jsonstat-io 0.1.0

package/dist/chunk-KFIKGBIJ.cjs

@@ -0,0 +1,200 @@
+'use strict';
+
+// src/sources/csv.ts
+var CsvSourceError = class extends Error {
+  constructor(message) {
+    super(message);
+    this.name = "CsvSourceError";
+  }
+};
+function parseCsv(input, delimiter = ",") {
+  const rows = [];
+  let row = [];
+  let field = "";
+  let inQuotes = false;
+  const delim = delimiter.charCodeAt(0);
+  for (let i = 0; i < input.length; i++) {
+    const c = input.charCodeAt(i);
+    if (inQuotes) {
+      if (c === 34) {
+        if (input.charCodeAt(i + 1) === 34) {
+          field += '"';
+          i++;
+        } else {
+          inQuotes = false;
+        }
+      } else {
+        field += input[i];
+      }
+    } else {
+      if (c === 34) {
+        inQuotes = true;
+      } else if (c === delim) {
+        row.push(field);
+        field = "";
+      } else if (c === 13) {
+        if (input.charCodeAt(i + 1) === 10) i++;
+        row.push(field);
+        rows.push(row);
+        row = [];
+        field = "";
+      } else if (c === 10) {
+        row.push(field);
+        rows.push(row);
+        row = [];
+        field = "";
+      } else {
+        field += input[i];
+      }
+    }
+  }
+  if (field.length > 0 || row.length > 0) {
+    row.push(field);
+    rows.push(row);
+  }
+  return rows;
+}
+function isAllNumeric(values) {
+  let any = false;
+  for (const v of values) {
+    if (v === "") continue;
+    any = true;
+    const n = Number(v);
+    if (!Number.isFinite(n)) return false;
+  }
+  return any;
+}
+function csvToCube(text, options = {}) {
+  const delimiter = options.delimiter ?? ",";
+  const header = options.header ?? true;
+  const rows = parseCsv(text, delimiter);
+  if (rows.length === 0) throw new CsvSourceError("CSV has no rows");
+  let headerNames;
+  let dataRows;
+  if (header) {
+    headerNames = rows[0];
+    dataRows = rows.slice(1);
+  } else {
+    headerNames = rows[0].map((_, i) => `col${i}`);
+    dataRows = rows;
+  }
+  if (headerNames.length === 0) throw new CsvSourceError("CSV header is empty");
+  if (dataRows.length === 0) throw new CsvSourceError("CSV has no data rows");
+  const colCount = headerNames.length;
+  const columns = headerNames.map(() => []);
+  for (const r of dataRows) {
+    for (let c = 0; c < colCount; c++) columns[c].push(r[c] ?? "");
+  }
+  let measureIdx = headerNames.findIndex((n2) => n2 === options.measure);
+  if (measureIdx === -1) {
+    measureIdx = columns.findIndex((col) => isAllNumeric(col));
+    if (measureIdx === -1) {
+      throw new CsvSourceError(
+        "No measure column found: no column is fully numeric. Pass options.measure to name one (values will be parsed with Number())."
+      );
+    }
+  }
+  let statusIdx = headerNames.findIndex((n2) => n2 === options.status);
+  if (statusIdx === -1 && !options.status) {
+    statusIdx = headerNames.findIndex((n2) => n2.toLowerCase() === "status");
+    if (statusIdx === measureIdx) statusIdx = -1;
+  }
+  let dimensionIdxs;
+  if (options.dimensions) {
+    dimensionIdxs = options.dimensions.map((n2) => {
+      const idx = headerNames.indexOf(n2);
+      if (idx === -1) throw new CsvSourceError(`Dimension column "${n2}" not found in CSV header`);
+      return idx;
+    });
+  } else {
+    dimensionIdxs = headerNames.map((_, i) => i).filter((i) => i !== measureIdx && i !== statusIdx);
+    if (dimensionIdxs.length === 0) {
+      throw new CsvSourceError("No dimension columns remain after measure/status selection");
+    }
+  }
+  dataRows.length;
+  const dimensions = {};
+  for (const idx of dimensionIdxs) {
+    const name = headerNames[idx];
+    dimensions[name] = {
+      id: name,
+      values: columns[idx].map((v) => v === "" ? "" : v)
+    };
+  }
+  const measure = {
+    name: headerNames[measureIdx],
+    values: columns[measureIdx].map((v) => v === "" ? null : Number(v))
+  };
+  let status;
+  if (statusIdx !== -1) {
+    status = { values: columns[statusIdx].map((v) => v) };
+  }
+  const observations = {
+    dimensions,
+    measure,
+    status,
+    model: {
+      dimensionIds: dimensionIdxs.map((i) => headerNames[i]),
+      roles: options.roles,
+      valueForm: options.valueForm ?? "auto",
