From 415bf720ed312bc07e5334c4fb9f7fa5c2b1522d Mon Sep 17 00:00:00 2001 From: "dougt%netscape.com" Date: Thu, 29 Aug 2002 03:13:18 +0000 Subject: [PATCH] Fixes 95590 FTP: SYST limitations. Intergrates Cyrus Patel LIST Parser. r=dougt, sr=darin@Netscape.com. git-svn-id: svn://10.0.0.236/trunk@128429 18797224-902f-48f8-a5cc-f745e15eee43 --- mozilla/netwerk/build/nsNetModule.cpp | 22 +- mozilla/netwerk/macbuild/netwerk.xml | 54 + .../ftp/src/nsFtpConnectionThread.cpp | 35 +- .../protocol/ftp/src/nsFtpConnectionThread.h | 1 - .../netwerk/streamconv/converters/Makefile.in | 1 + .../streamconv/converters/ParseFTPList.cpp | 1802 +++++++++++++++++ .../streamconv/converters/ParseFTPList.h | 124 ++ .../converters/nsFTPDirListingConv.cpp | 801 +------- .../converters/nsFTPDirListingConv.h | 94 - 9 files changed, 2035 insertions(+), 899 deletions(-) create mode 100644 mozilla/netwerk/streamconv/converters/ParseFTPList.cpp create mode 100644 mozilla/netwerk/streamconv/converters/ParseFTPList.h diff --git a/mozilla/netwerk/build/nsNetModule.cpp b/mozilla/netwerk/build/nsNetModule.cpp index 327352f7fd9..0c654b45d27 100644 --- a/mozilla/netwerk/build/nsNetModule.cpp +++ b/mozilla/netwerk/build/nsNetModule.cpp @@ -251,9 +251,7 @@ nsresult NS_NewHTTPCompressConv (nsHTTPCompressConv ** result); nsresult NS_NewNSTXTToHTMLConv(nsTXTToHTMLConv** result); nsresult NS_NewStreamConv(nsStreamConverterService **aStreamConv); -#define FTP_UNIX_TO_INDEX "?from=text/ftp-dir-unix&to=application/http-index-format" -#define FTP_NT_TO_INDEX "?from=text/ftp-dir-nt&to=application/http-index-format" -#define FTP_OS2_TO_INDEX "?from=text/ftp-dir-os2&to=application/http-index-format" +#define FTP_TO_INDEX "?from=text/ftp-dir&to=application/http-index-format" #define GOPHER_TO_INDEX "?from=text/gopher-dir&to=application/http-index-format" #define INDEX_TO_HTML "?from=application/http-index-format&to=text/html" #define MULTI_MIXED_X "?from=multipart/x-mixed-replace&to=*/*" @@ -281,9 +279,7 @@ static PRUint32 g_StreamConverterCount = 16; #endif static const char *const g_StreamConverterArray[] = { - FTP_UNIX_TO_INDEX, - FTP_NT_TO_INDEX, - FTP_OS2_TO_INDEX, + FTP_TO_INDEX, GOPHER_TO_INDEX, INDEX_TO_HTML, MULTI_MIXED_X, @@ -772,19 +768,7 @@ static const nsModuleComponentInfo gNetModuleInfo[] = { // from netwerk/streamconv/converters: { "FTPDirListingConverter", NS_FTPDIRLISTINGCONVERTER_CID, - NS_ISTREAMCONVERTER_KEY FTP_UNIX_TO_INDEX, - CreateNewFTPDirListingConv - }, - - { "FTPDirListingConverter", - NS_FTPDIRLISTINGCONVERTER_CID, - NS_ISTREAMCONVERTER_KEY FTP_NT_TO_INDEX, - CreateNewFTPDirListingConv - }, - - { "FTPDirListingConverter", - NS_FTPDIRLISTINGCONVERTER_CID, - NS_ISTREAMCONVERTER_KEY FTP_OS2_TO_INDEX, + NS_ISTREAMCONVERTER_KEY FTP_TO_INDEX, CreateNewFTPDirListingConv }, diff --git a/mozilla/netwerk/macbuild/netwerk.xml b/mozilla/netwerk/macbuild/netwerk.xml index a2d0b37e738..bcd2299d65a 100644 --- a/mozilla/netwerk/macbuild/netwerk.xml +++ b/mozilla/netwerk/macbuild/netwerk.xml @@ -1212,6 +1212,13 @@ Text Debug + + Name + ParseFTPList.cpp + MacOS + Text + Debug + Name mozTXTToHTMLConv.cpp @@ -1788,6 +1795,11 @@ nsFTPDirListingConv.cpp MacOS + + Name + ParseFTPList.cpp + MacOS + Name mozTXTToHTMLConv.cpp @@ -3234,6 +3246,13 @@ Text Debug + + Name + ParseFTPList.cpp + MacOS + Text + Debug + Name mozTXTToHTMLConv.cpp @@ -3810,6 +3829,11 @@ nsFTPDirListingConv.cpp MacOS + + Name + ParseFTPList.cpp + MacOS + Name mozTXTToHTMLConv.cpp @@ -5256,6 +5280,13 @@ Text Debug + + Name + ParseFTPList.cpp + MacOS + Text + Debug + Name mozTXTToHTMLConv.cpp @@ -5818,6 +5849,11 @@ nsFTPDirListingConv.cpp MacOS + + Name + ParseFTPList.cpp + MacOS + Name mozTXTToHTMLConv.cpp @@ -7254,6 +7290,13 @@ Text Debug + + Name + ParseFTPList.cpp + MacOS + Text + Debug + Name mozTXTToHTMLConv.cpp @@ -7816,6 +7859,11 @@ nsFTPDirListingConv.cpp MacOS + + Name + ParseFTPList.cpp + MacOS + Name mozTXTToHTMLConv.cpp @@ -8568,6 +8616,12 @@ nsFTPDirListingConv.cpp MacOS + + Necko.shlb + Name + ParseFTPList.cpp + MacOS + Necko.shlb Name diff --git a/mozilla/netwerk/protocol/ftp/src/nsFtpConnectionThread.cpp b/mozilla/netwerk/protocol/ftp/src/nsFtpConnectionThread.cpp index ab6dc693e1d..a9575a6258c 100644 --- a/mozilla/netwerk/protocol/ftp/src/nsFtpConnectionThread.cpp +++ b/mozilla/netwerk/protocol/ftp/src/nsFtpConnectionThread.cpp @@ -1116,9 +1116,11 @@ nsFtpState::S_pass() { nsIAuthPrompt::SAVE_PASSWORD_PERMANENTLY, &passwd, &retval); - // we want to fail if the user canceled or didn't enter a password. - if (!retval || (passwd && !*passwd) ) + // we want to fail if the user canceled. Note here that if they want + // a blank password, we will pass it along. + if (!retval) return NS_ERROR_FAILURE; + mPassword = passwd; } // XXX mPassword may contain non-ASCII characters! what do we do? @@ -1255,7 +1257,6 @@ nsFtpState::R_syst() { // since we just alerted the user, clear mResponseMsg, // which is displayed to the user. mResponseMsg = ""; - return FTP_ERROR; } @@ -1433,9 +1434,6 @@ nsFtpState::SetContentType() switch (mListFormat) { case nsIDirectoryListing::FORMAT_RAW: { - nsAutoString fromStr(NS_LITERAL_STRING("text/ftp-dir-")); - SetDirMIMEType(fromStr); - contentType = NS_LITERAL_CSTRING("text/ftp-dir-"); } break; @@ -2408,27 +2406,6 @@ nsFtpState::StopProcessing() { return NS_OK; } -void -nsFtpState::SetDirMIMEType(nsString& aString) { - // the from content type is a string of the form - // "text/ftp-dir-SERVER_TYPE" where SERVER_TYPE represents the server we're talking to. - switch (mServerType) { - case FTP_UNIX_TYPE: - aString.Append(NS_LITERAL_STRING("unix")); - break; - case FTP_NT_TYPE: - aString.Append(NS_LITERAL_STRING("nt")); - break; - case FTP_OS2_TYPE: - aString.Append(NS_LITERAL_STRING("os2")); - break; - case FTP_VMS_TYPE: - aString.Append(NS_LITERAL_STRING("vms")); - break; - default: - aString.Append(NS_LITERAL_STRING("generic")); - } -} nsresult nsFtpState::BuildStreamConverter(nsIStreamListener** convertStreamListener) @@ -2446,8 +2423,8 @@ nsFtpState::BuildStreamConverter(nsIStreamListener** convertStreamListener) if (NS_FAILED(rv)) return rv; - nsAutoString fromStr(NS_LITERAL_STRING("text/ftp-dir-")); - SetDirMIMEType(fromStr); + nsAutoString fromStr(NS_LITERAL_STRING("text/ftp-dir")); + switch (mListFormat) { case nsIDirectoryListing::FORMAT_RAW: converterListener = listener; diff --git a/mozilla/netwerk/protocol/ftp/src/nsFtpConnectionThread.h b/mozilla/netwerk/protocol/ftp/src/nsFtpConnectionThread.h index ae343a6b812..4814e88ec8c 100644 --- a/mozilla/netwerk/protocol/ftp/src/nsFtpConnectionThread.h +++ b/mozilla/netwerk/protocol/ftp/src/nsFtpConnectionThread.h @@ -157,7 +157,6 @@ private: /////////////////////////////////// // internal methods - void SetDirMIMEType(nsString& aString); void MoveToNextState(FTP_STATE nextState); nsresult Process(); diff --git a/mozilla/netwerk/streamconv/converters/Makefile.in b/mozilla/netwerk/streamconv/converters/Makefile.in index 17af6653ad1..149d3668dc9 100644 --- a/mozilla/netwerk/streamconv/converters/Makefile.in +++ b/mozilla/netwerk/streamconv/converters/Makefile.in @@ -47,6 +47,7 @@ EXPORTS = \ $(NULL) CPPSRCS = \ + ParseFTPList.cpp \ nsMultiMixedConv.cpp \ nsFTPDirListingConv.cpp \ nsGopherDirListingConv.cpp \ diff --git a/mozilla/netwerk/streamconv/converters/ParseFTPList.cpp b/mozilla/netwerk/streamconv/converters/ParseFTPList.cpp new file mode 100644 index 00000000000..ae575088898 --- /dev/null +++ b/mozilla/netwerk/streamconv/converters/ParseFTPList.cpp @@ -0,0 +1,1802 @@ +/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */ +/* ----- BEGIN LICENSE BLOCK ----- + * Version: MPL 1.1/GPL 2.0/LGPL 2.1 + * + * The contents of this file are subject to the Mozilla Public License Version + * 1.1 (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * http://www.mozilla.org/MPL/ + * + * Software distributed under the License is distributed on an "AS IS" basis, + * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License + * for the specific language governing rights and limitations under the + * License. + * + * The Initial Developer of the Original Code is + * Cyrus Patel + * Portions created by the Initial Developer are Copyright (C) 2002 + * the Initial Developer. All Rights Reserved. + * + * Contributor(s): Doug Turner + * + * Alternatively, the contents of this file may be used under the terms of + * either of the GNU General Public License Version 2 or later (the "GPL"), + * or the GNU Lesser General Public License Version 2.1 or later (the "LGPL"), + * in which case the provisions of the GPL or the LGPL are applicable instead + * of those above. If you wish to allow use of your version of this file only + * under the terms of either the GPL or the LGPL, and not to allow others to + * use your version of this file under the terms of the MPL, indicate your + * decision by deleting the provisions above and replace them with the notice + * and other provisions required by the LGPL or the GPL. If you do not delete + * the provisions above, a recipient may use your version of this file under + * the terms of any one of the MPL, the GPL or the LGPL. + * + * ----- END LICENSE BLOCK ----- */ + +#include +#include +#include + +#include "ParseFTPList.h" + +/* ==================================================================== */ + +int ParseFTPList(const char *line, struct list_state *state, + struct list_result *result ) +{ + unsigned int carry_buf_len; /* copy of state->carry_buf_len */ + unsigned int linelen, pos; + const char *p; + + if (!line || !state || !result) + return 0; + + memset( result, 0, sizeof(*result) ); + if (state->magic != ((void *)ParseFTPList)) + { + memset( state, 0, sizeof(*state) ); + state->magic = ((void *)ParseFTPList); + } + state->numlines++; + + /* carry buffer is only valid from one line to the next */ + carry_buf_len = state->carry_buf_len; + state->carry_buf_len = 0; + + linelen = 0; + + /* strip leading whitespace */ + while (*line == ' ' || *line == '\t') + line++; + + /* line is terminated at first '\0' or '\n' */ + p = line; + while (*p && *p != '\n') + p++; + linelen = p - line; + + if (linelen > 0 && *p == '\n' && *(p-1) == '\r') + linelen--; + + /* DONT strip trailing whitespace. */ + + if (linelen > 0) + { + static const char *month_names = "JanFebMarAprMayJunJulAugSepOctNovDec"; + const char *tokens[16]; /* 16 is more than enough */ + unsigned int toklen[16]; + unsigned int numtoks = 0; + unsigned int tokmarker = 0; /* extra info for lstyle handler */ + unsigned int month_num = 0; + char tbuf[4]; + int lstyle = 0; + + if (carry_buf_len) /* VMS long filename carryover buffer */ + { + tokens[0] = state->carry_buf; + toklen[0] = carry_buf_len; + numtoks++; + } + + pos = 0; + while (pos < linelen && numtoks < (sizeof(tokens)/sizeof(tokens[0])) ) + { + while (pos < linelen && + (line[pos] == ' ' || line[pos] == '\t' || line[pos] == '\r')) + pos++; + if (pos < linelen) + { + tokens[numtoks] = &line[pos]; + while (pos < linelen && + (line[pos] != ' ' && line[pos] != '\t' && line[pos] != '\r')) + pos++; + if (tokens[numtoks] != &line[pos]) + { + toklen[numtoks] = (&line[pos] - tokens[numtoks]); + numtoks++; + } + } + } + + /* +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ */ + +#if defined(SUPPORT_EPLF) + /* EPLF handling must come somewhere before /bin/dls