X-Git-Url: http://sjero.net/git/?p=wget;a=blobdiff_plain;f=src%2Fhttp.c;h=748b4e8142a99bf28eed6d3248fd89fb61b89241;hp=b304af76cde67c48ac3b05c864f56c4fd6b406e0;hb=19ef7249b8679e7a88067dca3db37e947408e15f;hpb=ad21f8e2b774367156fbe71f4318b5646059b724 diff --git a/src/http.c b/src/http.c index b304af76..748b4e81 100644 --- a/src/http.c +++ b/src/http.c @@ -1,6 +1,7 @@ /* HTTP support. - Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, - 2004, 2005, 2006, 2007, 2008 Free Software Foundation, Inc. + Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, + 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software Foundation, + Inc. This file is part of GNU Wget. @@ -33,9 +34,7 @@ as that of the covered work. */ #include #include #include -#ifdef HAVE_UNISTD_H -# include -#endif +#include #include #include #include @@ -56,9 +55,7 @@ as that of the covered work. */ # include "http-ntlm.h" #endif #include "cookies.h" -#ifdef ENABLE_DIGEST -# include "gen-md5.h" -#endif +#include "md5.h" #include "convert.h" #include "spider.h" @@ -66,14 +63,20 @@ as that of the covered work. */ #include "test.h" #endif +#ifdef __VMS +# include "vms.h" +#endif /* def __VMS */ + extern char *version_string; /* Forward decls. */ +struct http_stat; static char *create_authorization_line (const char *, const char *, const char *, const char *, const char *, bool *); static char *basic_authentication_encode (const char *, const char *); static bool known_authentication_scheme_p (const char *, const char *); +static void ensure_extension (struct http_stat *, const char *, int *); static void load_cookies (void); #ifndef MIN @@ -86,8 +89,10 @@ static struct cookie_jar *wget_cookie_jar; #define TEXTHTML_S "text/html" #define TEXTXHTML_S "application/xhtml+xml" +#define TEXTCSS_S "text/css" /* Some status code validation macros: */ +#define H_10X(x) (((x) >= 100) && ((x) < 200)) #define H_20X(x) (((x) >= 200) && ((x) < 300)) #define H_PARTIAL(x) ((x) == HTTP_STATUS_PARTIAL_CONTENTS) #define H_REDIRECTED(x) ((x) == HTTP_STATUS_MOVED_PERMANENTLY \ @@ -139,6 +144,8 @@ struct request { int hcount, hcapacity; }; +extern int numurls; + /* Create a new, empty request. At least request_set_method must be called before the request can be used. */ @@ -343,7 +350,7 @@ request_send (const struct request *req, int fd) APPEND (p, req->method); *p++ = ' '; APPEND (p, req->arg); *p++ = ' '; - memcpy (p, "HTTP/1.0\r\n", 10); p += 10; + memcpy (p, "HTTP/1.1\r\n", 10); p += 10; for (i = 0; i < req->hcount; i++) { @@ -397,19 +404,19 @@ maybe_send_basic_creds (const char *hostname, const char *user, if (opt.auth_without_challenge) { - DEBUGP(("Auth-without-challenge set, sending Basic credentials.\n")); + DEBUGP (("Auth-without-challenge set, sending Basic credentials.\n")); do_challenge = true; } else if (basic_authed_hosts && hash_table_contains(basic_authed_hosts, hostname)) { - DEBUGP(("Found `%s' in basic_authed_hosts.\n", hostname)); + DEBUGP (("Found %s in basic_authed_hosts.\n", quote (hostname))); do_challenge = true; } else { - DEBUGP(("Host `%s' has not issued a general basic challenge.\n", - hostname)); + DEBUGP (("Host %s has not issued a general basic challenge.