X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=index%2Ftrav.c;h=17f961d80a1c6837aa780b09b1f572303b32602a;hb=98d7e1e7378345d4d6b8b547ba25f2f0801b1a46;hp=4ac362934541e754af0b58da2628c3416b6e9b27;hpb=94134a7a5eff3cea5c2971836f7da89ac7ba7791;p=idzebra-moved-to-github.git diff --git a/index/trav.c b/index/trav.c index 4ac3629..17f961d 100644 --- a/index/trav.c +++ b/index/trav.c @@ -1,60 +1,40 @@ -/* - * Copyright (C) 1994-1995, Index Data I/S - * All rights reserved. - * Sebastian Hammer, Adam Dickmeiss - * - * $Log: trav.c,v $ - * Revision 1.12 1995-11-24 11:31:37 adam - * Commands add & del read filenames from stdin if source directory is - * empty. - * Match criteria supports 'constant' strings. - * - * Revision 1.11 1995/11/22 17:19:19 adam - * Record management uses the bfile system. - * - * Revision 1.10 1995/11/21 15:01:16 adam - * New general match criteria implemented. - * New feature: document groups. - * - * Revision 1.9 1995/11/21 09:20:32 adam - * Yet more work on record match. - * - * Revision 1.8 1995/11/20 16:59:46 adam - * New update method: the 'old' keys are saved for each records. - * - * Revision 1.7 1995/11/20 11:56:28 adam - * Work on new traversal. - * - * Revision 1.6 1995/11/17 15:54:42 adam - * Started work on virtual directory structure. - * - * Revision 1.5 1995/10/17 18:02:09 adam - * New feature: databases. Implemented as prefix to words in dictionary. - * - * Revision 1.4 1995/09/28 09:19:46 adam - * xfree/xmalloc used everywhere. - * Extract/retrieve method seems to work for text records. - * - * Revision 1.3 1995/09/06 16:11:18 adam - * Option: only one word key per file. - * - * Revision 1.2 1995/09/04 12:33:43 adam - * Various cleanup. YAZ util used instead. - * - * Revision 1.1 1995/09/01 14:06:36 adam - * Split of work into more files. - * - */ +/* $Id: trav.c,v 1.51 2006-05-10 08:13:22 adam Exp $ + Copyright (C) 1995-2005 + Index Data ApS + +This file is part of the Zebra server. + +Zebra is free software; you can redistribute it and/or modify it under +the terms of the GNU General Public License as published by the Free +Software Foundation; either version 2, or (at your option) any later +version. + +Zebra is distributed in the hope that it will be useful, but WITHOUT ANY +WARRANTY; without even the implied warranty of MERCHANTABILITY or +FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License +for more details. + +You should have received a copy of the GNU General Public License +along with Zebra; see the file LICENSE.zebra. If not, write to the +Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA +02111-1307, USA. +*/ + #include #include -#include -#include -#include #include +#ifdef WIN32 +#include +#define S_ISREG(x) (x & _S_IFREG) +#define S_ISDIR(x) (x & _S_IFDIR) +#endif +#if HAVE_UNISTD_H +#include +#endif +#include #include -#include +#include -#include #include "index.h" static int repComp (const char *a, const char *b, size_t len) @@ -64,30 +44,36 @@ static int repComp (const char *a, const char *b, size_t len) return memcmp (a, b, len); } -static void repositoryExtractR (int deleteFlag, char *rep, - struct recordGroup *rGroup) +static void repositoryExtractR (ZebraHandle zh, int deleteFlag, char *rep, + int level) { struct dir_entry *e; int i; size_t rep_len = strlen (rep); - e = dir_open (rep); + e = dir_open (rep, zh->path_reg, zh->m_follow_links); if (!e) return; + yaz_log (YLOG_LOG, "dir %s", rep); if (rep[rep_len-1] != '/') rep[rep_len] = '/'; else --rep_len; + for (i=0; e[i].name; i++) { + char *ecp; strcpy (rep +rep_len+1, e[i].name); + if ((ecp = strrchr (e[i].name, '/'))) + *ecp = '\0'; + switch (e[i].kind) { case dirs_file: - fileExtract (NULL, rep, rGroup, deleteFlag); + zebra_extract_file (zh, NULL, rep, deleteFlag); break; case dirs_dir: - repositoryExtractR (deleteFlag, rep, rGroup); + repositoryExtractR (zh, deleteFlag, rep, level+1); break; } } @@ -95,27 +81,51 @@ static void repositoryExtractR (int deleteFlag, char *rep, } -static void stdinExtractR (int deleteFlag, struct recordGroup *rGroup) +static void fileDeleteR (ZebraHandle zh, + struct dirs_info *di, struct dirs_entry *dst, + const char *base, char *src) { - char tmppath[256]; + char tmppath[1024]; + size_t src_len = strlen (src); - while (scanf ("%s", tmppath) == 1) - fileExtract (NULL, tmppath, rGroup, deleteFlag); + while (dst && !repComp (dst->path, src, src_len+1)) + { + switch (dst->kind) + { + case dirs_file: + sprintf (tmppath, "%s%s", base, dst->path); + zebra_extract_file (zh, &dst->sysno, tmppath, 1); + + strcpy (tmppath, dst->path); + dst = dirs_read (di); + dirs_del (di, tmppath); + break; + case dirs_dir: + strcpy (tmppath, dst->path); + dst = dirs_read (di); + dirs_rmdir (di, tmppath); + break; + default: + dst = dirs_read (di); + } + } } -static void repositoryUpdateR (struct dirs_info *di, struct dirs_entry *dst, - const char *base, char *src, - struct recordGroup *rGroup) +static void fileUpdateR (ZebraHandle zh, + struct dirs_info *di, struct dirs_entry *dst, + const char *base, char *src, + int level) { struct dir_entry *e_src; int i_src = 0; - static char tmppath[256]; + static char tmppath[1024]; size_t src_len = strlen (src); sprintf (tmppath, "%s%s", base, src); - e_src = dir_open (tmppath); + e_src = dir_open (tmppath, zh->path_reg, zh->m_follow_links); + yaz_log (YLOG_LOG, "dir %s", tmppath); -#if 1 +#if 0 if (!dst || repComp (dst->path, src, src_len)) #else if (!dst || strcmp (dst->path, src)) @@ -123,26 +133,29 @@ static void repositoryUpdateR (struct dirs_info *di, struct dirs_entry *dst, { if (!e_src) return; - if (src_len && src[src_len-1] == '/') - --src_len; - else + + if (src_len && src[src_len-1] != '/') + { src[src_len] = '/'; - src[src_len+1] = '\0'; + src[++src_len] = '\0'; + } dirs_mkdir (di, src, 0); - dst = NULL; + if (dst && repComp (dst->path, src, src_len)) + dst = NULL; } else if (!e_src) { - /* delete tree dst */ + strcpy (src, dst->path); + fileDeleteR (zh, di, dst, base, src); return; } else { - if (src_len && src[src_len-1] == '/') - --src_len; - else + if (src_len && src[src_len-1] != '/') + { src[src_len] = '/'; - src[src_len+1] = '\0'; + src[++src_len] = '\0'; + } dst = dirs_read (di); } dir_sort (e_src); @@ -151,13 +164,13 @@ static void repositoryUpdateR (struct dirs_info *di, struct dirs_entry *dst, { int sd; - if (dst && !repComp (dst->path, src, src_len+1)) + if (dst && !repComp (dst->path, src, src_len)) { if (e_src[i_src].name) { - logf (LOG_DEBUG, "dst=%s src=%s", dst->path + src_len+1, + yaz_log (YLOG_DEBUG, "dst=%s src=%s", dst->path + src_len, e_src[i_src].name); - sd = strcmp (dst->path + src_len+1, e_src[i_src].name); + sd = strcmp (dst->path + src_len, e_src[i_src].name); } else sd = -1; @@ -166,26 +179,31 @@ static void repositoryUpdateR (struct dirs_info *di, struct dirs_entry *dst, sd = 1; else break; - logf (LOG_DEBUG, "trav sd=%d", sd); + yaz_log (YLOG_DEBUG, "trav sd=%d", sd); + if (sd == 0) { - strcpy (src + src_len+1, e_src[i_src].name); + strcpy (src + src_len, e_src[i_src].name); sprintf (tmppath, "%s%s", base, src); switch (e_src[i_src].kind) { case dirs_file: - if (e_src[i_src].ctime > dst->ctime) + if (e_src[i_src].mtime > dst->mtime) { - if (fileExtract (&dst->sysno, tmppath, rGroup, 0)) - dirs_add (di, src, dst->sysno, e_src[i_src].ctime); + if (zebra_extract_file (zh, &dst->sysno, tmppath, 0) == ZEBRA_OK) + { + dirs_add (di, src, dst->sysno, e_src[i_src].mtime); + } + yaz_log (YLOG_DEBUG, "old: %s", ctime (&dst->mtime)); + yaz_log (YLOG_DEBUG, "new: %s", ctime (&e_src[i_src].mtime)); } dst = dirs_read (di); break; case dirs_dir: - repositoryUpdateR (di, dst, base, src, rGroup); + fileUpdateR (zh, di, dst, base, src, level+1); dst = dirs_last (di); - logf (LOG_DEBUG, "last is %s", dst ? dst->path : "null"); + yaz_log (YLOG_DEBUG, "last is %s", dst ? dst->path : "null"); break; default: dst = dirs_read (di); @@ -195,17 +213,17 @@ static void repositoryUpdateR (struct dirs_info *di, struct dirs_entry *dst, else if (sd > 0) { SYSNO sysno = 0; - strcpy (src + src_len+1, e_src[i_src].name); + strcpy (src + src_len, e_src[i_src].name); sprintf (tmppath, "%s%s", base, src); switch (e_src[i_src].kind) { case dirs_file: - if (fileExtract (&sysno, tmppath, rGroup, 0)) - dirs_add (di, src, sysno, e_src[i_src].ctime); + if (zebra_extract_file (zh, &sysno, tmppath, 0) == ZEBRA_OK) + dirs_add (di, src, sysno, e_src[i_src].mtime); break; case dirs_dir: - repositoryUpdateR (di, dst, base, src, rGroup); + fileUpdateR (zh, di, dst, base, src, level+1); if (dst) dst = dirs_last (di); break; @@ -214,50 +232,226 @@ static void repositoryUpdateR (struct dirs_info *di, struct dirs_entry *dst, } else /* sd < 0 */ { - assert (0); + strcpy (src, dst->path); + sprintf (tmppath, "%s%s", base, dst->path); + + switch (dst->kind) + { + case dirs_file: + zebra_extract_file (zh, &dst->sysno, tmppath, 1); + dirs_del (di, dst->path); + dst = dirs_read (di); + break; + case dirs_dir: + fileDeleteR (zh, di, dst, base, src); + dst = dirs_last (di); + } } } dir_free (&e_src); } -void repositoryUpdate (struct recordGroup *rGroup) +void repositoryShow (ZebraHandle zh, const char *path) { - struct dirs_info *di; - char src[256]; + char src[1024]; + int src_len; + struct dirs_entry *dst; Dict dict; + struct dirs_info *di; - dict = dict_open ("repdict", 40, 1); - - assert (rGroup->path); - di = dirs_open (dict, rGroup->path); - strcpy (src, ""); - repositoryUpdateR (di, dirs_read (di), rGroup->path, src, rGroup); + if (!(dict = dict_open_res (zh->reg->bfs, FMATCH_DICT, 50, 0, 0, zh->res))) + { + yaz_log (YLOG_FATAL, "dict_open fail of %s", FMATCH_DICT); + return; + } + + strncpy(src, path, sizeof(src)-1); + src[sizeof(src)-1]='\0'; + src_len = strlen (src); + + if (src_len && src[src_len-1] != '/') + { + src[src_len] = '/'; + src[++src_len] = '\0'; + } + + di = dirs_open (dict, src, zh->m_flag_rw); + + while ( (dst = dirs_read (di)) ) + yaz_log (YLOG_LOG, "%s", dst->path); dirs_free (&di); - dict_close (dict); } -void repositoryDelete (struct recordGroup *rGroup) +static void fileUpdate (ZebraHandle zh, Dict dict, const char *path) { - char src[256]; + struct dirs_info *di; + struct stat sbuf; + char src[1024]; + char dst[1024]; + int src_len, ret; + + assert (path); + + if (zh->path_reg && !yaz_is_abspath(path)) + { + strcpy (src, zh->path_reg); + strcat (src, "/"); + } + else + *src = '\0'; + strcat (src, path); + ret = zebra_file_stat (src, &sbuf, zh->m_follow_links); + + strcpy (src, path); + src_len = strlen (src); - assert (rGroup->path); - strcpy (src, rGroup->path); - if (*src == '\0') - stdinExtractR (1, rGroup); + if (ret == -1) + { + yaz_log (YLOG_WARN|YLOG_ERRNO, "Cannot access path %s", src); + } + else if (S_ISREG(sbuf.st_mode)) + { + struct dirs_entry *e_dst; + di = dirs_fopen (dict, src, zh->m_flag_rw); + + e_dst = dirs_read (di); + if (e_dst) + { + if (sbuf.st_mtime > e_dst->mtime) + if (zebra_extract_file (zh, &e_dst->sysno, src, 0) == ZEBRA_OK) + dirs_add (di, src, e_dst->sysno, sbuf.st_mtime); + } + else + { + SYSNO sysno = 0; + if (zebra_extract_file (zh, &sysno, src, 0) == ZEBRA_OK) + dirs_add (di, src, sysno, sbuf.st_mtime); + } + dirs_free (&di); + } + else if (S_ISDIR(sbuf.st_mode)) + { + if (src_len && src[src_len-1] != '/') + { + src[src_len] = '/'; + src[++src_len] = '\0'; + } + di = dirs_open (dict, src, zh->m_flag_rw); + *dst = '\0'; + fileUpdateR (zh, di, dirs_read (di), src, dst, 0); + dirs_free (&di); + } else - repositoryExtractR (1, src, rGroup); + { + yaz_log (YLOG_WARN, "Skipping path %s", src); + } } -void repositoryAdd (struct recordGroup *rGroup) +static void repositoryExtract (ZebraHandle zh, + int deleteFlag, const char *path) { - char src[256]; + struct stat sbuf; + char src[1024]; + int ret; + + assert (path); + + if (zh->path_reg && !yaz_is_abspath(path)) + { + strcpy (src, zh->path_reg); + strcat (src, "/"); + } + else + *src = '\0'; + strcat (src, path); + ret = zebra_file_stat (src, &sbuf, zh->m_follow_links); + + strcpy (src, path); + + if (ret == -1) + yaz_log (YLOG_WARN|YLOG_ERRNO, "Cannot access path %s", src); + else if (S_ISREG(sbuf.st_mode)) + zebra_extract_file (zh, NULL, src, deleteFlag); + else if (S_ISDIR(sbuf.st_mode)) + repositoryExtractR (zh, deleteFlag, src, 0); + else + yaz_log (YLOG_WARN, "Skipping path %s", src); +} - assert (rGroup->path); - strcpy (src, rGroup->path); - if (*src == '\0') - stdinExtractR (0, rGroup); +static void repositoryExtractG (ZebraHandle zh, const char *path, + int deleteFlag) +{ + if (!strcmp(path, "") || !strcmp(path, "-")) + { + char src[1024]; + + while (scanf ("%1020s", src) == 1) + repositoryExtract (zh, deleteFlag, src); + } else - repositoryExtractR (0, src, rGroup); + repositoryExtract (zh, deleteFlag, path); +} + +#if 0 +static int dump_file_dict_func(char *name, const char *info, int pos, + void *client) +{ + yaz_log(YLOG_LOG, "%s", name); + return 0; +} +static void dump_file_dict(Dict dict) +{ + int before = 10; + int after = 1000; + char term[1000]; + + strcpy(term, "0"); + dict_scan (dict, term, &before, &after, 0, dump_file_dict_func); +} +#endif + +void repositoryUpdate (ZebraHandle zh, const char *path) +{ + assert (path); + if (zh->m_record_id && !strcmp (zh->m_record_id, "file")) + { + Dict dict; + if (!(dict = dict_open_res (zh->reg->bfs, FMATCH_DICT, 50, + zh->m_flag_rw, 0, zh->res))) + { + yaz_log (YLOG_FATAL, "dict_open fail of %s", FMATCH_DICT); + return ; + } + if (!strcmp(path, "") || !strcmp(path, "-")) + { + char src[1024]; + while (scanf ("%s", src) == 1) + fileUpdate (zh, dict, src); + } + else + fileUpdate (zh, dict, path); +#if 0 + dump_file_dict(dict); +#endif + dict_close (dict); + + } + else + repositoryExtractG (zh, path, 0); } +void repositoryDelete (ZebraHandle zh, const char *path) +{ + assert (path); + repositoryExtractG (zh, path, 1); +} + +/* + * Local variables: + * c-basic-offset: 4 + * indent-tabs-mode: nil + * End: + * vim: shiftwidth=4 tabstop=8 expandtab + */ +