IMPORT_CSV destStruct, delimiter, fName, timeZone, retCodeIdent_Int[, timeMaskIdent][, lineFrom, lineTo][TIME] [UTF8 | ENCODING "@APP_DEFAULT@"] [COLMAP _colMap]
IMPORT_CSV destStruct, delimiter, handle, timeZone, retCodeIdent_Int[, timeMaskIdent], numLines [TIME] [COLMAP _colMap]
|destStruct||in||Column identifier or identifier of local variable of Record type.|
|delimiter||in||Identifier of Text type - delimiter used in CSV file.|
|fName||in||File name with extension.|
Parameter of Text, Int or Bool type - defines interpretation of timestamps.
|timeMaskIdent||in||Identifier of Text type - mask for reading a value of Absolute time type.|
|retCodeIdent||out||Return code of Int type - action success.|
|lineFrom||in||Parameter of Int type - begin line.|
|lineTo||in||Parameter of Int type - end line.|
|handle||in||Parameter of Int type - value returned by the function FIO_OpenRead.|
|numLines||in||Parameter of Int type - maximum number of rows to read.|
|ENCODING||in||Key word which is followed by a name of file encoding in the text format. The list of supported encodings in ESL.|
|COLMAP||in||Key word which is followed by identifier of TEXT type - _colMap.|
|_colMap||in||Identifier of TEXT type. It includes the column names of structure, into which data from csv file will be inserted.|
The CSV file format is described in the action EXPORT_CSV.
If the identifier timeMaskIdent is not defined, the function expects the "dd.mm.yyyy hh:mi:ss" format for items containing an absolute time. If the parameter is defined, the format of the time is to be imported according to that mask. In case that the hh, mi or ss time components are missing, they will be automatically replaced with 0.
When real numbers are imported the separator of decimal places may be both "." (dot) and "," (comma) symbols.
When integer and real numbers are imported from CSV file the thousand separator is a blank space.
If the value of the parameter timeMaskIdent is "" (empty mask), values of Absolute time and Real type are to be converted according to the settings in the Regional and Language Options of the current user on Windows systems. For different operating systems the empty mask is automatically replaced by the value of "dd.mm.rrrr hh:mi:ss".
When the item is converted to Absolute time and is an empty string, the action interprets such item as invalid value.
If the key word TIME is used, there is the assumption, that the occurrence times of values are saved in the CSV file (the double number of columns in this CSV file).
If the key word UTF8 is used the action read the text file and supposes that it is encoded in UTF-8. From that reason the action converts the file from UTF-8 -> WIN1250 at its reading.
In case of need, the local variable
destStruct will be resized. Its size is given by the number of values in the particular CSV file.
The parameter timeZone defines interpretation of timestamps. In case of empty text, timestamps are interpreted as local times. In case of defined time zone name (e.g. "Europe/London"), timestamps are interpreted as local times for given time zone. In case of integer value, timestamps are interpreted as times with fixed offset from UTC, where given value defines offset in seconds. For historical reasons boolean values are accepted: @FALSE - timestamps are interpreted as local times, @TRUE - timestamps are interpreted as times with fixed offset of 3600 seconds (1 hour) from UTC. Usage of boolean value @TRUE is not recommended and generates warning when ESL script is being saved. It is recommended to replace it with integer value.
If the parameters lineFrom and lineTo are defined, there will be imported all the lines from the line defined by LineFrom up to the line defined by the parameter lineTo. If the value of both the parameters is -1, there will be imported the whole CSV file.
The second variant of the action reads the next rows (number is defined by the parameter numLines) from the given CSV file. The file must be open by the function FIO_OpenRead.
It is useful to use this variant for large CSV files - we recommend you not to read them at the same time.
The CSV delimiter (the parameter delimiter) may be obtained by calling the function %GetCSVDelimiter.
The parameter _colMap enables to define the columns of destination structure and the order of inserting data from the imported file. The names of columns must be separated by a delimiter. If some columns form .csv file are ignored, they must be defined in the parameter _colMap with the help of delimiter.
If the column names in _colMap differ from the structure definition, script declares an error "Invalid name of Cols in ColMap: bad column names".
If parameter _colMap contains a zero-length string, script declares the error "ColMap is empty".
If .csv file contains less columns than is required, the missing columns are imported as invalid values!
Example of use:
IMPORT_CSV destStruct, delimiter, fName, bMonoTimes, retCodeIdent_Int, , lineFrom, lineTo
RECORD NOALIAS (SD.AZZD_Imp_Day) _Import_H INT _retCode INT _handle _handle := %FIO_OpenRead("c:\Application\Import\LP_0407_0408180810.csv") DO_LOOP IMPORT_CSV _Import_H, ";", _handle, @FALSE, _retCode, "dd.mm.rr", 1000 ; processing read data EXIT_LOOP _Import_H\DIM < 1000 END_LOOP _bOk := %FIO_Close(_handle)