@loaders.gl/parquet 3.4.11 → 3.4.13

package/dist/parquetjs/schema/shred.js

@@ -1,347 +0,0 @@
-"use strict";
-// Forked from https://github.com/kbajalc/parquets under MIT license (Copyright (c) 2017 ironSource Ltd.)
-var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
-    if (k2 === undefined) k2 = k;
-    var desc = Object.getOwnPropertyDescriptor(m, k);
-    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
-      desc = { enumerable: true, get: function() { return m[k]; } };
-    }
-    Object.defineProperty(o, k2, desc);
-}) : (function(o, m, k, k2) {
-    if (k2 === undefined) k2 = k;
-    o[k2] = m[k];
-}));
-var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
-    Object.defineProperty(o, "default", { enumerable: true, value: v });
-}) : function(o, v) {
-    o["default"] = v;
-});
-var __importStar = (this && this.__importStar) || function (mod) {
-    if (mod && mod.__esModule) return mod;
-    var result = {};
-    if (mod != null) for (var k in mod) if (k !== "default" && Object.prototype.hasOwnProperty.call(mod, k)) __createBinding(result, mod, k);
-    __setModuleDefault(result, mod);
-    return result;
-};
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.materializeRecords = exports.shredRecord = exports.shredBuffer = exports.ParquetBuffer = void 0;
-const declare_1 = require("./declare");
-Object.defineProperty(exports, "ParquetBuffer", { enumerable: true, get: function () { return declare_1.ParquetBuffer; } });
-const Types = __importStar(require("./types"));
-function shredBuffer(schema) {
-    const columnData = {};
-    for (const field of schema.fieldList) {
-        columnData[field.key] = {
-            dlevels: [],
-            rlevels: [],
-            values: [],
-            pageHeaders: [],
-            count: 0
-        };
-    }
-    return { rowCount: 0, columnData };
-}
-exports.shredBuffer = shredBuffer;
-/**
- * 'Shred' a record into a list of <value, repetition_level, definition_level>
- * tuples per column using the Google Dremel Algorithm..
- *
- * The buffer argument must point to an object into which the shredded record
- * will be returned. You may re-use the buffer for repeated calls to this function
- * to append to an existing buffer, as long as the schema is unchanged.
- *
- * The format in which the shredded records will be stored in the buffer is as
- * follows:
- *
- *   buffer = {
- *     columnData: [
- *       'my_col': {
- *          dlevels: [d1, d2, .. dN],
- *          rlevels: [r1, r2, .. rN],
- *          values: [v1, v2, .. vN],
- *        }, ...
- *      ],
- *      rowCount: X,
- *   }
- */
-function shredRecord(schema, record, buffer) {
-    /* shred the record, this may raise an exception */
-    const data = shredBuffer(schema).columnData;
-    shredRecordFields(schema.fields, record, data, 0, 0);
-    /* if no error during shredding, add the shredded record to the buffer */
-    if (buffer.rowCount === 0) {
-        buffer.rowCount = 1;
-        buffer.columnData = data;
-        return;
-    }
-    buffer.rowCount += 1;
-    for (const field of schema.fieldList) {
-        Array.prototype.push.apply(buffer.columnData[field.key].rlevels, data[field.key].rlevels);
-        Array.prototype.push.apply(buffer.columnData[field.key].dlevels, data[field.key].dlevels);
-        Array.prototype.push.apply(buffer.columnData[field.key].values, data[field.key].values);
-        buffer.columnData[field.key].count += data[field.key].count;
-    }
-}
-exports.shredRecord = shredRecord;
-// eslint-disable-next-line max-statements, complexity
-function shredRecordFields(fields, record, data, rLevel, dLevel) {
-    for (const name in fields) {
-        const field = fields[name];
-        // fetch values
-        let values = [];
-        if (record &&
-            field.name in record &&
-            record[field.name] !== undefined &&
-            record[field.name] !== null) {
-            if (record[field.name].constructor === Array) {
-                values = record[field.name];
-            }
-            else {
-                values.push(record[field.name]);
-            }
-        }
-        // check values
-        if (values.length === 0 && Boolean(record) && field.repetitionType === 'REQUIRED') {
-            throw new Error(`missing required field: ${field.name}`);
-        }
-        if (values.length > 1 && field.repetitionType !== 'REPEATED') {
-            throw new Error(`too many values for field: ${field.name}`);
-        }
-        // push null
-        if (values.length === 0) {
-            if (field.isNested) {
-                shredRecordFields(field.fields, null, data, rLevel, dLevel);