+      statusForm: status ? "auto" : "none"
+    }
+  };
+  return observations;
+}
+async function csvToDataset(text, options) {
+  const { toDataset } = await import('./cubeBuilder-ZU76VVEX.cjs');
+  return toDataset(csvToCube(text, options));
+}
+function cubeColumns(obs) {
+  const dimIds = obs.model.dimensionIds;
+  const measureName = obs.measure.name ?? "value";
+  const hasStatus = obs.status !== void 0;
+  const header = [...dimIds, measureName];
+  if (hasStatus) header.push("status");
+  const n = obs.measure.values.length;
+  const rows = new Array(n);
+  for (let i = 0; i < n; i++) {
+    const cells = [];
+    for (const id of dimIds) {
+      const col = obs.dimensions[id];
+      cells.push(col ? col.values[i] : "");
+    }
+    const v = obs.measure.values[i];
+    cells.push(v === null ? "" : String(v));
+    if (hasStatus) cells.push(obs.status.values[i] ?? "");
+    rows[i] = cells.join("\0");
+  }
+  return { header, rows };
+}
+function quoteCsvField(value, delimiter, force = false) {
+  const needsQuoting = force || value.includes(delimiter) || value.includes('"') || value.includes("\r") || value.includes("\n");
+  if (!needsQuoting) return value;
+  return `"${value.replace(/"/g, '""')}"`;
+}
+function serializeRows(header, rows, options = {}) {
+  const delimiter = options.delimiter ?? ",";
+  const eol = options.lineTerminator ?? "\r\n";
+  const parts = [];
+  parts.push(header.map((h) => quoteCsvField(h, delimiter)).join(delimiter));
+  for (const row of rows) {
+    const cells = row.split("\0");
+    parts.push(cells.map((c) => quoteCsvField(c, delimiter)).join(delimiter));
+  }
+  return parts.join(eol) + eol;
+}
+function cubeToCsv(obs, options = {}) {
+  const { header, rows } = cubeColumns(obs);
+  return serializeRows(header, rows, options);
+}
+
+exports.CsvSourceError = CsvSourceError;
+exports.csvToCube = csvToCube;
+exports.csvToDataset = csvToDataset;
+exports.cubeColumns = cubeColumns;
+exports.cubeToCsv = cubeToCsv;
+exports.parseCsv = parseCsv;
+exports.quoteCsvField = quoteCsvField;
+exports.serializeRows = serializeRows;
+//# sourceMappingURL=chunk-KFIKGBIJ.cjs.map
+//# sourceMappingURL=chunk-KFIKGBIJ.cjs.map

package/dist/chunk-KFIKGBIJ.cjs.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../src/sources/csv.ts"],"names":["n"],"mappings":";;;AAqCO,IAAM,cAAA,GAAN,cAA6B,KAAA,CAAM;AAAA,EACxC,YAAY,OAAA,EAAiB;AAC3B,IAAA,KAAA,CAAM,OAAO,CAAA;AACb,IAAA,IAAA,CAAK,IAAA,GAAO,gBAAA;AAAA,EACd;AACF;AA4BO,SAAS,QAAA,CAAS,KAAA,EAAe,SAAA,GAAY,GAAA,EAAiB;AACnE,EAAA,MAAM,OAAmB,EAAC;AAC1B,EAAA,IAAI,MAAgB,EAAC;AACrB,EAAA,IAAI,KAAA,GAAQ,EAAA;AACZ,EAAA,IAAI,QAAA,GAAW,KAAA;AACf,EAAA,MAAM,KAAA,GAAQ,SAAA,CAAU,UAAA,CAAW,CAAC,CAAA;AAEpC,EAAA,KAAA,IAAS,CAAA,GAAI,CAAA,EAAG,CAAA,GAAI,KAAA,CAAM,QAAQ,CAAA,EAAA,EAAK;AACrC,IAAA,MAAM,CAAA,GAAI,KAAA,CAAM,UAAA,CAAW,CAAC,CAAA;AAC5B,IAAA,IAAI,QAAA,EAAU;AACZ,MAAA,IAAI,MAAM,EAAA,EAAY;AAEpB,QAAA,IAAI,KAAA,CAAM,UAAA,CAAW,CAAA,GAAI,CAAC,MAAM,EAAA,EAAI;AAClC,UAAA,KAAA,IAAS,GAAA;AACT,UAAA,CAAA,EAAA;AAAA,QACF,CAAA,MAAO;AACL,UAAA,QAAA,GAAW,KAAA;AAAA,QACb;AAAA,MACF,CAAA,MAAO;AACL,QAAA,KAAA,IAAS,MAAM,CAAC,CAAA;AAAA,MAClB;AAAA,IACF,CAAA,MAAO;AACL,MAAA,IAAI,MAAM,EAAA,EAAI;AACZ,QAAA,QAAA,GAAW,IAAA;AAAA,MACb,CAAA,MAAA,IAAW,MAAM,KAAA,EAAO;AACtB,QAAA,GAAA,CAAI,KAAK,KAAK,CAAA;AACd,QAAA,KAAA,GAAQ,EAAA;AAAA,MACV,CAAA,MAAA,IAAW,MAAM,EAAA,EAAa;AAE5B,QAAA,IAAI,KAAA,CAAM,UAAA,CAAW,CAAA,GAAI,CAAC,MAAM,EAAA,EAAI,CAAA,EAAA;AACpC,QAAA,GAAA,CAAI,KAAK,KAAK,CAAA;AACd,QAAA,IAAA,CAAK,KAAK,GAAG,CAAA;AACb,QAAA,GAAA,GAAM,EAAC;AACP,QAAA,KAAA,GAAQ,EAAA;AAAA,MACV,CAAA,MAAA,IAAW,MAAM,EAAA,EAAa;AAC5B,QAAA,GAAA,CAAI,KAAK,KAAK,CAAA;AACd,QAAA,IAAA,CAAK,KAAK,GAAG,CAAA;AACb,QAAA,GAAA,GAAM,EAAC;AACP,QAAA,KAAA,GAAQ,EAAA;AAAA,MACV,CAAA,MAAO;AACL,QAAA,KAAA,IAAS,MAAM,CAAC,CAAA;AAAA,MAClB;AAAA,IACF;AAAA,EACF;AAEA,EAAA,IAAI,KAAA,CAAM,MAAA,GAAS,CAAA,IAAK,GAAA,CAAI,SAAS,CAAA,EAAG;AACtC,IAAA,GAAA,CAAI,KAAK,KAAK,CAAA;AACd,IAAA,IAAA,CAAK,KAAK,GAAG,CAAA;AAAA,EACf;AACA,EAAA,OAAO,IAAA;AACT;AAOA,SAAS,aAAa,MAAA,EAA2B;AAC/C,EAAA,IAAI,GAAA,GAAM,KAAA;AACV,EAAA,KAAA,MAAW,KAAK,MAAA,EAAQ;AACtB,IAAA,IAAI,MAAM,EAAA,EAAI;AACd,IAAA,GAAA,GAAM,IAAA;AACN,IAAA,MAAM,CAAA,GAAI,OAAO,CAAC,CAAA;AAClB,IAAA,IAAI,CAAC,MAAA,CAAO,QAAA,CAAS,CAAC,GAAG,OAAO,KAAA;AAAA,EAClC;AACA,EAAA,OAAO,GAAA