Revision history for Perl extension Data::Table. 1.54 Sun Feb 10 21:35:02 PST 2008 Modify fromFileGetTopLines method, remove dependency on bytes bytes::substr causes infinite loop in some older version of perl 1.53 Thu Jan 3 21:13:40 PST 2008 add "use bytes" to Table.pm Just patched test.pl, because some OS cannot open in-memory file. 1.52 Fri Dec 14 11:48:42 PST 2007 1.51 Wed Dec 12 15:36:22 PST 2007 1. Add a class methods Data::Table::fromFile(file_name), which can guess the file format and call fromCSV/fromTSV internally. fromFile relies on the following new methods fromFileGuessOS(file_name) fromFileGetTopLines($file_name, $OS, $lineNumber) fromFileIsHeader($string) fromFileGuessDelimiter($arrayRefToLines) to figure out if the input file is from UNIX/PC/MAC, whether its first row contains column headers, and whether it uses ",", "\t" or ":" as field delimiters. It then calls either fromCSV or fromTSV to return the table object. $t = Data::Table::fromFile("myFileName_CSVorTSV_HeaderOrNoHeader_UNIXorPCorMAC"); Please refers to the updated document for details. 2. When fromFile/fromCSV/fromTSV reads from an empty file, it returns an undef object, rather than quit. 3. Provide more informative error message, when invalid column header is found. 4. fixed a bug in 1.51 where fromFileGuessOS failed in Windows Thanks to patches provided by "whitebell". 1.50 Thu Sep 28 07:21:38 PDT 2006 Small modifications to sort subroutine example in the document, no bug. join method, if $cols2 is undefined, defaults @$cols2 to @$cols1 Update fromCSV, fromTSV, csv methods to be able to deal with certain delimiters correctly. (When the delimiter is a special symbol for regexp, it should be escaped, e.g., set delimiter to '\|' for pipe symbol). Thanks to suggestions from Michael Slaven. update fromCSV, fromTSV to take additional arguments: skip_lines and skip_pattern skip_lines lets user skip several lines in the beginning of the input file skip_pattern lets user skip all lines that match a regular expression Please read documents under fromCSV and fromTSV for details. Thanks to suggestions from Wenbin Ye. 1.49 Wed Aug 30 09:29:51 PDT 2006 Add %Data::Table::DEFAULTS to store the default settings for OS, CSV_DELIMITER and CSV_QUALIFIER Thanks to suggestions from Roman Filippov Patch sort method to deal with undef table element. undef value is considered to be larger than any other value, two undef values are considered equal during sorting. 1.48 Thu Jun 8 13:25:54 PDT 2006 Update fromCSV, parseCSV to enable user-specified delimiter and qualifier, see document and examples under fromCSV. csvEscape is modified accordingly. Thanks to help from Roman Filippov 1.47 Sun May 21 15:03:14 PDT 2006 Upload the wrong code in 1.46, re-upload 1.46 Sat May 13 05:44:09 PDT 2006 fromCSV, fromTSV, csv, tsv can all take either a file hander or a file name Notice: to leave rooms for future development, file handler is not closed by Data::Table. It's caller's responsibility to close it afterwards, if no longer used. table::sort code is replaced, the old sort method is renamed to sort_v0 and is deprecated. The new sort method allow user-defined sorting operators, please read manual on table::sort The new sort method also runs faster in some benchmark tests. A big thank to Wenbin Ye for suggestions, as well as contributing both the new sort code and test examples 1.45 Mon May 1 09:08:20 PDT 2006 Fix a bug in fromTSV last column name is truncated by one character (introduced in 1.44) Thanks to Albert V. Smith 1.44 Sat Apr 15 04:27:28 PDT 2006 Fix a bug in join (type=2 and 3) When right or full join, key fields are undef for right-only entries. modify fromCSV, fromTSV, tsv, csv subroutines to support read/write PC, Mac and UNIX files. csv and tsv can take a file name and directly writes to it. 1.43 Tue Nov 9 10:23:44 PST 2004 Patch html so that valid XHTML code is generated Several mispelled words were corrected Thanks all to Wolfgang Dautermann 1.42 Fri Oct 8 11:56:41 PDT 2004 Minor changes to group and pivot, not a bug 1.41 Thu Oct 7 14:04:17 CDT 2004 Add two useful methods: group and pivot group can make the records unique based on given key columns pivot is handy to transfer database table into a more user readable format group+pivot make accounting operations easy, please read the document for details. Due to the spam, please use the following for email contact easydatabase at gmail dot com 1.40 Wed Oct 15 12:11:11 CDT 2003 Patch colMap, as suggested by Jeff Janes 1.39 Wed Mar 26 09:33:01 CST 2003 Fix a bug in match_pattern, match_string and row_mask; When a new table was created by these methods, deleting columns had a side effect of the original parent table. 