lang/seed7: update to 05_20190407

20190407:

• The program db7.sd7 (Database Inspector) has been improved to use a scrollable list of tables instead of radio buttons for the selection of a table.
• The example program findchar.sd7 has been added. Findchar is an utility program to search for specific characters in files.
• The new library iobuffer.s7i has been added. This library supports buffering of files.
• The new library mixarith.s7i has been added. This library supports mixing of arithmethic operations between numeric types.
• The file type fileInTar has been added to the library tar.s7i. This file type allows that a specific file from a tar archive can be opened. The functions open, getc, gets, eof, hasNext, length, seek, tell, and openFileInTar have been added to support the file type fileInTar.
• The functions skipClassicComment and getSimpleStringLiteral have been added to the libraries scanfile.s7i and scanstri.s7i.
• The makefile mk_emccl.mak has been added. This makefile supports compiling with the emcc C compiler from Emscripten under Linux. The support of Emscripten is experimental (see also src/read_me.txt).
• The deprecated functions float2Bits and bits2Float have been removed from bin32.s7i and bin64.s7i. The functions bin32(), bin64() and float() can be used instead, to convert between float and bits in IEEE 754 representation.
• In the programs chkbig.sd7, chkovf.sd7, chkset.sd7 and chkstr.sd7 large functions have been split into smaller functions. This avoids Stack errors with Emscripten.
• Tests for allowed field names and CLOB fields have been added to chkdb.sd7. The functions create, testFieldNames and testClobField have been added to chkdb.sd7.
• The primitive actions BIG_PARSE, BST_PARSE, FLT_PARSE and INT_PARSE have been renamed to BIG_PARSE1, BST_PARSE1, FLT_PARSE1 and INT_PARSE1 respectively. This actions parse the first argument instead of the third argument. The libraries bigint.s7i, bstring.s7i, float.s7i and integer.s7i have been improved to use the new actions.
• In several libraries the 'parse' operator has been changed to call a function with the name of the type.
• In fil.unx.c the function setupFiles has been improved to correctly mount the file system root for emscripten and node.js.
• In sql_fire.c, sql_my.c the support for BLOB and CLOB fields has been improved.
• In sql_oci.c, sql_odbc.c and sql_post.c the function processStatementStri() has been improved to accept strings enclosed in single quotes (') and double quotes (").
• The files db_fire.h and sql_fire.c have been changed to define the database client functions with the calling convention __stdcall.
• The files db_post.h, db_lite.h, sql_post and sql_lite.c have been changed to define the database client functions with the calling convention __cdecl.
• The database driver sql_fire.c has been improved:
  • The function getClob() has been added. This function is used by sqlColumnStri() to get a CLOB.
  • The function sqlBindStri() has been improved to work for CLOBs and to raise RANGE_ERROR for BLOBs.
  • The function sqlColumnStri() has been improved to work for BLOBs and CLOBs.
  • The function sqlColumnBStri() has been improved to work for BLOBs and to raise RANGE_ERROR for CLOBs.
  • The function sqlStmtColumnName() has been improved to assume that the name of a column is encoded with UTF-8.
• The database driver sql_my.c has been improved:
  • The function processStatementStri() has been improved. In strings delimited with single quotes (') backslashes (\) are optionaly replaced with double backslashes (\\). Strings delimited with double quotes (") are converted into strings, that are delimited with a backtick (`).
  • The function setupResultColumn() has been improved to distinguish a BLOB from a CLOB.
  • The function sqlColumnStri() has been improved to work for BLOBs and CLOBs.
  • The function sqlColumnBStri() has been improved to work for BLOBs and to raise RANGE_ERROR for CLOBs.
  • The function determineIfBackslashEscapes has been introduced. This function determines, if a backslash is used as escape character in string literals. In this case processStatementStri() will double escaping backslashes in string literals.
• The database driver sql_odbc.c has been improved:
  • The management of buffer memory in bind functions has been improved for the data types SQL_DECIMAL, SQL_NUMERIC, SQL_CHAR, SQL_VARCHAR, SQL_LONGVARCHAR, SQL_WCHAR, SQL_WVARCHAR, SQL_WLONGVARCHAR, SQL_BINARY, SQL_VARBINARY and SQL_LONGVARBINARY.
• The database driver sql_post.c has been improved:
  • The function setupDll has been improved to open the dll libeay32.dll, when necessary.
  • In the function sqlBindStri() the binding of a field with the type BYTEAOID (used for BLOB data) has been improved, to work correct, when the string contains a zero byte ('\0;').
  • The function sqlColumnBigRat() has been improved to process fields with the type FLOAT8OID as double instead of float.
• The program chkccomp.c has been improved. Now the C compiler used to compile chkccomp.c can be different from the C compiler checked by chkccomp.c. E.g.: For Emscripten chkccomp.c is compiled with gcc, but the checked C compiler is emcc.
• In chkccomp.c the function determineFseekFunctions has been improved to check fseek() and ftell() without access to the local files.
• In chkccomp.c the functions to configure access to PostgreSql, Oracle and Firebird databases has been improved.
• In chkccomp.c code to determine NULL_DEVICE, HAS_CDECL, HAS_STDCALL, INTPTR_T_DEFINED, PRINTF_ROUNDING, LINE_DIRECTIVE_ACCEPTS_UTF8, HAS_SETJMP and HAS_SIGSETJMP has been added or improved.
• The file sigutl.c has been improved to work correct, when neither signal() nor sigaction() are present.
• Log functions have been added or improved in con_emc.c, con_inf.c, sql_my.c, sql_post.c and timlib.c.
• Exprerimental code has been added to con_emc.c.