;AACT;AAWO,SAAS,SAAA,CAAU,IAAA,EAAc,OAAA,GAA4B,EAAC,EAAiB;AACpF,EAAA,MAAM,SAAA,GAAY,QAAQ,SAAA,IAAa,GAAA;AACvC,EAAA,MAAM,MAAA,GAAS,QAAQ,MAAA,IAAU,IAAA;AACjC,EAAA,MAAM,IAAA,GAAO,QAAA,CAAS,IAAA,EAAM,SAAS,CAAA;AACrC,EAAA,IAAI,KAAK,MAAA,KAAW,CAAA,EAAG,MAAM,IAAI,eAAe,iBAAiB,CAAA;AAEjE,EAAA,IAAI,WAAA;AACJ,EAAA,IAAI,QAAA;AACJ,EAAA,IAAI,MAAA,EAAQ;AACV,IAAA,WAAA,GAAc,KAAK,CAAC,CAAA;AACpB,IAAA,QAAA,GAAW,IAAA,CAAK,MAAM,CAAC,CAAA;AAAA,EACzB,CAAA,MAAO;AACL,IAAA,WAAA,GAAc,IAAA,CAAK,CAAC,CAAA,CAAE,GAAA,CAAI,CAAC,CAAA,EAAG,CAAA,KAAM,CAAA,GAAA,EAAM,CAAC,CAAA,CAAE,CAAA;AAC7C,IAAA,QAAA,GAAW,IAAA;AAAA,EACb;AACA,EAAA,IAAI,YAAY,MAAA,KAAW,CAAA,EAAG,MAAM,IAAI,eAAe,qBAAqB,CAAA;AAC5E,EAAA,IAAI,SAAS,MAAA,KAAW,CAAA,EAAG,MAAM,IAAI,eAAe,sBAAsB,CAAA;AAE1E,EAAA,MAAM,WAAW,WAAA,CAAY,MAAA;AAG7B,EAAA,MAAM,OAAA,GAAsB,WAAA,CAAY,GAAA,CAAI,MAAM,EAAE,CAAA;AACpD,EAAA,KAAA,MAAW,KAAK,QAAA,EAAU;AAExB,IAAA,KAAA,IAAS,CAAA,GAAI,CAAA,EAAG,CAAA,GAAI,QAAA,EAAU,CAAA,EAAA,EAAK,OAAA,CAAQ,CAAC,CAAA,CAAE,IAAA,CAAK,CAAA,CAAE,CAAC,CAAA,IAAK,EAAE,CAAA;AAAA,EAC/D;AAGA,EAAA,IAAI,aAAa,WAAA,CAAY,SAAA,CAAU,CAACA,EAAAA,KAAMA,EAAAA,KAAM,QAAQ,OAAO,CAAA;AACnE,EAAA,IAAI,eAAe,EAAA,EAAI;AAErB,IAAA,UAAA,GAAa,QAAQ,SAAA,CAAU,CAAC,GAAA,KAAQ,YAAA,CAAa,GAAG,CAAC,CAAA;AACzD,IAAA,IAAI,eAAe,EAAA,EAAI;AACrB,MAAA,MAAM,IAAI,cAAA;AAAA,QACR;AAAA,OAEF;AAAA,IACF;AAAA,EACF;AAGA,EAAA,IAAI,YAAY,WAAA,CAAY,SAAA,CAAU,CAACA,EAAAA,KAAMA,EAAAA,KAAM,QAAQ,MAAM,CAAA;AACjE,EAAA,IAAI,SAAA,KAAc,EAAA,IAAM,CAAC,OAAA,CAAQ,MAAA,EAAQ;AACvC,IAAA,SAAA,GAAY,YAAY,SAAA,CAAU,CAACA,OAAMA,EAAAA,CAAE,WAAA,OAAkB,QAAQ,CAAA;AACrE,IAAA,IAAI,SAAA,KAAc,YAAY,SAAA,GAAY,EAAA;AAAA,EAC5C;AAGA,EAAA,IAAI,aAAA;AACJ,EAAA,IAAI,QAAQ,UAAA,EAAY;AACtB,IAAA,aAAA,GAAgB,OAAA,CAAQ,UAAA,CAAW,GAAA,CAAI,CAACA,EAAAA,KAAM;AAC5C,MAAA,MAAM,GAAA,GAAM,WAAA,CAAY,OAAA,CAAQA,EAAC,CAAA;AACjC,MAAA,IAAI,QAAQ,EAAA,EAAI,MAAM,IAAI,cAAA,CAAe,CAAA,kBAAA,EAAqBA,EAAC,CAAA,yBAAA,CAA2B,CAAA;AAC1F,MAAA,OAAO,GAAA;AAAA,IACT,CAAC,CAAA;AAAA,EACH,CAAA,MAAO;AACL,IAAA,aAAA,GAAgB,WAAA,CACb,GAAA,CAAI,CAAC,CAAA,EAAG,CAAA,KAAM,CAAC,CAAA,CACf,MAAA,CAAO,CAAC,CAAA,KAAM,CAAA,KAAM,UAAA,IAAc,MAAM,SAAS,CAAA;AACpD,IAAA,IAAI,aAAA,CAAc,WAAW,CAAA,EAAG;AAC9B,MAAA,MAAM,IAAI,eAAe,4DAA4D,CAAA;AAAA,IACvF;AAAA,EACF;AAEA,EAAU,QAAA,CAAS;AACnB,EAAA,MAAM,aAA8C,EAAC;AACrD,EAAA,KAAA,MAAW,OAAO,aAAA,EAAe;AAC/B,IAAA,MAAM,IAAA,GAAO,YAAY,GAAG,CAAA;AAC5B,IAAA,UAAA,CAAW,IAAI,CAAA,GAAI;AAAA,MACjB,EAAA,EAAI,IAAA;AAAA,MACJ,MAAA,EAAQ,OAAA,CAAQ,GAAG,CAAA,CAAE,GAAA,CAAI,CAAC,CAAA,KAAO,CAAA,KAAM,EAAA,GAAK,EAAA,GAAK,CAAE;AAAA,KACrD;AAAA,EACF;AAEA,EAAA,MAAM,OAAA,GAAyB;AAAA,IAC7B,IAAA,EAAM,YAAY,UAAU,CAAA;AAAA,IAC5B,MAAA,EAAQ,OAAA,CAAQ,UAAU,CAAA,CAAE,GAAA,CAAI,CAAC,CAAA,KAAO,CAAA,KAAM,EAAA,GAAK,IAAA,GAAO,MAAA,CAAO,CAAC,CAAE;AAAA,GACtE;AAEA,EAAA,IAAI,MAAA;AACJ,EAAA,IAAI,cAAc,EAAA,EAAI;AACpB,IAAA,MAAA,GAAS,EAAE,QAAQ,OAAA,CAAQ,SAAS,EAAE,GAAA,CAAI,CAAC,CAAA,KAAM,CAAC,CAAA,EAAE;AAAA,EACtD;AAEA,EAAA,MAAM,YAAA,GAA6B;AAAA,IACjC,UAAA;AAAA,IACA,OAAA;AAAA,IACA,MAAA;AAAA,IACA,KAAA,EAAO;AAAA,MACL,cAAc,aAAA,CAAc,GAAA,CAAI,CAAC,CAAA,KAAM,WAAA,CAAY,CAAC,CAAC,CAAA;AAAA,MACrD,OAAO,OAAA,CAAQ,KAAA;AAAA,MACf,SAAA,EAAW,QAAQ,SAAA,IAAa,MAAA;AAAA,MAChC,UAAA,EAAY,SAAS,MAAA,GAAS;AAAA;AAChC,GACF;AAIA,EAAA,OAAO,YAAA;AACT;AAGA,eAAsB,YAAA,CACpB,MACA,OAAA,EAC0B;AAC1B,EAAA,MAAM,EAAE,SAAA,EAAU,GAAI,MAAM,OAAO,4BAAqB,CAAA;AACxD,EAAA,OAAO,SAAA,CAAU,SAAA,CAAU,IAAA,EAAM,OAAO,CAAC,CAAA;AAC3C;AAuBO,SAAS,YAAY,GAAA,EAG1B;AACA,EAAA,MAAM,MAAA,GAAS,IAAI,KAAA,CAAM,YAAA;AACzB,EAAA,MAAM,WAAA,GAAc,GAAA,CAAI,OAAA,CAAQ,IAAA,IAAQ,OAAA;AACxC,EAAA,MAAM,SAAA,GAAY,IAAI,MAAA,KAAW,MAAA;AAEjC,EAAA,MAAM,MAAA,GAAmB,CAAC,GAAG,MAAA,EAAQ,WAAW,CAAA;AAChD,EAAA,IAAI,SAAA,EAAW,MAAA,CAAO,IAAA,CAAK,QAAQ,CAAA;AAEnC,EAAA,MAAM,CAAA,GAAI,GAAA,CAAI,OAAA,CAAQ,MAAA,CAAO,MAAA;AAC7B,EAAA,MAAM,IAAA,GAAiB,IAAI,KAAA,CAAM,CAAC,CAAA;AAClC,EAAA,KAAA,IAAS,CAAA,GAAI,CAAA,EAAG,CAAA,GAAI,CAAA,EAAG,CAAA,EAAA,EAAK;AAC1B,IAAA,MAAM,QAAkB,EAAC;AACzB,IAAA,KAAA,MAAW,MAAM,MAAA,EAAQ;AACvB,MAAA,MAAM,GAAA,GAAM,GAAA,CAAI,UAAA,CAAW,EAAE,CAAA;AAC7B,MAAA,KAAA,CAAM,KAAK,GAAA,GAAM,GAAA,CAAI,MAAA,CAAO,CAAC,IAAI,EAAE,CAAA;AAAA,IACrC;AACA,IAAA,MAAM,CAAA,GAAI,GAAA,CAAI,OAAA,CAAQ,MAAA,CAAO,CAAC,CAAA;AAC9B,IAAA,KAAA,CAAM,KAAK,CAAA,KAAM,IAAA,GAAO,EAAA,GAAK,MAAA,CAAO,CAAC,CAAC,CAAA;AACtC,IAAA,IAAI,SAAA,QAAiB,IAAA,CAAK,GAAA,CAAI,OAAQ,MAAA,CAAO,CAAC,KAAK,EAAE,CAAA;AACrD,IAAA,IAAA,CAAK,CAAC,CAAA,GAAI,KAAA,CAAM,IAAA,CAAK,IAAQ,CAAA;AAAA,EAC/B;AACA,EAAA,OAAO,EAAE,QAAQ,IAAA,EAAK;AACxB;AAMO,SAAS,aAAA,CACd,KAAA,EACA,SAAA,EACA,KAAA,GAAQ,KAAA,EACA;AACR,EAAA,MAAM,eACJ,KAAA,IACA,KAAA,CAAM,QAAA,CAAS,SAAS,KACxB,KAAA,CAAM,QAAA,CAAS,GAAG,CAAA,IAClB,MAAM,QAAA,CAAS,IAAI,CAAA,IACnB,KAAA,CAAM,SAAS,IAAI,CAAA;AACrB,EAAA,IAAI,CAAC,cAAc,OAAO,KAAA;AAC1B,EAAA,OAAO,CAAA,CAAA,EAAI,KAAA,CAAM,OAAA,CAAQ,IAAA,EAAM,IAAI,CAAC,CAAA,CAAA,CAAA;AACtC;AAQO,SAAS,aAAA,CACd,MAAA,EACA,IAAA,EACA,OAAA,GAA4B,EAAC,EACrB;AACR,EAAA,MAAM,SAAA,GAAY,QAAQ,SAAA,IAAa,GAAA;AACvC,EAAA,MAAM,GAAA,GAAM,QAAQ,cAAA,IAAkB,MAAA;AACtC,EAAA,MAAM,QAAkB,EAAC;AACzB,EAAA,KAAA,CAAM,IAAA,CAAK,MAAA,CAAO,GAAA,CAAI,CAAC,CAAA,KAAM,aAAA,CAAc,CAAA,EAAG,SAAS,CAAC,CAAA,CAAE,IAAA,CAAK,SAAS,CAAC,CAAA;AACzE,EAAA,KAAA,MAAW,OAAO,IAAA,EAAM;AACtB,IAAA,MAAM,KAAA,GAAQ,GAAA,CAAI,KAAA,CAAM,IAAQ,CAAA;AAChC,IAAA,KAAA,CAAM,IAAA,CAAK,KAAA,CAAM,GAAA,CAAI,CAAC,CAAA,KAAM,aAAA,CAAc,CAAA,EAAG,SAAS,CAAC,CAAA,CAAE,IAAA,CAAK,SAAS,CAAC,CAAA;AAAA,EAC1E;AACA,EAAA,OAAO,KAAA,CAAM,IAAA,CAAK,GAAG,CAAA,GAAI,GAAA;AAC3B;AAmBO,SAAS,SAAA,CACd,GAAA,EACA,OAAA,GAA4B,EAAC,EACrB;AACR,EAAA,MAAM,EAAE,MAAA,EAAQ,IAAA,EAAK,GAAI,YAAY,GAAG,CAAA;AACxC,EAAA,OAAO,aAAA,CAAc,MAAA,EAAQ,IAAA,EAAM,OAAO,CAAA;AAC5C","file":"chunk-KFIKGBIJ.cjs","sourcesContent":["/**\n * CSV adapter — `jsonstat-io/csv`.\n *\n * **Import:** unlike Parquet/DuckDB/Polars (which funnel through the Arrow\n * hub), plain CSV has no native Arrow producer in this package, so this adapter\n * builds the [`Observations`](../model/ir.ts) IR **directly** from parsed rows.\n *\n * **Export:** [`cubeToCsv`] writes the [`Observations`] IR to RFC-4180-style\n * CSV text (dimension columns + measure column + optional status column).\n *\n * ## Mapping heuristics (no metadata)\n *\n * Without a schema, the adapter infers:\n *  - The **measure** is the first column that parses as a number for every\n *    non-empty row (or the one named by `options.measure`).\n *  - Every other column is a **dimension** (string-typed), in file order.\n *  - A column named `status` (case-insensitive) is treated as the status\n *    column unless `options.status` overrides it.\n *\n * Empty cells become `null` in the measure and `\"\"` in dimensions (then\n * rejected by the builder; sources should pre-clean). For richer, lossless\n * mapping with declared types/labels/roles, use CSVW (see [`csvw`](./csvw.ts)).\n */\n\nimport type {\n  DimensionColumn,\n  MeasureColumn,\n  Observations,\n  RoleMap,\n  StatusColumn,\n} from \"../model/ir\";\nimport type { JsonStatDataset } from \"../model/jsonstat\";\n\n// ---------------------------------------------------------------------------\n// Errors & options\n// ---------------------------------------------------------------------------\n\nexport class CsvSourceError extends Error {\n  constructor(message: string) {\n    super(message);\n    this.name = \"CsvSourceError\";\n  }\n}\n\nexport interface CsvToCubeOptions {\n  /** Explicit measure column name. Overrides inference. */\n  measure?: string;\n  /** Explicit dimension column names, in order. */\n  dimensions?: string[];\n  /** Explicit status column name. Overrides the `status`-name heuristic. */\n  status?: string;\n  /** Role assignments. */\n  roles?: RoleMap;\n  /** Value-form hint. */\n  valueForm?: \"auto\" | \"dense\" | \"sparse\";\n  /** CSV delimiter (default \",\"). */\n  delimiter?: string;\n  /** Treat the first row as a header (default true). */\n  header?: boolean;\n}\n\n// ---------------------------------------------------------------------------\n// Minimal CSV parser\n// ---------------------------------------------------------------------------\n\n/**\n * A tiny, dependency-free RFC-4180-ish CSV parser. Handles quoted fields,\n * embedded quotes (doubled), and embedded newlines. Sufficient for the\n * importer's needs; for heavy CSV work, prefer piping through DuckDB/Polars.\n */\nexport function parseCsv(input: string, delimiter = \",\"): string[][] {\n  const rows: string[][] = [];\n  let row: string[] = [];\n  let field = \"\";\n  let inQuotes = false;\n  const delim = delimiter.charCodeAt(0);\n\n  for (let i = 0; i < input.length; i++) {\n    const c = input.charCodeAt(i);\n    if (inQuotes) {\n      if (c === 34 /* \" */) {\n        // Doubled quote → literal quote; otherwise end quote.