diff options
Diffstat (limited to '')
| -rw-r--r-- | WWW/Library/Implementation/HTFile.c | 3327 |
1 files changed, 3327 insertions, 0 deletions
diff --git a/WWW/Library/Implementation/HTFile.c b/WWW/Library/Implementation/HTFile.c new file mode 100644 index 0000000..9074574 --- /dev/null +++ b/WWW/Library/Implementation/HTFile.c @@ -0,0 +1,3327 @@ +/* + * $LynxId: HTFile.c,v 1.151 2018/05/11 23:20:35 tom Exp $ + * + * File Access HTFile.c + * =========== + * + * This is unix-specific code in general, with some VMS bits. + * These are routines for file access used by browsers. + * Development of this module for Unix DIRED_SUPPORT in Lynx + * regrettably has has been conducted in a manner with now + * creates a major impediment for hopes of adapting Lynx to + * a newer version of the library. + * + * History: + * Feb 91 Written Tim Berners-Lee CERN/CN + * Apr 91 vms-vms access included using DECnet syntax + * 26 Jun 92 (JFG) When running over DECnet, suppressed FTP. + * Fixed access bug for relative names on VMS. + * Sep 93 (MD) Access to VMS files allows sharing. + * 15 Nov 93 (MD) Moved HTVMSname to HTVMSUTILS.C + * 27 Dec 93 (FM) FTP now works with VMS hosts. + * FTP path must be Unix-style and cannot include + * the device or top directory. + */ + +#include <HTUtils.h> + +#ifndef VMS +#if defined(DOSPATH) +#undef LONG_LIST +#define LONG_LIST /* Define this for long style unix listings (ls -l), + the actual style is configurable from lynx.cfg */ +#endif +/* #define NO_PARENT_DIR_REFERENCE */ +/* Define this for no parent links */ +#endif /* !VMS */ + +#if defined(DOSPATH) +#define HAVE_READDIR 1 +#define USE_DIRENT +#endif + +#if defined(USE_DOS_DRIVES) +#include <HTDOS.h> +#endif + +#include <HTFile.h> /* Implemented here */ + +#ifdef VMS +#include <stat.h> +#endif /* VMS */ + +#if defined (USE_ZLIB) || defined (USE_BZLIB) +#include <GridText.h> +#endif + +#define MULTI_SUFFIX ".multi" /* Extension for scanning formats */ + +#include <HTParse.h> +#include <HTTCP.h> +#ifndef DECNET +#include <HTFTP.h> +#endif /* !DECNET */ +#include <HTAnchor.h> +#include <HTAtom.h> +#include <HTAAProt.h> +#include <HTFWriter.h> +#include <HTInit.h> +#include <HTBTree.h> +#include <HTAlert.h> +#include <HTCJK.h> +#include <UCDefs.h> +#include <UCMap.h> +#include <UCAux.h> + +#include <LYexit.h> +#include <LYCharSets.h> +#include <LYGlobalDefs.h> +#include <LYStrings.h> +#include <LYUtils.h> + +#ifdef USE_PRETTYSRC +# include <LYPrettySrc.h> +#endif + +#include <LYLeaks.h> + +typedef struct _HTSuffix { + char *suffix; + HTAtom *rep; + HTAtom *encoding; + char *desc; + float quality; +} HTSuffix; + +typedef struct { + struct stat file_info; + char sort_tags; + char file_name[1]; /* on the end of the struct, since its length varies */ +} DIRED; + +#ifndef NGROUPS +#ifdef NGROUPS_MAX +#define NGROUPS NGROUPS_MAX +#else +#define NGROUPS 32 +#endif /* NGROUPS_MAX */ +#endif /* NGROUPS */ + +#ifndef GETGROUPS_T +#define GETGROUPS_T int +#endif + +#include <HTML.h> /* For directory object building */ + +#define PUTC(c) (*target->isa->put_character)(target, c) +#define PUTS(s) (*target->isa->put_string)(target, s) +#define START(e) (*target->isa->start_element)(target, e, 0, 0, -1, 0) +#define END(e) (*target->isa->end_element)(target, e, 0) +#define MAYBE_END(e) if (HTML_dtd.tags[e].contents != SGML_EMPTY) \ + (*target->isa->end_element)(target, e, 0) +#define FREE_TARGET (*target->isa->_free)(target) +#define ABORT_TARGET (*targetClass._abort)(target, NULL); + +struct _HTStructured { + const HTStructuredClass *isa; + /* ... */ +}; + +/* + * Controlling globals. + */ +int HTDirAccess = HT_DIR_OK; + +#ifdef DIRED_SUPPORT +int HTDirReadme = HT_DIR_README_NONE; + +#else +int HTDirReadme = HT_DIR_README_TOP; +#endif /* DIRED_SUPPORT */ + +static const char *HTMountRoot = "/Net/"; /* Where to find mounts */ + +#ifdef VMS +static const char *HTCacheRoot = "/WWW$SCRATCH"; /* Where to cache things */ + +#else +static const char *HTCacheRoot = "/tmp/W3_Cache_"; /* Where to cache things */ +#endif /* VMS */ + +static char s_no_suffix[] = "*"; +static char s_unknown_suffix[] = "*.*"; + +/* + * Suffix registration. + */ +static HTList *HTSuffixes = 0; + +static HTSuffix no_suffix = +{ + s_no_suffix, NULL, NULL, NULL, 1.0 +}; + +static HTSuffix unknown_suffix = +{ + s_unknown_suffix, NULL, NULL, NULL, 1.0 +}; + +/* To free up the suffixes at program exit. + * ---------------------------------------- + */ +#ifdef LY_FIND_LEAKS +static void free_suffixes(void); +#endif + +static char *FindSearch(const char *filename) +{ + char *result = 0; + + if ((result = strchr(filename, '?')) == 0) { + result = strstr(filename, "%3F"); + } + return result; +} + +#ifdef LONG_LIST +static char *FormatStr(char **bufp, + char *start, + const char *entry) +{ + char fmt[512]; + + if (*start) { + sprintf(fmt, "%%%.*ss", (int) sizeof(fmt) - 3, start); + HTSprintf0(bufp, fmt, entry); + } else if (*bufp && !(entry && *entry)) { + **bufp = '\0'; + } else if (entry) { + StrAllocCopy(*bufp, entry); + } + return *bufp; +} + +static char *FormatSize(char **bufp, + char *start, + off_t entry) +{ + char fmt[512]; + + if (*start) { + sprintf(fmt, "%%%.*s" PRI_off_t, + (int) sizeof(fmt) - DigitsOf(start) - 3, start); + + HTSprintf0(bufp, fmt, entry); + } else { + sprintf(fmt, "%" PRI_off_t, CAST_off_t (entry)); + + StrAllocCopy(*bufp, fmt); + } + return *bufp; +} + +static char *FormatNum(char **bufp, + char *start, + int entry) +{ + char fmt[512]; + + if (*start) { + sprintf(fmt, "%%%.*sd", (int) sizeof(fmt) - 3, start); + HTSprintf0(bufp, fmt, entry); + } else { + sprintf(fmt, "%d", entry); + StrAllocCopy(*bufp, fmt); + } + return *bufp; +} + +static void LYListFmtParse(const char *fmtstr, + DIRED * data, + char *file, + HTStructured * target, + char *tail) +{ + char c; + char *s; + char *end; + char *start; + char *str = NULL; + char *buf = NULL; + char tmp[LY_MAXPATH]; + char type; + +#ifndef NOUSERS + const char *name; +#endif + time_t now; + char *datestr; + +#ifdef S_IFLNK + int len; +#endif +#define SEC_PER_YEAR (60 * 60 * 24 * 365) + +#ifdef _WINDOWS /* 1998/01/06 (Tue) 21:20:53 */ + static const char *pbits[] = + { + "---", "--x", "-w-", "-wx", + "r--", "r-x", "rw-", "rwx", + 0}; + +#define PBIT(a, n, s) pbits[((a) >> (n)) & 0x7] + +#else + static const char *pbits[] = + {"---", "--x", "-w-", "-wx", + "r--", "r-x", "rw-", "rwx", 0}; + static const char *psbits[] = + {"--S", "--s", "-wS", "-ws", + "r-S", "r-s", "rwS", "rws", 0}; + +#define PBIT(a, n, s) (s) ? psbits[((a) >> (n)) & 0x7] : \ + pbits[((a) >> (n)) & 0x7] +#endif +#if defined(S_ISVTX) && !defined(_WINDOWS) + static const char *ptbits[] = + {"--T", "--t", "-wT", "-wt", + "r-T", "r-t", "rwT", "rwt", 0}; + +#define PTBIT(a, s) (s) ? ptbits[(a) & 0x7] : pbits[(a) & 0x7] +#else +#define PTBIT(a, s) PBIT(a, 0, 0) +#endif + + if (data->file_info.st_mode == 0) + fmtstr = " %a"; /* can't stat so just do anchor */ + + StrAllocCopy(str, fmtstr); + s = str; + end = str + strlen(str); + while (*s) { + start = s; + while (*s) { + if (*s == '%') { + if (*(s + 1) == '%') /* literal % */ + s++; + else + break; + } + s++; + } + /* s is positioned either at a % or at \0 */ + *s = '\0'; + if (s > start) { /* some literal chars. */ + PUTS(start); + } + if (s == end) + break; + start = ++s; + while (isdigit(UCH(*s)) || *s == '.' || *s == '-' || *s == ' ' || + *s == '#' || *s == '+' || *s == '\'') + s++; + c = *s; /* the format char. or \0 */ + *s = '\0'; + + switch (c) { + case '\0': + PUTS(start); + continue; + + case 'A': + case 'a': /* anchor */ + HTDirEntry(target, tail, data->file_name); + FormatStr(&buf, start, data->file_name); + PUTS(buf); + END(HTML_A); + *buf = '\0'; +#ifdef S_IFLNK + if (c != 'A' && S_ISLNK(data->file_info.st_mode) && + (len = (int) readlink(file, tmp, sizeof(tmp) - 1)) >= 0) { + PUTS(" -> "); + tmp[len] = '\0'; + PUTS(tmp); + } +#endif + break; + + case 'T': /* MIME type */ + case 't': /* MIME type description */ + if (S_ISDIR(data->file_info.st_mode)) { + if (c != 'T') { + FormatStr(&buf, start, ENTRY_IS_DIRECTORY); + } else { + FormatStr(&buf, start, ""); + } + } else { + const char *cp2; + HTFormat format; + + format = HTFileFormat(file, NULL, &cp2); + + if (c != 'T') { + if (cp2 == NULL) { + if (!StrNCmp(HTAtom_name(format), + "application", 11)) { + cp2 = HTAtom_name(format) + 12; + if (!StrNCmp(cp2, "x-", 2)) + cp2 += 2; + } else { + cp2 = HTAtom_name(format); + } + } + FormatStr(&buf, start, cp2); + } else { + FormatStr(&buf, start, HTAtom_name(format)); + } + } + break; + + case 'd': /* date */ + now = time(0); + datestr = ctime(&data->file_info.st_mtime); + if ((now - data->file_info.st_mtime) < SEC_PER_YEAR / 2) + /* + * MMM DD HH:MM + */ + sprintf(tmp, "%.12s", datestr + 4); + else + /* + * MMM DD YYYY + */ + sprintf(tmp, "%.7s %.4s ", datestr + 4, + datestr + 20); + FormatStr(&buf, start, tmp); + break; + + case 's': /* size in bytes */ + FormatSize(&buf, start, data->file_info.st_size); + break; + + case 'K': /* size in Kilobytes but not for directories */ + if (S_ISDIR(data->file_info.st_mode)) { + FormatStr(&buf, start, ""); + StrAllocCat(buf, " "); + break; + } + /* FALL THROUGH */ + case 'k': /* size in Kilobytes */ + FormatSize(&buf, start, ((data->file_info.st_size + 1023) / 1024)); + StrAllocCat(buf, "K"); + break; + + case 'p': /* unix-style permission bits */ + switch (data->file_info.st_mode & S_IFMT) { +#if defined(_MSC_VER) && defined(_S_IFIFO) + case _S_IFIFO: + type = 'p'; + break; +#else + case S_IFIFO: + type = 'p'; + break; +#endif + case S_IFCHR: + type = 'c'; + break; + case S_IFDIR: + type = 'd'; + break; + case S_IFREG: + type = '-'; + break; +#ifdef S_IFBLK + case S_IFBLK: + type = 'b'; + break; +#endif +#ifdef S_IFLNK + case S_IFLNK: + type = 'l'; + break; +#endif +#ifdef S_IFSOCK +# ifdef S_IFIFO /* some older machines (e.g., apollo) have a conflict */ +# if S_IFIFO != S_IFSOCK + case