handling. */ + if (!lstyle && (!state->lstyle || state->lstyle == 'E')) + { + if (*line == '+' && linelen > 4 && numtoks >= 2) + { + pos = 1; + while (pos < (linelen-1)) + { + p = &line[pos++]; + if (*p == '/') + result->fe_type = 'd'; /* its a dir */ + else if (*p == 'r') + result->fe_type = 'f'; /* its a file */ + else if (*p == 'm') + { + if (isdigit(line[pos])) + { + while (pos < linelen && isdigit(line[pos])) + pos++; + if (pos < linelen && line[pos] == ',') + { + PRTime t; + PR_sscanf(p+1, "%llu", &t); + PR_ExplodeTime(t, PR_LocalTimeParameters, &(result->fe_time) ); + } + } + } + else if (*p == 's') + { + if (isdigit(line[pos])) + { + while (pos < linelen && isdigit(line[pos])) + pos++; + if (pos < linelen && line[pos] == ',' && + ((&line[pos]) - (p+1)) < (sizeof(result->fe_size)-1) ) + { + memcpy( result->fe_size, p+1, (unsigned)(&line[pos] - (p+1)) ); + result->fe_size[(&line[pos] - (p+1))] = '\0'; + } + } + } + else if (isalpha(*p)) /* 'i'/'up' or unknown "fact" (property) */ + { + while (pos < linelen && *++p != ',') + pos++; + } + else if (*p != '\t' || (p+1) != tokens[1]) + { + break; /* its not EPLF after all */ + } + else + { + state->parsed_one = 1; + state->lstyle = lstyle = 'E'; + + p = &(tokens[numtoks-1][toklen[numtoks-1]]); + result->fe_fname = tokens[1]; + result->fe_fnlen = p - tokens[1]; + + if (!result->fe_type) /* access denied */ + { + result->fe_type = 'f'; /* is assuming 'f'ile correct? */ + return '?'; /* NO! junk it. */ + } + return result->fe_type; + } + if (pos >= (linelen-1) || line[pos] != ',') + break; + pos++; + } /* while (pos < linelen) */ + memset( result, 0, sizeof(*result) ); + } /* if (*line == '+' && linelen > 4 && numtoks >= 2) */ + } /* if (!lstyle && (!state->lstyle || state->lstyle == 'E')) */ +#endif /* SUPPORT_EPLF */ + + /* +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ */ + +#if defined(SUPPORT_VMS) + if (!lstyle && (!state->lstyle || state->lstyle == 'V')) + { /* try VMS Multinet/UCX/CMS server */ + /* + * Legal characters in a VMS file/dir spec are [A-Z0-9$.-_~]. + * '$' cannot begin a filename and `-' cannot be used as the first + * or last character. '.' is only valid as a directory separator + * and . separator. A canonical filename spec might look + * like this: DISK$VOL:[DIR1.DIR2.DIR3]FILE.TYPE;123 + * All VMS FTP servers LIST in uppercase. + * + * We need to be picky about this in order to support + * multi-line listings correctly. + */ + if (!state->parsed_one && + (numtoks == 1 || (numtoks == 2 && toklen[0] == 9 && + memcmp(tokens[0], "Directory", 9)==0 ))) + { + /* If no dirstyle has been detected yet, and this line is a + * VMS list's dirname, then turn on VMS dirstyle. + * eg "ACA:[ANONYMOUS]", "DISK$FTP:[ANONYMOUS]", "SYS$ANONFTP:" + */ + p = tokens[0]; + pos = toklen[0]; + if (numtoks == 2) + { + p = tokens[1]; + pos = toklen[1]; + } + pos--; + if (pos >= 3) + { + while (pos > 0 && p[pos] != '[') + { + pos--; + if (p[pos] == '-' || p[pos] == '$') + { + if (pos == 0 || p[pos-1] == '[' || p[pos-1] == '.' || + (p[pos] == '-' && (p[pos+1] == ']' || p[pos+1] == '.'))) + break; + } + else if (p[pos] != '.' && p[pos] != '~' && + !isdigit(p[pos]) && !isalpha(p[pos])) + break; + else if (isalpha(p[pos]) && p[pos] != toupper(p[pos])) + break; + } + if (pos > 0) + { + pos--; + if (p[pos] != ':' || p[pos+1] != '[') + pos = 0; + } + } + if (pos > 0 && p[pos] == ':') + { + while (pos > 0) + { + pos--; + if (p[pos] != '$' && p[pos] != '_' && p[pos] != '-' && + p[pos] != '~' && !isdigit(p[pos]) && !isalpha(p[pos])) + break; + else if (isalpha(p[pos]) && p[pos] != toupper(p[pos])) + break; + } + if (pos == 0) + { + state->lstyle = 'V'; + return '?'; /* its junk */ + } + } + /* fallthrough */ + } + else if ((tokens[0][toklen[0]-1]) != ';') + { + if (numtoks == 1 && (state->lstyle == 'V' && !carry_buf_len)) + lstyle = 'V'; + else if (numtoks < 4) + ; + else if (toklen[1] >= 10 && memcmp(tokens[1], "%RMS-E-PRV", 10) == 0) + lstyle = 'V'; + else if ((&line[linelen] - tokens[1]) >= 22 && + memcmp(tokens[1], "insufficient privilege", 22) == 0) + lstyle = 'V'; + else if (numtoks != 4 && numtoks != 6) + ; + else if (numtoks == 6 && ( + toklen[4] < 3 || *tokens[4] != '[' || /* owner */ + (tokens[4][toklen[4]-1]) != ']' || + toklen[5] < 4 || *tokens[5] != '(' || /* perms */ + (tokens[5][toklen[5]-1]) != ')' )) + ; + else if ( (toklen[2] == 10 || toklen[2] == 11) && + (tokens[2][toklen[2]-5]) == '-' && + (tokens[2][toklen[2]-9]) == '-' && + (toklen[3]==4 || toklen[3]==5 || toklen[3]==7 || /* time */ + toklen[3]==8) && + (tokens[3][toklen[3]-3]) == ':' && + isdigit(*tokens[1]) && /* size */ + isdigit(*tokens[2]) && /* date */ + isdigit(*tokens[3]) /* time */ + ) + { + lstyle = 'V'; + } + if (lstyle == 'V') + { + /* + * MultiNet FTP: + * LOGIN.COM;2 1 4-NOV-1994 04:09 [ANONYMOUS] (RWE,RWE,,) + * PUB.DIR;1 1 27-JAN-1994 14:46 [ANONYMOUS] (RWE,RWE,RE,RWE) + * README.FTP;1 %RMS-E-PRV, insufficient privilege or file protection violation + * ROUSSOS.DIR;1 1 27-JAN-1994 14:48 [CS,ROUSSOS] (RWE,RWE,RE,R) + * S67-50903.JPG;1 328 22-SEP-1998 16:19 [ANONYMOUS] (RWED,RWED,,) + * UCX FTP: + * CII-MANUAL.TEX;1 213/216 29-JAN-1996 03:33:12 [ANONYMOU,ANONYMOUS] (RWED,RWED,,) + * CMU/VMS-IP FTP + * [VMSSERV.FILES]ALARM.DIR;1 1/3 5-MAR-1993 18:09 + * Long filename example: + * THIS-IS-A-LONG-VMS-FILENAME.AND-THIS-IS-A-LONG-VMS-FILETYPE\r\n + * 213[/nnn] 29-JAN-1996 03:33[:nn] [ANONYMOU,ANONYMOUS] (RWED,RWED,,) + */ + tokmarker = 0; + p = tokens[0]; + pos = 0; + if (*p == '[' && toklen[0] >= 4) /* CMU style */ + { + if (p[1] != ']') + { + p++; + pos++; + } + while (lstyle && pos < toklen[0] && *p != ']') + { + if (*p != '$' && *p != '.' && *p != '_' && *p != '-' && + *p != '~' && !isdigit(*p) && !isalpha(*p)) + lstyle = 0; + pos++; + p++; + } + if (lstyle && pos < (toklen[0]-1) && *p == ']') + { + pos++; + p++; + tokmarker = pos; /* length of leading "[DIR1.DIR2.etc]" */ + } + } + while (lstyle && pos < toklen[0] && *p != ';') + { + if (*p != '$' && *p != '.' && *p != '_' && *p != '-' && + *p != '~' && !isdigit(*p) && !isalpha(*p)) + lstyle = 0; + else if (isalpha(*p) && *p != toupper(*p)) + lstyle = 0; + p++; + pos++; + } + if (lstyle && *p == ';') + { + if (pos == 0 || pos == (toklen[0]-1)) + lstyle = 0; + for (pos++;lstyle && pos < toklen[0];pos++) + { + if (!isdigit(tokens[0][pos])) + lstyle = 0; + } + } + pos = (p - tokens[0]); /* => fnlength sans ";####" */ + pos -= tokmarker; /* => fnlength sans "[DIR1.DIR2.etc]" */ + p = &(tokens[0][tokmarker]); /* offset of basename */ + + if (!lstyle || pos > 80) /* VMS filenames can't be longer than that */ + { + lstyle = 0; + } + else if (numtoks == 1) + { + /* if VMS has been detected and there is only one token and that + * token was a VMS filename then this is a multiline VMS LIST entry. + */ + if (pos >= (sizeof(state->carry_buf)-1)) + pos = (sizeof(state->carry_buf)-1); /* shouldn't happen */ + memcpy( state->carry_buf, p, pos ); + state->carry_buf_len = pos; + return '?'; /* tell caller to treat as junk */ + } + else if (isdigit(*tokens[1])) /* not no-privs message */ + { + for (pos = 0; lstyle && pos < (toklen[1]); pos++) + { + if (!isdigit((tokens[1][pos])) && (tokens[1][pos]) != '/') + lstyle = 0; + } + if (lstyle && numtoks > 4) /* Multinet or UCX but not CMU */ + { + for (pos = 1; lstyle && pos < (toklen[5]-1); pos++) + { + p = &(tokens[5][pos]); + if (*p!='R' && *p!='W' && *p!='E' && *p!='D' && *p!=',') + lstyle = 0; + } + } + } + } /* passed initial tests */ + } /* else if ((tokens[0][toklen[0]-1]) != ';') */ + + if (lstyle == 'V') + { + state->parsed_one = 1; + state->lstyle = lstyle; + + if (isdigit(*tokens[1])) /* not permission denied etc */ + { + /* strip leading directory name */ + if (*tokens[0] == '[') /* CMU server */ + { + pos = toklen[0]-1; + p = tokens[0]+1; + while (*p != ']') + { + p++; + pos--; + } + toklen[0] = --pos; + tokens[0] = ++p; + } + pos = 0; + while (pos < toklen[0] && (tokens[0][pos]) != ';') + pos++; + + result->fe_cinfs = 1; + result->fe_type = 'f'; + result->fe_fname = tokens[0]; + result->fe_fnlen = pos; + + if (pos > 4) + { + p = &(tokens[0][pos-4]); + if (p[0] == '.' && p[1] == 'D' && p[2] == 'I' && p[3] == 'R') + { + result->fe_fnlen -= 4; + result->fe_type = 'd'; + } + } + + if (result->fe_type != 'd') + { + /* #### or used/allocated form. If used/allocated form, then + * 'used' is the size in bytes if and only if 'used'<=allocated. + * If 'used' is size in bytes then it can be > 2^32 + * If 'used' is not size in bytes then it is size in blocks. + */ + pos = 0; + while (pos < toklen[1] && (tokens[1][pos]) != '/') + pos++; + + if (pos < toklen[1] && ( (pos<<1) > (toklen[1]-1) || + (strtoul(tokens[1], (char **)0, 10) > + strtoul(tokens[1]+pos+1, (char **)0, 10)) )) + { /* size is in bytes */ + if (pos > (sizeof(result->fe_size)-1)) + pos = sizeof(result->fe_size)-1; + memcpy( result->fe_size, tokens[1], pos ); + result->fe_size[pos] = '\0'; + } + else /* size is in blocks */ + { + /* size requires multiplication by blocksize. + * + * We could assume blocksize is 512 (like Lynx does) and + * shift by 9, but that might not be right. Even if it + * were, doing that wouldn't reflect what the file's + * real size was. The sanest thing to do is not use the + * LISTing's filesize, so we won't (like ftpmirror). + * + * ulltoa(((unsigned long long)fsz)<<9, result->fe_size, 10); + */ + result->fe_size[0] = '\0'; + } + + } /* if (result->fe_type != 'd') */ + + p = tokens[2] + 2; + if (*p == '-') + p++; + tbuf[0] = p[0]; + tbuf[1] = tolower(p[1]); + tbuf[2] = tolower(p[2]); + month_num = 0; + for (pos = 0; pos < (12*3); pos+=3) + { + if (tbuf[0] == month_names[pos+0] && + tbuf[1] == month_names[pos+1] && + tbuf[2] == month_names[pos+2]) + break; + month_num++; + } + if (month_num >= 12) + month_num = 0; + result->fe_time.tm_month = month_num; + result->fe_time.tm_mday = atoi(tokens[2]); + result->fe_time.tm_year = atoi(p+4) - 1900; + + p = tokens[3] + 2; + if (*p == ':') + p++; + if (p[2] == ':') + result->fe_time.tm_sec = atoi(p+3); + result->fe_time.tm_hour = atoi(tokens[3]); + result->fe_time.tm_min = atoi(p); + + return result->fe_type; + + } /* if (isdigit(*tokens[1])) */ + + return '?'; /* junk */ + + } /* if (lstyle == 'V') */ + } /* if (!lstyle && (!state->lstyle || state->lstyle == 'V')) */ +#endif + + /* +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ */ + +#if defined(SUPPORT_CMS) + /* Virtual Machine/Conversational Monitor System (IBM Mainframe) */ + if (!lstyle && (!state->lstyle || state->lstyle == 'C')) /* VM/CMS */ + { + /* LISTing according to mirror.pl + * Filename FileType Fm Format Lrecl Records Blocks Date Time + * LASTING GLOBALV A1 V 41 21 1 9/16/91 15:10:32 + * J43401 NETLOG A0 V 77 1 1 9/12/91 12:36:04 + * PROFILE EXEC A1 V 17 3 1 9/12/91 12:39:07 + * DIRUNIX SCRIPT A1 V 77 1216 17 1/04/93 20:30:47 + * MAIL PROFILE A2 F 80 1 1 10/14/92 16:12:27 + * BADY2K TEXT A0 V 1 1 1 1/03/102 10:11:12 + * AUTHORS A1 DIR - - - 9/20/99 10:31:11 + * + * LISTing from vm.marist.edu and vm.sc.edu + * 220-FTPSERVE IBM VM Level 420 at VM.MARIST.EDU, 04:58:12 EDT WEDNESDAY 2002-07-10 + * AUTHORS DIR - - - 1999-09-20 10:31:11 - + * HARRINGTON DIR - - - 1997-02-12 15:33:28 - + * PICS DIR - - - 2000-10-12 15:43:23 - + * SYSFILE DIR - - - 2000-07-20 17:48:01 - + * WELCNVT EXEC V 72 9 1 1999-09-20 17:16:18 - + * WELCOME EREADME F 80 21 1 1999-12-27 16:19:00 - + * WELCOME README V 82 21 1 1999-12-27 16:19:04 - + * README ANONYMOU V 71 26 1 1997-04-02 12:33:20 TCP291 + * README ANONYOLD V 71 15 1 1995-08-25 16:04:27 TCP291 + */ + if (numtoks >= 7 && (toklen[0]+toklen[1]) <= 16) + { + for (pos = 1; !lstyle && (pos+5) < numtoks; pos++) + { + p = tokens[pos]; + if ((toklen[pos] == 1 && (*p == 'F' || *p == 'V')) || + (toklen[pos] == 3 && *p == 'D' && p[1] == 'I' && p[2] == 'R')) + { + if (toklen[pos+5] == 8 && (tokens[pos+5][2]) == ':' && + (tokens[pos+5][5]) == ':' ) + { + p = tokens[pos+4]; + if ((toklen[pos+4] == 10 && p[4] == '-' && p[7] == '-') || + (toklen[pos+4] >= 7 && toklen[pos+4] <= 9 && + p[((p[1]!