Release Notes for JRButils for Micro Focus v25

Changes to Multiple Programs

  • Added two new values for environment variable JRBDATETIME. A value of ISO displays date and times using local time in ISO format e.g. 20180524132145. A value of ISOZ displays dates and times using GMT time in ISO format e.g. 20180524012145Z.
  • Added the ability to accept date/times in ISO format, either in local, or GMT time if ‘Z’ is appended. Not all variations on the ISO format are accepted but the following are:
    • A date may be given in the form yyyymmdd or yyyy-mm-dd e.g. 2018-08-23.
    • A time may be given in the form hhmmss or hh:mm:ss e.g. 21:19:03.
    • The minutes and seconds may be omitted as per the traditional date/time formats.
    • A date and time may be concatenated together without a separator e.g. 2018-08-2321:19:03 or separated by ‘T’ e.g. 2018-08-23T21:19:03.
    • ‘Z’ may be appended to a date and time to indicate that the value is in GMT time rather than local time.
  • Made changes to the code for copying directory structures, used by programs such as netcopy and fsupdate, to avoid an apparent NSS/NCP bug where on rare occasions, not all subdirectories in a directory were copied. The NCP request for retrieving entries (NCP 89 (20)) could return “no more entries” when the previous request correctly indicated not all entries had yet been returned. Copying via Windows File Manager worked correctly. Changing the fields requested (attributes, creation date, etc) to closely match those in NCP requests issued by the Client for OES Enterprise Server, appears to have bypassed the issue. An error is now reported when NCP 89 (20) returns zero entries, but the previous request indicated there were more entries.
  • Made changes to the code for copying directory structures, used by programs such as netcopy and fsupdate, to detect an unusual situation where the long name of a file is a valid DOS name (e.g. HUBERT~1.DOC), and another file has that actual DOS name. It is not known how this arises but NSS/NCP does not allow creation of a file with a long name that matches either another long name or a DOS name in the same directory. The file with the long name equalling a DOS name must be renamed before it can be copied to another NSS location. The programs now detect this situation and give an appropriate error message.
  • Changed the size of the buffer to the maximum allowed when retrieving group members. This was in response to an apparent eDirectory bug where not all members were returned for a complex nested group structure of 90 groups and 1900 members. As not all 1900 members fitted into a single buffer whatever the size, the buffer size should have been irrelevant. But the problem no longer occurred when using the maximum buffer size. The issue could not be reproduced here with a simple nested structure and more than 2000 members.
  • Corrected an issue with name expansion created several years back, for some forms of a relative eDirectory name when the object did not yet exist. This affected some programs such as grpadd when /c was used to create the group.
  • Updated all programs to accept input files containing unicode (big-endian or little-endian), or UTF-8 as indicated by a byte order marker (BOM). The unicode options may be useful when processing files created from a spreadsheet.
  • Updated all programs accepting input files, either as a command line parameter, or as an option value, to check for a comma and keyword appended to the file name indicating that the file contains single byte ANSI characters in OEM or CHAR format e.g. p:\temp\users.txt,char. Files containing unicode or utf-8 can be identified via a BOM at the start of the file but there is no such marker for single byte files, and no way to programmatically determine if the contents use the OEM or CHAR character set. Programs in JRButils assume OEM by default, but will now correctly process files containing CHAR providing “,char” is appended to the name. Note that distinguishing between OEM and CHAR is relevant only when the file contains extended characters such as .
  • Added the ability to create error and log files in UTF-8 format.
  • Fixed an issue in the Linux versions of various programs where the last character was lost from some lines of in-built help due to the narrower default console width compared to a Windows command prompt.