\n", + quote (hostname))); } if (do_challenge) { @@ -430,7 +437,7 @@ register_basic_auth_host (const char *hostname) if (!hash_table_contains(basic_authed_hosts, hostname)) { hash_table_put (basic_authed_hosts, xstrdup(hostname), NULL); - DEBUGP(("Inserted `%s' into basic_authed_hosts\n", hostname)); + DEBUGP (("Inserted %s into basic_authed_hosts\n", quote (hostname))); } } @@ -765,7 +772,7 @@ resp_status (const struct response *resp, char **message) while (p < end && c_isdigit (*p)) ++p; if (p < end && *p == '.') - ++p; + ++p; while (p < end && c_isdigit (*p)) ++p; } @@ -810,7 +817,8 @@ print_response_line(const char *prefix, const char *b, const char *e) { char *copy; BOUNDED_TO_ALLOCA(b, e, copy); - logprintf (LOG_ALWAYS, "%s%s\n", prefix, escnonprint(copy)); + logprintf (LOG_ALWAYS, "%s%s\n", prefix, + quotearg_style (escape_quoting_style, copy)); } /* Print the server response, line by line, omitting the trailing CRLF @@ -891,29 +899,49 @@ parse_content_range (const char *hdr, wgint *first_byte_ptr, mode, the body is displayed for debugging purposes. */ static bool -skip_short_body (int fd, wgint contlen) +skip_short_body (int fd, wgint contlen, bool chunked) { enum { SKIP_SIZE = 512, /* size of the download buffer */ SKIP_THRESHOLD = 4096 /* the largest size we read */ }; + wgint remaining_chunk_size = 0; char dlbuf[SKIP_SIZE + 1]; dlbuf[SKIP_SIZE] = '\0'; /* so DEBUGP can safely print it */ - /* We shouldn't get here with unknown contlen. (This will change - with HTTP/1.1, which supports "chunked" transfer.) */ - assert (contlen != -1); + assert (contlen != -1 || contlen); /* If the body is too large, it makes more sense to simply close the connection than to try to read the body. */ if (contlen > SKIP_THRESHOLD) return false; - DEBUGP (("Skipping %s bytes of body: [", number_to_static_string (contlen))); - - while (contlen > 0) + while (contlen > 0 || chunked) { - int ret = fd_read (fd, dlbuf, MIN (contlen, SKIP_SIZE), -1); + int ret; + if (chunked) + { + if (remaining_chunk_size == 0) + { + char *line = fd_read_line (fd); + char *endl; + if (line == NULL) + break; + + remaining_chunk_size = strtol (line, &endl, 16); + if (remaining_chunk_size == 0) + { + fd_read_line (fd); + break; + } + } + + contlen = MIN (remaining_chunk_size, SKIP_SIZE); + } + + DEBUGP (("Skipping %s bytes of body: [", number_to_static_string (contlen))); + + ret = fd_read (fd, dlbuf, MIN (contlen, SKIP_SIZE), -1); if (ret <= 0) { /* Don't normally report the error since this is an @@ -923,6 +951,15 @@ skip_short_body (int fd, wgint contlen) return false; } contlen -= ret; + + if (chunked) + { + remaining_chunk_size -= ret; + if (remaining_chunk_size == 0) + if (fd_read_line (fd) == NULL) + return false; + } + /* Safe even if %.*s bogusly expects terminating \0 because we've zero-terminated dlbuf above. */ DEBUGP (("%.*s", ret, dlbuf)); @@ -932,6 +969,66 @@ skip_short_body (int fd, wgint contlen) return true; } +#define NOT_RFC2231 0 +#define RFC2231_NOENCODING 1 +#define RFC2231_ENCODING 2 + +/* extract_param extracts the parameter name into NAME. + However, if the parameter name is in RFC2231 format then + this function adjusts NAME by stripping of the trailing + characters that are not part of the name but are present to + indicate the presence of encoding information in the value + or a fragment of a long parameter value +*/ +static int +modify_param_name(param_token *name) +{ + const char *delim1 = memchr (name->b, '*', name->e - name->b); + const char *delim2 = memrchr (name->b, '*', name->e - name->b); + + int result; + + if(delim1 == NULL) + { + result = NOT_RFC2231; + } + else if(delim1 == delim2) + { + if ((name->e - 1) == delim1) + { + result = RFC2231_ENCODING; + } + else + { + result = RFC2231_NOENCODING; + } + name->e = delim1; + } + else + { + name->e = delim1; + result = RFC2231_ENCODING; + } + return result; +} + +/* extract_param extract the paramater