='/')?(2):(1))] == '/' && + p[((p[1]!='/')?(5):(4))] == '/')) + /* Y2K bugs possible ("7/06/102" or "13/02/101") */ + { + if ( (*tokens[pos+1] == '-' && + *tokens[pos+2] == '-' && + *tokens[pos+3] == '-') || + (isdigit(*tokens[pos+1]) && + isdigit(*tokens[pos+2]) && + isdigit(*tokens[pos+3])) ) + { + lstyle = 'C'; + tokmarker = pos; + } + } + } + } + } /* for (pos = 1; !lstyle && (pos+5) < numtoks; pos++) */ + } /* if (numtoks >= 7) */ + + /* extra checking if first pass */ + if (lstyle && !state->lstyle) + { + for (pos = 0, p = tokens[0]; lstyle && pos < toklen[0]; pos++, p++) + { + if (isalpha(*p) && toupper(*p) != *p) + lstyle = 0; + } + for (pos = tokmarker+1; pos <= tokmarker+3; pos++) + { + if (!(toklen[pos] == 1 && *tokens[pos] == '-')) + { + for (p = tokens[pos]; lstyle && p<(tokens[pos]+toklen[pos]); p++) + { + if (!isdigit(*p)) + lstyle = 0; + } + } + } + for (pos = 0, p = tokens[tokmarker+4]; + lstyle && pos < toklen[tokmarker+4]; pos++, p++) + { + if (*p == '/') + { + /* There may be Y2K bugs in the date. Don't simplify to + * pos != (len-3) && pos != (len-6) like time is done. + */ + if ((tokens[tokmarker+4][1]) == '/') + { + if (pos != 1 && pos != 4) + lstyle = 0; + } + else if (pos != 2 && pos != 5) + lstyle = 0; + } + else if (*p != '-' && !isdigit(*p)) + lstyle = 0; + else if (*p == '-' && pos != 4 && pos != 7) + lstyle = 0; + } + for (pos = 0, p = tokens[tokmarker+5]; + lstyle && pos < toklen[tokmarker+5]; pos++, p++) + { + if (*p != ':' && !isdigit(*p)) + lstyle = 0; + else if (*p == ':' && pos != (toklen[tokmarker+5]-3) + && pos != (toklen[tokmarker+5]-6)) + lstyle = 0; + } + } /* initial if() */ + + if (lstyle == 'C') + { + state->parsed_one = 1; + state->lstyle = lstyle; + + p = tokens[tokmarker+4]; + if (toklen[tokmarker+4] == 10) /* newstyle: YYYY-MM-DD format */ + { + result->fe_time.tm_year = atoi(p+0) - 1900; + result->fe_time.tm_month = atoi(p+5) - 1; + result->fe_time.tm_mday = atoi(p+8); + } + else /* oldstyle: [M]M/DD/YY format */ + { + pos = toklen[tokmarker+4]; + result->fe_time.tm_month = atoi(p) - 1; + result->fe_time.tm_mday = atoi((p+pos)-5); + result->fe_time.tm_year = atoi((p+pos)-2); + if (result->fe_time.tm_year < 70) + result->fe_time.tm_year += 100; + } + + p = tokens[tokmarker+5]; + pos = toklen[tokmarker+5]; + result->fe_time.tm_hour = atoi(p); + result->fe_time.tm_min = atoi((p+pos)-5); + result->fe_time.tm_sec = atoi((p+pos)-2); + + result->fe_cinfs = 1; + result->fe_fname = tokens[0]; + result->fe_fnlen = toklen[0]; + result->fe_type = 'f'; + + p = tokens[tokmarker]; + if (toklen[tokmarker] == 3 && *p=='D' && p[1]=='I' && p[2]=='R') + result->fe_type = 'd'; + + if ((/*newstyle*/ toklen[tokmarker+4] == 10 && tokmarker > 1) || + (/*oldstyle*/ toklen[tokmarker+4] != 10 && tokmarker > 2)) + { /* have a filetype column */ + char *dot; + p = &(tokens[0][toklen[0]]); + memcpy( &dot, &p, sizeof(dot) ); /* NASTY! */ + *dot++ = '.'; + p = tokens[1]; + for (pos = 0; pos < toklen[1]; pos++) + *dot++ = *p++; + result->fe_fnlen += 1 + toklen[1]; + } + + /* oldstyle LISTING: + * files/dirs not on the 'A' minidisk are not RETRievable/CHDIRable + if (toklen[tokmarker+4] != 10 && *tokens[tokmarker-1] != 'A') + return '?'; + */ + + /* VM/CMS LISTings have no usable filesize field. + * Have to use the 'SIZE' command for that. + */ + return result->fe_type; + + } /* if (lstyle == 'C' && (!state->lstyle || state->lstyle == lstyle)) */ + } /* VM/CMS */ +#endif + + /* +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ */ + +#if defined(SUPPORT_DOS) /* WinNT DOS dirstyle */ + if (!lstyle && (!state->lstyle || state->lstyle == 'W')) + { + /* + * "10-23-00 01:27PM veronist" + * "06-15-00 07:37AM zoe" + * "07-14-00 01:35PM 2094926 canprankdesk.tif" + * "07-21-00 01:19PM 95077 Jon Kauffman Enjoys the Good Life.jpg" + * "07-21-00 01:19PM 52275 Name Plate.jpg" + * "07-14-00 01:38PM 2250540 Valentineoffprank-HiRes.jpg" + */ + if ((numtoks >= 4) && toklen[0] == 8 && toklen[1] == 7 && + (*tokens[2] == '<' || isdigit(*tokens[2])) ) + { + p = tokens[0]; + if ( isdigit(p[0]) && isdigit(p[1]) && p[2]=='-' && + isdigit(p[3]) && isdigit(p[4]) && p[5]=='-' && + isdigit(p[6]) && isdigit(p[7]) ) + { + p = tokens[1]; + if ( isdigit(p[0]) && isdigit(p[1]) && p[2]==':' && + isdigit(p[3]) && isdigit(p[4]) && + (p[5]=='A' || p[5]=='P') && p[6]=='M') + { + lstyle = 'W'; + if (!state->lstyle) + { + p = tokens[2]; + /* or */ + if (*p != '<' || p[toklen[2]-1] != '>') + { + for (pos = 1; (lstyle && pos < toklen[2]); pos++) + { + if (!isdigit(*++p)) + lstyle = 0; + } + } + } + } + } + } + + if (lstyle == 'W') + { + state->parsed_one = 1; + state->lstyle = lstyle; + + p = &(tokens[numtoks-1][toklen[numtoks-1]]); /* line end sans wsp */ + result->fe_cinfs = 1; + result->fe_fname = tokens[3]; + result->fe_fnlen = p - tokens[3]; + result->fe_type = 'd'; + + if (*tokens[2] != '<') /* not or */ + { + result->fe_type = 'f'; + pos = toklen[2]; + while (pos > (sizeof(result->fe_size)-1)) + pos = (sizeof(result->fe_size)-1); + memcpy( result->fe_size, tokens[2], pos ); + result->fe_size[pos] = '\0'; + } + else if ((tokens[2][1]) != 'D') /* not */ + { + result->fe_type = 'l'; + for (pos = 4; (pos+1) < numtoks; pos++) + { + if (toklen[pos] == 2 && (tokens[pos][1]) == '>' && + (*tokens[pos] == '=' || *tokens[pos] == '-')) + { + p = &(tokens[pos-1][toklen[pos-1]]); + result->fe_fnlen = p - tokens[3]; + p = &(tokens[numtoks-1][toklen[numtoks-1]]); + result->fe_lname = tokens[pos+1]; + result->fe_lnlen = p - tokens[pos+1]; + break; + } + } + } + + result->fe_time.tm_month = atoi(tokens[0]+0); + if (result->fe_time.tm_month != 0) + { + result->fe_time.tm_month--; + result->fe_time.tm_mday = atoi(tokens[0]+3); + result->fe_time.tm_year = atoi(tokens[0]+6); + if (result->fe_time.tm_year < 80) + result->fe_time.tm_year += 100; + } + + result->fe_time.tm_hour = atoi(tokens[1]+0); + result->fe_time.tm_min = atoi(tokens[1]+3); + if ((tokens[1][5]) == 'P' && result->fe_time.tm_hour < 12) + result->fe_time.tm_hour += 12; + + /* the caller should do this (if dropping "." and ".." is desired) + if (result->fe_type == 'd' && result->fe_fname[0] == '.' && + (result->fe_fnlen == 1 || (result->fe_fnlen == 2 && + result->fe_fname[1] == '.'))) + return '?'; + */ + + return result->fe_type; + } /* if (lstyle == 'W' && (!state->lstyle || state->lstyle == lstyle)) */ + } /* if (!lstyle && (!state->lstyle || state->lstyle == 'W')) */ +#endif + + /* +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ */ + +#if defined(SUPPORT_OS2) + if (!lstyle && (!state->lstyle || state->lstyle == 'O')) /* OS/2 test */ + { + /* 220 server IBM TCP/IP for OS/2 - FTP Server ver 23:04:36 on Jan 15 1997 ready. + * fixed position, space padded columns. I have only a vague idea + * of what the contents between col 18 and 34 might be: All I can infer + * is that there may be attribute flags in there and there may be + * a " DIR" in there. + * + * 1 2 3 4 5 6 + *0123456789012345678901234567890123456789012345678901234567890123456789 + *----- size -------|??????????????? MM-DD-YY| HH:MM| nnnnnnnnn.... + * 0 DIR 04-11-95 16:26 . + * 0 DIR 04-11-95 16:26 .. + * 0 DIR 04-11-95 16:26 ADDRESS + * 612 RHSA 07-28-95 16:45 air_tra1.bag + * 195 A 08-09-95 10:23 Alfa1.bag + * 0 RHS DIR 04-11-95 16:26 ATTACH + * 372 A 08-09-95 10:26 Aussie_1.bag + * 310992 06-28-94 09:56 INSTALL.EXE + * 1 2 3 4 + * 01234567890123456789012345678901234567890123456789 + * dirlist from the mirror.pl project, col positions from Mozilla. + */ + p = &(line[toklen[0]]); + /* \s(\d\d-\d\d-\d\d)\s+(\d\d:\d\d)\s */ + if (numtoks >= 4 && toklen[0] <= 18 && isdigit(*tokens[0]) && + (linelen - toklen[0]) >= (53-18) && + p[18-18] == ' ' && p[34-18] == ' ' && + p[37-18] == '-' && p[40-18] == '-' && p[43-18] == ' ' && + p[45-18] == ' ' && p[48-18] == ':' && p[51-18] == ' ' && + isdigit(p[35-18]) && isdigit(p[36-18]) && + isdigit(p[38-18]) && isdigit(p[39-18]) && + isdigit(p[41-18]) && isdigit(p[42-18]) && + isdigit(p[46-18]) && isdigit(p[47-18]) && + isdigit(p[49-18]) && isdigit(p[50-18]) + ) + { + lstyle = 'O'; /* OS/2 */ + if (!state->lstyle) + { + for (pos = 1; lstyle && pos < toklen[0]; pos++) + { + if (!isdigit(tokens[0][pos])) + lstyle = 0; + } + } + } + + if (lstyle == 'O') + { + state->parsed_one = 1; + state->lstyle = lstyle; + + p = &(line[toklen[0]]); + + result->fe_cinfs = 1; + result->fe_fname = &p[53-18]; + result->fe_fnlen = (&(tokens[numtoks-1][toklen[numtoks-1]])) + - (result->fe_fname); + result->fe_type = 'f'; + + /* I don't have a real listing to determine exact pos, so scan. */ + for (pos = (18-18); pos < ((35-18)-4); pos++) + { + if (p[pos+0] == ' ' && p[pos+1] == 'D' && + p[pos+2] == 'I' && p[pos+3] == 'R') + { + result->fe_type = 'd'; + break; + } + } + + if (result->fe_type != 'd') + { + pos = toklen[0]; + if (pos > (sizeof(result->fe_size)-1)) + pos = (sizeof(result->fe_size)-1); + memcpy( result->fe_size, tokens[0], pos ); + result->fe_size[pos] = '\0'; + } + + result->fe_time.tm_month = atoi(&p[35-18]) - 1; + result->fe_time.tm_mday = atoi(&p[38-18]); + result->fe_time.tm_year = atoi(&p[41-18]); + if (result->fe_time.tm_year < 80) + result->fe_time.tm_year += 100; + result->fe_time.tm_hour = atoi(&p[46-18]); + result->fe_time.tm_min = atoi(&p[49-18]); + + /* the caller should do this (if dropping "." and ".." is desired) + if (result->fe_type == 'd' && result->fe_fname[0] == '.' && + (result->fe_fnlen == 1 || (result->fe_fnlen == 2 && + result->fe_fname[1] == '.'))) + return '?'; + */ + + return result->fe_type; + } /* if (lstyle == 'O') */ + + } /* if (!lstyle && (!state->lstyle || state->lstyle == 'O')) */ +#endif + + /* +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ */ + +#if defined(SUPPORT_LSL) + if (!lstyle && (!state->lstyle || state->lstyle == 'U')) /* /bin/ls & co. */ + { + /* UNIX-style listing, without inum and without blocks + * "-rw-r--r-- 1 root other 531 Jan 29 03:26 README" + * "dr-xr-xr-x 2 root other 512 Apr 8 1994 etc" + * "dr-xr-xr-x 2 root 512 Apr 8 1994 etc" + * "lrwxrwxrwx 1 root other 7 Jan 25 00:17 bin -> usr/bin" + * Also produced by Microsoft's FTP servers for Windows: + * "---------- 1 owner group 1803128 Jul 10 10:18 ls-lR.Z" + * "d--------- 1 owner group 0 May 9 19:45 Softlib" + * Also WFTPD for MSDOS: + * "-rwxrwxrwx 1 noone nogroup 322 Aug 19 1996 message.ftp" + * Hellsoft for NetWare: + * "d[RWCEMFA] supervisor 512 Jan 16 18:53 login" + * "-[RWCEMFA] rhesus 214059 Oct 20 15:27 cx.exe" + * Newer Hellsoft for NetWare: (netlab2.usu.edu) + * - [RWCEAFMS] NFAUUser 192 Apr 27 15:21 HEADER.html + * d [RWCEAFMS] jrd 512 Jul 11 03:01 allupdates + * Also NetPresenz for the Mac: + * "-------r-- 326 1391972 1392298 Nov 22 1995 MegaPhone.sit" + * "drwxrwxr-x folder 2 May 10 1996 network" + * Protected directory: + * "drwx-wx-wt 2 root wheel 512 Jul 1 02:15 incoming" + * uid/gid instead of username/groupname: + * "drwxr-xr-x 2 0 0 512 May 28 22:17 etc" + */ + + if (numtoks >= 6) + { + /* there are two perm formats (Hellsoft/NetWare and *IX strmode(3)). + * Scan for size column only if the perm format is one or the other. + */ + if (toklen[0] == 1 || (tokens[0][1]) == '[') + { + if (*tokens[0] == 'd' || *tokens[0] == '-') + { + pos = toklen[0]-1; + p = tokens[0] + 1; + if (pos == 0) + { + p = tokens[1]; + pos = toklen[1]; + } + if ((pos == 9 || pos == 10) && + (*p == '[' && p[pos-1] == ']') && + (p[1] == 'R' || p[1] == '-') && + (p[2] == 'W' || p[2] == '-') && + (p[3] == 'C' || p[3] == '-') && + (p[4] == 'E' || p[4] == '-')) + { + /* rest is FMA[S] or AFM[S] */ + lstyle = 'U'; /* very likely one of the NetWare servers */ + } + } + } + else if ((toklen[0] == 10 || toklen[0] == 11) + && strchr("-bcdlpsw?DFam", *tokens[0])) + { + p = &(tokens[0][1]); + if ((p[0] == 'r' || p[0] == '-') && + (p[1] == 'w' || p[1] == '-') && + (p[3] == 'r' || p[3] == '-') && + (p[4] == 'w' || p[4] == '-') && + (p[6] == 'r' || p[6] == '-') && + (p[7] == 'w' || p[7] == '-')) + /* 'x'/p[9] can be S|s|x|-|T|t or implementation specific */ + { + lstyle = 'U'; /* very likely /bin/ls */ + } + } + } + if (lstyle == 'U') /* first token checks out */ + { + lstyle = 0; + for (pos = (numtoks-5); !lstyle && pos > 1; pos--) + { + /* scan for: (\d+)\s+([A-Z][a-z][a-z])\s+ + * (\d\d\d\d|\d\:\d\d|\d\d\:\d\d|\d\:\d\d\:\d\d|\d\d\:\d\d\:\d\d) + * \s+(.