-            }
-            else {
-                data[field.key].count += 1;
-                data[field.key].rlevels.push(rLevel);
-                data[field.key].dlevels.push(dLevel);
-            }
-            continue; // eslint-disable-line no-continue
-        }
-        // push values
-        for (let i = 0; i < values.length; i++) {
-            const rlvl = i === 0 ? rLevel : field.rLevelMax;
-            if (field.isNested) {
-                shredRecordFields(field.fields, values[i], data, rlvl, field.dLevelMax);
-            }
-            else {
-                data[field.key].count += 1;
-                data[field.key].rlevels.push(rlvl);
-                data[field.key].dlevels.push(field.dLevelMax);
-                data[field.key].values.push(Types.toPrimitive((field.originalType || field.primitiveType), values[i]));
-            }
-        }
-    }
-}
-/**
- * 'Materialize' a list of <value, repetition_level, definition_level>
- * tuples back to nested records (objects/arrays) using the Google Dremel
- * Algorithm..
- *
- * The buffer argument must point to an object with the following structure (i.e.
- * the same structure that is returned by shredRecords):
- *
- *   buffer = {
- *     columnData: [
- *       'my_col': {
- *          dlevels: [d1, d2, .. dN],
- *          rlevels: [r1, r2, .. rN],
- *          values: [v1, v2, .. vN],
- *        }, ...
- *      ],
- *      rowCount: X,
- *   }
- */
-function materializeRecords(schema, buffer) {
-    const records = [];
-    for (let i = 0; i < buffer.rowCount; i++) {
-        records.push({});
-    }
-    for (const key in buffer.columnData) {
-        const columnData = buffer.columnData[key];
-        if (columnData.count) {
-            materializeColumn(schema, columnData, key, records);
-        }
-    }
-    return records;
-}
-exports.materializeRecords = materializeRecords;
-// eslint-disable-next-line max-statements, complexity
-function materializeColumn(schema, columnData, key, records) {
-    const field = schema.findField(key);
-    const branch = schema.findFieldBranch(key);
-    // tslint:disable-next-line:prefer-array-literal
-    const rLevels = new Array(field.rLevelMax + 1).fill(0);
-    let vIndex = 0;
-    for (let i = 0; i < columnData.count; i++) {
-        const dLevel = columnData.dlevels[i];
-        const rLevel = columnData.rlevels[i];
-        rLevels[rLevel]++;
-        rLevels.fill(0, rLevel + 1);
-        let rIndex = 0;
-        let record = records[rLevels[rIndex++] - 1];
-        // Internal nodes - Build a nested row object
-        for (const step of branch) {
-            if (step === field || dLevel < step.dLevelMax) {
-                break;
-            }
-            switch (step.repetitionType) {
-                case 'REPEATED':
-                    if (!(step.name in record)) {
-                        // eslint-disable max-depth
-                        record[step.name] = [];
-                    }
-                    const ix = rLevels[rIndex++];
-                    while (record[step.name].length <= ix) {
-                        // eslint-disable max-depth
-                        record[step.name].push({});
-                    }
-                    record = record[step.name][ix];
-                    break;
-                default:
-                    record[step.name] = record[step.name] || {};
-                    record = record[step.name];
-            }
-        }
-        // Leaf node - Add the value
-        if (dLevel === field.dLevelMax) {
-            const value = Types.fromPrimitive(
-            // @ts-ignore
-            field.originalType || field.primitiveType, columnData.values[vIndex], field);
-            vIndex++;
-            switch (field.repetitionType) {
-                case 'REPEATED':
-                    if (!(field.name in record)) {
-                        // eslint-disable max-depth
-                        record[field.name] = [];
-                    }
-                    const ix = rLevels[rIndex];
-                    while (record[field.name].length <= ix) {
-                        // eslint-disable max-depth
-                        record[field.name].push(null);
-                    }
-                    record[field.name][ix] = value;
-                    break;
-                default:
-                    record[field.name] = value;
-            }
-        }
-    }
-}
-// Columnar export
-/**
- * 'Materialize' a list of <value, repetition_level, definition_level>
- * tuples back to nested records (objects/arrays) using the Google Dremel
- * Algorithm..