S_IFSOCK: + type = 's'; + break; +# endif +# else + case S_IFSOCK: + type = 's'; + break; +# endif +#endif /* S_IFSOCK */ + default: + type = '?'; + break; + } +#ifdef _WINDOWS + sprintf(tmp, "%c%s", type, + PBIT(data->file_info.st_mode, 6, data->file_info.st_mode & S_IRWXU)); +#else + sprintf(tmp, "%c%s%s%s", type, + PBIT(data->file_info.st_mode, 6, data->file_info.st_mode & S_ISUID), + PBIT(data->file_info.st_mode, 3, data->file_info.st_mode & S_ISGID), + PTBIT(data->file_info.st_mode, data->file_info.st_mode & S_ISVTX)); +#endif + FormatStr(&buf, start, tmp); + break; + + case 'o': /* owner */ +#ifndef NOUSERS + name = HTAA_UidToName((int) data->file_info.st_uid); + if (*name) { + FormatStr(&buf, start, name); + } else { + FormatNum(&buf, start, (int) data->file_info.st_uid); + } +#endif + break; + + case 'g': /* group */ +#ifndef NOUSERS + name = HTAA_GidToName((int) data->file_info.st_gid); + if (*name) { + FormatStr(&buf, start, name); + } else { + FormatNum(&buf, start, (int) data->file_info.st_gid); + } +#endif + break; + + case 'l': /* link count */ + FormatNum(&buf, start, (int) data->file_info.st_nlink); + break; + + case '%': /* literal % with flags/width */ + FormatStr(&buf, start, "%"); + break; + + default: + fprintf(stderr, + "Unknown format character `%c' in list format\n", c); + break; + } + if (buf) + PUTS(buf); + + s++; + } + FREE(buf); + PUTC('\n'); + FREE(str); +} +#endif /* LONG_LIST */ + +/* Define the representation associated with a file suffix. + * -------------------------------------------------------- + * + * Calling this with suffix set to "*" will set the default + * representation. + * Calling this with suffix set to "*.*" will set the default + * representation for unknown suffix files which contain a ".". + * + * The encoding parameter can give a trivial (8bit, 7bit, binary) + * or real (gzip, compress) encoding. + * + * If filename suffix is already defined with the same encoding + * its previous definition is overridden. + */ +void HTSetSuffix5(const char *suffix, + const char *representation, + const char *encoding, + const char *desc, + double value) +{ + HTSuffix *suff; + BOOL trivial_enc = (BOOL) IsUnityEncStr(encoding); + + if (strcmp(suffix, s_no_suffix) == 0) + suff = &no_suffix; + else if (strcmp(suffix, s_unknown_suffix) == 0) + suff = &unknown_suffix; + else { + HTList *cur = HTSuffixes; + + while (NULL != (suff = (HTSuffix *) HTList_nextObject(cur))) { + if (suff->suffix && 0 == strcmp(suff->suffix, suffix) && + ((trivial_enc && IsUnityEnc(suff->encoding)) || + (!trivial_enc && !IsUnityEnc(suff->encoding) && + strcmp(encoding, HTAtom_name(suff->encoding)) == 0))) + break; + } + if (!suff) { /* Not found -- create a new node */ + suff = typecalloc(HTSuffix); + if (suff == NULL) + outofmem(__FILE__, "HTSetSuffix"); + + if (!HTSuffixes) { + HTSuffixes = HTList_new(); +#ifdef LY_FIND_LEAKS + atexit(free_suffixes); +#endif + } + + HTList_addObject(HTSuffixes, suff); + + StrAllocCopy(suff->suffix, suffix); + } + } + + if (representation) + suff->rep = HTAtom_for(representation); + + /* + * Memory leak fixed. + * 05-28-94 Lynx 2-3-1 Garrett Arch Blythe + * Invariant code removed. + */ + suff->encoding = HTAtom_for(encoding); + + StrAllocCopy(suff->desc, desc); + + suff->quality = (float) value; +} + +#ifdef LY_FIND_LEAKS +/* + * Purpose: Free all added suffixes. + * Arguments: void + * Return Value: void + * Remarks/Portability/Dependencies/Restrictions: + * To be used at program exit. + * Revision History: + * 05-28-94 created Lynx 2-3-1 Garrett Arch Blythe + */ +static void free_suffixes(void) +{ + HTSuffix *suff = NULL; + + /* + * Loop through all suffixes. + */ + while (!HTList_isEmpty(HTSuffixes)) { + /* + * Free off each item and its members if need be. + */ + suff = (HTSuffix *) HTList_removeLastObject(HTSuffixes); + FREE(suff->suffix); + FREE(suff->desc); + FREE(suff); + } + /* + * Free off the list itself. + */ + HTList_delete(HTSuffixes); + HTSuffixes = NULL; +} +#endif /* LY_FIND_LEAKS */ + +/* Make the cache file name for a W3 document. + * ------------------------------------------- + * Make up a suitable name for saving the node in + * + * E.g. /tmp/WWW_Cache_news/1234@cernvax.cern.ch + * /tmp/WWW_Cache_http/crnvmc/FIND/xx.xxx.xx + * + * On exit: + * Returns a malloc'ed string which must be freed by the caller. + */ +char *HTCacheFileName(const char *name) +{ + char *acc_method = HTParse(name, "", PARSE_ACCESS); + char *host = HTParse(name, "", PARSE_HOST); + char *path = HTParse(name, "", PARSE_PATH + PARSE_PUNCTUATION); + char *result = NULL; + + HTSprintf0(&result, "%s/WWW/%s/%s%s", HTCacheRoot, acc_method, host, path); + + FREE(path); + FREE(acc_method); + FREE(host); + return result; +} + +/* Open a file for write, creating the path. + * ----------------------------------------- + */ +#ifdef NOT_IMPLEMENTED +static int HTCreatePath(const char *path) +{ + return -1; +} +#endif /* NOT_IMPLEMENTED */ + +/* Convert filename from URL-path syntax to local path format + * ---------------------------------------------------------- + * Input name is assumed to be the URL-path of a local file + * URL, i.e. what comes after the "file://localhost". + * '#'-fragments to be treated as such must already be stripped. + * If expand_all is FALSE, unescape only escaped '/'. - kw + * + * On exit: + * Returns a malloc'ed string which must be freed by the caller. + */ +char *HTURLPath_toFile(const char *name, + int expand_all, + int is_remote GCC_UNUSED) +{ + char *path = NULL; + char *result = NULL; + + StrAllocCopy(path, name); + if (expand_all) + HTUnEscape(path); /* Interpret all % signs */ + else + HTUnEscapeSome(path, "/"); /* Interpret % signs for path delims */ + + CTRACE((tfp, "URLPath `%s' means path `%s'\n", name, path)); +#if defined(USE_DOS_DRIVES) + StrAllocCopy(result, is_remote ? path : HTDOS_name(path)); +#else + StrAllocCopy(result, path); +#endif + + FREE(path); + + return result; +} +/* Convert filenames between local and WWW formats. + * ------------------------------------------------ + * Make up a suitable name for saving the node in + * + * E.g. $(HOME)/WWW/news/1234@cernvax.cern.ch + * $(HOME)/WWW/http/crnvmc/FIND/xx.xxx.xx + * + * On exit: + * Returns a malloc'ed string which must be freed by the caller. + */ +/* NOTE: Don't use this function if you know that the input is a URL path + rather than a full URL, use HTURLPath_toFile instead. Otherwise + this function will return the wrong thing for some unusual + paths (like ones containing "//", possibly escaped). - kw +*/ +char *HTnameOfFile_WWW(const char *name, + int WWW_prefix, + int expand_all) +{ + char *acc_method = HTParse(name, "", PARSE_ACCESS); + char *host = HTParse(name, "", PARSE_HOST); + char *path = HTParse(name, "", PARSE_PATH + PARSE_PUNCTUATION); + const char *home; + char *result = NULL; + + if (expand_all) { + HTUnEscape(path); /* Interpret all % signs */ + } else + HTUnEscapeSome(path, "/"); /* Interpret % signs for path delims */ + + if (0 == strcmp(acc_method, "file") /* local file */ + ||!*acc_method) { /* implicitly local? */ + if ((0 == strcasecomp(host, HTHostName())) || + (0 == strcasecomp(host, "localhost")) || !*host) { + CTRACE((tfp, "Node `%s' means path `%s'\n", name, path)); + StrAllocCopy(result, HTSYS_name(path)); + } else if (WWW_prefix) { + HTSprintf0(&result, "%s%s%s", "/Net/", host, path); + CTRACE((tfp, "Node `%s' means file `%s'\n", name, result)); + } else { + StrAllocCopy(result, path); + } + } else if (WWW_prefix) { /* other access */ +#ifdef VMS + if ((home = LYGetEnv("HOME")) == NULL) + home = HTCacheRoot; + else + home = HTVMS_wwwName(home); +#else +#if defined(_WINDOWS) /* 1997/10/16 (Thu) 20:42:51 */ + home = Home_Dir(); +#else + home = LYGetEnv("HOME"); +#endif + if (home == NULL) + home = "/tmp"; +#endif /* VMS */ + HTSprintf0(&result, "%s/WWW/%s/%s%s", home, acc_method, host, path); + } else { + StrAllocCopy(result, path); + } + + FREE(host); + FREE(path); + FREE(acc_method); + + CTRACE((tfp, "HTnameOfFile_WWW(%s,%d,%d) = %s\n", + name, WWW_prefix, expand_all, result)); + + return result; +} + +/* Make a WWW name from a full local path name. + * -------------------------------------------- + * + * Bugs: + * At present, only the names of two network root nodes are hand-coded + * in and valid for the NeXT only. This should be configurable in + * the general case. + */ +char *WWW_nameOfFile(const char *name) +{ + char *result = NULL; + +#ifdef NeXT + if (0 == StrNCmp("/private/Net/", name, 13)) { + HTSprintf0(&result, "%s//%s", STR_FILE_URL, name + 13); + } else +#endif /* NeXT */ + if (0 == StrNCmp(HTMountRoot, name, 5)) { + HTSprintf0(&result, "%s//%s", STR_FILE_URL, name + 5); + } else { + HTSprintf0(&result, "%s//%s%s", STR_FILE_URL, HTHostName(), name); + } + CTRACE((tfp, "File `%s'\n\tmeans node `%s'\n", name, result)); + return result; +} + +/* Determine a suitable suffix, given the representation. + * ------------------------------------------------------ + * + * On entry, + * rep is the atomized MIME style representation + * enc is an encoding, trivial (8bit, binary, etc.) or gzip etc. + * + * On exit: + * Returns a pointer to a suitable suffix string if one has been + * found, else "". + */ +const char *HTFileSuffix(HTAtom *rep, + const char *enc) +{ + HTSuffix *suff; + +#ifdef FNAMES_8_3 + HTSuffix *first_found = NULL; +#endif + BOOL trivial_enc; + int n; + int i; + +#define NO_INIT /* don't init anymore since I do it in Lynx at startup */ +#ifndef NO_INIT + if (!HTSuffixes) + HTFileInit(); +#endif /* !NO_INIT */ + + trivial_enc = (BOOL) IsUnityEncStr(enc); + n = HTList_count(HTSuffixes); + for (i = 0; i < n; i++) { + suff = (HTSuffix *) HTList_objectAt(HTSuffixes, i); + if (suff->rep == rep && +#if defined(VMS) || defined(FNAMES_8_3) + /* Don't return a suffix whose first char is a dot, and which + has more dots or asterisks after that, for + these systems - kw */ + (!suff->suffix || !suff->suffix[0] || suff->suffix[0] != '.' || + (StrChr(suff->suffix + 1, '.') == NULL && + StrChr(suff->suffix + 1, '*') == NULL)) && +#endif + ((trivial_enc && IsUnityEnc(suff->encoding)) || + (!trivial_enc && !IsUnityEnc(suff->encoding) && + strcmp(enc, HTAtom_name(suff->encoding)) == 0))) { +#ifdef FNAMES_8_3 + if (suff->suffix && (strlen(suff->suffix) <= 4)) { + /* + * If length of suffix (including dot) is 4 or smaller, return + * this one even if we found a longer one earlier - kw + */ + return suff->suffix; + } else if (!first_found) { + first_found = suff; /* remember this one */ + } +#else + return suff->suffix; /* OK -- found */ +#endif + } + } +#ifdef FNAMES_8_3 + if (first_found) + return first_found->suffix; +#endif + return ""; /* Dunno */ +} + +/* + * Trim version from VMS filenames to avoid confusing comparisons. + */ +#ifdef VMS +static const char *VMS_trim_version(const char *filename) +{ + const char *result = filename; + const char *version = StrChr(filename, ';'); + + if (version != 0) { + static char *stripped; + + StrAllocCopy(stripped, filename); + stripped[version - filename] = '\0'; + result = (const char *) stripped; + } + return result; +} +#define VMS_DEL_VERSION(name) name = VMS_trim_version(name) +#else +#define VMS_DEL_VERSION(name) /* nothing */ +#endif + +/* Determine file format from file name. + * ------------------------------------- + * + * This version will return the representation and also set + * a variable for the encoding. + * + * Encoding may be a unity encoding (binary, 8bit, etc.) or + * a content-coding like gzip, compress. + * + * It will handle for example x.txt, x.txt,Z, x.Z + */ +HTFormat HTFileFormat(const char *filename, + HTAtom **pencoding, + const char **pdesc) +{ + HTSuffix *suff; + int n; + int i; + int lf; + char *search; + + VMS_DEL_VERSION(filename); + + if ((search = FindSearch(filename)) != 0) { + char *newname = NULL; + HTFormat result; + + StrAllocCopy(newname, filename); + newname[((const char *) search) - filename] = '\0'; + result = HTFileFormat(newname, pencoding, pdesc); + free(newname); + return result; + } + + if (pencoding) + *pencoding = NULL; + if (pdesc) + *pdesc = NULL; + if (LYforce_HTML_mode) { + if (pencoding) + *pencoding = WWW_ENC_8BIT; + return WWW_HTML; + } +#ifndef NO_INIT + if (!HTSuffixes) + HTFileInit(); +#endif /* !NO_INIT */ + lf = (int) strlen(filename); + n = HTList_count(HTSuffixes); + for (i = 0; i < n; i++) { + int ls; + + suff = (HTSuffix *) HTList_objectAt(HTSuffixes, i); + ls = (int) strlen(suff->suffix); + if ((ls <= lf) && 0 == strcasecomp(suff->suffix, filename + lf - ls)) { + int j; + + if (pencoding) + *pencoding = suff->encoding; + if (pdesc) + *pdesc = suff->desc; + if (suff->rep) { + return suff->rep; /* OK -- found */ + } + for (j = 0; j < n; j++) { /* Got encoding, need representation */ + int ls2; + + suff = (HTSuffix *) HTList_objectAt(HTSuffixes, j); + ls2 = (int) strlen(suff->suffix); + if ((ls + ls2 <= lf) && + !strncasecomp(suff->suffix, + filename + lf - ls - ls2, ls2)) { + if (suff->rep) { + if (pdesc && !(*pdesc)) + *pdesc = suff->desc; + if (pencoding && IsUnityEnc(*pencoding) && + *pencoding != WWW_ENC_7BIT && + !IsUnityEnc(suff->encoding)) + *pencoding = suff->encoding; + return suff->rep; + } + } + } + + } + } + + /* defaults tree */ + + suff = (StrChr(filename, '.') + ? (unknown_suffix.rep + ? &unknown_suffix + : &no_suffix) + : &no_suffix); + + /* + * Set default encoding unless found with suffix already. + */ + if (pencoding && !*pencoding) { + *pencoding = (suff->encoding + ? suff->encoding + : HTAtom_for("binary")); + } + return suff->rep ? suff->rep : WWW_BINARY; +} + +/* Revise the file format in relation to the Lynx charset. - FM + * ------------------------------------------------------- + * + * This checks the format associated with an anchor for + * an extended MIME Content-Type, and if a charset is + * indicated, sets Lynx up for proper handling in relation + * to the currently selected character set. - FM + */ +HTFormat HTCharsetFormat(HTFormat format, + HTParentAnchor *anchor, + int default_LYhndl) +{ + char *cp = NULL, *cp1, *cp2, *cp3 = NULL, *cp4; + BOOL chartrans_ok = FALSE; + int chndl = -1; + const char *format_name = format->name; + + FREE(anchor->charset); + if (format_name == 0) + format_name = ""; + StrAllocCopy(cp, format_name); + LYLowerCase(cp); + if (((cp1 = StrChr(cp, ';')) != NULL) && + (cp2 = strstr(cp1, "charset")) != NULL) { + CTRACE((tfp, "HTCharsetFormat: Extended MIME Content-Type is %s\n", + format_name)); + cp2 += 7; + while (*cp2 == ' ' || *cp2 == '=') + cp2++; + StrAllocCopy(cp3, cp2); /* copy to mutilate more */ + for (cp4 = cp3; (*cp4 != '\0' && *cp4 != '"' && + *cp4 != ';' && *cp4 != ':' && + !WHITE(*cp4)); cp4++) { + ; /* do nothing */ + } + *cp4 = '\0'; + cp4 = cp3; + chndl = UCGetLYhndl_byMIME(cp3); + if (UCCanTranslateFromTo(chndl, current_char_set)) { + chartrans_ok = YES; + *cp1 = '\0'; + format = HTAtom_for(cp); + StrAllocCopy(anchor->charset, cp4); + HTAnchor_setUCInfoStage(anchor, chndl, + UCT_STAGE_MIME, + UCT_SETBY_MIME); + } else if (chndl < 0) { + /* + * Got something but we don't recognize it. + */ + chndl = UCLYhndl_for_unrec; + if (chndl < 0) + /* + * UCLYhndl_for_unrec not defined :-( fallback to + * UCLYhndl_for_unspec which always valid. + */ + chndl = UCLYhndl_for_unspec; /* always >= 0 */ + if (UCCanTranslateFromTo(chndl, current_char_set)) { + chartrans_ok = YES; + HTAnchor_setUCInfoStage(anchor, chndl, + UCT_STAGE_MIME, + UCT_SETBY_DEFAULT); + } + } + if (chartrans_ok) { + LYUCcharset *p_in = HTAnchor_getUCInfoStage(anchor, + UCT_STAGE_MIME); + LYUCcharset *p_out = HTAnchor_setUCInfoStage(anchor, + current_char_set, + UCT_STAGE_HTEXT, + UCT_SETBY_DEFAULT); + + if (!p_out) { + /* + * Try again. + */ + p_out = HTAnchor_getUCInfoStage(anchor, UCT_STAGE_HTEXT); + } + if (!strcmp(p_in->MIMEname, "x-transparent")) { + HTPassEightBitRaw = TRUE; + HTAnchor_setUCInfoStage(anchor, + HTAnchor_getUCLYhndl(anchor, + UCT_STAGE_HTEXT), + UCT_STAGE_MIME, + UCT_SETBY_DEFAULT); + } + if (!strcmp(p_out->MIMEname, "x-transparent")) { + HTPassEightBitRaw = TRUE; + HTAnchor_setUCInfoStage(anchor, + HTAnchor_getUCLYhndl(anchor, + UCT_STAGE_MIME), + UCT_STAGE_HTEXT, + UCT_SETBY_DEFAULT); + } + if (p_in->enc != UCT_ENC_CJK) { + HTCJK = NOCJK; + if (!(p_in->codepoints & + UCT_CP_SUBSETOF_LAT1) && + chndl == current_char_set) { + HTPassEightBitRaw = TRUE; + } + } else if (p_out->enc == UCT_ENC_CJK) { + Set_HTCJK(p_in->MIMEname, p_out->MIMEname); + } + } else { + /* + * Cannot translate. If according to some heuristic the given + * charset and the current display character both are likely to be + * like ISO-8859 in structure, pretend we have some kind of match. + */ + BOOL given_is_8859 = (BOOL) (!StrNCmp(cp4, "iso-8859-", 9) && + isdigit(UCH(cp4[9]))); + BOOL given_is_8859like = (BOOL) (given_is_8859 || + !StrNCmp(cp4, "windows-", 8) || + !StrNCmp(cp4, "cp12", 4) || + !StrNCmp(cp4, "cp-12", 5)); + BOOL given_and_display_8859like = (BOOL) (given_is_8859like && + (strstr(LYchar_set_names[current_char_set], + "ISO-8859") || + strstr(LYchar_set_names[current_char_set], + "windows-"))); + + if (given_and_display_8859like) { + *cp1 = '\0'; + format = HTAtom_for(cp); + } + if (given_is_8859) { + cp1 = &cp4[10]; + while (*cp1 && + isdigit(UCH(*cp1))) + cp1++; + *cp1 = '\0'; + } + if (given_and_display_8859like) { + StrAllocCopy(anchor->charset, cp4); + HTPassEightBitRaw = TRUE; + } + HTAlert(*cp4 ? cp4 : anchor->charset); + } + FREE(cp3); + } else if (cp1 != NULL) { + /* + * No charset parameter is present. Ignore all other parameters, as we + * do when charset is present. - FM + */ + *cp1 = '\0'; + format = HTAtom_for(cp); + } + FREE(cp); + + /* + * Set up defaults, if needed. - FM + */ + if (!chartrans_ok && !anchor->charset && default_LYhndl >= 0) { + HTAnchor_setUCInfoStage(anchor, default_LYhndl, + UCT_STAGE_MIME, + UCT_SETBY_DEFAULT); + } + HTAnchor_copyUCInfoStage(anchor, + UCT_STAGE_PARSER, + UCT_STAGE_MIME, + -1); + + return format; +} + +/* Get various pieces of meta info from file name. + * ----------------------------------------------- + * + * LYGetFileInfo fills in information that can be determined without + * an actual (new) access to the filesystem, based on current suffix + * and character set configuration. If the file has been loaded and + * parsed before (with the same URL generated here!) and the anchor + * is still around, some results may be influenced by that (in + * particular, charset info from a META tag - this is not actually + * tested!). + * The caller should not keep pointers to the returned objects around + * for too long, the valid lifetimes vary. In particular, the returned + * charset string should be copied if necessary. If return of the + * file_anchor is requested, that one can be used to retrieve + * additional bits of info that are stored in the anchor object and + * are not covered here; as usual, don't keep pointers to the + * file_anchor longer than necessary since the object may disappear + * through HTuncache_current_document or at the next document load. + * - kw + */ +void LYGetFileInfo(const char *filename, + HTParentAnchor **pfile_anchor, + HTFormat *pformat, + HTAtom **pencoding, + const char **pdesc, + const char **pcharset, + int *pfile_cs) +{ + char *Afn; + char *Aname = NULL; + HTFormat format; + HTAtom *myEnc = NULL; + HTParentAnchor *file_anchor; + const char *file_csname; + int file_cs; + + /* + * Convert filename to URL. Note that it is always supposed to be a + * filename, not maybe-filename-maybe-URL, so we don't use + * LYFillLocalFileURL and LYEnsureAbsoluteURL. - kw + */ + Afn = HTEscape(filename, URL_PATH); + LYLocalFileToURL(&Aname, Afn); + file_anchor = HTAnchor_findSimpleAddress(Aname); + + format = HTFileFormat(filename, &myEnc, pdesc); + format = HTCharsetFormat(format, file_anchor, UCLYhndl_HTFile_for_unspec); + file_cs = HTAnchor_getUCLYhndl(file_anchor, UCT_STAGE_MIME); + file_csname = file_anchor->charset; + if (!file_csname) { + if (file_cs >= 0) + file_csname = LYCharSet_UC[file_cs].MIMEname; + else + file_csname = "display character set"; + } + CTRACE((tfp, "GetFileInfo: '%s' is a%s %s %s file, charset=%s (%d).