\n        if (input.charCodeAt(i + 1) === 34) {\n          field += '\"';\n          i++;\n        } else {\n          inQuotes = false;\n        }\n      } else {\n        field += input[i];\n      }\n    } else {\n      if (c === 34) {\n        inQuotes = true;\n      } else if (c === delim) {\n        row.push(field);\n        field = \"\";\n      } else if (c === 13 /* \\r */) {\n        // swallow; handle \\n\n        if (input.charCodeAt(i + 1) === 10) i++;\n        row.push(field);\n        rows.push(row);\n        row = [];\n        field = \"\";\n      } else if (c === 10 /* \\n */) {\n        row.push(field);\n        rows.push(row);\n        row = [];\n        field = \"\";\n      } else {\n        field += input[i];\n      }\n    }\n  }\n  // Flush trailing field/row.\n  if (field.length > 0 || row.length > 0) {\n    row.push(field);\n    rows.push(row);\n  }\n  return rows;\n}\n\n// ---------------------------------------------------------------------------\n// Type inference\n// ---------------------------------------------------------------------------\n\n/** Does every non-empty value parse as a finite number? */\nfunction isAllNumeric(values: string[]): boolean {\n  let any = false;\n  for (const v of values) {\n    if (v === \"\") continue;\n    any = true;\n    const n = Number(v);\n    if (!Number.isFinite(n)) return false;\n  }\n  return any;\n}\n\n// ---------------------------------------------------------------------------\n// Public API\n// ---------------------------------------------------------------------------\n\n/**\n * Parse CSV text into the [`Observations`] IR.\n *\n * @throws [`CsvSourceError`] on structural problems.\n */\nexport function csvToCube(text: string, options: CsvToCubeOptions = {}): Observations {\n  const delimiter = options.delimiter ?? \",\";\n  const header = options.header ?? true;\n  const rows = parseCsv(text, delimiter);\n  if (rows.length === 0) throw new CsvSourceError(\"CSV has no rows\");\n\n  let headerNames: string[];\n  let dataRows: string[][];\n  if (header) {\n    headerNames = rows[0];\n    dataRows = rows.slice(1);\n  } else {\n    headerNames = rows[0].map((_, i) => `col${i}`);\n    dataRows = rows;\n  }\n  if (headerNames.length === 0) throw new CsvSourceError(\"CSV header is empty\");\n  if (dataRows.length === 0) throw new CsvSourceError(\"CSV has no data rows\");\n\n  const colCount = headerNames.length;\n\n  // Transpose into columnar arrays for inference.\n  const columns: string[][] = headerNames.map(() => []);\n  for (const r of dataRows) {\n    // Pad/truncate to header length.\n    for (let c = 0; c < colCount; c++) columns[c].push(r[c] ?? \"\");\n  }\n\n  // Resolve measure column.\n  let measureIdx = headerNames.findIndex((n) => n === options.measure);\n  if (measureIdx === -1) {\n    // Inference: first all-numeric column.\n    measureIdx = columns.findIndex((col) => isAllNumeric(col));\n    if (measureIdx === -1) {\n      throw new CsvSourceError(\n        \"No measure column found: no column is fully numeric. \" +\n          \"Pass options.measure to name one (values will be parsed with Number()).\",\n      );\n    }\n  }\n\n  // Resolve status column.\n  let statusIdx = headerNames.findIndex((n) => n === options.status);\n  if (statusIdx === -1 && !options.status) {\n    statusIdx = headerNames.findIndex((n) => n.toLowerCase() === \"status\");\n    if (statusIdx === measureIdx) statusIdx = -1;\n  }\n\n  // Resolve dimensions: everything that's not measure or status.\n  let dimensionIdxs: number[];\n  if (options.dimensions) {\n    dimensionIdxs = options.dimensions.map((n) => {\n      const idx = headerNames.indexOf(n);\n      if (idx === -1) throw new CsvSourceError(`Dimension column \"${n}\" not found in CSV header`);\n      return idx;\n    });\n  } else {\n    dimensionIdxs = headerNames\n      .map((_, i) => i)\n      .filter((i) => i !== measureIdx && i !