1 /*
2  * Copyright (C) 2002 Cyrus Patel <cyp@fb14.uni-mainz.de>
3  *           (C) 2007 Apple Inc. All rights reserved.
4  *
5  * This library is free software; you can redistribute it and/or
6  * modify it under the terms of the GNU Lesser General Public
7  * License 2.1 as published by the Free Software Foundation.
8  *
9  * This library is distributed in the hope that it will be useful,
10  * but WITHOUT ANY WARRANTY; without even the implied warranty of
11  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
12  * Library General Public License for more details.
13  *
14  * You should have received a copy of the GNU Library General Public License
15  * along with this library; see the file COPYING.LIB.  If not, write to
16  * the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
17  * Boston, MA 02110-1301, USA.
18  */
19 
20 /* ParseFTPList() parses lines from an FTP LIST command.
21 **
22 ** Written July 2002 by Cyrus Patel <cyp@fb14.uni-mainz.de>
23 ** with acknowledgements to squid, lynx, wget and ftpmirror.
24 **
25 ** Arguments:
26 **   'line':       line of FTP data connection output. The line is assumed
27 **                 to end at the first '\0' or '\n' or '\r\n'.
28 **   'state':      a structure used internally to track state between
29 **                 lines. Needs to be bzero()'d at LIST begin.
30 **   'result':     where ParseFTPList will store the results of the parse
31 **                 if 'line' is not a comment and is not junk.
32 **
33 ** Returns one of the following:
34 **    'd' - LIST line is a directory entry ('result' is valid)
35 **    'f' - LIST line is a file's entry ('result' is valid)
36 **    'l' - LIST line is a symlink's entry ('result' is valid)
37 **    '?' - LIST line is junk. (cwd, non-file/dir/link, etc)
38 **    '"' - its not a LIST line (its a "comment")
39 **
40 ** It may be advisable to let the end-user see "comments" (particularly when
41 ** the listing results in ONLY such lines) because such a listing may be:
42 ** - an unknown LIST format (NLST or "custom" format for example)
43 ** - an error msg (EPERM,ENOENT,ENFILE,EMFILE,ENOTDIR,ENOTBLK,EEXDEV etc).
44 ** - an empty directory and the 'comment' is a "total 0" line or similar.
45 **   (warning: a "total 0" can also mean the total size is unknown).
46 **
47 ** ParseFTPList() supports all known FTP LISTing formats:
48 ** - '/bin/ls -l' and all variants (including Hellsoft FTP for NetWare);
49 ** - EPLF (Easily Parsable List Format);
50 ** - Windows NT's default "DOS-dirstyle";
51 ** - OS/2 basic server format LIST format;
52 ** - VMS (MultiNet, UCX, and CMU) LIST format (including multi-line format);
53 ** - IBM VM/CMS, VM/ESA LIST format (two known variants);
54 ** - SuperTCP FTP Server for Win16 LIST format;
55 ** - NetManage Chameleon (NEWT) for Win16 LIST format;
56 ** - '/bin/dls' (two known variants, plus multi-line) LIST format;
57 ** If there are others, then I'd like to hear about them (send me a sample).
58 **
59 ** NLSTings are not supported explicitely because they cannot be machine
60 ** parsed consistantly: NLSTings do not have unique characteristics - even
61 ** the assumption that there won't be whitespace on the line does not hold
62 ** because some nlistings have more than one filename per line and/or
63 ** may have filenames that have spaces in them. Moreover, distinguishing
64 ** between an error message and an NLST line would require ParseList() to
65 ** recognize all the possible strerror() messages in the world.
66 */
67 
68 // This was originally Mozilla code, titled ParseFTPList.h
69 // Original version of this file can currently be found at: http://mxr.mozilla.org/mozilla1.8/source/netwerk/streamconv/converters/ParseFTPList.h
70 
71 #ifndef FTPDirectoryParser_h
72 #define FTPDirectoryParser_h
73 
74 #include "PlatformString.h"
75 
76 #include <time.h>
77 
78 #define SUPPORT_LSL  /* Support for /bin/ls -l and dozens of variations therof */
79 #define SUPPORT_DLS  /* Support for /bin/dls format (very, Very, VERY rare) */
80 #define SUPPORT_EPLF /* Support for Extraordinarily Pathetic List Format */
81 #define SUPPORT_DOS  /* Support for WinNT server in 'site dirstyle' dos */
82 #define SUPPORT_VMS  /* Support for VMS (all: MultiNet, UCX, CMU-IP) */
83 #define SUPPORT_CMS  /* Support for IBM VM/CMS,VM/ESA (z/VM and LISTING forms) */
84 #define SUPPORT_OS2  /* Support for IBM TCP/IP for OS/2 - FTP Server */
85 #define SUPPORT_W16  /* Support for win16 hosts: SuperTCP or NetManage Chameleon */
86 
87 namespace WebCore {
88 
89 typedef struct tm FTPTime;
90 
91 struct ListState {
ListStateListState92     ListState()
93         : now(0)
94         , listStyle(0)
95         , parsedOne(false)
96         , carryBufferLength(0)
97         , numLines(0)
98     {
99         memset(&nowFTPTime, 0, sizeof(FTPTime));
100     }
101 
102     double      now;               /* needed for year determination */
103     FTPTime     nowFTPTime;
104     char        listStyle;         /* LISTing style */
105     bool        parsedOne;         /* returned anything yet? */
106     char        carryBuffer[84];   /* for VMS multiline */
107     int         carryBufferLength; /* length of name in carry_buf */
108     int64_t     numLines;          /* number of lines seen */
109 };
110 
111 enum FTPEntryType {
112     FTPDirectoryEntry,
113     FTPFileEntry,
114     FTPLinkEntry,
115     FTPMiscEntry,
116     FTPJunkEntry
117 };
118 
119 struct ListResult
120 {
ListResultListResult121     ListResult()
122     {
123         clear();
124     }
125 
clearListResult126     void clear()
127     {
128         valid = false;
129         type = FTPJunkEntry;
130         filename = 0;
131         filenameLength = 0;
132         linkname = 0;
133         linknameLength = 0;
134         fileSize.truncate(0);
135         caseSensitive = false;
136         memset(&modifiedTime, 0, sizeof(FTPTime));
137     }
138 
139     bool valid;
140     FTPEntryType type;
141 
142     const char* filename;
143     uint32_t filenameLength;
144 
145     const char* linkname;
146     uint32_t linknameLength;
147 
148     String fileSize;
149     FTPTime modifiedTime;
150     bool caseSensitive; // file system is definitely case insensitive
151 };
152 
153 FTPEntryType parseOneFTPLine(const char* inputLine, ListState&, ListResult&);
154 
155 } // namespace WebCore
156 
157 #endif // FTPDirectoryParser_h
158