\n", + filename, + ((myEnc && *HTAtom_name(myEnc) == '8') ? "n" : myEnc ? "" : + *HTAtom_name(format) == 'a' ? "n" : ""), + myEnc ? HTAtom_name(myEnc) : "", + HTAtom_name(format), + file_csname, + file_cs)); + FREE(Afn); + FREE(Aname); + if (pfile_anchor) + *pfile_anchor = file_anchor; + if (pformat) + *pformat = format; + if (pencoding) + *pencoding = myEnc; + if (pcharset) + *pcharset = file_csname; + if (pfile_cs) + *pfile_cs = file_cs; +} + +/* Determine value from file name. + * ------------------------------- + * + */ +float HTFileValue(const char *filename) +{ + HTSuffix *suff; + int n; + int i; + int lf = (int) strlen(filename); + +#ifndef NO_INIT + if (!HTSuffixes) + HTFileInit(); +#endif /* !NO_INIT */ + n = HTList_count(HTSuffixes); + for (i = 0; i < n; i++) { + int ls; + + suff = (HTSuffix *) HTList_objectAt(HTSuffixes, i); + ls = (int) strlen(suff->suffix); + if ((ls <= lf) && 0 == strcmp(suff->suffix, filename + lf - ls)) { + CTRACE((tfp, "File: Value of %s is %.3f\n", + filename, suff->quality)); + return suff->quality; /* OK -- found */ + } + } + return (float) 0.3; /* Dunno! */ +} + +/* + * Determine compression type from file name, by looking at its suffix. + * Sets as side-effect a pointer to the "dot" that begins the suffix. + */ +CompressFileType HTCompressFileType(const char *filename, + const char *dots, + int *rootlen) +{ + CompressFileType result = cftNone; + char *search; + + if ((search = FindSearch(filename)) != 0) { + char *newname = NULL; + + StrAllocCopy(newname, filename); + newname[((const char *) search) - filename] = '\0'; + result = HTCompressFileType(newname, dots, rootlen); + free(newname); + } else { + size_t len; + const char *ftype; + + VMS_DEL_VERSION(filename); + len = strlen(filename); + ftype = filename + len; + + if ((len > 4) + && !strcasecomp((ftype - 3), "bz2") + && StrChr(dots, ftype[-4]) != 0) { + result = cftBzip2; + ftype -= 4; + } else if ((len > 3) + && !strcasecomp((ftype - 2), "gz") + && StrChr(dots, ftype[-3]) != 0) { + result = cftGzip; + ftype -= 3; + } else if ((len > 3) + && !strcasecomp((ftype - 2), "zz") + && StrChr(dots, ftype[-3]) != 0) { + result = cftDeflate; + ftype -= 3; + } else if ((len > 2) + && !strcmp((ftype - 1), "Z") + && StrChr(dots, ftype[-2]) != 0) { + result = cftCompress; + ftype -= 2; + } + + *rootlen = (int) (ftype - filename); + + CTRACE((tfp, "HTCompressFileType(%s) returns %d:%s\n", + filename, (int) result, filename + *rootlen)); + } + return result; +} + +/* + * Determine expected file-suffix from the compression method. + */ +const char *HTCompressTypeToSuffix(CompressFileType method) +{ + const char *result = ""; + + switch (method) { + default: + case cftNone: + result = ""; + break; + case cftGzip: + result = ".gz"; + break; + case cftCompress: + result = ".Z"; + break; + case cftBzip2: + result = ".bz2"; + break; + case cftDeflate: + result = ".zz"; + break; + } + return result; +} + +/* + * Determine compression encoding from the compression method. + */ +const char *HTCompressTypeToEncoding(CompressFileType method) +{ + const char *result = NULL; + + switch (method) { + default: + case cftNone: + result = NULL; + break; + case cftGzip: + result = "gzip"; + break; + case cftCompress: + result = "compress"; + break; + case cftBzip2: + result = "bzip2"; + break; + case cftDeflate: + result = "deflate"; + break; + } + return result; +} + +/* + * Check if the token from "Content-Encoding" corresponds to a compression + * type. RFC 2068 (and cut/paste into RFC 2616) lists these: + * gzip + * compress + * deflate + * as well as "identity" (but that does nothing). + */ +CompressFileType HTEncodingToCompressType(const char *coding) +{ + CompressFileType result = cftNone; + + if (coding == NULL) { + result = cftNone; + } else if (!strcasecomp(coding, "gzip") || + !strcasecomp(coding, "x-gzip")) { + result = cftGzip; + } else if (!strcasecomp(coding, "compress") || + !strcasecomp(coding, "x-compress")) { + result = cftCompress; + } else if (!strcasecomp(coding, "bzip2") || + !strcasecomp(coding, "x-bzip2")) { + result = cftBzip2; + } else if (!strcasecomp(coding, "deflate") || + !strcasecomp(coding, "x-deflate")) { + result = cftDeflate; + } + return result; +} + +CompressFileType HTContentTypeToCompressType(const char *ct) +{ + CompressFileType method = cftNone; + + if (ct == NULL) { + method = cftNone; + } else if (!strncasecomp(ct, "application/gzip", 16) || + !strncasecomp(ct, "application/x-gzip", 18)) { + method = cftGzip; + } else if (!strncasecomp(ct, "application/compress", 20) || + !strncasecomp(ct, "application/x-compress", 22)) { + method = cftCompress; + } else if (!strncasecomp(ct, "application/bzip2", 17) || + !strncasecomp(ct, "application/x-bzip2", 19)) { + method = cftBzip2; + } + return method; +} + +/* + * Check the anchor's content_type and content_encoding elements for a gzip or + * Unix compressed file -FM, TD + */ +CompressFileType HTContentToCompressType(HTParentAnchor *anchor) +{ + CompressFileType method = cftNone; + const char *ct = HTAnchor_content_type(anchor); + const char *ce = HTAnchor_content_encoding(anchor); + + if (ct != 0) { + method = HTContentTypeToCompressType(ct); + } else if (ce != 0) { + method = HTEncodingToCompressType(ce); + } + return method; +} + +/* Determine write access to a file. + * --------------------------------- + * + * On exit: + * Returns YES if file can be accessed and can be written to. + * + * Bugs: + * 1. No code for non-unix systems. + * 2. Isn't there a quicker way? + */ +BOOL HTEditable(const char *filename GCC_UNUSED) +{ +#ifndef NO_GROUPS + GETGROUPS_T groups[NGROUPS]; + uid_t myUid; + int ngroups; /* The number of groups */ + struct stat fileStatus; + int i; + + if (stat(filename, &fileStatus)) /* Get details of filename */ + return NO; /* Can't even access file! */ + + ngroups = getgroups(NGROUPS, groups); /* Groups to which I belong */ + myUid = geteuid(); /* Get my user identifier */ + + if (TRACE) { + int i2; + + fprintf(tfp, + "File mode is 0%o, uid=%d, gid=%d. My uid=%d, %d groups (", + (unsigned int) fileStatus.st_mode, + (int) fileStatus.st_uid, + (int) fileStatus.st_gid, + (int) myUid, + (int) ngroups); + for (i2 = 0; i2 < ngroups; i2++) + fprintf(tfp, " %d", (int) groups[i2]); + fprintf(tfp, ")\n"); + } + + if (fileStatus.st_mode & 0002) /* I can write anyway? */ + return YES; + + if ((fileStatus.st_mode & 0200) /* I can write my own file? */ + &&(fileStatus.st_uid == myUid)) + return YES; + + if (fileStatus.st_mode & 0020) /* Group I am in can write? */ + { + for (i = 0; i < ngroups; i++) { + if (groups[i] == fileStatus.st_gid) + return YES; + } + } + CTRACE((tfp, "\tFile is not editable.\n")); +#endif /* NO_GROUPS */ + return NO; /* If no excuse, can't do */ +} + +/* Make a save stream. + * ------------------- + * + * The stream must be used for writing back the file. + * @@@ no backup done + */ +HTStream *HTFileSaveStream(HTParentAnchor *anchor) +{ + const char *addr = anchor->address; + char *localname = HTLocalName(addr); + FILE *fp = fopen(localname, BIN_W); + + FREE(localname); + if (!fp) + return NULL; + + return HTFWriter_new(fp); +} + +/* Output one directory entry. + * --------------------------- + */ +void HTDirEntry(HTStructured * target, const char *tail, const char *entry) +{ + char *relative = NULL; + char *stripped = NULL; + char *escaped = NULL; + int len; + + if (entry == NULL) + entry = ""; + StrAllocCopy(escaped, entry); + LYTrimPathSep(escaped); + if (strcmp(escaped, "..") != 0) { + stripped = escaped; + escaped = HTEscape(stripped, URL_XPALPHAS); + if (((len = (int) strlen(escaped)) > 2) && + escaped[(len - 3)] == '%' && + escaped[(len - 2)] == '2' && + TOUPPER(escaped[(len - 1)]) == 'F') { + escaped[(len - 3)] = '\0'; + } + } + + if (isEmpty(tail)) { + /* + * Handle extra slash at end of path. + */ + HTStartAnchor(target, NULL, (escaped[0] != '\0' ? escaped : "/")); + } else { + /* + * If empty tail, gives absolute ref below. + */ + relative = 0; + HTSprintf0(&relative, "%s%s%s", + tail, + (*escaped != '\0' ? "/" : ""), + escaped); + HTStartAnchor(target, NULL, relative); + FREE(relative); + } + FREE(stripped); + FREE(escaped); +} + +static BOOL view_structured(HTFormat format_out) +{ + BOOL result = FALSE; + +#ifdef USE_PRETTYSRC + if (psrc_view + || (format_out == HTAtom_for("www/dump"))) + result = TRUE; +#else + if (format_out == WWW_SOURCE) + result = TRUE; +#endif + return result; +} + +/* + * Write a DOCTYPE to the given stream if we happen to want to see the + * source view, or are dumping source. This is not needed when the source + * is not visible, since the document is rendered from a HTStructured object. + */ +void HTStructured_doctype(HTStructured * target, HTFormat format_out) +{ + if (view_structured(format_out)) + PUTS("<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 4.01 Transitional//EN\">\n"); +} + +void HTStructured_meta(HTStructured * target, HTFormat format_out) +{ + if (view_structured(format_out)) + PUTS("<meta http-equiv=\"Content-Type\" content=\"" STR_HTML + "; charset=iso-8859-1\">\n"); +} +/* Output parent directory entry. + * ------------------------------ + * + * This gives the TITLE and H1 header, and also a link + * to the parent directory if appropriate. + * + * On exit: + * Returns TRUE if an "Up to <parent>" link was not created + * for a readable local directory because LONG_LIST is defined + * and NO_PARENT_DIR_REFERENCE is not defined, so that the + * calling function should use LYListFmtParse() to create a link + * to the parent directory. Otherwise, it returns FALSE. - FM + */ +BOOL HTDirTitles(HTStructured * target, HTParentAnchor *anchor, + HTFormat format_out, + int tildeIsTop) +{ + const char *logical = anchor->address; + char *path = HTParse(logical, "", PARSE_PATH + PARSE_PUNCTUATION); + char *current; + char *cp = NULL; + BOOL need_parent_link = FALSE; + int i; + +#if defined(USE_DOS_DRIVES) + BOOL local_link = (strlen(logical) > 18 + && !strncasecomp(logical, "file://localhost/", 17) + && LYIsDosDrive(logical + 17)); + BOOL is_remote = !local_link; + +#else +#define is_remote TRUE +#endif + + /* + * Check tildeIsTop for treating home directory as Welcome (assume the + * tilde is not followed by a username). - FM + */ + if (tildeIsTop && !StrNCmp(path, "/~", 2)) { + if (path[2] == '\0') { + path[1] = '\0'; + } else { + for (i = 0; path[(i + 2)]; i++) { + path[i] = path[(i + 2)]; + } + path[i] = '\0'; + } + } + + /* + * Trim out the ;type= parameter, if present. - FM + */ + if ((cp = strrchr(path, ';')) != NULL) { + if (!strncasecomp((cp + 1), "type=", 5)) { + if (TOUPPER(*(cp + 6)) == 'D' || + TOUPPER(*(cp + 6)) == 'A' || + TOUPPER(*(cp + 6)) == 'I') + *cp = '\0'; + } + cp = NULL; + } + current = LYPathLeaf(path); /* last part or "" */ + + { + char *printable = NULL; + +#ifdef DIRED_SUPPORT + printable = HTURLPath_toFile(((!strncasecomp(path, "/%2F", 4)) /* "//" ? */ + ? (path + 1) + : path), + TRUE, + is_remote); + if (0 == strncasecomp(printable, "/vmsysu:", 8) || + 0 == strncasecomp(printable, "/anonymou.", 10)) { + StrAllocCopy(cp, (printable + 1)); + StrAllocCopy(printable, cp); + FREE(cp); + } +#else + StrAllocCopy(printable, current); + HTUnEscape(printable); +#endif /* DIRED_SUPPORT */ + + HTStructured_doctype(target, format_out); + + START(HTML_HEAD); + PUTC('\n'); + START(HTML_TITLE); + PUTS(*printable ? printable : WELCOME_MSG); + PUTS(SEGMENT_DIRECTORY); + END(HTML_TITLE); + PUTC('\n'); + HTStructured_meta(target, format_out); + END(HTML_HEAD); + PUTC('\n'); + + START(HTML_BODY); + PUTC('\n'); + +#ifdef DIRED_SUPPORT + START(HTML_H2); + PUTS(*printable ? SEGMENT_CURRENT_DIR : ""); + PUTS(*printable ? printable : WELCOME_MSG); + END(HTML_H2); + PUTC('\n'); +#else + START(HTML_H1); + PUTS(*printable ? printable : WELCOME_MSG); + END(HTML_H1); + PUTC('\n'); +#endif /* DIRED_SUPPORT */ + if (((0 == strncasecomp(printable, "vmsysu:", 7)) && + (cp = StrChr(printable, '.')) != NULL && + StrChr(cp, '/') == NULL) || + (0 == strncasecomp(printable, "anonymou.", 9) && + StrChr(printable, '/') == NULL)) { + FREE(printable); + FREE(path); + return (need_parent_link); + } + FREE(printable); + } + +#ifndef NO_PARENT_DIR_REFERENCE + /* + * Make link back to parent directory. + */ + if (current - path > 0 + && LYIsPathSep(current[-1]) + && current[0] != '\0') { /* was a slash AND something else too */ + char *parent = NULL; + char *relative = NULL; + + current[-1] = '\0'; + parent = strrchr(path, '/'); /* penultimate slash */ + + if ((parent && + (!strcmp(parent, "/..") || + !strncasecomp(parent, "/%2F", 4))) || + !strncasecomp(current, "%2F", 3)) { + FREE(path); + return (need_parent_link); + } + + relative = 0; + HTSprintf0(&relative, "%s/..", current); + +#if defined(DOSPATH) || defined(__EMX__) + if (local_link) { + if (parent != 0 && strlen(parent) == 3) { + StrAllocCat(relative, "/."); + } + } else +#endif + +#if !defined (VMS) + { + /* + * On Unix, if it's not ftp and the directory cannot be read, don't + * put out a link. + * + * On VMS, this problem is dealt with internally by + * HTVMSBrowseDir(). + */ + DIR *dp = NULL; + + if (LYisLocalFile(logical)) { + /* + * We need an absolute file path for the opendir. We also need + * to unescape for this test. Don't worry about %2F now, they + * presumably have been dealt with above, and shouldn't appear + * for local files anyway... Assume OS / filesystem will just + * ignore superfluous slashes. - KW + */ + char *fullparentpath = NULL; + + /* + * Path has been shortened above. + */ + StrAllocCopy(fullparentpath, *path ? path : "/"); + + /* + * Guard against weirdness. + */ + if (0 == strcmp(current, "..")) { + StrAllocCat(fullparentpath, "/../.."); + } else if (0 == strcmp(current, ".")) { + StrAllocCat(fullparentpath, "/.."); + } + + HTUnEscape(fullparentpath); + if ((dp = opendir(fullparentpath)) == NULL) { + FREE(fullparentpath); + FREE(relative); + FREE(path); + return (need_parent_link); + } + closedir(dp); + FREE(fullparentpath); +#ifdef LONG_LIST + need_parent_link = TRUE; + FREE(path); + FREE(relative); + return (need_parent_link); +#endif /* LONG_LIST */ + } + } +#endif /* !VMS */ + HTStartAnchor(target, "", relative); + FREE(relative); + + PUTS(SEGMENT_UP_TO); + if (parent) { + if ((0 == strcmp(current, ".")) || + (0 == strcmp(current, ".."))) { + /* + * Should not happen, but if it does, at least avoid giving + * misleading info. - KW + */ + PUTS(".."); + } else { + char *printable = NULL; + + StrAllocCopy(printable, parent + 1); + HTUnEscape(printable); + PUTS(printable); + FREE(printable); + } + } else { + PUTC('/'); + } + END(HTML_A); + PUTC('\n'); + } +#endif /* !NO_PARENT_DIR_REFERENCE */ + + FREE(path); + return (need_parent_link); +} + +#if defined HAVE_READDIR +/* Send README file. + * ----------------- + * + * If a README file exists, then it is inserted into the document here. + */ +static void do_readme(HTStructured * target, const char *localname) +{ + FILE *fp; + char *readme_file_name = NULL; + int ch; + + HTSprintf0(&readme_file_name, "%s/%s", localname, HT_DIR_README_FILE); + + fp = fopen(readme_file_name, "r"); + + if (fp) { + START(HTML_PRE); + while ((ch = fgetc(fp)) != EOF) { + PUTC((char) ch); + } + END(HTML_PRE); + HTDisplayPartial(); + fclose(fp); + } + FREE(readme_file_name); +} + +#define DIRED_BLOK(obj) (((DIRED *)(obj))->sort_tags) +#define DIRED_NAME(obj) (((DIRED *)(obj))->file_name) + +#define NM_cmp(a,b) ((a) < (b) ? -1 : ((a) > (b) ? 1 : 0)) + +#if defined(LONG_LIST) && defined(DIRED_SUPPORT) +static const char *file_type(const char *path) +{ + const char *type; + + while (*path == '.') + ++path; + type = StrChr(path, '.'); + if (type == NULL) + type = ""; + return type; +} +#endif /* LONG_LIST && DIRED_SUPPORT */ + +static int dired_cmp(void *a, void *b) +{ + DIRED *p = (DIRED *) a; + DIRED *q = (DIRED *) b; + int code = p->sort_tags - q->sort_tags; + +#if defined(LONG_LIST) && defined(DIRED_SUPPORT) + if (code == 0) { + switch (dir_list_order) { + case ORDER_BY_SIZE: + code = -NM_cmp(p->file_info.st_size, q->file_info.st_size); + break; + case ORDER_BY_DATE: + code = -NM_cmp(p->file_info.st_mtime, q->file_info.st_mtime); + break; + case ORDER_BY_MODE: + code = NM_cmp(p->file_info.st_mode, q->file_info.st_mode); + break; + case ORDER_BY_USER: + code = NM_cmp(p->file_info.st_uid, q->file_info.st_uid); + break; + case ORDER_BY_GROUP: + code = NM_cmp(p->file_info.st_gid, q->file_info.st_gid); + break; + case ORDER_BY_TYPE: + code = AS_cmp(file_type(p->file_name), file_type(q->file_name)); + break; + default: + code = 0; + break; + } + } +#endif /* LONG_LIST && DIRED_SUPPORT */ + if (code == 0) + code = AS_cmp(p->file_name, q->file_name); +#if 0 + CTRACE((tfp, "dired_cmp(%d) ->%d\n\t%c:%s (%s)\n\t%c:%s (%s)\n", + dir_list_order, + code, + p->sort_tags, p->file_name, file_type(p->file_name), + q->sort_tags, q->file_name, file_type(q->file_name))); +#endif + return code; +} + +static int print_local_dir(DIR *dp, char *localname, + HTParentAnchor *anchor, + HTFormat format_out, + HTStream *sink) +{ + HTStructured *target; /* HTML object */ + HTBTree *bt; + HTStructuredClass targetClass; + STRUCT_DIRENT *dirbuf; + char *pathname = NULL; + char *tail = NULL; + const char *p; + char *tmpfilename = NULL; + BOOL need_parent_link = FALSE; + BOOL preformatted = FALSE; + int status; + struct stat *actual_info; + +#ifdef DISP_PARTIAL + int num_of_entries = 0; /* lines counter */ +#endif + +#ifdef S_IFLNK + struct stat link_info; +#endif + + CTRACE((tfp, "print_local_dir() started\n")); + + pathname = HTParse(anchor->address, "", + PARSE_PATH + PARSE_PUNCTUATION); + + if ((p = strrchr(pathname, '/')) == NULL) + p = "/"; + StrAllocCopy(tail, (p + 1)); + FREE(pathname); + + if (UCLYhndl_HTFile_for_unspec >= 0) { + HTAnchor_setUCInfoStage(anchor, + UCLYhndl_HTFile_for_unspec, + UCT_STAGE_PARSER, + UCT_SETBY_DEFAULT); + } + + target = HTML_new(anchor, format_out, sink); + targetClass = *target->isa; /* Copy routine entry points */ + + /* + * The need_parent_link flag will be set if an "Up to <parent>" link was + * not created for a readable parent in HTDirTitles() because LONG_LIST is + * defined and NO_PARENT_DIR_REFERENCE is not defined so that need we to + * create the link via an LYListFmtParse() call. - FM + */ + need_parent_link = HTDirTitles(target, anchor, format_out, FALSE); + +#ifdef DIRED_SUPPORT + if (!isLYNXCGI(anchor->address)) { + HTAnchor_setFormat(anchor, WWW_DIRED); + lynx_edit_mode = TRUE; + } +#endif /* DIRED_SUPPORT */ + if (HTDirReadme == HT_DIR_README_TOP) + do_readme(target, localname); + + bt = HTBTree_new(dired_cmp); + + _HTProgress(READING_DIRECTORY); + status = HT_LOADED; /* assume we don't get interrupted */ + while ((dirbuf = readdir(dp)) != NULL) { + /* + * While there are directory entries to be read... + */ + DIRED *data = NULL; + +#ifdef STRUCT_DIRENT__D_INO + if (dirbuf->d_ino == 0) + /* + * If the entry is not being used, skip it. + */ + continue; +#endif + /* + * Skip self, parent if handled in HTDirTitles() or if + * NO_PARENT_DIR_REFERENCE is not defined, and any dot files if + * no_dotfiles is set or show_dotfiles is not set. - FM + */ + if (!strcmp(dirbuf->d_name, ".") /* self */ || + (!strcmp(dirbuf->d_name, "..") /* parent */ && + need_parent_link == FALSE) || + ((strcmp(dirbuf->d_name, "..")) && + (dirbuf->d_name[0] == '.' && + (no_dotfiles || !show_dotfiles)))) + continue; + + StrAllocCopy(tmpfilename, localname); + /* + * If filename is not root directory, add trailing