== statusIdx);\n    if (dimensionIdxs.length === 0) {\n      throw new CsvSourceError(\"No dimension columns remain after measure/status selection\");\n    }\n  }\n\n  const n = dataRows.length;\n  const dimensions: Record<string, DimensionColumn> = {};\n  for (const idx of dimensionIdxs) {\n    const name = headerNames[idx];\n    dimensions[name] = {\n      id: name,\n      values: columns[idx].map((v) => (v === \"\" ? \"\" : v)),\n    };\n  }\n\n  const measure: MeasureColumn = {\n    name: headerNames[measureIdx],\n    values: columns[measureIdx].map((v) => (v === \"\" ? null : Number(v))),\n  };\n\n  let status: StatusColumn | undefined;\n  if (statusIdx !== -1) {\n    status = { values: columns[statusIdx].map((v) => v) };\n  }\n\n  const observations: Observations = {\n    dimensions,\n    measure,\n    status,\n    model: {\n      dimensionIds: dimensionIdxs.map((i) => headerNames[i]),\n      roles: options.roles,\n      valueForm: options.valueForm ?? \"auto\",\n      statusForm: status ? \"auto\" : \"none\",\n    },\n  };\n  // n unused guard (kept for clarity; columns already encode length).\n  void n;\n\n  return observations;\n}\n\n/** Convenience: CSV text → JSON-stat [`Dataset`]. */\nexport async function csvToDataset(\n  text: string,\n  options?: CsvToCubeOptions,\n): Promise<JsonStatDataset> {\n  const { toDataset } = await import(\"../core/cubeBuilder\");\n  return toDataset(csvToCube(text, options));\n}\n\n// ---------------------------------------------------------------------------\n// Export: Observations IR → CSV text\n// ---------------------------------------------------------------------------\n\nexport interface CubeToCsvOptions {\n  /** CSV delimiter (default \",\"). */\n  delimiter?: string;\n  /** Line terminator: \"\\r\\n\" (RFC-4180 default) or \"\\n\". */\n  lineTerminator?: string;\n}\n\n/**\n * Resolve the ordered list of (column name, row value extractor) pairs for a\n * dataset IR. Shared by [`cubeToCsv`] and [`cubeToCsvw`] so the two writers\n * agree on column layout.\n *\n * Column order: dimension IDs (in `model.dimensionIds` order) → measure\n * (named `measure.name ?? \"value\"`) → status (`\"status\"`, if present).\n *\n * @internal\n */\nexport function cubeColumns(obs: Observations): {\n  header: string[];\n  rows: string[];\n} {\n  const dimIds = obs.model.dimensionIds;\n  const measureName = obs.measure.name ?? \"value\";\n  const hasStatus = obs.status !== undefined;\n\n  const header: string[] = [...dimIds, measureName];\n  if (hasStatus) header.push(\"status\");\n\n  const n = obs.measure.values.length;\n  const rows: string[] = new Array(n);\n  for (let i = 0; i < n; i++) {\n    const cells: string[] = [];\n    for (const id of dimIds) {\n      const col = obs.dimensions[id];\n      cells.push(col ? col.values[i] : \"\");\n    }\n    const v = obs.measure.values[i];\n    cells.push(v === null ? \"\" : String(v));\n    if (hasStatus) cells.push(obs.status!.values[i] ?? \"\");\n    rows[i] = cells.join(\"\\u0000\"); // placeholder; real delimiter applied later\n  }\n  return { header, rows };\n}\n\n/**\n * Quote a single CSV field per RFC-4180. A field needs quoting if it contains\n * the delimiter, a double quote, a CR, or an LF. Embedded quotes are doubled.\n */\nexport function quoteCsvField(\n  value: string,\n  delimiter: string,\n  force = false,\n): string {\n  const needsQuoting =\n    force ||\n    value.includes(delimiter) ||\n    value.includes('\"') ||\n    value.includes(\"\\r\") ||\n    value.includes(\"\\n\");\n  if (!needsQuoting) return value;\n  return `\"${value.replace(/\"/g, '\"\"')}\"`;\n}\n\n/**\n * Serialize a CSV header + rows (as produced by [`cubeColumns`]) into RFC-4180\n * text. Extracted so CSV and CSVW writers share one escaping path.\n *\n * @internal\n */\nexport function serializeRows(\n  header: string[],\n  rows: string[],\n  options: CubeToCsvOptions = {},\n): string {\n  const delimiter = options.delimiter ?? \",\";\n  const eol = options.lineTerminator ?? \"\\r\\n\";\n  const parts: string[] = [];\n  parts.push(header.map((h) => quoteCsvField(h, delimiter)).join(delimiter));\n  for (const row of rows) {\n    const cells = row.split(\"\\u0000\");\n    parts.push(cells.map((c) => quoteCsvField(c, delimiter)).join(delimiter));\n  }\n  return parts.join(eol) + eol;\n}\n\n/**\n * Write the [`Observations`](../model/ir.ts) IR to RFC-4180-style CSV text.\n *\n * The output is a tidy long table: one header row + one row per observation.