARAGRAPH_BREAK))
+        return stripped;
+    // Block tags leave runs of breaks and stray edge whitespace; keep one break between paragraphs.
+    return stripped
+        .split(PARAGRAPH_BREAK)
+        .map((paragraph) => paragraph.trim())
+        .filter(Boolean)
+        .join(PARAGRAPH_BREAK);
+}
+/**
+ * Normalize one provider string to plain text: decode character references once, then handle
+ * markup. Decoding first is what lets an encoded tag (`&lt;i&gt;x&lt;/i&gt;`) unwrap; decoding
+ * never runs again afterward, so double-encoded text keeps its remaining level.
+ */
+export function normalizeProviderText(text) {
+    return stripMarkup(decodeCharacterReferences(text));
+}
+function isRecord(value) {
+    return value !== null && typeof value === 'object' && !Array.isArray(value);
+}
+/** Normalize every string leaf of a JSON value; keys, numbers, booleans, and nulls pass through. */
+export function normalizeProviderValue(value) {
+    if (typeof value === 'string')
+        return normalizeProviderText(value);
+    if (Array.isArray(value))
+        return value.map(normalizeProviderValue);
+    if (isRecord(value)) {
+        const next = {};
+        for (const [k, v] of Object.entries(value))
+            next[k] = normalizeProviderValue(v);
+        return next;
+    }
+    return value;
+}
+//# sourceMappingURL=provider-text.js.map

package/dist/services/openalex/provider-text.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"provider-text.js","sourceRoot":"","sources":["../../../src/services/openalex/provider-text.ts"],"names":[],"mappings":"AAAA;;;;;;;;GAQG;AAEH,OAAO,WAAW,MAAM,sBAAsB,CAAC,OAAO,IAAI,EAAE,MAAM,EAAE,CAAC;AAErE;;;;;GAKG;AACH,MAAM,gBAAgB,GAAgC,IAAI,GAAG,CAAC,MAAM,CAAC,OAAO,CAAC,WAAW,CAAC,KAAK,CAAC,CAAC,CAAC;AAEjG,kGAAkG;AAClG,MAAM,iBAAiB,GAAwB,IAAI,GAAG,CAAC,WAAW,CAAC,MAAM,CAAC,CAAC;AAE3E;;;;;GAKG;AACH,MAAM,sBAAsB,GAAwB,IAAI,GAAG,CAAC;IAC1D,KAAK;IACL,MAAM;IACN,IAAI;IACJ,IAAI;IACJ,MAAM;IACN,MAAM;CACP,CAAC,CAAC;AAEH,SAAS,cAAc,CAAC,IAAY;IAClC,MAAM,KAAK,GAAG,gBAAgB,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;IACzC,IAAI,KAAK,KAAK,SAAS;QAAE,OAAO,KAAK,CAAC;IACtC,MAAM,MAAM,GAAG,IAAI,CAAC,WAAW,EAAE,CAAC;IAClC,OAAO,sBAAsB,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,gBAAgB,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;AACvF,CAAC;AAED,MAAM,sBAAsB,GAAG,QAAQ,CAAC;AAExC;;;;GAIG;AACH,MAAM,mBAAmB,GAAG,6DAA6D,CAAC;AAE1F,iGAAiG;AACjG,MAAM,UAAU,yBAAyB,CAAC,IAAY;IACpD,OAAO,gBAAgB,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;AACpC,CAAC;AAED,SAAS,iBAAiB,CAAC,IAAY,EAAE,QAAgB;IACvD,OAAO,MAAM,CAAC,SAAS,CAAC,IAAI,CAAC,IAAI,IAAI,IAAI,CAAC,IAAI,IAAI,IAAI,sBAAsB;QAC1E,CAAC,CAAC,MAAM,CAAC,aAAa,CAAC,IAAI,CAAC;QAC5B,CAAC,CAAC,QAAQ,CAAC;AACf,CAAC;AAED;;;;;;;;;;GAUG;AACH,SAAS,yBAAyB,CAAC,IAAY;IAC7C,IAAI,CAAC,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC;QAAE,OAAO,IAAI,CAAC;IACrC,OAAO,IAAI,CAAC,OAAO,CACjB,mBAAmB,EACnB,CACE,KAAK,EACL,OAA2B,EAC3B,GAAuB,EACvB,IAAwB,EACxB,SAAiB,EACjB,MAAc,EACd,EAAE;QACF,IAAI,OAAO,KAAK,SAAS;YAAE,OAAO,iBAAiB,CAAC,MAAM,CAAC,QAAQ,CAAC,OAAO,EAAE,EAAE,CAAC,EAAE,KAAK,CAAC,CAAC;QACzF,IAAI,GAAG,KAAK,SAAS;YAAE,OAAO,iBAAiB,CAAC,MAAM,CAAC,QAAQ,CAAC,GAAG,EAAE,EAAE,CAAC,EAAE,KAAK,CAAC,CAAC;QACjF,MAAM,SAAS,GAAG,IAAc,CAAC;QACjC,IAAI,SAAS;YAAE,OAAO,cAAc,CAAC,SAAS,CAAC,IAAI,KAAK,CAAC;QACzD,MAAM,IAAI,GAAG,IAAI,CAAC,MAAM,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC;QACzC,OAAO,iBAAiB,CAAC,GAAG,CAAC,SAAS,CAAC,IAAI,IAAI,KAAK,GAAG;YACrD,CAAC,CAAC,CAAC,gBAAgB,CAAC,GAAG,CAAC,SAAS,CAAC,IAAI,KAAK,CAAC;YAC5C,CAAC,CAAC,KAAK,CAAC;IACZ,CAAC,CACF,CAAC;AACJ,CAAC;AAED,qEAAqE;AACrE,MAAM,eAAe,GAAwB,IAAI,GAAG,CAAC;IACnD,GAAG;IACH,IAAI;IACJ,GAAG;IACH,gBAAgB;IAChB,WAAW;IACX,aAAa;IACb,IAAI;IACJ,KAAK;IACL,MAAM;IACN,QAAQ;IACR,UAAU;IACV,iFAAiF;IACjF,MAAM;IACN,UAAU;IACV,SAAS;IACT,OAAO;IACP,IAAI;IACJ,eAAe;IACf,IAAI;IACJ,IAAI;IACJ,OAAO;IACP,SAAS;IACT,UAAU;IACV,OAAO;IACP,MAAM;IACN,QAAQ;IACR,OAAO;IACP,QAAQ;IACR,MAAM;IACN,SAAS;IACT,MAAM;IACN,QAAQ;IACR,KAAK;IACL,OAAO;IACP,KAAK;IACL,QAAQ;IACR,YAAY;IACZ,WAAW;CACZ,CAAC,CAAC;AAEH,0DAA0D;AAC1D,MAAM,cAAc,GAAwB,IAAI,GAAG,CAAC;IAClD,IAAI;IACJ,QAAQ;IACR,UAAU;IACV,YAAY;IACZ,IAAI;IACJ,GAAG;CACJ,CAAC,CAAC;AAEH,MAAM,eAAe,GAAG,MAAM,CAAC;AAE/B;;;;GAIG;AACH,MAAM,cAAc,GAAgC,IAAI,GAAG,CAAC;IAC1D,CAAC,KAAK,EAAE,IAAI,CAAC;IACb,CAAC,UAAU,EAAE,IAAI,CAAC;IAClB,CAAC,KAAK,EAAE,IAAI,CAAC;IACb,CAAC,UAAU,EAAE,IAAI,CAAC;CACnB,CAAC,CAAC;AAEH,0FAA0F;AAC1F,MAAM,GAAG,GAAG,6CAA6C,CAAC;AAE1D;;;GAGG;AACH,SAAS,UAAU,CAAC,SAAiB;IACnC,MAAM,GAAG,GAAG,GAAG,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;IAChC,IAAI,CAAC,GAAG;QAAE,OAAO;IACjB,MAAM,OAAO,GAAG,GAAG,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC;IAC/B,MAAM,IAAI,GAAI,GAAG,CAAC,CAAC,CAAY,CAAC,WAAW,EAAE,CAAC;IAC9C,IAAI,eAAe,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,IAAI,CAAC,UAAU,CAAC,MAAM,CAAC;QAAE,OAAO,EAAE,CAAC;IACpE,IAAI,cAAc,CAAC,GAAG,CAAC,IAAI,CAAC;QAAE,OAAO,eAAe,CAAC;IACrD,MAAM,MAAM,GAAG,cAAc,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;IACxC,IAAI,MAAM,KAAK,SAAS;QAAE,OAAO;IACjC,IAAI,SAAS,CAAC,QAAQ,CAAC,IAAI,CAAC;QAAE,OAAO,EAAE,CAAC;IACxC,OAAO,OAAO,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,MAAM,CAAC;AAChC,CAAC;AAED,MAAM,YAAY,GAAG,MAAM,CAAC;AAC5B,qEAAqE;AACrE,MAAM,kBAAkB,GAAG,CAAC,CAAC;AAE7B,SAAS,QAAQ,CAAC,GAAa,EAAE,MAAc;IAC7C,IAAI,GAAG,CAAC,MAAM,GAAG,MAAM,CAAC,MAAM;QAAE,OAAO,KAAK,CAAC;IAC7C,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,MAAM,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACvC,IAAI,GAAG,CAAC,GAAG,CAAC,MAAM,GAAG,MAAM,CAAC,MAAM,GAAG,CAAC,CAAC,KAAK,MAAM,CAAC,CAAC,CAAC;YAAE,OAAO,KAAK,CAAC;IACtE,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED;;;;;;;;;;;;;GAaG;AACH,SAAS,WAAW,CAAC,IAAY;IAC/B,IAAI,CAAC,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC;QAAE,OAAO,IAAI,CAAC;IACrC,MAAM,GAAG,GAAa,EAAE,CAAC;IACzB,MAAM,KAAK,GAAa,EAAE,CAAC;IAC3B,IAAI,YAAY,GAAG,CAAC,CAAC,CAAC;IAEtB,KAAK,MAAM,EAAE,IAAI,IAAI,EAAE,CAAC;QACtB,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QACb,IAAI,EAAE,KAAK,GAAG,EAAE,CAAC;YACf,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;YAC3B,SAAS;QACX,CAAC;QACD,IAAI,EAAE,KAAK,GAAG,EAAE,CAAC;YACf,IAAI,YAAY,GAAG,CAAC,IAAI,QAAQ,CAAC,GAAG,EAAE,YAAY,CAAC;gBAAE,YAAY,GAAG,GAAG,CAAC,MAAM,GAAG,CAAC,CAAC;YACnF,SAAS;QACX,CAAC;QACD,IAAI,EAAE,KAAK,GAAG;YAAE,SAAS;QAEzB,IACE,YAAY,IAAI,CAAC;YACjB,GAAG,CAAC,MAAM,GAAG,YAAY,IAAI,kBAAkB;YAC/C,QAAQ,CAAC,GAAG,EAAE,KAAK,CAAC,EACpB,CAAC;YACD,GAAG,CAAC,MAAM,GAAG,YAAY,CAAC;YAC1B,YAAY,GAAG,CAAC,CAAC,CAAC;YAClB,OAAO,KAAK,CAAC,MAAM,GAAG,CAAC,IAAK,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC,CAAY,IAAI,GAAG,CAAC,MAAM;gBAAE,KAAK,CAAC,GAAG,EAAE,CAAC;YAC/E,SAAS;QACX,CAAC;QAED,yEAAyE;QACzE,MAAM,GAAG,GAAG,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;QACzB,IAAI,GAAG,KAAK,SAAS,IAAI,GAAG,IAAI,YAAY;YAAE,SAAS;QACvD,KAAK,CAAC,GAAG,EAAE,CAAC;QACZ,MAAM,WAAW,GAAG,UAAU,CAAC,GAAG,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,CAAC;QACxD,IAAI,WAAW,KAAK,SAAS,EAAE,CAAC;YAC9B,OAAO,KAAK,CAAC,MAAM,GAAG,CAAC,IAAK,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC,CAAY,GAAG,YAAY;gBAAE,KAAK,CAAC,GAAG,EAAE,CAAC;YAChF,SAAS;QACX,CAAC;QACD,GAAG,CAAC,MAAM,GAAG,GAAG,CAAC;QACjB,IAAI,WAAW;YAAE,GAAG,CAAC,IAAI,CAAC,WAAW,CAAC,CAAC;IACzC,CAAC;IAED,4FAA4F;IAC5F,kDAAkD;IAClD,MAAM,QAAQ,GAAG,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAC9B,IAAI,CAAC,GAAG,CAAC,QAAQ,CAAC,eAAe,CAAC;QAAE,OAAO,QAAQ,CAAC;IACpD,gGAAgG;IAChG,OAAO,QAAQ;SACZ,KAAK,CAAC,eAAe,CAAC;SACtB,GAAG,CAAC,CAAC,SAAS,EAAE,EAAE,CAAC,SAAS,CAAC,IAAI,EAAE,CAAC;SACpC,MAAM,CAAC,OAAO,CAAC;SACf,IAAI,CAAC,eAAe,CAAC,CAAC;AAC3B,CAAC;AAED;;;;GAIG;AACH,MAAM,UAAU,qBAAqB,CAAC,IAAY;IAChD,OAAO,WAAW,CAAC,yBAAyB,CAAC,IAAI,CAAC,CAAC,CAAC;AACtD,CAAC;AAED,SAAS,QAAQ,CAAC,KAAc;IAC9B,OAAO,KAAK,KAAK,IAAI,IAAI,OAAO,KAAK,KAAK,QAAQ,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC;AAC9E,CAAC;AAED,oGAAoG;AACpG,MAAM,UAAU,sBAAsB,CAAI,KAAQ;IAChD,IAAI,OAAO,KAAK,KAAK,QAAQ;QAAE,OAAO,qBAAqB,CAAC,KAAK,CAAM,CAAC;IACxE,IAAI,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC;QAAE,OAAO,KAAK,CAAC,GAAG,CAAC,sBAAsB,CAAM,CAAC;IACxE,IAAI,QAAQ,CAAC,KAAK,CAAC,EAAE,CAAC;QACpB,MAAM,IAAI,GAA4B,EAAE,CAAC;QACzC,KAAK,MAAM,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,MAAM,CAAC,OAAO,CAAC,KAAK,CAAC;YAAE,IAAI,CAAC,CAAC,CAAC,GAAG,sBAAsB,CAAC,CAAC,CAAC,CAAC;QAChF,OAAO,IAAS,CAAC;IACnB,CAAC;IACD,OAAO,KAAK,CAAC;AACf,CAAC"}

package/dist/services/openalex/types.d.ts

@@ -83,6 +83,11 @@ export interface EntityRecord {
 }
 export interface GroupRecord {
     count: number;
+    /**
+     * Set only on the missing-value bucket `include_unknown` adds, whose key is an OpenAlex
+     * sentinel (`-111`, `unknown`, `…/unknown`) rather than a measured value.
+     */
+    is_unknown?: true;
     key: string;
     key_display_name: string;
 }

package/dist/services/openalex/types.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"types.d.ts","sourceRoot":"","sources":["../../../src/services/openalex/types.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAEH,eAAO,MAAM,YAAY,YACvB,OAAO,EACP,SAAS,EACT,SAAS,EACT,cAAc,EACd,QAAQ,EACR,UAAU,EACV,YAAY,EACZ,SAAS,CACD,CAAC;AAEX,MAAM,MAAM,UAAU,GAAG,CAAC,OAAO,YAAY,CAAC,CAAC,MAAM,CAAC,CAAC;AAEvD;;;;;;;;;GASG;AACH,eAAO,MAAM,cAAc,EAAE,MAAM,CAAC,UAAU,EAAE,MAAM,EAAE,CAwDvD,CAAC;AAEF,MAAM,WAAW,YAAY;IAC3B,MAAM,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC5B,UAAU,EAAE,UAAU,CAAC;IACvB,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,GAAG,SAAS,CAAC;IAC7C,EAAE,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IACxB;;;OAGG;IACH,IAAI,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC1B,OAAO,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC7B,KAAK,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC3B,MAAM,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC5B,UAAU,CAAC,EAAE,SAAS,GAAG,OAAO,GAAG,UAAU,GAAG,SAAS,CAAC;IAC1D,IAAI,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC1B,MAAM,CAAC,EAAE,MAAM,EAAE,GAAG,SAAS,CAAC;IAC9B,IAAI,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;CAC3B;AAED,MAAM,WAAW,YAAY;IAC3B,IAAI,EAAE;QACJ,KAAK,EAAE,MAAM,CAAC;QACd,QAAQ,EAAE,MAAM,CAAC;QACjB,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;KAC5B,CAAC;IACF,OAAO,EAAE,YAAY,EAAE,CAAC;CACzB;AAED,MAAM,WAAW,aAAa;IAC5B,MAAM,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC5B,UAAU,EAAE,UAAU,CAAC;IACvB,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,GAAG,SAAS,CAAC;IAC7C,OAAO,EAAE,MAAM,CAAC;IAChB,cAAc,CAAC,EAAE,OAAO,GAAG,SAAS,CAAC;IACrC;;;;OAIG;IACH,KAAK,CAAC,EAAE,OAAO,GAAG,KAAK,GAAG,SAAS,CAAC;IACpC,OAAO,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