shared-snippets/matlab/csvconverter/csv2catalog.m

function [Catalog] = csv2catalog(csvFilePath, column_desc, idPrefix, sortCatalogByTime, useCsvHeader)
% DESCRIPTION: Program to create the catalogue v2.0x in the
%              Matlab format file from csv file.
% INPUTS:
% - csvFilePath :           path to text file with csv data. The separator between columns is defined by ‘,'.
%     Description of the most popular catalog fields can be found at https://docs.cyfronet.pl/display/ISDOC/Catalog+-+description
% - column_desc :           structure containing information how the catalog should be created
%    - include: whether the field should be included in the resulting catalog
%    - nameInCsv: name of the column in CSV file
%    - inputType: type of the column in CSV file (REAL/INTEGER/TEXT/DATE_TIME/DATE/DATE_DAY/DATE_MONTH/DATE_YEAR/TIME)
%    - inputTimeFormat: input format for reading time (if 'inputType' is one of time options)
%    - inputTimeGroup: input group for merging dates
%    - nameInCatalog: name of the column that should be insterted into resulting catalog
%    - description: description of the column in the resulting catalog
%    - format: format (display format) of the column in the resulting catalog
%    - unit: unit of the column in the resulting catalog
%    - fieldType: type of the column in the resulting catalog (e.g. Magnitude, Energy)
% - idPrefix :            prefix of the ID column if the IDs should be generated (if the catalog doesn't contain
%     the id or the column is not included ('include' is set to false)

  %TODO handle multiple time columns
  %TODO the script fails if any of the rows has an empty value at the end (but empty quotes is ok)

  data = readAndCheckHeaders(csvFilePath, column_desc, useCsvHeader);
  colCount = length(column_desc);
  noCsvHeaderModifier = 1;
  if useCsvHeader
    noCsvHeaderModifier = 0;
  end
  rowCount = length(data) / colCount - 1 + noCsvHeaderModifier;
  k = 1;    % column number in the generated catalog

  if ~contains_id(column_desc)
    if isempty(idPrefix)
      [~, idPrefix] = fileparts(csvFilePath);
    end
    Catalog(k).field = 'ID';
    Catalog(k).type = 3;
    for row = 1 : rowCount
      Catalog(k).val(row, 1) = { strcat(idPrefix, '_', num2str(row,'%04.f')) };
    end
    Catalog(k).unit = [];
    Catalog(k).description = 'Event ID';
    Catalog(k).fieldType = [];
    k = 2;
  end

  [columnsInTimeGroup, columnsNotToInclude] = getTimeGroups(column_desc);

  for col=1:colCount
      current_col = column_desc(col);
      if ~current_col.include || ismember(col, columnsNotToInclude)
          continue;
      end

      inputTimeGroup = current_col.inputTimeGroup;
      if ~isempty(current_col.inputTimeGroup)
        timeGroupArray = columnsInTimeGroup(inputTimeGroup);
        for columnInGroup = 2 : length(timeGroupArray)
          current_col.inputTimeFormat = [current_col.inputTimeFormat "-" column_desc(timeGroupArray(columnInGroup)).inputTimeFormat];
        end
      end

      Catalog(k).field = current_col.nameInCatalog;
      Catalog(k).type = current_col.format;
      for row = 1 : rowCount
          rawValue = data{(colCount * (row - noCsvHeaderModifier)) + col, 1};
          if ~isempty(current_col.inputTimeGroup)
            timeGroupArray = columnsInTimeGroup(inputTimeGroup);
            for columnInGroup = 2 : length(timeGroupArray)
               rawValue = [rawValue "-" data{(colCount * (row - noCsvHeaderModifier)) + timeGroupArray(columnInGroup), 1}];
            end
          end
          if isempty(rawValue)
            if strcmp(current_col.nameInCatalog, 'ID')
              error('ID of the event cannot be empty (row: %d)', row)
            elseif isText(current_col)
              Catalog(k).val(row, 1) = {''};
            else
              Catalog(k).val(row, 1) = NaN;
            end
          else
            parsedValue = parseTextValue(rawValue, current_col.inputType, current_col.inputTimeFormat);
            if strcmp(current_col.format, '5a')
                  Catalog(k).val(row, 1) = { datestr(parsedValue, 'yyyy') };
              elseif strcmp(current_col.format, '5b')
                  Catalog(k).val(row, 1) = { datestr(parsedValue, 'yyyy-mm') };
              else
                  Catalog(k).val(row, 1) = parsedValue;
              end
          end
      end
      Catalog(k).unit = current_col.unit;
      Catalog(k).description = current_col.description;
      Catalog(k).fieldType = current_col.fieldType;
      k=k+1;
  end
  if sortCatalogByTime
    Catalog = sortByTime(Catalog);
  end
end

function containsId = contains_id(column_desc)
  idIdxs = find(strcmp(column_desc(1).nameInCatalog, 'ID'));
  if isempty(idIdxs)
    containsId = 0;
  else
    containsId = column_desc(idIdxs).include;
  end
end