- *
- * The buffer argument must point to an object with the following structure (i.e.
- * the same structure that is returned by shredRecords):
- *
- *   buffer = {
- *     columnData: [
- *       'my_col': {
- *          dlevels: [d1, d2, .. dN],
- *          rlevels: [r1, r2, .. rN],
- *          values: [v1, v2, .. vN],
- *        }, ...
- *      ],
- *      rowCount: X,
- *   }
- *
-export function extractColumns(schema: ParquetSchema, buffer: ParquetBuffer): Record<string, unknown> {
-  const columns: ParquetRecord = {};
-  for (const key in buffer.columnData) {
-    const columnData = buffer.columnData[key];
-    if (columnData.count) {
-      extractColumn(schema, columnData, key, columns);
-    }
-  }
-  return columns;
-}
-
-// eslint-disable-next-line max-statements, complexity
-function extractColumn(
-  schema: ParquetSchema,
-  columnData: ParquetData,
-  key: string,
-  columns: Record<string, unknown>
-) {
-  if (columnData.count <= 0) {
-    return;
-  }
-
-  const record = columns;
-
-  const field = schema.findField(key);
-  const branch = schema.findFieldBranch(key);
-
-  // tslint:disable-next-line:prefer-array-literal
-  const rLevels: number[] = new Array(field.rLevelMax + 1).fill(0);
-  let vIndex = 0;
-
-  let i = 0;
-  const dLevel = columnData.dlevels[i];
-  const rLevel = columnData.rlevels[i];
-  rLevels[rLevel]++;
-  rLevels.fill(0, rLevel + 1);
-
-  let rIndex = 0;
-  let record = records[rLevels[rIndex++] - 1];
-
-  // Internal nodes
-  for (const step of branch) {
-    if (step === field || dLevel < step.dLevelMax) {
-      break;
-    }
-
-    switch (step.repetitionType) {
-      case 'REPEATED':
-        if (!(step.name in record)) {
-          // eslint-disable max-depth
-          record[step.name] = [];
-        }
-        const ix = rLevels[rIndex++];
-        while (record[step.name].length <= ix) {
-          // eslint-disable max-depth
-          record[step.name].push({});
-        }
-        record = record[step.name][ix];
-        break;
-
-      default:
-        record[step.name] = record[step.name] || {};
-        record = record[step.name];
-    }
-  }
-
-  // Leaf node
-  if (dLevel === field.dLevelMax) {
-    const value = Types.fromPrimitive(
-      // @ts-ignore
-      field.originalType || field.primitiveType,
-      columnData.values[vIndex],
-      field
-    );
-    vIndex++;
-
-    switch (field.repetitionType) {
-      case 'REPEATED':
-        if (!(field.name in record)) {
-          // eslint-disable max-depth
-          record[field.name] = [];
-        }
-        const ix = rLevels[rIndex];
-        while (record[field.name].length <= ix) {
-          // eslint-disable max-depth
-          record[field.name].push(null);
-        }
-        record[field.name][ix] = value;
-        break;
-
-      default:
-        record[field.name] = value;
-    }
-  }
-}
-*/