separator. + */ + LYAddPathSep(&tmpfilename); + + StrAllocCat(tmpfilename, dirbuf->d_name); + data = (DIRED *) malloc(sizeof(DIRED) + strlen(dirbuf->d_name) + 4); + if (data == NULL) { + status = HT_PARTIAL_CONTENT; + break; + } + LYTrimPathSep(tmpfilename); + + actual_info = &(data->file_info); +#ifdef S_IFLNK + if (lstat(tmpfilename, actual_info) < 0) { + actual_info->st_mode = 0; + } else { + if (S_ISLNK(actual_info->st_mode)) { + actual_info = &link_info; + if (stat(tmpfilename, actual_info) < 0) + actual_info->st_mode = 0; + } + } +#else + if (stat(tmpfilename, actual_info) < 0) + actual_info->st_mode = 0; +#endif + + strcpy(data->file_name, dirbuf->d_name); +#ifndef DIRED_SUPPORT + if (S_ISDIR(actual_info->st_mode)) { + data->sort_tags = 'D'; + } else { + data->sort_tags = 'F'; + /* D & F to have first directories, then files */ + } +#else + if (S_ISDIR(actual_info->st_mode)) { + if (dir_list_style == MIXED_STYLE) { + data->sort_tags = ' '; + LYAddPathSep0(data->file_name); + } else if (!strcmp(dirbuf->d_name, "..")) { + data->sort_tags = 'A'; + } else { + data->sort_tags = 'D'; + } + } else if (dir_list_style == MIXED_STYLE) { + data->sort_tags = ' '; + } else if (dir_list_style == FILES_FIRST) { + data->sort_tags = 'C'; + /* C & D to have first files, then directories */ + } else { + data->sort_tags = 'F'; + } +#endif /* !DIRED_SUPPORT */ + /* + * Sort dirname in the tree bt. + */ + HTBTree_add(bt, data); + +#ifdef DISP_PARTIAL + /* optimize for expensive operation: */ + if (num_of_entries % (partial_threshold > 0 ? + partial_threshold : display_lines) == 0) { + if (HTCheckForInterrupt()) { + status = HT_PARTIAL_CONTENT; + break; + } + } + num_of_entries++; +#endif /* DISP_PARTIAL */ + + } /* end while directory entries left to read */ + + if (status != HT_PARTIAL_CONTENT) + _HTProgress(OPERATION_OK); + else + CTRACE((tfp, "Reading the directory interrupted by user\n")); + + /* + * Run through tree printing out in order. + */ + { + HTBTElement *next_element = HTBTree_next(bt, NULL); + + /* pick up the first element of the list */ + int num_of_entries_output = 0; /* lines counter */ + + char state; + + /* I for initial (.. file), + D for directory file, + F for file */ + +#ifdef DIRED_SUPPORT + char test; +#endif /* DIRED_SUPPORT */ + state = 'I'; + + while (next_element != NULL) { + DIRED *entry; + +#ifndef DISP_PARTIAL + if (num_of_entries_output % HTMAX(display_lines, 10) == 0) { + if (HTCheckForInterrupt()) { + _HTProgress(TRANSFER_INTERRUPTED); + status = HT_PARTIAL_CONTENT; + break; + } + } +#endif + StrAllocCopy(tmpfilename, localname); + /* + * If filename is not root directory. + */ + LYAddPathSep(&tmpfilename); + + entry = (DIRED *) (HTBTree_object(next_element)); + /* + * Append the current entry's filename to the path. + */ + StrAllocCat(tmpfilename, entry->file_name); + HTSimplify(tmpfilename); + /* + * Output the directory entry. + */ + if (strcmp(DIRED_NAME(HTBTree_object(next_element)), "..")) { +#ifdef DIRED_SUPPORT + test = + (char) (DIRED_BLOK(HTBTree_object(next_element)) + == 'D' ? 'D' : 'F'); + if (state != test) { +#ifndef LONG_LIST + if (dir_list_style == FILES_FIRST) { + if (state == 'F') { + END(HTML_DIR); + PUTC('\n'); + } + } else if (dir_list_style != MIXED_STYLE) + if (state == 'D') { + END(HTML_DIR); + PUTC('\n'); + } +#endif /* !LONG_LIST */ + state = + (char) (DIRED_BLOK(HTBTree_object(next_element)) + == 'D' ? 'D' : 'F'); + if (preformatted) { + END(HTML_PRE); + PUTC('\n'); + preformatted = FALSE; + } + START(HTML_H2); + if (dir_list_style != MIXED_STYLE) { + START(HTML_EM); + PUTS(state == 'D' + ? LABEL_SUBDIRECTORIES + : LABEL_FILES); + END(HTML_EM); + } + END(HTML_H2); + PUTC('\n'); +#ifndef LONG_LIST + START(HTML_DIR); + PUTC('\n'); +#endif /* !LONG_LIST */ + } +#else + if (state != DIRED_BLOK(HTBTree_object(next_element))) { +#ifndef LONG_LIST + if (state == 'D') { + END(HTML_DIR); + PUTC('\n'); + } +#endif /* !LONG_LIST */ + state = + (char) (DIRED_BLOK(HTBTree_object(next_element)) + == 'D' ? 'D' : 'F'); + if (preformatted) { + END(HTML_PRE); + PUTC('\n'); + preformatted = FALSE; + } + START(HTML_H2); + START(HTML_EM); + PUTS(state == 'D' + ? LABEL_SUBDIRECTORIES + : LABEL_FILES); + END(HTML_EM); + END(HTML_H2); + PUTC('\n'); +#ifndef LONG_LIST + START(HTML_DIR); + PUTC('\n'); +#endif /* !LONG_LIST */ + } +#endif /* DIRED_SUPPORT */ +#ifndef LONG_LIST + START(HTML_LI); +#endif /* !LONG_LIST */ + } + if (!preformatted) { + START(HTML_PRE); + PUTC('\n'); + preformatted = TRUE; + } +#ifdef LONG_LIST + LYListFmtParse(list_format, entry, tmpfilename, target, tail); +#else + HTDirEntry(target, tail, entry->file_name); + PUTS(entry->file_name); + END(HTML_A); + MAYBE_END(HTML_LI); + PUTC('\n'); +#endif /* LONG_LIST */ + + next_element = HTBTree_next(bt, next_element); + /* pick up the next element of the list; + if none, return NULL */ + + /* optimize for expensive operation: */ +#ifdef DISP_PARTIAL + if (num_of_entries_output % + ((partial_threshold > 0) + ? partial_threshold + : display_lines) == 0) { + /* num_of_entries, num_of_entries_output... */ + HTDisplayPartial(); + + if (HTCheckForInterrupt()) { + _HTProgress(TRANSFER_INTERRUPTED); + status = HT_PARTIAL_CONTENT; + break; + } + } + num_of_entries_output++; +#endif /* DISP_PARTIAL */ + + } /* end while next_element */ + + if (status == HT_LOADED) { + if (state == 'I') { + START(HTML_P); + PUTS("Empty Directory"); + } +#ifndef LONG_LIST + else + END(HTML_DIR); +#endif /* !LONG_LIST */ + } + } /* end printing out the tree in order */ + if (preformatted) { + END(HTML_PRE); + PUTC('\n'); + } + END(HTML_BODY); + PUTC('\n'); + + FREE(tmpfilename); + FREE(tail); + HTBTreeAndObject_free(bt); + + if (status == HT_LOADED) { + if (HTDirReadme == HT_DIR_README_BOTTOM) + do_readme(target, localname); + FREE_TARGET; + } else { + ABORT_TARGET; + } + HTFinishDisplayPartial(); + return status; /* document loaded, maybe partial */ +} +#endif /* HAVE_READDIR */ + +#ifndef VMS +int HTStat(const char *filename, + struct stat *data) +{ + int result = -1; + size_t len = strlen(filename); + + if (len != 0 && LYIsPathSep(filename[len - 1])) { + char *temp_name = NULL; + + HTSprintf0(&temp_name, "%s.", filename); + result = HTStat(temp_name, data); + FREE(temp_name); + } else { + result = stat(filename, data); +#ifdef _WINDOWS + /* + * Someone claims that stat() doesn't give the proper result for a + * directory on Windows. + */ + if (result == -1 + && access(filename, 0) == 0) { + data->st_mode = S_IFDIR; + result = 0; + } +#endif + } + return result; +} +#endif + +#if defined(USE_ZLIB) || defined(USE_BZLIB) +static BOOL sniffStream(FILE *fp, char *buffer, size_t needed) +{ + long offset = ftell(fp); + BOOL result = FALSE; + + if (offset >= 0) { + if (fread(buffer, sizeof(char), needed, fp) == needed) { + result = TRUE; + } + if (fseek(fp, offset, SEEK_SET) < 0) { + CTRACE((tfp, "error seeking in stream\n")); + result = FALSE; + } + } + return result; +} +#endif + +#ifdef USE_ZLIB +static BOOL isGzipStream(FILE *fp) +{ + char buffer[3]; + BOOL result; + + if (sniffStream(fp, buffer, sizeof(buffer)) + && !MemCmp(buffer, "\037\213", sizeof(buffer) - 1)) { + result = TRUE; + } else { + CTRACE((tfp, "not a gzip-stream\n")); + result = FALSE; + } + return result; +} + +/* + * Strictly speaking, DEFLATE has no header bytes. But decode what we can, + * (to eliminate the one "reserved" pattern) and provide a trace. See RFC-1951 + * discussion of BFINAL and BTYPE. + */ +static BOOL isDeflateStream(FILE *fp) +{ + char buffer[3]; + BOOL result = FALSE; + + if (sniffStream(fp, buffer, sizeof(buffer))) { + int bit1 = ((buffer[0] >> 0) & 1); + int bit2 = ((buffer[0] >> 1) & 1); + int bit3 = ((buffer[0] >> 2) & 1); + int btype = ((bit3 << 1) + bit2); + + if (!MemCmp(buffer, "\170\234", sizeof(buffer) - 1)) { + result = TRUE; + CTRACE((tfp, "isDeflate: assume zlib-wrapped deflate\n")); + } else if (btype == 3) { + CTRACE((tfp, "isDeflate: not a deflate-stream\n")); + } else { + CTRACE((tfp, "isDeflate: %send block, %s compression\n", + (bit1 ? "" : "non-"), + (btype == 0 + ? "no" + : (btype == 1 + ? "static Huffman" + : "dynamic Huffman")))); + result = TRUE; + } + } + return result; +} +#endif + +#ifdef USE_BZLIB +static BOOL isBzip2Stream(FILE *fp) +{ + char buffer[6]; + BOOL result; + + if (sniffStream(fp, buffer, sizeof(buffer)) + && !MemCmp(buffer, "BZh", 3) + && isdigit(UCH(buffer[3])) + && isdigit(UCH(buffer[4]))) { + result = TRUE; + } else { + CTRACE((tfp, "not a bzip2-stream\n")); + result = FALSE; + } + return result; +} +#endif + +#ifdef VMS +#define FOPEN_MODE(bin) "r", "shr=put", "shr=upd" +#define DOT_STRING "._-" /* FIXME: should we check if suffix is after ']' or ':' ? */ +#else +#define FOPEN_MODE(bin) (bin ? BIN_R : "r") +#define DOT_STRING "." +#endif + +static int decompressAndParse(HTParentAnchor *anchor, + HTFormat format_out, + HTStream *sink, + char *nodename GCC_UNUSED, + char *filename, + HTAtom *myEncoding, + HTFormat format, + int *statusp) +{ + HTAtom *encoding = 0; + +#ifdef USE_ZLIB + FILE *zzfp = 0; + gzFile gzfp = 0; +#endif /* USE_ZLIB */ +#ifdef USE_BZLIB + BZFILE *bzfp = 0; +#endif /* USE_ZLIB */ +#if defined(USE_ZLIB) || defined(USE_BZLIB) + CompressFileType internal_decompress = cftNone; + BOOL failed_decompress = NO; +#endif + int rootlen = 0; + char *localname = filename; + int bin; + FILE *fp; + int result = FALSE; + +#ifdef VMS + /* + * Assume that the file is in Unix-style syntax if it contains a '/' after + * the leading one. @@ + */ + localname = (StrChr(localname + 1, '/') + ? HTVMS_name(nodename, localname) + : localname + 1); +#endif /* VMS */ + + bin = HTCompressFileType(filename, ".", &rootlen) != cftNone; + fp = fopen(localname, FOPEN_MODE(bin)); + +#ifdef VMS + /* + * If the file wasn't VMS syntax, then perhaps it is Ultrix. + */ + if (!fp) { + char *ultrixname = 0; + + CTRACE((tfp, "HTLoadFile: Can't open as %s\n", localname)); + HTSprintf0(&ultrixname, "%s::\"%s\"", nodename, filename); + fp = fopen(ultrixname, FOPEN_MODE(bin)); + if (!fp) { + CTRACE((tfp, "HTLoadFile: Can't open as %s\n", ultrixname)); + } + FREE(ultrixname); + } +#endif /* VMS */ + CTRACE((tfp, "HTLoadFile: Opening `%s' gives %p\n", localname, (void *) fp)); + if (fp) { /* Good! */ + if (HTEditable(localname)) { + HTAtom *put = HTAtom_for("PUT"); + HTList *methods = HTAnchor_methods(anchor); + + if (HTList_indexOf(methods, put) == (-1)) { + HTList_addObject(methods, put); + } + } + /* + * Fake a Content-Encoding for compressed files. - FM + */ + if (!IsUnityEnc(myEncoding)) { + /* + * We already know from the call to HTFileFormat that + * this is a compressed file, no need to look at the filename + * again. - kw + */ +#if defined(USE_ZLIB) || defined(USE_BZLIB) + CompressFileType method = HTEncodingToCompressType(HTAtom_name(myEncoding)); +#endif + +#define isDOWNLOAD(m) (strcmp(format_out->name, "www/download") && (method == m)) +#ifdef USE_ZLIB + if (isDOWNLOAD(cftGzip)) { + if (isGzipStream(fp)) { + fclose(fp); + fp = 0; + gzfp = gzopen(localname, BIN_R); + + CTRACE((tfp, "HTLoadFile: gzopen of `%s' gives %p\n", + localname, (void *) gzfp)); + } + internal_decompress = cftGzip; + } else if (isDOWNLOAD(cftDeflate)) { + if (isDeflateStream(fp)) { + zzfp = fp; + fp = 0; + + CTRACE((tfp, "HTLoadFile: zzopen of `%s' gives %p\n", + localname, (void *) zzfp)); + } + internal_decompress = cftDeflate; + } else +#endif /* USE_ZLIB */ +#ifdef USE_BZLIB + if (isDOWNLOAD(cftBzip2)) { + if (isBzip2Stream(fp)) { + fclose(fp); + fp = 0; + bzfp = BZ2_bzopen(localname, BIN_R); + + CTRACE((tfp, "HTLoadFile: bzopen of `%s' gives %p\n", + localname, bzfp)); + } + internal_decompress = cftBzip2; + } else +#endif /* USE_BZLIB */ + { + StrAllocCopy(anchor->content_type, format->name); + StrAllocCopy(anchor->content_encoding, HTAtom_name(myEncoding)); + format = HTAtom_for("www/compressed"); + } + } else { + CompressFileType cft = HTCompressFileType(localname, DOT_STRING, &rootlen); + + if (cft != cftNone) { + char *cp = NULL; + + StrAllocCopy(cp, localname); + cp[rootlen] = '\0'; + format = HTFileFormat(cp, &encoding, NULL); + FREE(cp); + format = HTCharsetFormat(format, anchor, + UCLYhndl_HTFile_for_unspec); + StrAllocCopy(anchor->content_type, format->name); + } + + switch (cft) { + case cftCompress: + StrAllocCopy(anchor->content_encoding, "x-compress"); + format = HTAtom_for("www/compressed"); + break; + case cftDeflate: + StrAllocCopy(anchor->content_encoding, "x-deflate"); +#ifdef USE_ZLIB + if (strcmp(format_out->name, "www/download") != 0) { + if (isDeflateStream(fp)) { + zzfp = fp; + fp = 0; + + CTRACE((tfp, "HTLoadFile: zzopen of `%s' gives %p\n", + localname, (void *) zzfp)); + } + internal_decompress = cftDeflate; + } +#else /* USE_ZLIB */ + format = HTAtom_for("www/compressed"); +#endif /* USE_ZLIB */ + break; + case cftGzip: + StrAllocCopy(anchor->content_encoding, "x-gzip"); +#ifdef USE_ZLIB + if (strcmp(format_out->name, "www/download") != 0) { + if (isGzipStream(fp)) { + fclose(fp); + fp = 0; + gzfp = gzopen(localname, BIN_R); + + CTRACE((tfp, "HTLoadFile: gzopen of `%s' gives %p\n", + localname, (void *) gzfp)); + } + internal_decompress = cftGzip; + } +#else /* USE_ZLIB */ + format = HTAtom_for("www/compressed"); +#endif /* USE_ZLIB */ + break; + case cftBzip2: + StrAllocCopy(anchor->content_encoding, "x-bzip2"); +#ifdef USE_BZLIB + if (strcmp(format_out->name, "www/download") != 0) { + if (isBzip2Stream(fp)) { + fclose(fp); + fp = 0; + bzfp = BZ2_bzopen(localname, BIN_R); + + CTRACE((tfp, "HTLoadFile: bzopen of `%s' gives %p\n", + localname, bzfp)); + } + internal_decompress = cftBzip2; + } +#else /* USE_BZLIB */ + format = HTAtom_for("www/compressed"); +#endif /* USE_BZLIB */ + break; + case cftNone: + break; + } + } +#if defined(USE_ZLIB) || defined(USE_BZLIB) + if (internal_decompress != cftNone) { + switch (internal_decompress) { +#ifdef USE_ZLIB + case cftDeflate: + failed_decompress = (BOOLEAN) (zzfp == NULL); + break; + case cftCompress: + case cftGzip: + failed_decompress = (BOOLEAN) (gzfp == NULL); + break; +#endif +#ifdef USE_BZLIB + case cftBzip2: + failed_decompress = (BOOLEAN) (bzfp == NULL); + break; +#endif + default: + failed_decompress = YES; + break; + } + if (failed_decompress) { + *statusp = HTLoadError(NULL, + -(HT_ERROR), + FAILED_OPEN_COMPRESSED_FILE); + } else { + char *sugfname = NULL; + + if (anchor->SugFname) { + StrAllocCopy(sugfname, anchor->SugFname); + } else { + char *anchor_path = HTParse(anchor->address, "", + PARSE_PATH + PARSE_PUNCTUATION); + char *lastslash; + + HTUnEscape(anchor_path); + lastslash = strrchr(anchor_path, '/'); + if (lastslash) + StrAllocCopy(sugfname, lastslash + 1); + FREE(anchor_path); + } + FREE(anchor->content_encoding); + if (sugfname && *sugfname) + HTCheckFnameForCompression(&sugfname, anchor, + TRUE); + if (sugfname && *sugfname) + StrAllocCopy(anchor->SugFname, sugfname); + FREE(sugfname); +#ifdef USE_BZLIB + if (bzfp) + *statusp = HTParseBzFile(format, format_out, + anchor, + bzfp, sink); +#endif +#ifdef USE_ZLIB + if (gzfp) + *statusp = HTParseGzFile(format, format_out, + anchor, + gzfp, sink); + else if (zzfp) + *statusp = HTParseZzFile(format, format_out, + anchor, + zzfp, sink); +#endif + } + } else +#endif /* USE_ZLIB || USE_BZLIB */ + { + *statusp = HTParseFile(format, format_out, anchor, fp, sink); + } + if (fp != 0) { + fclose(fp); + fp = 0; + } + result = TRUE; + } /* If successful open */ + return result; +} + +/* Load a document. + * ---------------- + * + * On entry: + * addr must point to the fully qualified hypertext reference. + * This is the physical address of the file + * + * On exit: + * returns <0 Error has occurred. + * HTLOADED OK + * + */ +int HTLoadFile(const char *addr, + HTParentAnchor *anchor, + HTFormat format_out, + HTStream *sink) +{ + char *filename = NULL; + char *acc_method = NULL; + HTFormat format; + char *nodename = NULL; + char *newname = NULL; /* Simplified name of file */ + HTAtom *myEncoding = NULL; /* enc of this file, may be gzip etc. */ + int status = -1; + +#ifndef DISABLE_FTP + char *ftp_newhost; +#endif + +#ifdef VMS + struct stat stat_info; +#endif /* VMS */ + + /* + * Reduce the filename to a basic form (hopefully unique!). + */ + StrAllocCopy(newname, addr); + filename = HTParse(newname, "", PARSE_PATH | PARSE_PUNCTUATION); + nodename = HTParse(newname, "", PARSE_HOST); + + /* + * If access is ftp, or file is on another host, invoke ftp now. + */ + acc_method = HTParse(newname, "", PARSE_ACCESS); + if (strcmp("ftp", acc_method) == 0 || + (!LYSameHostname("localhost", nodename) && + !LYSameHostname(nodename, HTHostName()))) { + status = -1; + FREE(newname); + FREE(filename); + FREE(nodename); + FREE(acc_method); +#ifndef DISABLE_FTP + ftp_newhost = HTParse(addr, "", PARSE_HOST); + if (strcmp(ftp_lasthost, ftp_newhost)) + ftp_local_passive = ftp_passive; + + status = HTFTPLoad(addr, anchor, format_out, sink); + + if (ftp_passive == ftp_local_passive) { + if ((status >= 400) || (status < 0)) { + ftp_local_passive = (BOOLEAN) !ftp_passive; + status = HTFTPLoad(addr, anchor, format_out, sink); + } + } + + free(ftp_lasthost); + ftp_lasthost = ftp_newhost; +#endif /* DISABLE_FTP */ + return status; + } else { + FREE(newname); + FREE(acc_method); + } +#if defined(VMS) || defined(USE_DOS_DRIVES) + HTUnEscape(filename); +#endif /* VMS */ + + /* + * Determine the format and encoding mapped to any suffix. + */ + if (anchor->content_type && anchor->content_encoding) { + /* + * If content_type and content_encoding are BOTH already set in the + * anchor object, we believe it and don't try to derive format and + * encoding from the filename. - kw + */ + format = HTAtom_for(anchor->content_type); + myEncoding = HTAtom_for(anchor->content_encoding); + } else { + int default_UCLYhndl = UCLYhndl_HTFile_for_unspec; + + if (force_old_UCLYhndl_on_reload) { + force_old_UCLYhndl_on_reload = FALSE; + default_UCLYhndl = forced_UCLYhdnl; + } + + format = HTFileFormat(filename, &myEncoding, NULL); + + /* + * Check the format for an extended MIME charset value, and act on it + * if present. Otherwise, assume what is indicated by the last + * parameter (fallback will effectively be UCLYhndl_for_unspec, by + * default ISO-8859-1). - kw + */ + format = HTCharsetFormat(format, anchor, default_UCLYhndl); + } + +#ifdef VMS + /* + * Check to see if the 'filename' is in fact a directory. If it is create + * a new hypertext object containing a list of files and subdirectories + * contained in the directory. All of these are links to the directories + * or files listed. + */ + if (HTStat(filename, &stat_info) == -1) { + CTRACE((tfp, "HTLoadFile: Can't stat %s\n", filename)); + } else { + if (S_ISDIR(stat_info.st_mode)) { + if (HTDirAccess == HT_DIR_FORBID) { + FREE(filename); + FREE(nodename); + return HTLoadError(sink, 403, DISALLOWED_DIR_SCAN); + } + + if (HTDirAccess == HT_DIR_SELECTIVE) { + char *enable_file_name = NULL; + + HTSprintf0(&enable_file_name, "%s/%s", filename, HT_DIR_ENABLE_FILE); + if (HTStat(enable_file_name, &stat_info) == -1) { + FREE(filename); + FREE(nodename); + FREE(enable_file_name); + return HTLoadError(sink, 403, DISALLOWED_SELECTIVE_ACCESS); + } + } + + FREE(filename); + FREE(nodename); + return HTVMSBrowseDir(addr, anchor, format_out, sink); + } + } + + if (decompressAndParse(anchor, + format_out, + sink, + nodename, + filename, + myEncoding, + format, + &status)) { + FREE(nodename); + FREE(filename); + return status; + } + FREE(filename); + +#else /* not VMS: */ + + FREE(filename); + + /* + * For unix, we try to translate the name into the name of a transparently + * mounted file. + * + * Not allowed in secure (HTClientHost) situations. TBL 921019 + */ +#ifndef NO_UNIX_IO + /* Need protection here for telnet server but not httpd server. */ + + if (!HTSecure) { /* try local file