+)$ + */ + if (isdigit(*tokens[pos]) /* size */ + /* (\w\w\w) */ + && toklen[pos+1] == 3 && isalpha(*tokens[pos+1]) && + isalpha(tokens[pos+1][1]) && isalpha(tokens[pos+1][2]) + /* (\d|\d\d) */ + && isdigit(*tokens[pos+2]) && + (toklen[pos+2] == 1 || + (toklen[pos+2] == 2 && isdigit(tokens[pos+2][1]))) + && toklen[pos+3] >= 4 && isdigit(*tokens[pos+3]) + /* (\d\:\d\d\:\d\d|\d\d\:\d\d\:\d\d) */ + && (toklen[pos+3] <= 5 || ( + (toklen[pos+3] == 7 || toklen[pos+3] == 8) && + (tokens[pos+3][toklen[pos+3]-3]) == ':')) + && isdigit(tokens[pos+3][toklen[pos+3]-2]) + && isdigit(tokens[pos+3][toklen[pos+3]-1]) + && ( + /* (\d\d\d\d) */ + ((toklen[pos+3] == 4 || toklen[pos+3] == 5) && + isdigit(tokens[pos+3][1]) && + isdigit(tokens[pos+3][2]) ) + /* (\d\:\d\d|\d\:\d\d\:\d\d) */ + || ((toklen[pos+3] == 4 || toklen[pos+3] == 7) && + (tokens[pos+3][1]) == ':' && + isdigit(tokens[pos+3][2]) && isdigit(tokens[pos+3][3])) + /* (\d\d\:\d\d|\d\d\:\d\d\:\d\d) */ + || ((toklen[pos+3] == 5 || toklen[pos+3] == 8) && + isdigit(tokens[pos+3][1]) && (tokens[pos+3][2]) == ':' && + isdigit(tokens[pos+3][3]) && isdigit(tokens[pos+3][4])) + ) + ) + { + lstyle = 'U'; /* assume /bin/ls or variant format */ + tokmarker = pos; + + /* check that size is numeric */ + p = tokens[tokmarker]; + for (pos = 0; lstyle && pos < toklen[tokmarker]; pos++) + { + if (!isdigit(*p++)) + lstyle = 0; + } + if (lstyle) + { + month_num = 0; + p = tokens[tokmarker+1]; + for (pos = 0;pos < (12*3); pos+=3) + { + if (p[0] == month_names[pos+0] && + p[1] == month_names[pos+1] && + p[2] == month_names[pos+2]) + break; + month_num++; + } + if (month_num >= 12) + lstyle = 0; + } + } /* relative position test */ + } /* while (pos+5) < numtoks */ + } /* if (numtoks >= 4) */ + + if (lstyle == 'U') + { + state->parsed_one = 1; + state->lstyle = lstyle; + + result->fe_cinfs = 0; + result->fe_type = '?'; + if (*tokens[0] == 'd' || *tokens[0] == 'l') + result->fe_type = *tokens[0]; + else if (*tokens[0] == 'D') + result->fe_type = 'd'; + else if (*tokens[0] == '-' || *tokens[0] == 'F') + result->fe_type = 'f'; /* (hopefully a regular file) */ + + if (result->fe_type != 'd') + { + pos = toklen[tokmarker]; + if (pos > (sizeof(result->fe_size)-1)) + pos = (sizeof(result->fe_size)-1); + memcpy( result->fe_size, tokens[tokmarker], pos ); + result->fe_size[pos] = '\0'; + } + + result->fe_time.tm_month = month_num; + result->fe_time.tm_mday = atoi(tokens[tokmarker+2]); + if (result->fe_time.tm_mday == 0) + result->fe_time.tm_mday++; + + p = tokens[tokmarker+3]; + pos = (unsigned int)atoi(p); + if (p[1] == ':') /* one digit hour */ + p--; + if (p[2] != ':') /* year */ + { + if (pos >= 19100u) /* Y2K bug */ + pos %= 1000u; + else if (pos >= 1900u) + pos -= 1900; + else + pos = 0; + result->fe_time.tm_year = pos; + } + else + { + result->fe_time.tm_hour = pos; + result->fe_time.tm_min = atoi(p+3); + if (p[5] == ':') + result->fe_time.tm_sec = atoi(p+6); + + if (!state->now_time) + { + state->now_time = PR_Now(); + PR_ExplodeTime((state->now_time), PR_LocalTimeParameters, &(state->now_tm) ); + } + + result->fe_time.tm_year = state->now_tm.tm_year; + if ( (( state->now_tm.tm_month << 4) + state->now_tm.tm_mday) < + ((result->fe_time.tm_month << 4) + result->fe_time.tm_mday) ) + result->fe_time.tm_year--; + + } /* time/year */ + + result->fe_fname = tokens[tokmarker+4]; + result->fe_fnlen = (&(tokens[numtoks-1][toklen[numtoks-1]])) + - (result->fe_fname); + + if (result->fe_type == 'l' && result->fe_fnlen > 4) + { + p = tokens[tokmarker+4] + 1; + for (pos = 1; pos < (result->fe_fnlen - 4); pos++) + { + if (*p == ' ' && p[1] == '-' && p[2] == '>' && p[3] == ' ') + { + result->fe_lname = p + 4; + result->fe_lnlen = (&(tokens[numtoks-1][toklen[numtoks-1]])) + - (result->fe_lname); + result->fe_fnlen = pos; + break; + } + } + } + +#if defined(SUPPORT_LSLF) /* some (very rare) servers return ls -lF */ + if (result->fe_fnlen > 1) + { + p = result->fe_fname[result->fe_fnlen-1]; + pos = result->fe_type; + if (pos == 'd') { + if (*p == '/') result->fe_fnlen--; /* directory */ + } else if (pos == 'l') { + if (*p == '@') result->fe_fnlen--; /* symlink */ + } else if (pos == 'f') { + if (*p == '*') result->fe_fnlen--; /* executable */ + } else if (*p == '=' || *p == '%' || *p == '|') { + result->fe_fnlen--; /* socket, whiteout, fifo */ + } + } +#endif + + /* the caller should do this (if dropping "." and ".." is desired) + if (result->fe_type == 'd' && result->fe_fname[0] == '.' && + (result->fe_fnlen == 1 || (result->fe_fnlen == 2 && + result->fe_fname[1] == '.'))) + return '?'; + */ + + return result->fe_type; + + } /* if (lstyle == 'U') */ + + } /* if (!lstyle && (!state->lstyle || state->lstyle == 'U')) */ +#endif + + /* +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ */ + +#if defined(SUPPORT_W16) /* 16bit Windows */ + if (!lstyle && (!state->lstyle || state->lstyle == 'w')) + { /* old SuperTCP suite FTP server for Win3.1 */ + /* old NetManage Chameleon TCP/IP suite FTP server for Win3.1 */ + /* + * SuperTCP dirlist from the mirror.pl project + * mon/day/year separator may be '/' or '-'. + * . 11-16-94 17:16 + * .. 11-16-94 17:16 + * INSTALL 11-16-94 17:17 + * CMT 11-21-94 10:17 + * DESIGN1.DOC 11264 05-11-95 14:20 + * README.TXT 1045 05-10-95 11:01 + * WPKIT1.EXE 960338 06-21-95 17:01 + * CMT.CSV 0 07-06-95 14:56 + * + * Chameleon dirlist guessed from lynx + * . Nov 16 1994 17:16 + * .. Nov 16 1994 17:16 + * INSTALL Nov 16 1994 17:17 + * CMT Nov 21 1994 10:17 + * DESIGN1.DOC 11264 May 11 1995 14:20 A + * README.TXT 1045 May 10 1995 11:01 + * WPKIT1.EXE 960338 Jun 21 1995 17:01 R + * CMT.CSV 0 Jul 06 1995 14:56 RHA + */ + if (numtoks >= 4 && toklen[0] < 13 && + ((toklen[1] == 5 && *tokens[1] == '<') || isdigit(*tokens[1])) ) + { + if (numtoks == 4 + && (toklen[2] == 8 || toklen[2] == 9) + && (((tokens[2][2]) == '/' && (tokens[2][5]) == '/') || + ((tokens[2][2]) == '-' && (tokens[2][5]) == '-')) + && (toklen[3] == 4 || toklen[3] == 5) + && (tokens[3][toklen[3]-3]) == ':' + && isdigit(tokens[2][0]) && isdigit(tokens[2][1]) + && isdigit(tokens[2][3]) && isdigit(tokens[2][4]) + && isdigit(tokens[2][6]) && isdigit(tokens[2][7]) + && (toklen[2] < 9 || isdigit(tokens[2][8])) + && isdigit(tokens[3][toklen[3]-1]) && isdigit(tokens[3][toklen[3]-2]) + && isdigit(tokens[3][toklen[3]-4]) && isdigit(*tokens[3]) + ) + { + lstyle = 'w'; + } + else if ((numtoks == 6 || numtoks == 7) + && toklen[2] == 3 && toklen[3] == 2 + && toklen[4] == 4 && toklen[5] == 5 + && (tokens[5][2]) == ':' + && isalpha(tokens[2][0]) && isalpha(tokens[2][1]) + && isalpha(tokens[2][2]) + && isdigit(tokens[3][0]) && isdigit(tokens[3][1]) + && isdigit(tokens[4][0]) && isdigit(tokens[4][1]) + && isdigit(tokens[4][2]) && isdigit(tokens[4][3]) + && isdigit(tokens[5][0]) && isdigit(tokens[5][1]) + && isdigit(tokens[5][3]) && isdigit(tokens[5][4]) + /* could also check that (&(tokens[5][5]) - tokens[2]) == 17 */ + ) + { + lstyle = 'w'; + } + if (lstyle && state->lstyle != lstyle) /* first time */ + { + p = tokens[1]; + if (toklen[1] != 5 || p[0] != '<' || p[1] != 'D' || + p[2] != 'I' || p[3] != 'R' || p[4] != '>') + { + for (pos = 0; lstyle && pos < toklen[1]; pos++) + { + if (!isdigit(*p++)) + lstyle = 0; + } + } /* not */ + } /* if (first time) */ + } /* if (numtoks == ...) */ + + if (lstyle == 'w') + { + state->parsed_one = 1; + state->lstyle = lstyle; + + result->fe_cinfs = 1; + result->fe_fname = tokens[0]; + result->fe_fnlen = toklen[0]; + result->fe_type = 'd'; + + p = tokens[1]; + if (isdigit(*p)) + { + result->fe_type = 'f'; + pos = toklen[1]; + if (pos > (sizeof(result->fe_size)-1)) + pos = sizeof(result->fe_size)-1; + memcpy( result->fe_size, p, pos ); + result->fe_size[pos] = '\0'; + } + + p = tokens[2]; + if (toklen[2] == 3) /* Chameleon */ + { + tbuf[0] = toupper(p[0]); + tbuf[1] = tolower(p[1]); + tbuf[2] = tolower(p[2]); + for (pos = 0; pos < (12*3); pos+=3) + { + if (tbuf[0] == month_names[pos+0] && + tbuf[1] == month_names[pos+1] && + tbuf[2] == month_names[pos+2]) + { + result->fe_time.tm_month = pos/3; + result->fe_time.tm_mday = atoi(tokens[3]); + result->fe_time.tm_year = atoi(tokens[4]) - 1900; + break; + } + } + pos = 5; /* Chameleon toknum of date field */ + } + else + { + result->fe_time.tm_month = atoi(p+0)-1; + result->fe_time.tm_mday = atoi(p+3); + result->fe_time.tm_year = atoi(p+6); + if (result->fe_time.tm_year < 80) /* SuperTCP */ + result->fe_time.tm_year += 100; + + pos = 3; /* SuperTCP toknum of date field */ + } + + result->fe_time.tm_hour = atoi(tokens[pos]); + result->fe_time.tm_min = atoi(&(tokens[pos][toklen[pos]-2])); + + /* the caller should do this (if dropping "." and ".." is desired) + if (result->fe_type == 'd' && result->fe_fname[0] == '.' && + (result->fe_fnlen == 1 || (result->fe_fnlen == 2 && + result->fe_fname[1] == '.'))) + return '?'; + */ + + return result->fe_type; + } /* (lstyle == 'w') */ + + } /* if (!lstyle && (!state->lstyle || state->lstyle == 'w')) */ +#endif + + /* +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ */ + +#if defined(SUPPORT_DLS) /* dls -dtR */ + if (!lstyle && + (state->lstyle == 'D' || (!state->lstyle && state->numlines == 1))) + /* /bin/dls lines have to be immediately recognizable (first line) */ + { + /* I haven't seen an FTP server that delivers a /bin/dls listing, + * but can infer the format from the lynx and mirror.pl projects. + * Both formats are supported. + * + * Lynx says: + * README 763 Information about this server\0 + * bin/ - \0 + * etc/ = \0 + * ls-lR 0 \0 + * ls-lR.Z 3 \0 + * pub/ = Public area\0 + * usr/ - \0 + * morgan 14 -> ../real/morgan\0 + * TIMIT.mostlikely.Z\0 + * 79215 \0 + * + * mirror.pl says: + * filename: ^(\S*)\s+ + * size: (\-|\=|\d+)\s+ + * month/day: ((\w\w\w\s+\d+|\d+\s+\w\w\w)\s+ + * time/year: (\d+:\d+|\d\d\d\d))\s+ + * rest: (.