/* mkfinfo.c -- Copyright 1995, 1996 Liam R. Quin.
 * All Rights Reserved.
 * This code is NOT in the public domain.
 * See the file COPYRIGHT for full details.
 */

/*
 * mkfinfo -- make a new FileInfo struct given the file name
 *
 * $Id: mkfinfo.c,v 1.7 2001/05/31 03:50:13 liam Exp $
 *
 */

#include "globals.h" /* defines and declarations for database filenames */
#include "error.h"

#include <stdio.h>
#include <unistd.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <errno.h>

#ifndef FILE
# include <stdio.h>
#endif

#ifdef HAVE_FCNTL_H
# include <fcntl.h>
#endif

#ifdef HAVE_STRING_H
# include <string.h>
#else
# include <strings.h>
#endif

#ifdef HAVE_STDLIB_H
# include <stdlib.h>
#else
# include <malloc.h>
#endif

#include "fileinfo.h"
#include "lqutil.h"
#include "liblqtext.h"
#include "emalloc.h"
#include "lqtrace.h"
#include "filter.h"

/** Unix system calls that need to be declared: **/
/** library functions that need to be declared: */
/** **/

/* <Function>
 *   <Name>LQT_MakeFileInfo
 *   <Class>Database/Update, Database/Documents
 *   <Purpose>
 *      <P>Creates a t_FileInfo structure to describe the given FileName.
 *	This routine should only be used if you are going to add the
 *	given FileName to the given lq-text Database db; to get a
 *	FileInfo describing a file already in the index,
 *	use LQT_NameToFID and LQT_FIDToFileInfo.
 *   <Returns>
 *	<P>If the file is not already in the database, a new FID is allocated,
 *	and a newly malloc'd t_FileInfo object is returned, complete with
 *	a stdio FILE pointer already opened, either as a file or as a pipe,
 *	depending on the file type and filter table;
 *	it is the caller's responsibility to call LQT_DestroyFileInfo to
 *	free the memory and close the stdio stream.</P>
 *	<P>On error, or if the file is already in the database and has not
 *	changed since it was last indexed, a warning is issued
 *	and a NULL pointer is returned.</P>
 *   <Errors>
 *      If the file can't be found, or can't be opened, a warning is
 *	produced.
 *   <SeeAlso>
 *	LQT_DestroyFileInfo
 *	LQT_NameToFID
 *	LQT_FIDToFileInfo
 *	LQT_GetFilterType
 *	LQT_MakeInput
 * </Function>
 */
API t_FileInfo *
LQT_MakeFileInfo(db, FileName)
    t_LQTEXT_Database *db;
    char *FileName;
{
#ifdef BSD
    extern time_t time();
#else
    extern long time();
#endif
    struct stat StatBuf;

    t_FileInfo *FileInfo = 0;
    t_FID FID;

    if (!FileName) {
	Error(E_INTERNAL, "LQT_MakeFileInfo called with NULL FileName");
	return (t_FileInfo *) 0;
    } else if (!*FileName) {
	Error(E_INTERNAL, "LQT_MakeFileInfo called with empty FileName");
	return (t_FileInfo *) 0;
    }

    if (stat(FileName, &StatBuf) < 0) {
	char *doc;

	if ((doc = LQT_FindFile(db, FileName)) == (char *) 0) {
	    Error(E_WARN, "Can't find document file \"%s\"", FileName);
	    return (t_FileInfo *) 0;
	}

	if (stat(doc, &StatBuf) < 0) {
	    Error(E_WARN, "%s: can't stat document file \"%s\"",
		FileName, doc
	    );
	    return (t_FileInfo *) 0;
	}
	FileName = doc;
    }

    if ((StatBuf.st_mode & S_IFMT) == S_IFDIR) {
	LQT_Trace(LQTRACE_VERBOSE|LQTRACE_DEBUG,
	    "%s is a directory -- not indexed",
	    FileName
	);
	return (t_FileInfo *) 0;
    }

    if (StatBuf.st_size == 0L) {
	LQT_Trace(LQTRACE_VERBOSE|LQTRACE_DEBUG,
	    "%s empty -- not indexed",
	    FileName
	);
	return (t_FileInfo *) 0;
    }

    /* See if it's in the index already: */
    if ((FID = LQT_NameToFID(db, FileName)) != (t_FID) 0) {

	if ((FileInfo = LQT_FIDToFileInfo(db, FID)) != (t_FileInfo *) 0) {
	    /* Check to see if the file has changed since it was last
	     * indexed.  If it has, we should delete the old one from
	     * the database and give this one a new FID, but I have
	     * not done that yet -- that's /usr/local/lib/lqtextd or
	     * something, I suppose!
	     */
	    if (FileInfo->Date >= StatBuf.st_mtime &&
				FileInfo->FileSize == StatBuf.st_size) {
		LQT_Trace(LQTRACE_VERBOSE|LQTRACE_DEBUG,
		    "%s unchanged since last run -- not indexed",
		    FileName
		);
	    } else if (LQT_TraceFlagsSet(LQTRACE_VERBOSE|LQTRACE_DEBUG)) {
		Error(E_WARN,
		    "%s changed since last run -- use lqupdate instead",
		    FileName
		);
	    }
	    LQT_DestroyFileInfo(db, FileInfo);
	    return (t_FileInfo *) 0;
	}
    } else {
	FID = LQT_GetMaxOrAllocateFID(db, (long) StatBuf.st_size);
    }

    if (FileInfo == (t_FileInfo *) 0) {
	/* Allocate Structure */
	FileInfo = (t_FileInfo *) emalloc("MakeFileInfo", sizeof(t_FileInfo));

	/* Although not always necessary, call emalloc here so that a
	 * FileInfo can always be deleted with LQT_DestroyFileInfo()
	 */
	FileInfo->Name = emalloc(
	    "MakeFileInfo.Name",
	    (unsigned)(strlen(FileName) + 1)
	);
	(void) strcpy(FileInfo->Name, FileName);

	/* Other bits to set: */

	FileInfo->Date = StatBuf.st_mtime;
	FileInfo->FileSize = StatBuf.st_size;
	FileInfo->Stream = 0;

	/* file type */
	if ((FileInfo->FilterType = LQT_GetFilterType(db, FileInfo, &StatBuf)) < 0) {
	    LQT_Trace(LQTRACE_VERBOSE|LQTRACE_DEBUG,
		"%s unknown file type -- not indexed",
		FileName
	    );
	    LQT_DestroyFileInfo(db, FileInfo);
	    return (t_FileInfo *) 0;
	}
    }

    FileInfo->FID = FID;
    FileInfo->Date = (long) time((long *) 0); /* it's a time_t on BSD */

    FileInfo->Stream = LQT_MakeInput(db, FileInfo);
    if (!FileInfo->Stream) {
	Error(E_WARN|E_SYS,
	    "couldn't open input filter for %s -- not indexed",
	    FileInfo->Name
	);
	LQT_DestroyFileInfo(db, FileInfo);
	return (t_FileInfo *) 0;
    }

    return FileInfo;
}