system */ + char *localname = HTLocalName(addr); + struct stat dir_info; + +#ifdef HAVE_READDIR + /* + * Multiformat handling. + * + * If needed, scan directory to find a good file. Bug: We don't stat + * the file to find the length. + */ + if ((strlen(localname) > strlen(MULTI_SUFFIX)) && + (0 == strcmp(localname + strlen(localname) - strlen(MULTI_SUFFIX), + MULTI_SUFFIX))) { + DIR *dp = 0; + BOOL forget_multi = NO; + + STRUCT_DIRENT *dirbuf; + float best = (float) NO_VALUE_FOUND; /* So far best is bad */ + HTFormat best_rep = NULL; /* Set when rep found */ + HTAtom *best_enc = NULL; + char *best_name = NULL; /* Best dir entry so far */ + + char *base = strrchr(localname, '/'); + size_t baselen = 0; + + if (!base || base == localname) { + forget_multi = YES; + } else { + *base++ = '\0'; /* Just got directory name */ + baselen = strlen(base) - strlen(MULTI_SUFFIX); + base[baselen] = '\0'; /* Chop off suffix */ + + dp = opendir(localname); + } + if (forget_multi || !dp) { + FREE(localname); + FREE(nodename); + return HTLoadError(sink, 500, FAILED_DIR_SCAN); + } + + while ((dirbuf = readdir(dp)) != NULL) { + /* + * While there are directory entries to be read... + */ +#ifdef STRUCT_DIRENT__D_INO + if (dirbuf->d_ino == 0) + continue; /* if the entry is not being used, skip it */ +#endif + if (strlen(dirbuf->d_name) > baselen && /* Match? */ + !StrNCmp(dirbuf->d_name, base, baselen)) { + HTAtom *enc; + HTFormat rep = HTFileFormat(dirbuf->d_name, &enc, NULL); + float filevalue = HTFileValue(dirbuf->d_name); + float value = HTStackValue(rep, format_out, + filevalue, + 0L /* @@@@@@ */ ); + + if (value <= 0.0) { + int rootlen = 0; + const char *atomname = NULL; + CompressFileType cft = + HTCompressFileType(dirbuf->d_name, ".", &rootlen); + char *cp = NULL; + + enc = NULL; + if (cft != cftNone) { + StrAllocCopy(cp, dirbuf->d_name); + cp[rootlen] = '\0'; + format = HTFileFormat(cp, NULL, NULL); + FREE(cp); + value = HTStackValue(format, format_out, + filevalue, 0L); + } + switch (cft) { + case cftCompress: + atomname = "application/x-compressed"; + break; + case cftGzip: + atomname = "application/x-gzip"; + break; + case cftDeflate: + atomname = "application/x-deflate"; + break; + case cftBzip2: + atomname = "application/x-bzip2"; + break; + case cftNone: + break; + } + + if (atomname != NULL) { + value = HTStackValue(format, format_out, + filevalue, 0L); + if (value <= 0.0) { + format = HTAtom_for(atomname); + value = HTStackValue(format, format_out, + filevalue, 0L); + } + if (value <= 0.0) { + format = HTAtom_for("www/compressed"); + value = HTStackValue(format, format_out, + filevalue, 0L); + } + } + } + if (value < NO_VALUE_FOUND) { + CTRACE((tfp, + "HTLoadFile: value of presenting %s is %f\n", + HTAtom_name(rep), value)); + if (value > best) { + best_rep = rep; + best_enc = enc; + best = value; + StrAllocCopy(best_name, dirbuf->d_name); + } + } /* if best so far */ + } + /* if match */ + } /* end while directory entries left to read */ + closedir(dp); + + if (best_rep) { + format = best_rep; + myEncoding = best_enc; + base[-1] = '/'; /* Restore directory name */ + base[0] = '\0'; + StrAllocCat(localname, best_name); + FREE(best_name); + } else { /* If not found suitable file */ + FREE(localname); + FREE(nodename); + return HTLoadError(sink, 403, FAILED_NO_REPRESENTATION); + } + /*NOTREACHED */ + } + /* if multi suffix */ + /* + * Check to see if the 'localname' is in fact a directory. If it is + * create a new hypertext object containing a list of files and + * subdirectories contained in the directory. All of these are links + * to the directories or files listed. NB This assumes the existence + * of a type 'STRUCT_DIRENT', which will hold the directory entry, and + * a type 'DIR' which is used to point to the current directory being + * read. + */ +#if defined(USE_DOS_DRIVES) + if (strlen(localname) == 2 && LYIsDosDrive(localname)) + LYAddPathSep(&localname); +#endif + if (HTStat(localname, &dir_info) == -1) /* get file information */ + { + /* if can't read file information */ + CTRACE((tfp, "HTLoadFile: can't stat %s\n", localname)); + + } else { /* Stat was OK */ + + if (S_ISDIR(dir_info.st_mode)) { + /* + * If localname is a directory. + */ + DIR *dp; + struct stat file_info; + + CTRACE((tfp, "%s is a directory\n", localname)); + + /* + * Check directory access. Selective access means only those + * directories containing a marker file can be browsed. + */ + if (HTDirAccess == HT_DIR_FORBID) { + FREE(localname); + FREE(nodename); + return HTLoadError(sink, 403, DISALLOWED_DIR_SCAN); + } + + if (HTDirAccess == HT_DIR_SELECTIVE) { + char *enable_file_name = NULL; + + HTSprintf0(&enable_file_name, "%s/%s", localname, HT_DIR_ENABLE_FILE); + if (stat(enable_file_name, &file_info) != 0) { + FREE(localname); + FREE(nodename); + FREE(enable_file_name); + return HTLoadError(sink, 403, DISALLOWED_SELECTIVE_ACCESS); + } + } + + CTRACE((tfp, "Opening directory %s\n", localname)); + dp = opendir(localname); + if (!dp) { + FREE(localname); + FREE(nodename); + return HTLoadError(sink, 403, FAILED_DIR_UNREADABLE); + } + + /* + * Directory access is allowed and possible. + */ + + status = print_local_dir(dp, localname, + anchor, format_out, sink); + closedir(dp); + FREE(localname); + FREE(nodename); + return status; /* document loaded, maybe partial */ + + } + /* end if localname is a directory */ + if (S_ISREG(dir_info.st_mode)) { +#ifdef LONG_MAX + if (dir_info.st_size <= LONG_MAX) +#endif + anchor->content_length = (long) dir_info.st_size; + } + + } /* end if file stat worked */ + +/* End of directory reading section +*/ +#endif /* HAVE_READDIR */ + if (decompressAndParse(anchor, + format_out, + sink, + nodename, + localname, + myEncoding, + format, + &status)) { + FREE(nodename); + FREE(localname); + return status; + } + FREE(localname); + } /* local unix file system */ +#endif /* !NO_UNIX_IO */ +#endif /* VMS */ + +#ifndef DECNET + /* + * Now, as transparently mounted access has failed, we try FTP. + */ + { + /* + * Deal with case-sensitivity differences on VMS versus Unix. + */ +#ifdef VMS + if (strcasecomp(nodename, HTHostName()) != 0) +#else + if (strcmp(nodename, HTHostName()) != 0) +#endif /* VMS */ + { + status = -1; + FREE(nodename); + if (StrNCmp(addr, "file://localhost", 16)) { + /* never go to ftp site when URL + * is file://localhost + */ +#ifndef DISABLE_FTP + status = HTFTPLoad(addr, anchor, format_out, sink); +#endif /* DISABLE_FTP */ + } + return status; + } + FREE(nodename); + } +#endif /* !DECNET */ + + /* + * All attempts have failed. + */ + { + CTRACE((tfp, "Can't open `%s', errno=%d\n", addr, SOCKET_ERRNO)); + + return HTLoadError(sink, 403, FAILED_FILE_UNREADABLE); + } +} + +static const char *program_paths[pp_Last]; + +/* + * Given a program number, return its path + */ +const char *HTGetProgramPath(ProgramPaths code) +{ + const char *result = NULL; + + if (code > ppUnknown && code < pp_Last) + result = program_paths[code]; + return result; +} + +/* + * Store a program's path. The caller must allocate the string used for 'path', + * since HTInitProgramPaths() may free it. + */ +void HTSetProgramPath(ProgramPaths code, const char *path) +{ + if (code > ppUnknown && code < pp_Last) { + program_paths[code] = isEmpty(path) ? 0 : path; + } +} + +/* + * Reset the list of known program paths to the ones that are compiled-in + */ +void HTInitProgramPaths(BOOL init) +{ + ProgramPaths code; + int n; + const char *path; + const char *test; + + for (n = (int) ppUnknown + 1; n < (int) pp_Last; ++n) { + switch (code = (ProgramPaths) n) { +#ifdef BZIP2_PATH + case ppBZIP2: + path = BZIP2_PATH; + break; +#endif +#ifdef CHMOD_PATH + case ppCHMOD: + path = CHMOD_PATH; + break; +#endif +#ifdef COMPRESS_PATH + case ppCOMPRESS: + path = COMPRESS_PATH; + break; +#endif +#ifdef COPY_PATH + case ppCOPY: + path = COPY_PATH; + break; +#endif +#ifdef CSWING_PATH + case ppCSWING: + path = CSWING_PATH; + break; +#endif +#ifdef GZIP_PATH + case ppGZIP: + path = GZIP_PATH; + break; +#endif +#ifdef INFLATE_PATH + case ppINFLATE: + path = INFLATE_PATH; + break; +#endif +#ifdef INSTALL_PATH + case ppINSTALL: + path = INSTALL_PATH; + break; +#endif +#ifdef MKDIR_PATH + case ppMKDIR: + path = MKDIR_PATH; + break; +#endif +#ifdef MV_PATH + case ppMV: + path = MV_PATH; + break; +#endif +#ifdef RLOGIN_PATH + case ppRLOGIN: + path = RLOGIN_PATH; + break; +#endif +#ifdef RM_PATH + case ppRM: + path = RM_PATH; + break; +#endif +#ifdef RMDIR_PATH + case ppRMDIR: + path = RMDIR_PATH; + break; +#endif +#ifdef SETFONT_PATH + case ppSETFONT: + path = SETFONT_PATH; + break; +#endif +#ifdef TAR_PATH + case ppTAR: + path = TAR_PATH; + break; +#endif +#ifdef TELNET_PATH + case ppTELNET: + path = TELNET_PATH; + break; +#endif +#ifdef TN3270_PATH + case ppTN3270: + path = TN3270_PATH; + break; +#endif +#ifdef TOUCH_PATH + case ppTOUCH: + path = TOUCH_PATH; + break; +#endif +#ifdef UNCOMPRESS_PATH + case ppUNCOMPRESS: + path = UNCOMPRESS_PATH; + break; +#endif +#ifdef UNZIP_PATH + case ppUNZIP: + path = UNZIP_PATH; + break; +#endif +#ifdef UUDECODE_PATH + case ppUUDECODE: + path = UUDECODE_PATH; + break; +#endif +#ifdef ZCAT_PATH + case ppZCAT: + path = ZCAT_PATH; + break; +#endif +#ifdef ZIP_PATH + case ppZIP: + path = ZIP_PATH; + break; +#endif + default: + path = NULL; + break; + } + test = HTGetProgramPath(code); + if (test != NULL && test != path) { + free(DeConst(test)); + } + if (init) { + HTSetProgramPath(code, path); + } + } +} + +/* + * Protocol descriptors + */ +#ifdef GLOBALDEF_IS_MACRO +#define _HTFILE_C_1_INIT { "ftp", HTLoadFile, 0 } +GLOBALDEF(HTProtocol, HTFTP, _HTFILE_C_1_INIT); +#define _HTFILE_C_2_INIT { "file", HTLoadFile, HTFileSaveStream } +GLOBALDEF(HTProtocol, HTFile, _HTFILE_C_2_INIT); +#else +GLOBALDEF HTProtocol HTFTP = +{"ftp", HTLoadFile, 0}; +GLOBALDEF HTProtocol HTFile = +{"file", HTLoadFile, HTFileSaveStream}; +#endif /* GLOBALDEF_IS_MACRO */ |