+) + * + * README 763 Jul 11 21:05 Information about this server + * bin/ - Apr 28 1994 + * etc/ = 11 Jul 21:04 + * ls-lR 0 6 Aug 17:14 + * ls-lR.Z 3 05 Sep 1994 + * pub/ = Jul 11 21:04 Public area + * usr/ - Sep 7 09:39 + * morgan 14 Apr 18 09:39 -> ../real/morgan + * TIMIT.mostlikely.Z + * 79215 Jul 11 21:04 + */ + if (!state->lstyle && line[linelen-1] == ':' && + linelen >= 2 && toklen[numtoks-1] != 1) + { + /* code in mirror.pl suggests that a listing may be preceded + * by a PWD line in the form "/some/dir/names/here:" + * but does not necessarily begin with '/'. *sigh* + */ + pos = 0; + p = line; + while (pos < (linelen-1)) + { + /* illegal (or extremely unusual) chars in a dirspec */ + if (*p == '<' || *p == '|' || *p == '>' || + *p == '?' || *p == '*' || *p == '\\') + break; + if (*p == '/' && pos < (linelen-2) && p[1] == '/') + break; + pos++; + p++; + } + if (pos == (linelen-1)) + { + state->lstyle = 'D'; + return '?'; + } + } + + if (!lstyle && numtoks >= 2) + { + pos = 22; /* pos of (\d+|-|=) if this is not part of a multiline */ + if (state->lstyle && carry_buf_len) /* first is from previous line */ + pos = toklen[1]-1; /* and is 'as-is' (may contain whitespace) */ + + if (linelen > pos) + { + p = &line[pos]; + if ((*p == '-' || *p == '=' || isdigit(*p)) && + ((linelen == (pos+1)) || + (linelen >= (pos+3) && p[1] == ' ' && p[2] == ' ')) ) + { + tokmarker = 1; + if (!carry_buf_len) + { + pos = 1; + while (pos < numtoks && (tokens[pos]+toklen[pos]) < (&line[23])) + pos++; + tokmarker = 0; + if ((tokens[pos]+toklen[pos]) == (&line[23])) + tokmarker = pos; + } + if (tokmarker) + { + lstyle = 'D'; + if (*tokens[tokmarker] == '-' || *tokens[tokmarker] == '=') + { + if (toklen[tokmarker] != 1 || + (tokens[tokmarker-1][toklen[tokmarker-1]-1]) != '/') + lstyle = 0; + } + else + { + for (pos = 0; lstyle && pos < toklen[tokmarker]; pos++) + { + if (!isdigit(tokens[tokmarker][pos])) + lstyle = 0; + } + } + if (lstyle && !state->lstyle) /* first time */ + { + /* scan for illegal (or incredibly unusual) chars in fname */ + for (p = tokens[0]; lstyle && + p < &(tokens[tokmarker-1][toklen[tokmarker-1]]); p++) + { + if (*p == '<' || *p == '|' || *p == '>' || + *p == '?' || *p == '*' || *p == '/' || *p == '\\') + lstyle = 0; + } + } + + } /* size token found */ + } /* expected chars behind expected size token */ + } /* if (linelen > pos) */ + } /* if (!lstyle && numtoks >= 2) */ + + if (!lstyle && state->lstyle == 'D' && !carry_buf_len) + { + /* the filename of a multi-line entry can be identified + * correctly only if dls format had been previously established. + * This should always be true because there should be entries + * for '.' and/or '..' and/or CWD that precede the rest of the + * listing. + */ + pos = linelen; + if (pos > (sizeof(state->carry_buf)-1)) + pos = sizeof(state->carry_buf)-1; + memcpy( state->carry_buf, line, pos ); + state->carry_buf_len = pos; + return '?'; + } + + if (lstyle == 'D') + { + state->parsed_one = 1; + state->lstyle = lstyle; + + p = &(tokens[tokmarker-1][toklen[tokmarker-1]]); + result->fe_fname = tokens[0]; + result->fe_fnlen = p - tokens[0]; + result->fe_type = 'f'; + + if (result->fe_fname[result->fe_fnlen-1] == '/') + { + if (result->fe_lnlen == 1) + result->fe_type = '?'; + else + { + result->fe_fnlen--; + result->fe_type = 'd'; + } + } + else if (isdigit(*tokens[tokmarker])) + { + pos = toklen[tokmarker]; + if (pos > (sizeof(result->fe_size)-1)) + pos = sizeof(result->fe_size)-1; + memcpy( result->fe_size, tokens[tokmarker], pos ); + result->fe_size[pos] = '\0'; + } + + if ((tokmarker+3) < numtoks && + (&(tokens[numtoks-1][toklen[numtoks-1]]) - + tokens[tokmarker+1]) >= (1+1+3+1+4) ) + { + pos = (tokmarker+3); + p = tokens[pos]; + pos = toklen[pos]; + + if ((pos == 4 || pos == 5) + && isdigit(*p) && isdigit(p[pos-1]) && isdigit(p[pos-2]) + && ((pos == 5 && p[2] == ':') || + (pos == 4 && (isdigit(p[1]) || p[1] == ':'))) + ) + { + month_num = tokmarker+1; /* assumed position of month field */ + pos = tokmarker+2; /* assumed position of mday field */ + if (isdigit(*tokens[month_num])) /* positions are reversed */ + { + month_num++; + pos--; + } + p = tokens[month_num]; + if (isdigit(*tokens[pos]) + && (toklen[pos] == 1 || + (toklen[pos] == 2 && isdigit(tokens[pos][1]))) + && toklen[month_num] == 3 + && isalpha(*p) && isalpha(p[1]) && isalpha(p[2]) ) + { + pos = atoi(tokens[pos]); + if (pos > 0 && pos <= 31) + { + result->fe_time.tm_mday = pos; + month_num = 1; + for (pos = 0; pos < (12*3); pos+=3) + { + if (p[0] == month_names[pos+0] && + p[1] == month_names[pos+1] && + p[2] == month_names[pos+2]) + break; + month_num++; + } + if (month_num > 12) + result->fe_time.tm_mday = 0; + else + result->fe_time.tm_month = month_num - 1; + } + } + if (result->fe_time.tm_mday) + { + tokmarker += 3; /* skip mday/mon/yrtime (to find " -> ") */ + p = tokens[tokmarker]; + + pos = atoi(p); + if (pos > 24) + result->fe_time.tm_year = pos-1900; + else + { + if (p[1] == ':') + p--; + result->fe_time.tm_hour = pos; + result->fe_time.tm_min = atoi(p+3); + if (!state->now_time) + { + state->now_time = PR_Now(); + PR_ExplodeTime((state->now_time), PR_LocalTimeParameters, &(state->now_tm) ); + } + result->fe_time.tm_year = state->now_tm.tm_year; + if ( (( state->now_tm.tm_month << 4) + state->now_tm.tm_mday) < + ((result->fe_time.tm_month << 4) + result->fe_time.tm_mday) ) + result->fe_time.tm_year--; + } /* got year or time */ + } /* got month/mday */ + } /* may have year or time */ + } /* enough remaining to possibly have date/time */ + + if (numtoks > (tokmarker+2)) + { + pos = tokmarker+1; + p = tokens[pos]; + if (toklen[pos] == 2 && *p == '-' && p[1] == '>') + { + p = &(tokens[numtoks-1][toklen[numtoks-1]]); + result->fe_type = 'l'; + result->fe_lname = tokens[pos+1]; + result->fe_lnlen = p - result->fe_lname; + if (result->fe_lnlen > 1 && + result->fe_lname[result->fe_lnlen-1] == '/') + result->fe_lnlen--; + } + } /* if (numtoks > (tokmarker+2)) */ + + /* the caller should do this (if dropping "." and ".." is desired) + if (result->fe_type == 'd' && result->fe_fname[0] == '.' && + (result->fe_fnlen == 1 || (result->fe_fnlen == 2 && + result->fe_fname[1] == '.'))) + return '?'; + */ + + return result->fe_type; + + } /* if (lstyle == 'D') */ + } /* if (!lstyle && (!state->lstyle || state->lstyle == 'D')) */ +#endif + + /* +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ */ + + } /* if (linelen > 0) */ + + if (state->parsed_one || state->lstyle) /* junk if we fail to parse */ + return '?'; /* this time but had previously parsed sucessfully */ + return '"'; /* its part of a comment or error message */ +} + +/* ==================================================================== */ +/* standalone testing */ +/* ==================================================================== */ +#if 0 + +#include + +static int do_it(FILE *outfile, + char *line, size_t linelen, struct list_state *state, + char **cmnt_buf, unsigned int *cmnt_buf_sz, + char **list_buf, unsigned int *list_buf_sz ) +{ + struct list_result result; + char *p; + int rc; + + rc = ParseFTPLIST( line, state, &result ); + + if (!outfile) + { + outfile = stdout; + if (rc == '?') + fprintf(outfile, "junk: %.*s\n", (int)linelen, line ); + else if (rc == '"') + fprintf(outfile, "cmnt: %.*s\n", (int)linelen, line ); + else + fprintf(outfile, + "list: %02u-%02u-%02u %02u:%02u%cM %20s %.*s%s%.*s\n", + (result.fe_time.tm_mday ? (result.fe_time.tm_month + 1) : 0), + result.fe_time.tm_mday, + (result.fe_time.tm_mday ? (result.fe_time.tm_year % 100) : 0), + result.fe_time.tm_hour - + ((result.fe_time.tm_hour > 12)?(12):(0)), + result.fe_time.tm_min, + ((result.fe_time.tm_hour >= 12) ? 'P' : 'A'), + (rc == 'd' ? " " : + (rc == 'l' ? " " : result.fe_size)), + (int)result.fe_fnlen, result.fe_fname, + ((rc == 'l' && result.fe_lnlen) ? " -> " : ""), + (int)((rc == 'l' && result.fe_lnlen) ? result.fe_lnlen : 0), + ((rc == 'l' && result.fe_lnlen) ? result.fe_lname : "") ); + } + else if (rc != '?') /* NOT junk */ + { + char **bufp = list_buf; + unsigned int *bufz = list_buf_sz; + + if (rc == '"') /* comment - make it a 'result' */ + { + memset( &result, 0, sizeof(result)); + result.fe_fname = line; + result.fe_fnlen = linelen; + result.fe_type = 'f'; + if (line[linelen-1] == '/') + { + result.fe_type = 'd'; + result.fe_fnlen--; + } + bufp = cmnt_buf; + bufz = cmnt_buf_sz; + rc = result.fe_type; + } + + linelen = 80 + result.fe_fnlen + result.fe_lnlen; + p = (char *)realloc( *bufp, *bufz + linelen ); + if (!p) + return -1; + sprintf( &p[*bufz], + "%02u-%02u-%04u %02u:%02u:%02u %20s %.*s%s%.*s\n", + (result.fe_time.tm_mday ? (result.fe_time.tm_month + 1) : 0), + result.fe_time.tm_mday, + (result.fe_time.tm_mday ? (result.fe_time.tm_year + 1900) : 0), + result.fe_time.tm_hour, + result.fe_time.tm_min, + result.fe_time.tm_sec, + (rc == 'd' ? " " : + (rc == 'l' ? " " : result.fe_size)), + (int)result.fe_fnlen, result.fe_fname, + ((rc == 'l' && result.fe_lnlen) ? " -> " : ""), + (int)((rc == 'l' && result.fe_lnlen) ? result.fe_lnlen : 0), + ((rc == 'l' && result.fe_lnlen) ? result.fe_lname : "") ); + linelen = strlen(&p[*bufz]); + *bufp = p; + *bufz = *bufz + linelen; + } + return 0; +} + +int main(int argc, char *argv[]) +{ + FILE *infile = (FILE *)0; + FILE *outfile = (FILE *)0; + int need_close_in = 0; + int need_close_out = 0; + + if (argc > 1) + { + infile = stdin; + if (strcmp(argv[1], "-") == 0) + need_close_in = 0; + else if ((infile = fopen(argv[1], "r")) != ((FILE *)0)) + need_close_in = 1; + else + fprintf(stderr, "Unable to open input file '%s'\n", argv[1]); + } + if (infile && argc > 2) + { + outfile = stdout; + if (strcmp(argv[2], "-") == 0) + need_close_out = 0; + else if ((outfile = fopen(argv[2], "w")) != ((FILE *)0)) + need_close_out = 1; + else + { + fprintf(stderr, "Unable to open output file '%s'\n", argv[2]); + fclose(infile); + infile = (FILE *)0; + } + } + + if (!infile) + { + char *appname = &(argv[0][strlen(argv[0])]); + while (appname > argv[0]) + { + appname--; + if (*appname == '/' || *appname == '\\' || *appname == ':') + { + appname++; + break; + } + } + fprintf(stderr, + "Usage: %s []\n" + "\nIf an outout file is specified the results will be" + "\nbe post-processed, and only the file entries will appear" + "\n(or all comments if there are no file entries)." + "\nNot specifying an output file causes %s to run in \"debug\"" + "\nmode, ie results are printed as lines are parsed." + "\nIf a filename is a single dash ('-'), stdin/stdout is used." + "\n", appname, appname ); + } + else + { + char *cmnt_buf = (char *)0; + unsigned int cmnt_buf_sz = 0; + char *list_buf = (char *)0; + unsigned int list_buf_sz = 0; + + struct list_state state; + char line[512]; + + memset( &state, 0, sizeof(state) ); + while (fgets(line, sizeof(line), infile)) + { + size_t linelen = strlen(line); + if (linelen < (sizeof(line)-1)) + { + if (linelen > 0 && line[linelen-1] == '\n') + linelen--; + if (do_it( outfile, line, linelen, &state, + &cmnt_buf, &cmnt_buf_sz, &list_buf, &list_buf_sz) != 0) + { + fprintf(stderr, "Insufficient memory. Listing may be incomplete.