gnats/gnats ChangeLog Makefile.in adm.c adm.h ...

[bmanson at sourceware dot cygnus dot com]

CVSROOT:	/cvs/gnats
Module name:	gnats
Changes by:	bmanson@sourceware.cygnus.com	00/01/13 23:32:47

Modified files:
	gnats          : ChangeLog Makefile.in adm.c adm.h check-db.sh 
	                 client.c cmds.c config.h configure configure.in 
	                 database.c database.h delete-pr.sh edit-pr.sh 
	                 edit.c fconfig.y fconfigl.l field-config.in 
	                 field.h file-pr.c files.c files.h 
	                 gen-closed-date.c getclose.c gnats.h gnatsd.c 
	                 gnatsd.h index.c index.h internal.c mail.c 
	                 misc.c pr-edit.c pr-init.c pr.c pr.h query-pr.c 
	                 query.c query.h responsible 
Added files:
	gnats          : diff-prs mail.h 

Log message:
	* pr.c (getFieldFlags): New function.
	(read_header): Fix problem with overwriting array boundaries,
	and make sure the header value has a terminating NUL.
	
	* gnatsd.c (freeArgs): New function.
	(main): Restructure a bit; make sure we're freeing up the arguments
	we parse.
	
	* field.h (struct field_def): Add requireChangeReason flag.
	(The flags should probably be collapsed into a single bitfield.)
	
	* misc.c (init_gnats): Call clearCurrDatabase ().
	
	* database.c: Encapsulate the database state into a struct
	(tho there are still static variables scattered throughout
	the code).
	(clearCurrDatabase): New function.
	
	* configure.in: Did some remaining cleanup; deleted several
	unused options.
	
	* cmds.c (GNATS_fieldflags): New function.
	
	* client.c (netFieldFlags): New function.
	
	* query-pr.c (long_options): Add --field-flags option.
	
	* edit-pr.sh: Use diff-prs; ask about fields that require a
	reason for change, if they've changed.
	
	* diff-prs: New script to list the fields that have changed
	between two PRs.
	
	* delete-pr.sh: Don't send mail about deleted PRs--the server does
	it now.
	
	* Makefile.in, configure.in: Deleted GNATS_ADMIN; it's now wired
	to gnats-admin. (To change gnats-admin's mail address, change it
	in the responsible file.)
	
	* config.h: Delete INDEX define, no longer used.
	
	* index.c: Use the "path" entry in the index descriptor.
	(build_index_entry): Fix index allocation code.
	(free_pr_index): Free the individual field entries if we didn't
	allocate them with a buffer.
	
	* query.c: Change the funny '[' and '%' designators to real words
	(builtinfield: and fieldtype: respectively).  Use FieldList
	instead of FieldIndex for referring to which fields to search;
	this makes it possible to do even more interesting queries.
	Remove the FieldType search ('#' operator), it's no longer needed.
	
	* files.c: Delete a bunch of unused functions.
	
	* query.c (append_string, append_char, do_print): New function.
	(format_pr_field, process_printf_format, process_format): Add the
	ability to produce output into a char* instead of a FILE pointer.
	(allocateNamedParameter, getNamedParameterValue,
	freeFormatParameterList):  New functions, for supporting parameters
	to format statements.
	
	* pr.c (get_field_value): Add PARAMS argument, for format
	parameters.
	
	* edit.c: Add parameters for the email address of the user performing
	the various edit actions.
	(addAuditTrailEnt, sendAuditMail): New functions.
	
	* edit-pr.sh: Simplify--the server now takes care of Audit-Trail:
	entries and sending email for edits.
	
	* cmds.c: Add editEmailAddr static variable (this belongs
	somewhere else...)
	(GNATS_user): Use get_responsible_addr () to set a default edit
	email address for the user.
	(GNATS_editaddr): New function.
	
	* client.c (netSetEditEmailAddr): New function.
	(netEditField): Set the email address of the editing user before
	doing the edit.
	(netModifyPR): Ditto.
	
	* database.c (setAuditTrailFormat, getAuditTrailFormat): New
	functions.
	
	Gack.  This is way too much to check in at once.
	
	Adds formatted mail and audit-trail support.  Fixes several memory
	leaks as well.  (The three checkers I've been using claim there aren't
	any more leaks for the limited testing I've done...I'm sure there are
	more waiting.)