;CAC9B;AAED,MAAM,WAAW,aAAa;IAC5B,MAAM,EAAE,WAAW,EAAE,CAAC;IACtB,IAAI,EAAE;QACJ,KAAK,EAAE,MAAM,CAAC;QACd,YAAY,EAAE,MAAM,GAAG,IAAI,CAAC;QAC5B,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;KAC5B,CAAC;CACH;AAED,MAAM,WAAW,kBAAkB;IACjC,UAAU,CAAC,EAAE,UAAU,GAAG,SAAS,CAAC;IACpC,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,GAAG,SAAS,CAAC;IAC7C,KAAK,EAAE,MAAM,CAAC;CACf;AAED,MAAM,WAAW,kBAAkB;IACjC,OAAO,EAAE,kBAAkB,EAAE,CAAC;CAC/B;AAED,MAAM,WAAW,YAAY;IAC3B;;;;OAIG;IACH,YAAY,EAAE,MAAM,GAAG,IAAI,CAAC;IAC5B,EAAE,EAAE,MAAM,CAAC;IACX,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAC;CACxB;AAED,MAAM,WAAW,WAAW;IAC1B,KAAK,EAAE,MAAM,CAAC;IACd,GAAG,EAAE,MAAM,CAAC;IACZ,gBAAgB,EAAE,MAAM,CAAC;CAC1B;AAED,MAAM,WAAW,kBAAkB;IACjC,cAAc,EAAE,MAAM,CAAC;IACvB;;;;OAIG;IACH,YAAY,EAAE,MAAM,GAAG,IAAI,CAAC;IAC5B,WAAW,EAAE,MAAM,CAAC;IACpB,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;IAC3B,IAAI,EAAE,MAAM,GAAG,IAAI,CAAC;IACpB,EAAE,EAAE,MAAM,CAAC;IACX,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;CAC5B"}
+{"version":3,"file":"types.d.ts","sourceRoot":"","sources":["../../../src/services/openalex/types.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAEH,eAAO,MAAM,YAAY,YACvB,OAAO,EACP,SAAS,EACT,SAAS,EACT,cAAc,EACd,QAAQ,EACR,UAAU,EACV,YAAY,EACZ,SAAS,CACD,CAAC;AAEX,MAAM,MAAM,UAAU,GAAG,CAAC,OAAO,YAAY,CAAC,CAAC,MAAM,CAAC,CAAC;AAEvD;;;;;;;;;GASG;AACH,eAAO,MAAM,cAAc,EAAE,MAAM,CAAC,UAAU,EAAE,MAAM,EAAE,CAwDvD,CAAC;AAEF,MAAM,WAAW,YAAY;IAC3B,MAAM,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC5B,UAAU,EAAE,UAAU,CAAC;IACvB,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,GAAG,SAAS,CAAC;IAC7C,EAAE,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IACxB;;;OAGG;IACH,IAAI,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC1B,OAAO,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC7B,KAAK,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC3B,MAAM,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC5B,UAAU,CAAC,EAAE,SAAS,GAAG,OAAO,GAAG,UAAU,GAAG,SAAS,CAAC;IAC1D,IAAI,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC1B,MAAM,CAAC,EAAE,MAAM,EAAE,GAAG,SAAS,CAAC;IAC9B,IAAI,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;CAC3B;AAED,MAAM,WAAW,YAAY;IAC3B,IAAI,EAAE;QACJ,KAAK,EAAE,MAAM,CAAC;QACd,QAAQ,EAAE,MAAM,CAAC;QACjB,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;KAC5B,CAAC;IACF,OAAO,EAAE,YAAY,EAAE,CAAC;CACzB;AAED,MAAM,WAAW,aAAa;IAC5B,MAAM,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC5B,UAAU,EAAE,UAAU,CAAC;IACvB,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,GAAG,SAAS,CAAC;IAC7C,OAAO,EAAE,MAAM,CAAC;IAChB,cAAc,CAAC,EAAE,OAAO,GAAG,SAAS,CAAC;IACrC;;;;OAIG;IACH,KAAK,CAAC,EAAE,OAAO,GAAG,KAAK,GAAG,SAAS,CAAC;IACpC,OAAO,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;CAC9B;AAED,MAAM,WAAW,aAAa;IAC5B,MAAM,EAAE,WAAW,EAAE,CAAC;IACtB,IAAI,EAAE;QACJ,KAAK,EAAE,MAAM,CAAC;QACd,YAAY,EAAE,MAAM,GAAG,IAAI,CAAC;QAC5B,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;KAC5B,CAAC;CACH;AAED,MAAM,WAAW,kBAAkB;IACjC,UAAU,CAAC,EAAE,UAAU,GAAG,SAAS,CAAC;IACpC,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,GAAG,SAAS,CAAC;IAC7C,KAAK,EAAE,MAAM,CAAC;CACf;AAED,MAAM,WAAW,kBAAkB;IACjC,OAAO,EAAE,kBAAkB,EAAE,CAAC;CAC/B;AAED,MAAM,WAAW,YAAY;IAC3B;;;;OAIG;IACH,YAAY,EAAE,MAAM,GAAG,IAAI,CAAC;IAC5B,EAAE,EAAE,MAAM,CAAC;IACX,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAC;CACxB;AAED,MAAM,WAAW,WAAW;IAC1B,KAAK,EAAE,MAAM,CAAC;IACd;;;OAGG;IACH,UAAU,CAAC,EAAE,IAAI,CAAC;IAClB,GAAG,EAAE,MAAM,CAAC;IACZ,gBAAgB,EAAE,MAAM,CAAC;CAC1B;AAED,MAAM,WAAW,kBAAkB;IACjC,cAAc,EAAE,MAAM,CAAC;IACvB;;;;OAIG;IACH,YAAY,EAAE,MAAM,GAAG,IAAI,CAAC;IAC5B,WAAW,EAAE,MAAM,CAAC;IACpB,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;IAC3B,IAAI,EAAE,MAAM,GAAG,IAAI,CAAC;IACpB,EAAE,EAAE,MAAM,CAAC;IACX,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;CAC5B"}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@cyanheads/openalex-mcp-server",
-  "version": "0.7.15",
+  "version": "0.7.16",
   "description": "Access the OpenAlex academic research catalog - 270M+ publications through MCP. STDIO & Streamable HTTP.",
   "mcpName": "io.github.cyanheads/openalex-mcp-server",
   "type": "module",
@@ -99,6 +99,9 @@
     "@types/node": "^26.6.2",
     "depcheck": "^1.4.7",
     "ignore": "^7.0.9",
+    "mdast-util-from-markdown": "^2.0.3",
+    "mdast-util-gfm": "^3.1.0",
+    "micromark-extension-gfm": "^3.0.0",
     "tsc-alias": "^1.9.5",
     "typescript": "^7.0.2",
     "vitest": "^5.0.1"

package/server.json

@@ -6,7 +6,7 @@
     "url": "https://github.com/cyanheads/openalex-mcp-server",
     "source": "github"
   },
-  "version": "0.7.15",
+  "version": "0.7.16",
   "remotes": [
     {
       "type": "streamable-http",
@@ -19,7 +19,7 @@
       "registryBaseUrl": "https://registry.npmjs.org",
       "identifier": "@cyanheads/openalex-mcp-server",
       "runtimeHint": "node",
-      "version": "0.7.15",
+      "version": "0.7.16",
       "packageArguments": [
         {
           "type": "positional",
@@ -67,7 +67,7 @@
       "registryBaseUrl": "https://registry.npmjs.org",
       "identifier": "@cyanheads/openalex-mcp-server",
       "runtimeHint": "node",
-      "version": "0.7.15",
+      "version": "0.7.16",
       "packageArguments": [
         {
           "type": "positional",