\n"); + break; + } + } + else + { + /* no '\n' found. drop this and everything up to the next '\n' */ + fprintf(stderr, "drop: %.*s", (int)linelen, line ); + while (linelen == sizeof(line)) + { + if (!fgets(line, sizeof(line), infile)) + break; + linelen = 0; + while (linelen < sizeof(line) && line[linelen] != '\n') + linelen++; + fprintf(stderr, "%.*s", (int)linelen, line ); + } + fprintf(stderr, "\n"); + } + } + if (outfile) + { + if (list_buf) + fwrite( list_buf, 1, list_buf_sz, outfile ); + else if (cmnt_buf) + fwrite( cmnt_buf, 1, cmnt_buf_sz, outfile ); + } + if (list_buf) + free(list_buf); + if (cmnt_buf) + free(cmnt_buf); + + if (need_close_in) + fclose(infile); + if (outfile && need_close_out) + fclose(outfile); + } + + return 0; +} +#endif diff --git a/mozilla/netwerk/streamconv/converters/ParseFTPList.h b/mozilla/netwerk/streamconv/converters/ParseFTPList.h new file mode 100644 index 00000000000..3e2f5b27f15 --- /dev/null +++ b/mozilla/netwerk/streamconv/converters/ParseFTPList.h @@ -0,0 +1,124 @@ +/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */ +/* ----- BEGIN LICENSE BLOCK ----- + * Version: MPL 1.1/GPL 2.0/LGPL 2.1 + * + * The contents of this file are subject to the Mozilla Public License Version + * 1.1 (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * http://www.mozilla.org/MPL/ + * + * Software distributed under the License is distributed on an "AS IS" basis, + * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License + * for the specific language governing rights and limitations under the + * License. + * + * The Initial Developer of the Original Code is + * Cyrus Patel + * Portions created by the Initial Developer are Copyright (C) 2002 + * the Initial Developer. All Rights Reserved. + * + * Contributor(s): Doug Turner + * + * Alternatively, the contents of this file may be used under the terms of + * either of the GNU General Public License Version 2 or later (the "GPL"), + * or the GNU Lesser General Public License Version 2.1 or later (the "LGPL"), + * in which case the provisions of the GPL or the LGPL are applicable instead + * of those above. If you wish to allow use of your version of this file only + * under the terms of either the GPL or the LGPL, and not to allow others to + * use your version of this file under the terms of the MPL, indicate your + * decision by deleting the provisions above and replace them with the notice + * and other provisions required by the LGPL or the GPL. If you do not delete + * the provisions above, a recipient may use your version of this file under + * the terms of any one of the MPL, the GPL or the LGPL. + * + * ----- END LICENSE BLOCK ----- */ +#include "nspr.h" + +/* ParseFTPList() parses lines from an FTP LIST command. +** +** Written July 2002 by Cyrus Patel +** with acknowledgements to squid, lynx, wget and ftpmirror. +** +** Arguments: +** 'line': line of FTP data connection output. The line is assumed +** to end at the first '\0' or '\n' or '\r\n'. +** 'state': a structure used internally to track state between +** lines. Needs to be bzero()'d at LIST begin. +** 'result': where ParseFTPList will store the results of the parse +** if 'line' is not a comment and is not junk. +** +** Returns one of the following: +** 'd' - LIST line is a directory entry ('result' is valid) +** 'f' - LIST line is a file's entry ('result' is valid) +** 'l' - LIST line is a symlink's entry ('result' is valid) +** '?' - LIST line is junk. (cwd, non-file/dir/link, etc) +** '"' - its not a LIST line (its a "comment") +** +** It may be advisable to let the end-user see "comments" (particularly when +** the listing results in ONLY such lines) because such a listing may be: +** - an unknown LIST format (NLST or "custom" format for example) +** - an error msg (EPERM,ENOENT,ENFILE,EMFILE,ENOTDIR,ENOTBLK,EEXDEV etc). +** - an empty directory and the 'comment' is a "total 0" line or similar. +** (warning: a "total 0" can also mean the total size is unknown). +** +** ParseFTPList() supports all known FTP LISTing formats: +** - '/bin/ls -l' and all variants (including Hellsoft FTP for NetWare); +** - EPLF (Easily Parsable List Format); +** - Windows NT's default "DOS-dirstyle"; +** - OS/2 basic server format LIST format; +** - VMS (MultiNet, UCX, and CMU) LIST format (including multi-line format); +** - IBM VM/CMS, VM/ESA LIST format (two known variants); +** - SuperTCP FTP Server for Win16 LIST format; +** - NetManage Chameleon (NEWT) for Win16 LIST format; +** - '/bin/dls' (two known variants, plus multi-line) LIST format; +** If there are others, then I'd like to hear about them (send me a sample). +** +** NLSTings are not supported explicitely because they cannot be machine +** parsed consistantly: NLSTings do not have unique characteristics - even +** the assumption that there won't be whitespace on the line does not hold +** because some nlistings have more than one filename per line and/or +** may have filenames that have spaces in them. Moreover, distinguishing +** between an error message and an NLST line would require ParseList() to +** recognize all the possible strerror() messages in the world. +*/ + + +/* #undef anything you don't want to support */ +#define SUPPORT_LSL /* /bin/ls -l and dozens of variations therof */ +#define SUPPORT_DLS /* /bin/dls format (very, Very, VERY rare) */ +#define SUPPORT_EPLF /* Extraordinarily Pathetic List Format */ +#define SUPPORT_DOS /* WinNT server in 'site dirstyle' dos */ +#define SUPPORT_VMS /* VMS (all: MultiNet, UCX, CMU-IP) */ +#define SUPPORT_CMS /* IBM VM/CMS,VM/ESA (z/VM and LISTING forms) */ +#define SUPPORT_OS2 /* IBM TCP/IP for OS/2 - FTP Server */ +#define SUPPORT_W16 /* win16 hosts: SuperTCP or NetManage Chameleon */ + +struct list_state +{ + void *magic; /* to determine if previously initialized */ + PRTime now_time; /* needed for year determination */ + PRExplodedTime now_tm; /* needed for year determination */ + PRInt32 lstyle; /* LISTing style */ + PRInt32 parsed_one; /* returned anything yet? */ + char carry_buf[84]; /* for VMS multiline */ + PRUint32 carry_buf_len; /* length of name in carry_buf */ + PRUint32 numlines; /* number of lines seen */ +}; + +struct list_result +{ + PRInt32 fe_type; /* 'd'(dir) or 'l'(link) or 'f'(file) */ + const char * fe_fname; /* pointer to filename */ + PRUint32 fe_fnlen; /* length of filename */ + const char * fe_lname; /* pointer to symlink name */ + PRUint32 fe_lnlen; /* length of symlink name */ + char fe_size[40]; /* size of file in bytes (<= (2^128 - 1)) */ + PRExplodedTime fe_time; /* last-modified time */ + PRInt32 fe_cinfs; /* file system is definitely case insensitive */ + /* (converting all-upcase names may be desirable) */ +}; + +int ParseFTPList(const char *line, + struct list_state *state, + struct list_result *result ); + diff --git a/mozilla/netwerk/streamconv/converters/nsFTPDirListingConv.cpp b/mozilla/netwerk/streamconv/converters/nsFTPDirListingConv.cpp index 83f0ef23d23..27c87c5988c 100644 --- a/mozilla/netwerk/streamconv/converters/nsFTPDirListingConv.cpp +++ b/mozilla/netwerk/streamconv/converters/nsFTPDirListingConv.cpp @@ -55,15 +55,7 @@ #include "nsCRT.h" #include "nsMimeTypes.h" -#define IS_LWS(c) (PL_strchr(" \t\r\n",c) != 0) -#define IS_FTYPE(c) (PL_strchr("-dlbcsp",c) != 0) -#define IS_RPERM(c) (PL_strchr("r-",c) != 0) -#define IS_WPERM(c) (PL_strchr("w-",c) != 0) -#define IS_SPERM(c) (PL_strchr("sSx-",c) != 0) -#define IS_TPERM(c) (PL_strchr("tTx-",c) != 0) - -static NS_DEFINE_CID(kLocaleServiceCID, NS_LOCALESERVICE_CID); -static NS_DEFINE_CID(kDateTimeCID, NS_DATETIMEFORMAT_CID); +#include "ParseFTPList.h" #if defined(PR_LOGGING) // @@ -87,111 +79,17 @@ NS_IMPL_THREADSAFE_ISUPPORTS3(nsFTPDirListingConv, nsIStreamListener, nsIRequestObserver); -// Common code of Convert and AsyncConvertData function - -static FTP_Server_Type -DetermineServerType (nsCString &fromMIMEString, const PRUnichar *aFromType) -{ - fromMIMEString.AssignWithConversion(aFromType); - const char *from = fromMIMEString.get(); - NS_ASSERTION(from, "nsCString/PRUnichar acceptance failed."); - - from = PL_strstr(from, "/ftp-dir-"); - if (!from) return ERROR_TYPE; - from += 9; - fromMIMEString = from; - - if (-1 != fromMIMEString.Find("unix")) { - return UNIX; - } else if (-1 != fromMIMEString.Find("nt")) { - return NT; - } else if (-1 != fromMIMEString.Find("dcts")) { - return DCTS; - } else if (-1 != fromMIMEString.Find("ncsa")) { - return NCSA; - } else if (-1 != fromMIMEString.Find("peter_lewis")) { - return PETER_LEWIS; - } else if (-1 != fromMIMEString.Find("machten")) { - return MACHTEN; - } else if (-1 != fromMIMEString.Find("cms")) { - return CMS; - } else if (-1 != fromMIMEString.Find("tcpc")) { - return TCPC; - } else if (-1 != fromMIMEString.Find("os2")) { - return OS_2; - } - - return GENERIC; -} // nsIStreamConverter implementation - -#define CONV_BUF_SIZE (4*1024) NS_IMETHODIMP nsFTPDirListingConv::Convert(nsIInputStream *aFromStream, const PRUnichar *aFromType, const PRUnichar *aToType, nsISupports *aCtxt, nsIInputStream **_retval) { - nsresult rv; - - // set our internal state to reflect the server type - nsCString fromMIMEString; - - mServerType = DetermineServerType(fromMIMEString, aFromType); - if (mServerType == ERROR_TYPE) return NS_ERROR_FAILURE; - - char buffer[CONV_BUF_SIZE]; - int i = 0; - while (i < CONV_BUF_SIZE) { - buffer[i] = '\0'; - i++; - } - CBufDescriptor desc(buffer, PR_TRUE, CONV_BUF_SIZE); - nsCAutoString aBuffer(desc); - nsCString convertedData; - - NS_ASSERTION(aCtxt, "FTP dir conversion needs the context"); - - nsCOMPtr uri(do_QueryInterface(aCtxt, &rv)); - if (NS_FAILED(rv)) return rv; - - rv = GetHeaders(convertedData, uri); - if (NS_FAILED(rv)) return rv; - - // build up the body - while (1) { - PRUint32 amtRead = 0; - - rv = aFromStream->Read(buffer+aBuffer.Length(), - CONV_BUF_SIZE-aBuffer.Length(), &amtRead); - if (NS_FAILED(rv)) return rv; - - if (!amtRead) { - // EOF - break; - } - - aBuffer = DigestBufferLines(buffer, convertedData); - } - // end body building - -#ifndef DEBUG_valeski - PR_LOG(gFTPDirListConvLog, PR_LOG_DEBUG, ("::OnData() sending the following %d bytes...\n\n%s\n\n", - convertedData.Length(), convertedData.get()) ); -#else - char *unescData = ToNewCString(convertedData); - nsUnescape(unescData); - printf("::OnData() sending the following %d bytes...\n\n%s\n\n", convertedData.Length(), unescData); - nsMemory::Free(unescData); -#endif // DEBUG_valeski - - // send the converted data out. - return NS_NewCStringInputStream(_retval, convertedData); + return NS_ERROR_NOT_IMPLEMENTED; } - - // Stream converter service calls this to initialize the actual stream converter (us). NS_IMETHODIMP nsFTPDirListingConv::AsyncConvertData(const PRUnichar *aFromType, const PRUnichar *aToType, @@ -204,13 +102,6 @@ nsFTPDirListingConv::AsyncConvertData(const PRUnichar *aFromType, const PRUnicha mFinalListener = aListener; NS_ADDREF(mFinalListener); - // set our internal state to reflect the server type - nsCString fromMIMEString; - - mServerType = DetermineServerType(fromMIMEString, aFromType); - if (mServerType == ERROR_TYPE) return NS_ERROR_FAILURE; - - // we need our own channel that represents the content-type of the // converted data. NS_ASSERTION(aCtxt, "FTP dir listing needs a context (the uri)"); @@ -228,7 +119,7 @@ nsFTPDirListingConv::AsyncConvertData(const PRUnichar *aFromType, const PRUnicha if (NS_FAILED(rv)) return rv; PR_LOG(gFTPDirListConvLog, PR_LOG_DEBUG, - ("nsFTPDirListingConv::AsyncConvertData() converting FROM raw %s, TO application/http-index-format\n", fromMIMEString.get())); + ("nsFTPDirListingConv::AsyncConvertData() converting FROM raw, TO application/http-index-format\n")); return NS_OK; } @@ -268,11 +159,11 @@ nsFTPDirListingConv::OnDataAvailable(nsIRequest* request, nsISupports *ctxt, mBuffer.Truncate(); } -#ifndef DEBUG_valeski +#ifndef DEBUG_dougt PR_LOG(gFTPDirListConvLog, PR_LOG_DEBUG, ("::OnData() received the following %d bytes...