\n * Columns are the dimension IDs (in `model.dimensionIds` order), then the\n * measure column (`measure.name ?? \"value\"`), then an optional `status`\n * column. Null measures are emitted as empty cells.\n *\n * This is the lossy inverse of [`csvToCube`]: labels, roles, and category\n * metadata are not preserved (use [`cubeToCsvw`] for a lossless export).\n *\n * @example\n * ```ts\n * import { cubeToCsv } from \"jsonstat-io/csv\";\n * const csv = cubeToCsv(observations);\n * ```\n */\nexport function cubeToCsv(\n  obs: Observations,\n  options: CubeToCsvOptions = {},\n): string {\n  const { header, rows } = cubeColumns(obs);\n  return serializeRows(header, rows, options);\n}\n\n"]}

package/dist/chunk-L65CRXTV.js

@@ -0,0 +1,64 @@
+// src/core/strides.ts
+function strides(size) {
+  const out = new Array(size.length);
+  let acc = 1;
+  for (let d = size.length - 1; d >= 0; d--) {
+    out[d] = acc;
+    acc *= size[d];
+  }
+  return out;
+}
+function totalCells(size) {
+  return size.reduce((a, b) => a * b, 1);
+}
+function flatPosition(indices, size) {
+  if (indices.length !== size.length) {
+    throw new Error(
+      `flatPosition: indices length ${indices.length} !== size length ${size.length}`
+    );
+  }
+  const s = strides(size);
+  let pos = 0;
+  for (let d = 0; d < indices.length; d++) {
+    const i = indices[d];
+    if (i < 0 || i >= size[d]) {
+      throw new Error(
+        `flatPosition: index ${i} out of range [0,${size[d]}) for dimension ${d}`
+      );
+    }
+    pos += i * s[d];
+  }
+  return pos;
+}
+function multiIndex(pos, size) {
+  const s = strides(size);
+  const total = totalCells(size);
+  if (pos < 0 || pos >= total) {
+    throw new Error(`multiIndex: position ${pos} out of range [0,${total})`);
+  }
+  const out = new Array(size.length);
+  let rem = pos;
+  for (let d = 0; d < size.length; d++) {
+    out[d] = Math.floor(rem / s[d]);
+    rem = rem % s[d];
+  }
+  return out;
+}
+function enumerateCells(size, visit) {
+  const k = size.length;
+  const total = totalCells(size);
+  if (total === 0) return;
+  const idx = new Array(k).fill(0);
+  for (let pos = 0; pos < total; pos++) {
+    visit(idx.slice(), pos);
+    for (let d = k - 1; d >= 0; d--) {
+      idx[d]++;
+      if (idx[d] < size[d]) break;
+      idx[d] = 0;
+    }
+  }
+}
+
+export { enumerateCells, flatPosition, multiIndex, strides, totalCells };
+//# sourceMappingURL=chunk-L65CRXTV.js.map
+//# sourceMappingURL=chunk-L65CRXTV.js.map

package/dist/chunk-L65CRXTV.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../src/core/strides.ts"],"names":[],"mappings":";AAmBO,SAAS,QAAQ,IAAA,EAA0B;AAChD,EAAA,MAAM,GAAA,GAAM,IAAI,KAAA,CAAc,IAAA,CAAK,MAAM,CAAA;AACzC,EAAA,IAAI,GAAA,GAAM,CAAA;AAEV,EAAA,KAAA,IAAS,IAAI,IAAA,CAAK,MAAA,GAAS,CAAA,EAAG,CAAA,IAAK,GAAG,CAAA,EAAA,EAAK;AACzC,IAAA,GAAA,CAAI,CAAC,CAAA,GAAI,GAAA;AACT,IAAA,GAAA,IAAO,KAAK,CAAC,CAAA;AAAA,EACf;AACA,EAAA,OAAO,GAAA;AACT;AAGO,SAAS,WAAW,IAAA,EAAwB;AACjD,EAAA,OAAO,KAAK,MAAA,CAAO,CAAC,GAAG,CAAA,KAAM,CAAA,GAAI,GAAG,CAAC,CAAA;AACvC;AAQO,SAAS,YAAA,CAAa,SAAmB,IAAA,EAAwB;AACtE,EAAA,IAAI,OAAA,CAAQ,MAAA,KAAW,IAAA,CAAK,MAAA,EAAQ;AAClC,IAAA,MAAM,IAAI,KAAA;AAAA,MACR,CAAA,6BAAA,EAAgC,OAAA,CAAQ,MAAM,CAAA,iBAAA,EAAoB,KAAK,MAAM,CAAA;AAAA,KAC/E;AAAA,EACF;AACA,EAAA,MAAM,CAAA,GAAI,QAAQ,IAAI,CAAA;AACtB,EAAA,IAAI,GAAA,GAAM,CAAA;AACV,EAAA,KAAA,IAAS,CAAA,GAAI,CAAA,EAAG,CAAA,GAAI,OAAA,CAAQ,QAAQ,CAAA,EAAA,EAAK;AACvC,IAAA,MAAM,CAAA,GAAI,QAAQ,CAAC,CAAA;AACnB,IAAA,IAAI,CAAA,GAAI,CAAA,IAAK,CAAA,IAAK,IAAA,CAAK,CAAC,CAAA,EAAG;AACzB,MAAA,MAAM,IAAI,KAAA;AAAA,QACR,uBAAuB,CAAC,CAAA,iBAAA,EAAoB,KAAK,CAAC,CAAC,mBAAmB,CAAC,CAAA;AAAA,OACzE;AAAA,IACF;AACA,IAAA,GAAA,IAAO,CAAA,GAAI,EAAE,CAAC,CAAA;AAAA,EAChB;AACA,EAAA,OAAO,GAAA;AACT;AAMO,SAAS,UAAA,CAAW,KAAa,IAAA,EAA0B;AAChE,EAAA,MAAM,CAAA,GAAI,QAAQ,IAAI,CAAA;AACtB,EAAA,MAAM,KAAA,GAAQ,WAAW,IAAI,CAAA;AAC7B,EAAA,IAAI,GAAA,GAAM,CAAA,IAAK,GAAA,IAAO,KAAA,EAAO;AAC3B,IAAA,MAAM,IAAI,KAAA,CAAM,CAAA,qBAAA,EAAwB,GAAG,CAAA,iBAAA,EAAoB,KAAK,CAAA,CAAA,CAAG,CAAA;AAAA,EACzE;AACA,EAAA,MAAM,GAAA,GAAM,IAAI,KAAA,CAAc,IAAA,CAAK,MAAM,CAAA;AACzC,EAAA,IAAI,GAAA,GAAM,GAAA;AACV,EAAA,KAAA,IAAS,CAAA,GAAI,CAAA,EAAG,CAAA,GAAI,IAAA,CAAK,QAAQ,CAAA,EAAA,EAAK;AACpC,IAAA,GAAA,CAAI,CAAC,CAAA,GAAI,IAAA,CAAK,MAAM,GAAA,GAAM,CAAA,CAAE,CAAC,CAAC,CAAA;AAC9B,IAAA,GAAA,GAAM,GAAA,GAAM,EAAE,CAAC,CAAA;AAAA,EACjB;AACA,EAAA,OAAO,GAAA;AACT;AASO,SAAS,cAAA,CACd,MACA,KAAA,EACM;AACN,EAAA,MAAM,IAAI,IAAA,CAAK,MAAA;AACf,EAAA,MAAM,KAAA,GAAQ,WAAW,IAAI,CAAA;AAC7B,EAAA,IAAI,UAAU,CAAA,EAAG;AACjB,EAAA,MAAM,MAAM,IAAI,KAAA,CAAc,CAAC,CAAA,CAAE,KAAK,CAAC,CAAA;AACvC,EAAA,KAAA,IAAS,GAAA,GAAM,CAAA,EAAG,GAAA,GAAM,KAAA,EAAO,GAAA,EAAA,EAAO;AACpC,IAAA,KAAA,CAAM,GAAA,CAAI,KAAA,EAAM,EAAG,GAAG,CAAA;AAEtB,IAAA,KAAA,IAAS,CAAA,GAAI,CAAA,GAAI,CAAA,EAAG,CAAA,IAAK,GAAG,CAAA,EAAA,EAAK;AAC/B,MAAA,GAAA,CAAI,CAAC,CAAA,EAAA;AACL,MAAA,IAAI,GAAA,CAAI,CAAC,CAAA,GAAI,IAAA,CAAK,CAAC,CAAA,EAAG;AACtB,MAAA,GAAA,CAAI,CAAC,CAAA,GAAI,CAAA;AAAA,IACX;AAAA,EACF;AACF","file":"chunk-L65CRXTV.js","sourcesContent":["/**\n * Row-major index math — the correctness backbone of the importer.\n *\n * JSON-stat stores cell values in a flat array laid out in **row-major order**,\n * described by the spec as \"what does not change, first\": the *last* dimension\n * in `id` changes fastest (see wiki/format-specification.md and\n * wiki/sparse-cubes.md).\n *\n * Concretely, for dimensions with sizes `s₀, s₁, …, sₖ₋₁`, the flat position of\n * a cell at multi-index `(i₀, i₁, …, iₖ₋₁)` is:\n *\n *     pos = i₀·(s₁·…·sₖ₋₁) + i₁·(s₂·…·sₖ₋₁) + … + iₖ₋₂·sₖ₋₁ + iₖ₋₁\n *\n * i.e. the stride of dimension `d` is the product of the sizes of all\n * dimensions that come *after* it. This module centralizes that arithmetic so\n * every converter and the round-trip tests share one source of truth.\n */\n\n/** Compute row-major strides from a `size` array. strides[d] = ∏ size[d+1..]. */\nexport function strides(size: number[]): number[] {\n  const out = new Array<number>(size.length);\n  let acc = 1;\n  // Walk right-to-left so each stride is the running product of later sizes.\n  for (let d = size.length - 1; d >= 0; d--) {\n    out[d] = acc;\n    acc *= size[d];\n  }\n  return out;\n}\n\n/** Total number of cells = product of all sizes (the dense cell count). */\nexport function totalCells(size: number[]): number {\n  return size.reduce((a, b) => a * b, 1);\n}\n\n/**\n * Convert a multi-index to a flat row-major position.\n *\n * `indices` length must equal `size` length; each `indices[d]` must satisfy\n * `0 ≤ indices[d] < size[d]`.\n */\nexport function flatPosition(indices: number[], size: number[]): number {\n  if (indices.length !== size.length) {\n    throw new Error(\n      `flatPosition: indices length ${indices.length} !== size length ${size.length}`,\n    );\n  }\n  const s = strides(size);\n  let pos = 0;\n  for (let d = 0; d < indices.length; d++) {\n    const i = indices[d];\n    if (i < 0 || i >= size[d]) {\n      throw new Error(\n        `flatPosition: index ${i} out of range [0,${size[d]}) for dimension ${d}`,\n      );\n    }\n    pos += i * s[d];\n  }\n  return pos;\n}\n\n/**\n * Inverse of [`flatPosition`](#flatposition): decompose a flat position back\n * into a multi-index. Used by the cube reader (Phase-2 seam) and by tests.\n */\nexport function multiIndex(pos: number, size: number[]): number[] {\n  const s = strides(size);\n  const total = totalCells(size);\n  if (pos < 0 || pos >= total) {\n    throw new Error(`multiIndex: position ${pos} out of range [0,${total})`);\n  }\n  const out = new Array<number>(size.length);\n  let rem = pos;\n  for (let d = 0; d < size.length; d++) {\n    out[d] = Math.floor(rem / s[d]);\n    rem = rem % s[d];\n  }\n  return out;\n}\n\n/**\n * Enumerate every multi-index in row-major order, yielding flat positions\n * 0..(total-1). Useful for materializing a dense value array from a sparse map.\n *\n * Calls `visit(indices, flatPos)` for each cell. Iterative (no recursion) to\n * avoid stack overflow on large cubes.\n */\nexport function enumerateCells(\n  size: number[],\n  visit: (indices: number[], flatPos: number) => void,\n): void {\n  const k = size.length;\n  const total = totalCells(size);\n  if (total === 0) return;\n  const idx = new Array<number>(k).fill(0);\n  for (let pos = 0; pos < total; pos++) {\n    visit(idx.slice(), pos);\n    // Increment the last dimension first (it changes fastest), carrying over.\n    for (let d = k - 1; d >= 0; d--) {\n      idx[d]++;\n      if (idx[d] < size[d]) break;\n      idx[d] = 0; // carry\n    }\n  }\n}\n"]}