buoy-analysis/server/noaa.js

/**
 * NOAA-specific filtering.
 */
module.exports = {
    /**
     * Different years have added new columns.
     * 46054h1998 --> 16 cols, Missing mm (col 5).  Missing TIDE (col 18)
     * 46054h2004 --> 17 cols. Added TIDE. Missing mm (col 5)
     * 46054h2006 --> 18 cols, Added mm.
     */
    parseLine: function(str) {
        var arr = str.trim().split(/\s+/);
        // REMOVE THIS LINE? Ben 290116
        arr.filter(function(val) { return (val.length > 0); })

        // Add TIDE value if not there
        if (arr.length === 16) {
            arr.push('99.00');
        }

        // Add mm value if not there
        if (arr.length === 17) {
            arr.splice(4, 0, '00');
        }

        return arr;
    },

    /**
     * Receives a stream from a file read event.
     */
    parseTxt: function(str) {
        console.log('Parsing NOAA space-delimited columnar data into JSON.');

        var arr = [];
        var cols = null;
        var lines = str.split('\n');
        var len = lines.length;

        if (len > 8) {
            for (var i = 0; i < len; i++) {
                cols = module.exports.parseLine(lines[i]);
                cols.length > 0 ? arr.push(cols) : null;
            }
        }

        return arr;
    },

    /**
     * After all files have been parsed, Promises.all passes them all as an array.
     * This function does filtering on them and finalizes a JSON string.
     */
    convert: function(arr) {
        console.log('Converting aggregated month files to JSON.');

        // Sort.
        var sorted = arr.sort(function(a, b) {
            var dateA = parseInt([a[0], a[1], a[2], a[3], ('00' + a[4]).substr(-2)].join('')) || 0;
            var dateB = parseInt([b[0], b[1], b[2], b[3], ('00' + b[4]).substr(-2)].join('')) || 0;

            return dateA - dateB;
        });

        // Flag violations: headings, units, empty.
        // Prepare average to trim incidental head/tail data points from different years.
        var sum = 0;
        var count = 0;

        sorted.forEach(function(row) {
            if (row[0] === '#YY' || row[0] === 'YYYY' || row[0] === 'YY' || row[0] === '#yr' || row.length === 1) {
                row[0] = null;
            }
            else {
                row[0] = ('19' + row[0]).substr(-4);
                sum += parseInt(row[0]);
                count++;
            }
        });

        // Filter rows that have been flagged or that are the wrong year.
        var year = Math.round(sum / count);
        var result = sorted.filter(function(row, i) {
            if (row[0] === null || row[0] != year) {
                return false;
            }

            return true;
        });

        // Convert to JSON that can later be read easily.
        var str = null;
        if (result.length > 0) {
            str = JSON.stringify(result)
            str = str.replace(/\],\[/g, '],\n[');
        }

        return str;
    },

    /**
     * Used to aggregate month files after they have been split into a lines array.
     * Each line has been split into individual elements.
     * The array passed to this function is therefore an array of two dimensional arrays.
     *
     * This function adds non-empty lines to a common result set.
     */
    aggregate: function(arr) {
        console.log('Aggregating month files for the year.');

        var tmp = [];

        arr.forEach(function(rows) {
            if (rows.length === 0) {
                return;
            }

            tmp = tmp.concat(rows);
        });

        return tmp;
    }
};