\n\n%s\n\n", streamLen, buffer) ); #else printf("::OnData() received the following %d bytes...\n\n%s\n\n", streamLen, buffer); -#endif // DEBUG_valeski +#endif // DEBUG_dougt nsCString indexFormat; if (!mSentHeading) { @@ -290,7 +181,7 @@ nsFTPDirListingConv::OnDataAvailable(nsIRequest* request, nsISupports *ctxt, char *line = buffer; line = DigestBufferLines(line, indexFormat); -#ifndef DEBUG_valeski +#ifndef DEBUG_dougt PR_LOG(gFTPDirListConvLog, PR_LOG_DEBUG, ("::OnData() sending the following %d bytes...\n\n%s\n\n", indexFormat.Length(), indexFormat.get()) ); #else @@ -298,7 +189,7 @@ nsFTPDirListingConv::OnDataAvailable(nsIRequest* request, nsISupports *ctxt, nsUnescape(unescData); printf("::OnData() sending the following %d bytes...\n\n%s\n\n", indexFormat.Length(), unescData); nsMemory::Free(unescData); -#endif // DEBUG_valeski +#endif // DEBUG_dougt // if there's any data left over, buffer it. if (line && *line) { @@ -356,7 +247,6 @@ nsFTPDirListingConv::OnStopRequest(nsIRequest* request, nsISupports *ctxt, nsFTPDirListingConv::nsFTPDirListingConv() { NS_INIT_ISUPPORTS(); mFinalListener = nsnull; - mServerType = GENERIC; mPartChannel = nsnull; mSentHeading = PR_FALSE; } @@ -416,368 +306,8 @@ nsFTPDirListingConv::GetHeaders(nsACString& headers, return rv; } -PRInt8 -nsFTPDirListingConv::MonthNumber(const char *month) { - NS_ASSERTION(month && month[1] && month[2], "bad month"); - if (!month || !month[0] || !month[1] || !month[2]) - return -1; - - char c1 = month[1], c2 = month[2]; - PRInt8 rv = -1; - - //PR_LOG(gFTPDirListConvLog, PR_LOG_DEBUG, ("nsFTPDirListingConv::MonthNumber(month = %s) ", month) ); - - switch (*month) { - case 'f': case 'F': - rv = 1; break; - case 'm': case 'M': - // c1 == 'a' || c1 == 'A' - if (c2 == 'r' || c2 == 'R') - rv = 2; - else - // c2 == 'y' || c2 == 'Y' - rv = 4; - break; - case 'a': case 'A': - if (c1 == 'p' || c1 == 'P') - rv = 3; - else - // c1 == 'u' || c1 == 'U' - rv = 7; - break; - case 'j': case 'J': - if (c1 == 'u' || c1 == 'U') { - if (c2 == 'n' || c2 == 'N') - rv = 5; - else - // c2 == 'l' || c2 == 'L' - rv = 6; - } else { - // c1 == 'a' || c1 == 'A' - rv = 0; - } - break; - case 's': case 'S': - rv = 8; break; - case 'o': case 'O': - rv = 9; break; - case 'n': case 'N': - rv = 10; break; - case 'd': case 'D': - rv = 11; break; - default: - rv = -1; - } - - //PR_LOG(gFTPDirListConvLog, PR_LOG_DEBUG, ("returning %d\n", rv) ); - - return rv; -} - - -// Return true if the string is of the form: -// "Sep 1 1990 " or -// "Sep 11 11:59 " or -// "Dec 12 1989 " or -// "FCv 23 1990 " ... -PRBool -nsFTPDirListingConv::IsLSDate(char *aCStr) { - - /* must start with three alpha characters */ - if (!nsCRT::IsAsciiAlpha(*aCStr++) || !nsCRT::IsAsciiAlpha(*aCStr++) || !nsCRT::IsAsciiAlpha(*aCStr++)) - return PR_FALSE; - - /* space */ - if (*aCStr != ' ') - return PR_FALSE; - aCStr++; - - /* space or digit */ - if ((*aCStr != ' ') && !nsCRT::IsAsciiDigit(*aCStr)) - return PR_FALSE; - aCStr++; - - /* digit */ - if (!nsCRT::IsAsciiDigit(*aCStr)) - return PR_FALSE; - aCStr++; - - /* space */ - if (*aCStr != ' ') - return PR_FALSE; - aCStr++; - - /* space or digit */ - if ((*aCStr != ' ') && !nsCRT::IsAsciiDigit(*aCStr)) - return PR_FALSE; - aCStr++; - - /* digit */ - if (!nsCRT::IsAsciiDigit(*aCStr)) - return PR_FALSE; - aCStr++; - - /* colon or digit */ - if ((*aCStr != ':') && !nsCRT::IsAsciiDigit(*aCStr)) - return PR_FALSE; - aCStr++; - - /* digit */ - if (!nsCRT::IsAsciiDigit(*aCStr)) - return PR_FALSE; - aCStr++; - - /* space or digit */ - if ((*aCStr != ' ') && !nsCRT::IsAsciiDigit(*aCStr)) - return PR_FALSE; - aCStr++; - - /* space */ - if (*aCStr != ' ') - return PR_FALSE; - aCStr++; - - return PR_TRUE; -} - - -// Converts a date string from 'ls -l' to a PRTime number -// "Sep 1 1990 " or -// "Sep 11 11:59 " or -// "Dec 12 1989 " or -// "FCv 23 1990 " ... -// Returns 0 on error. -PRBool -nsFTPDirListingConv::ConvertUNIXDate(char *aCStr, PRExplodedTime& outDate) { - - PRExplodedTime curTime; - InitPRExplodedTime(curTime); - - char *bcol = aCStr; /* Column begin */ - char *ecol; /* Column end */ - - // MONTH - char tmpChar = bcol[3]; - bcol[3] = '\0'; - - if ((curTime.tm_month = MonthNumber(bcol)) < 0) - return PR_FALSE; - bcol[3] = tmpChar; - - // DAY - ecol = &bcol[3]; - while (*(++ecol) == ' ') ; - while (*(++ecol) != ' ') ; - *ecol = '\0'; - bcol = ecol+1; - while (*(--ecol) != ' ') ; - - PRInt32 error; - nsCAutoString day(ecol); - curTime.tm_mday = day.ToInteger(&error, 10); - - // YEAR - if ((ecol = PL_strchr(bcol, ':')) == NULL) { - nsCAutoString intStr(bcol); - curTime.tm_year = intStr.ToInteger(&error, 10); - } else { - // TIME - /* If the time is given as hh:mm, then the file is less than 1 year - * old, but we might shift calandar year. This is avoided by checking - * if the date parsed is future or not. - */ - *ecol = '\0'; - nsCAutoString intStr(++ecol); - curTime.tm_min = intStr.ToInteger(&error, 10); // Right side of ':' - - intStr = bcol; - curTime.tm_hour = intStr.ToInteger(&error, 10); // Left side of ':' - - PRExplodedTime nowETime; - PR_ExplodeTime(PR_Now(), PR_LocalTimeParameters, &nowETime); - curTime.tm_year = nowETime.tm_year; - - PRBool thisCalendarYear = PR_FALSE; - if (nowETime.tm_month > curTime.tm_month) { - thisCalendarYear = PR_TRUE; - } else if (nowETime.tm_month == curTime.tm_month - && nowETime.tm_mday > curTime.tm_mday) { - thisCalendarYear = PR_TRUE; - } else if (nowETime.tm_month == curTime.tm_month - && nowETime.tm_mday == curTime.tm_mday - && nowETime.tm_hour > curTime.tm_hour) { - thisCalendarYear = PR_TRUE; - } else if (nowETime.tm_month == curTime.tm_month - && nowETime.tm_mday == curTime.tm_mday - && nowETime.tm_hour == curTime.tm_hour - && nowETime.tm_min >= curTime.tm_min) { - thisCalendarYear = PR_TRUE; - } - - if (!thisCalendarYear) curTime.tm_year--; - } - - // set the out param - outDate = curTime; - return PR_TRUE; -} - -PRBool -nsFTPDirListingConv::ConvertDOSDate(char *aCStr, PRExplodedTime& outDate) { - - PRExplodedTime curTime, nowTime; - PR_ExplodeTime(PR_Now(), PR_LocalTimeParameters, &nowTime); - PRInt16 currCentury = nowTime.tm_year / 100; - - InitPRExplodedTime(curTime); - - curTime.tm_month = ((aCStr[0]-'0')*10 + (aCStr[1]-'0')) - 1; - - curTime.tm_mday = (((aCStr[3]-'0')*10) + aCStr[4]-'0'); - curTime.tm_year = currCentury*100 + ((aCStr[6]-'0')*10) + aCStr[7]-'0'; - // year is only 2-digit, so we've converted. If its in the future, - // it must have been last century - if (curTime.tm_year > nowTime.tm_year) { - curTime.tm_year -= 100; - } - - curTime.tm_hour = (((aCStr[10]-'0')*10) + aCStr[11]-'0'); - - if (aCStr[15] == 'P') - curTime.tm_hour += 12; - - curTime.tm_min = (((aCStr[13]-'0')*10) + aCStr[14]-'0'); - - outDate = curTime; - return PR_TRUE; -} - -nsresult -nsFTPDirListingConv::ParseLSLine(char *aLine, indexEntry *aEntry) { - - PRInt32 base=1; - PRInt32 size_num=0; - char save_char; - char *ptr, *escName; - - // insure that we have enough data to parse here - // using 27 for a minimal LIST line - if (PL_strlen(aLine) <= 27) { - NS_WARNING("ls -l is incorrectly formatted"); - aEntry->mName.Adopt(nsEscape(aLine, url_Path)); - // initialize the time struct to 0 - InitPRExplodedTime(aEntry->mMDTM); - return NS_OK; - } - - for (ptr = &aLine[PL_strlen(aLine) - 1]; - (ptr > aLine+13) && (!nsCRT::IsAsciiSpace(*ptr) || !IsLSDate(ptr-12)); ptr--) - ; /* null body */ - save_char = *ptr; - *ptr = '\0'; - if (ptr > aLine+13) { - ConvertUNIXDate(ptr-12, aEntry->mMDTM); - } else { - // must be a dl listing - // unterminate the line - *ptr = save_char; - // find the first whitespace and terminate - for(ptr=aLine; *ptr != '\0'; ptr++) - if (nsCRT::IsAsciiSpace(*ptr)) { - *ptr = '\0'; - break; - } - escName = nsEscape(aLine, url_Path); - aEntry->mName.Adopt(escName); - - // initialize the time struct to 0 to be safe - InitPRExplodedTime(aEntry->mMDTM); - - return NS_OK; - } - - escName = nsEscape(ptr+1, url_Path); - aEntry->mName.Adopt(escName); - - // parse size - if (ptr > aLine+15) { - ptr -= 14; - while (nsCRT::IsAsciiDigit(*ptr)) { - size_num += ((PRInt32) (*ptr - '0')) * base; - base *= 10; - ptr--; - } - - aEntry->mContentLen = size_num; - } - return NS_OK; -} - -void -nsFTPDirListingConv::InitPRExplodedTime(PRExplodedTime& aTime) { - aTime.tm_usec = 0; - aTime.tm_sec = 0; - aTime.tm_min = 0; - aTime.tm_hour = 0; - aTime.tm_mday = 0; - aTime.tm_month= 0; - aTime.tm_year = 0; - aTime.tm_wday = 0; - aTime.tm_yday = 0; - aTime.tm_params.tp_gmt_offset = 0; - aTime.tm_params.tp_dst_offset = 0; -} - -/** - * Can this line possibly be a ls-l line? - * Ideally, one should parse the line into tokens, etc. but - * that's for later. At the moment, just test the line for - * length and examine the first token. An absolutely minimal - * line has the form - * -rwxrwxrwx 0 Dec 31 23:59 F - * which is 27 characters. If the line is at least 27 bytes - * long and the first token matches the regular expression - * [-dlbcps][r-][w-][sSx-][r-][w-][sSx-][r-][w-][tTx-] - * then accept the line *provided* there is another token on - * the line. - - * As written, this function *assumes* leading whitespace has - * been removed. That's what the rest of the code assumes. It - * may not be a reasonable assumption. - - * DigestBufferLines hints that the regular expression test should - * be done case insensitively. That would relax this test so it is - * *not* done. If definitive evidence exists of a server that does - * this, then it can be changed. - - * This function uses a lot of macros for clarity. They probably - * should be improved. - */ - -PRBool -nsFTPDirListingConv::ls_lCandidate(const char *lsLine) { - - const char *cp; - - if (PL_strlen(lsLine) < 27) return PR_FALSE; - if (!IS_FTYPE(lsLine[0])) return PR_FALSE; - // shorter tests first - if (!IS_RPERM(lsLine[1])) return PR_FALSE; - if (!IS_WPERM(lsLine[2])) return PR_FALSE; - if (!IS_RPERM(lsLine[4])) return PR_FALSE; - if (!IS_WPERM(lsLine[5])) return PR_FALSE; - if (!IS_RPERM(lsLine[7])) return PR_FALSE; - if (!IS_WPERM(lsLine[8])) return PR_FALSE; - if (!IS_SPERM(lsLine[3])) return PR_FALSE; - if (!IS_SPERM(lsLine[6])) return PR_FALSE; - for (cp = &lsLine[10]; *cp; ++cp) - if (!IS_LWS(*cp)) return PR_TRUE; - return PR_FALSE; -} - char * nsFTPDirListingConv::DigestBufferLines(char *aBuffer, nsCString &aString) { - nsresult rv; char *line = aBuffer; char *eol; PRBool cr = PR_FALSE; @@ -793,281 +323,53 @@ nsFTPDirListingConv::DigestBufferLines(char *aBuffer, nsCString &aString) { *eol = '\0'; cr = PR_FALSE; } - indexEntry *thisEntry = nsnull; - NS_NEWXPCOM(thisEntry, indexEntry); - if (!thisEntry) return nsnull; - // XXX we need to