1.38 Sun Jan 19 18:26:23 CST 2003 Change die to croak as suggested by Jeff Janes. 1.37 Tue Sep 17 17:53:55 CDT 2002 Add $countOnly to match_string and match_pattern, thanks to Serge Batalov. 1.36 Thu Sep 12 14:47:59 CDT 2002 Add close() to both fromTSV and fromCSV, thanks to Brian Coon. 1.35 Mon Jul 1 13:04:43 PDT 2002 Optimization in parseCSV, thanks to Jeff Janes. 1.34 Wed May 1 12:13:33 CDT 2002 Fix a bug in colMerge 1.33 Wed Jan 16 17:55:34 CST 2002 Small patches to join method. Not a bug. Thanks to Xiao-Jun Ma 1.32 Sun Sep 30 16:21:02 CDT 2001 No change, just update Table.html (forgot in version 1.31) 1.31 Wed Sep 20 21:22:22 PDT 2001 add colsMap($fun), which does more than colMap can; Unlike colMap, $fun here have access to multiple columns. Read document for details. 1.30 Wed Sep 19 20:02:52 PDT 2001 Improve header method, which can now take a new header argument. Improve fromTSV and fromCSV, which now can take the 3rd argument -- if header is supplied, it will always be used (despite the 2nd argument). Read document for details. Fix a bug in adding a new column to a empty table. Thanks to Serge Batalov. 1.29 Mon Sep 17 18:18:13 CDT 2001 a bug fixed in fromTSV The first line was skipped when header==0 is specified in fromTSV. 1.28 Wed Sep 5 17:59:36 CDT 2001 a bug fixed in fromCSV, where \c or \\ apears in the file. Fix provided by Jeff Janes. 1.27 Mon Jul 9 00:04:54 PDT 2001 accept more formatting parameters for html, combine html2 with html. 1.26 Mon May 21 09:35:41 PDT 2001 A typo bug in swap fixed Thanks to Jeff Janes 1.25 Thu May 3 00:50:03 PDT 2001 add BEGIN, check perl version, update README. We realize Table.pm requires 5.005 at least. See README for the patch for older perl. Thanks to Jeffery Cann 1.24 Sat Apr 21 20:30:18 PDT 2001 a bug in match_pattern fixed (important!) Thanks to Robson Francisco de Souza 1.23 Thu Apr 12 14:36:45 PDT 2001 a bug in html, html2 fixed, where table element "" displayed ugly introduced in version 1.21 1.22 Sat Mar 25 14:30:06 PST 2001 join method added support four join types: inner, left outer, right outer, and full outer. a small bug in html2 fixed, thanks to Fred Lovine 1.21 Fri Mar 9 21:04:30 PST 2001 rowMask method added A bug in html, html2 fixed, where table element 0 is not displayed Thanks to Sven Neuhaus 1.20 Wed Feb 28 12:38:53 PST 2001 A bug in match_string is fixed. This will affect results if you change your "string" value in the program; Also add a caseIgnore control argument to match_string method. Thanks to Bryan Coon. 1.19 Sat Feb 24 23:23:37 PST 2001 A bug in fromSQL is fixed (caused by typo) This happens when user use the third argument (a reference to an array). The old package will show an error message in some cases. Add $header option for csv and tsv, output header or not Add the following instant methods fromSQLi fromCSVi fromTSVi so that these methods can be inherited. Thanks to Michael Schlueter Update new method, so that it can be used as an instant method as well. Add method rowHashRef which returns a copy of a table row in hash reference. Officially support TSV format via the following three methods Data::Table::fromTSV fromTSVi tsv Read "TSV FORMAT" section for details. 1.18 Fix the problem in Data::Table::fromCSV caused by null trailing fields. E.g., a line "a,b,," in a csv file was split into two fields before. Thanks to Karsten Fix the warning message in Data::Table::match_string, when table contains an undef element. Add Data::Table::fromTSV and Data::Table:tsv TSV - tab-deliminated file format. TSV preserves NULL element and line-break chars in a table. \0, \\, \r, \b, \n, \t are slash-escaped. undef is escaped into \N. This is based on MySQL specification. 1.16 Fri Sep 29 22:18:06 PDT 2000 Package name changed from Table to Data::Table, due to name collision with PerlQt first official release version 1.15 Tue Sep 26 18:32:52 PDT 2000 submitted to CPAN