handle comments in the raw stream. + list_state state; + list_result result; - // special case windows servers who masquerade as unix servers - if (NT == mServerType && !nsCRT::IsAsciiSpace(line[8])) - mServerType = UNIX; + int type = ParseFTPList(line, &state, &result ); - // check for an eplf response - if (line[0] == '+') - mServerType = EPLF; - - char *escName = nsnull; - switch (mServerType) { - - case UNIX: - case PETER_LEWIS: - case MACHTEN: + // if it is other than a directory, file, or link -OR- if it is a + // directory named . or .., skip over this line. + if ((type != 'd' && type != 'f' && type != 'l') || + (result.fe_type == 'd' && result.fe_fname[0] == '.' && + (result.fe_fnlen == 1 || (result.fe_fnlen == 2 && result.fe_fname[1] == '.'))) ) { - // don't bother w/ these lines. - if (!PL_strncmp(line, "total ", 6) - || !PL_strncmp(line, "ls: total", 9) - || (PL_strstr(line, "Permission denied") != NULL) - || (PL_strstr(line, "not available") != NULL)) { - NS_DELETEXPCOM(thisEntry); - if (cr) - line = eol+2; - else - line = eol+1; - continue; - } - - PRInt32 len = PL_strlen(line); - - // check first character of ls -l output - // For example: "dr-x--x--x" is what we're starting with. - // sanity check for dir permission bits - if ((line[0] == 'D' || line[0] == 'd') && ls_lCandidate(line)) { - /* it's a directory */ - thisEntry->mType = Dir; - thisEntry->mSupressSize = PR_TRUE; - } else if ((line[0] == 'L' || line[0] == 'l') && ls_lCandidate(line)) { - /** - * Dir Links are not displayed properly - * we need a more robust implementation - */ - thisEntry->mType = Link; - thisEntry->mSupressSize = PR_TRUE; - - /* strip off " -> pathname" */ - PRInt32 i; - for (i = len - 1; (i > 3) && (!nsCRT::IsAsciiSpace(line[i]) - || (line[i-1] != '>') - || (line[i-2] != '-') - || (line[i-3] != ' ')); i--) - ; /* null body */ - if (i > 3) { - line[i-3] = '\0'; - len = i - 3; - } - } - - rv = ParseLSLine(line, thisEntry); - if ( NS_FAILED(rv) || (thisEntry->mName.Equals("..")) || (thisEntry->mName.Equals(".")) ) { - NS_DELETEXPCOM(thisEntry); - if (cr) - line = eol+2; - else - line = eol+1; - continue; - } - - break; // END UNIX, PETER_LEWIS, MACHTEN - } - - case NCSA: - case TCPC: - { - escName = nsEscape(line, url_Path); - thisEntry->mName.Adopt(escName); - - if (thisEntry->mName.Last() == '/') { - thisEntry->mType = Dir; - thisEntry->mName.Truncate(thisEntry->mName.Length()-1); - } - - break; // END NCSA, TCPC - } - - case CMS: - { - escName = nsEscape(line, url_Path); - thisEntry->mName.Adopt(escName); - break; // END CMS - } - case NT: - { - // don't bother w/ these lines. - if (!PL_strncmp(line, "total ", 6) - || !PL_strncmp(line, "ls: total", 9) - || (PL_strstr(line, "Permission denied") != NULL) - || (PL_strstr(line, "not available") != NULL)) { - NS_DELETEXPCOM(thisEntry); - if (cr) - line = eol+2; - else - line = eol+1; - continue; - } - - char *date, *size_s, *name; - - if (PL_strlen(line) > 37) { - date = line; - line[17] = '\0'; - size_s = &line[18]; - line[38] = '\0'; - name = &line[39]; - - if (PL_strstr(size_s, "")) { - thisEntry->mType = Dir; - } else { - nsCAutoString size(size_s); - size.StripWhitespace(); - thisEntry->mContentLen = atol(size.get()); - } - - ConvertDOSDate(date, thisEntry->mMDTM); - - escName = nsEscape(name, url_Path); - thisEntry->mName.Adopt(escName); - } else { - escName = nsEscape(line, url_Path); - thisEntry->mName.Adopt(escName); - } - break; // END NT - } - case EPLF: - { - - int flagcwd = 0; - int when = 0; - int flagsize = 0; - unsigned long size = 0; - PRBool processing = PR_TRUE; - while (*line && processing) - switch (*line) { - case '\t': - { - if (flagcwd) { - thisEntry->mType = Dir; - thisEntry->mContentLen = 0; - } else { - thisEntry->mType = File; - thisEntry->mContentLen = size; - } - - escName = nsEscape(line+1, url_Path); - thisEntry->mName.Adopt(escName); - thisEntry->mSupressSize = !flagsize; - - // Mutiply what the last modification date to get usecs. - PRInt64 usecs = LL_Zero(); - PRInt64 seconds = LL_Zero(); - PRInt64 multiplier = LL_Zero(); - LL_I2L(seconds, when); - LL_I2L(multiplier, PR_USEC_PER_SEC); - LL_MUL(usecs, seconds, multiplier); - PR_ExplodeTime(usecs, PR_LocalTimeParameters, &thisEntry->mMDTM); - - processing = PR_FALSE; - } - break; - case 's': - flagsize = 1; - size = 0; - while (*++line && nsCRT::IsAsciiDigit(*line)) - size = size * 10 + (*line - '0'); - break; - case 'm': - while (*++line && nsCRT::IsAsciiDigit(*line)) - when = when * 10 + (*line - '0'); - break; - case '/': - flagcwd = 1; - default: - while (*line) if (*line++ == ',') break; - } - break; //END EPLF - } - - case OS_2: - { - if (!PL_strncmp(line, "total ", 6) - || (PL_strstr(line, "not authorized") != NULL) - || (PL_strstr(line, "Path not found") != NULL) - || (PL_strstr(line, "No Files") != NULL)) { - NS_DELETEXPCOM(thisEntry); - if (cr) - line = eol+2; - else - line = eol+1; - continue; - } - - char *name; - nsCAutoString str; - - if (PL_strstr(line, "DIR")) { - thisEntry->mType = Dir; - thisEntry->mSupressSize = PR_TRUE; - } + if (cr) + line = eol+2; else - thisEntry->mType = File; - - PRInt32 error; - line[18] = '\0'; - str = line; - str.StripWhitespace(); - thisEntry->mContentLen = str.ToInteger(&error, 10); - - InitPRExplodedTime(thisEntry->mMDTM); - line[37] = '\0'; - str = &line[35]; - thisEntry->mMDTM.tm_month = str.ToInteger(&error, 10) - 1; - - line[40] = '\0'; - str = &line[38]; - thisEntry->mMDTM.tm_mday = str.ToInteger(&error, 10); - - line[43] = '\0'; - str = &line[41]; - thisEntry->mMDTM.tm_year = str.ToInteger(&error, 10); - - line[48] = '\0'; - str = &line[46]; - thisEntry->mMDTM.tm_hour = str.ToInteger(&error, 10); - - line[51] = '\0'; - str = &line[49]; - thisEntry->mMDTM.tm_min = str.ToInteger(&error, 10); - - name = &line[53]; - escName = nsEscape(name, url_Path); - thisEntry->mName.Adopt(escName); - - break; - } - - default: - { - escName = nsEscape(line, url_Path); - thisEntry->mName.Adopt(escName); - break; // END default (catches GENERIC, DCTS) + line = eol+1; + + continue; } - } // end switch (mServerType) - // blast the index entry into the indexFormat buffer as a 201: line. aString.Append("201: "); // FILENAME - aString.Append(thisEntry->mName); - if (thisEntry->mType == Dir) + + aString.Append(nsDependentCString(result.fe_fname, result.fe_fnlen)); + + /* Bug 143770 + if (type == 'd') aString.Append('/'); + */ aString.Append(' '); // CONTENT LENGTH - if (!thisEntry->mSupressSize) { - aString.AppendInt(thisEntry->mContentLen); - } else { - aString.Append('0'); + + if (type == 'f') + { + for (int i = 0; i < sizeof(result.fe_size); i++) + { + if (result.fe_size[i] != '\0') + aString.Append((const char*)&result.fe_size[i], 1); + } + + aString.Append(' '); } - aString.Append(' '); + else + aString.Append("0 "); + // MODIFIED DATE char buffer[256] = ""; @@ -1075,40 +377,27 @@ nsFTPDirListingConv::DigestBufferLines(char *aBuffer, nsCString &aString) { // the application/http-index-format specs // viewers of such a format can then reformat this into the // current locale (or anything else they choose) - - // make sure we don't have a null time struct - if ((thisEntry->mMDTM.tm_month + thisEntry->mMDTM.tm_mday + - thisEntry->mMDTM.tm_year + thisEntry->mMDTM.tm_hour + - thisEntry->mMDTM.tm_min) != nsnull) { - PR_FormatTimeUSEnglish(buffer, sizeof(buffer), - "%a, %d %b %Y %H:%M:%S", &thisEntry->mMDTM ); - } + PR_FormatTimeUSEnglish(buffer, sizeof(buffer), + "%a, %d %b %Y %H:%M:%S", &result.fe_time ); char *escapedDate = nsEscape(buffer, url_Path); - aString.Append(escapedDate); nsMemory::Free(escapedDate); aString.Append(' '); - // ENTRY TYPE - switch (thisEntry->mType) { - case Dir: + if (type == 'd') aString.Append("DIRECTORY"); - break; - case Link: + else if (type == 'l') aString.Append("SYMBOLIC-LINK"); - break; - default: + else aString.Append("FILE"); - } + aString.Append(' '); aString.Append(char(nsCRT::LF)); // complete this line // END 201: - NS_DELETEXPCOM(thisEntry); - if (cr) line = eol+2; else diff --git a/mozilla/netwerk/streamconv/converters/nsFTPDirListingConv.h b/mozilla/netwerk/streamconv/converters/nsFTPDirListingConv.h index 31023209d7f..2d93a06ffe6 100644 --- a/mozilla/netwerk/streamconv/converters/nsFTPDirListingConv.h +++ b/mozilla/netwerk/streamconv/converters/nsFTPDirListingConv.h @@ -53,62 +53,6 @@ } static NS_DEFINE_CID(kFTPDirListingConverterCID, NS_FTPDIRLISTINGCONVERTER_CID); -// The nsFTPDirListingConv stream converter converts a stream of type "text/ftp-dir-SERVER_TYPE" -// (where SERVER_TYPE is one of the following): -// -// SERVER TYPES: -// generic -// unix -// dcts -// ncsa -// peter_lewis -// machten -// cms -// tcpc -// vms -// nt -// eplf -// -// nsFTPDirListingConv converts the raw ascii text directory generated via a FTP -// LIST or NLST command, to the application/http-index-format MIME-type. -// For more info see: http://www.area.com/~roeber/file_format.html - -typedef enum _FTP_Server_Type { - GENERIC, - UNIX, - DCTS, - NCSA, - PETER_LEWIS, - MACHTEN, - CMS, - TCPC, - VMS, - NT, - EPLF, - OS_2, - ERROR_TYPE -} FTP_Server_Type; - -typedef enum _FTPentryType { - Dir, - File, - Link -} FTPentryType; - - -// indexEntry is the data structure used to maintain directory entry information. -class indexEntry { -public: - indexEntry() { mContentLen = 0; mType = File; mSupressSize = PR_FALSE; }; - - nsCString mName; // the file or dir name - FTPentryType mType; - PRInt32 mContentLen; // length of the file - nsCString mContentType; // type of the file - PRExplodedTime mMDTM; // modified time - PRBool mSupressSize; // supress the size info from display -}; - class nsFTPDirListingConv : public nsIStreamConverter { public: // nsISupports methods @@ -128,53 +72,15 @@ public: virtual ~nsFTPDirListingConv(); nsresult Init(); - // For factory creation. - static NS_METHOD - Create(nsISupports *aOuter, REFNSIID aIID, void **aResult) { - nsresult rv; - if (aOuter) - return NS_ERROR_NO_AGGREGATION; - - nsFTPDirListingConv* _s = new nsFTPDirListingConv(); - if (_s == nsnull) - return NS_ERROR_OUT_OF_MEMORY; - NS_ADDREF(_s); - rv = _s->Init(); - if (NS_FAILED(rv)) { - delete _s; - return rv; - } - rv = _s->QueryInterface(aIID, aResult); - NS_RELEASE(_s); - return rv; - } - private: // Get the application/http-index-format headers nsresult GetHeaders(nsACString& str, nsIURI* uri); - - // util parsing methods - PRInt8 MonthNumber(const char *aCStr); - PRBool IsLSDate(char *aCStr); - - // date conversion/parsing methods - PRBool ConvertUNIXDate(char *aCStr, PRExplodedTime& outDate); - PRBool ConvertDOSDate(char *aCStr, PRExplodedTime& outDate); - - // line parsing methods - nsresult ParseLSLine(char *aLine, indexEntry *aEntry); - nsresult ParseVMSLine(char *aLine, indexEntry *aEntry); - - void InitPRExplodedTime(PRExplodedTime& aTime); - PRBool ls_lCandidate(const char *aLine); char* DigestBufferLines(char *aBuffer, nsCString &aString); // member data - FTP_Server_Type mServerType; // what kind of server is the data coming from? nsCAutoString mBuffer; // buffered data. PRBool mSentHeading; // have we sent 100, 101, 200, and 300 lines yet? - nsIStreamListener *mFinalListener; // this guy gets the converted data via his OnDataAvailable() nsIChannel *mPartChannel; // the channel for the given part we're processing. // one channel per part.