value into VALUE. + Like modify_param_name this function modifies VALUE by + stripping off the encoding information from the actual value +*/ +static void +modify_param_value (param_token *value, int encoding_type ) +{ + if (RFC2231_ENCODING == encoding_type) + { + const char *delim = memrchr (value->b, '\'', value->e - value->b); + if ( delim != NULL ) + { + value->b = (delim+1); + } + } +} + /* Extract a parameter from the string (typically an HTTP header) at **SOURCE and advance SOURCE to the next parameter. Return false when there are no more parameters to extract. The name of the @@ -1003,9 +1100,31 @@ extract_param (const char **source, param_token *name, param_token *value, if (*p == separator) ++p; } *source = p; + + int param_type = modify_param_name(name); + if (NOT_RFC2231 != param_type) + { + modify_param_value(value, param_type); + } return true; } +#undef NOT_RFC2231 +#undef RFC2231_NOENCODING +#undef RFC2231_ENCODING + +/* Appends the string represented by VALUE to FILENAME */ + +static void +append_value_to_filename (char **filename, param_token const * const value) +{ + int original_length = strlen(*filename); + int new_length = strlen(*filename) + (value->e - value->b); + *filename = xrealloc (*filename, new_length+1); + memcpy (*filename + original_length, value->b, (value->e - value->b)); + (*filename)[new_length] = '\0'; +} + #undef MAX #define MAX(p, q) ((p) > (q) ? (p) : (q)) @@ -1024,47 +1143,46 @@ extract_param (const char **source, param_token *name, param_token *value, false. The file name is stripped of directory components and must not be - empty. */ + empty. + + Historically, this function returned filename prefixed with opt.dir_prefix, + now that logic is handled by the caller, new code should pay attention, + changed by crq, Sep 2010. +*/ static bool parse_content_disposition (const char *hdr, char **filename) { param_token name, value; + *filename = NULL; while (extract_param (&hdr, &name, &value, ';')) - if (BOUNDED_EQUAL_NO_CASE (name.b, name.e, "filename") && value.b != NULL) - { - /* Make the file name begin at the last slash or backslash. */ - const char *last_slash = memrchr (value.b, '/', value.e - value.b); - const char *last_bs = memrchr (value.b, '\\', value.e - value.b); - if (last_slash && last_bs) - value.b = 1 + MAX (last_slash, last_bs); - else if (last_slash || last_bs) - value.b = 1 + (last_slash ? last_slash : last_bs); - if (value.b == value.e) - continue; - /* Start with the directory prefix, if specified. */ - if (opt.dir_prefix) - { - int prefix_length = strlen (opt.dir_prefix); - bool add_slash = (opt.dir_prefix[prefix_length - 1] != '/'); - int total_length; - - if (add_slash) - ++prefix_length; - total_length = prefix_length + (value.e - value.b); - *filename = xmalloc (total_length + 1); - strcpy (*filename, opt.dir_prefix); - if (add_slash) - (*filename)[prefix_length - 1] = '/'; - memcpy (*filename + prefix_length, value.b, (value.e - value.b)); - (*filename)[total_length] = '\0'; - } - else - *filename = strdupdelim (value.b, value.e); - return true; - } - return false; + { + int isFilename = BOUNDED_EQUAL_NO_CASE ( name.b, name.e, "filename" ); + if ( isFilename && value.b != NULL) + { + /* Make the file name begin at the last slash or backslash. */ + const char *last_slash = memrchr (value.b, '/', value.e - value.b); + const char *last_bs = memrchr (value.b, '\\', value.e - value.b); + if (last_slash && last_bs) + value.b = 1 + MAX (last_slash, last_bs); + else if (last_slash || last_bs) + value.b = 1 + (last_slash ? last_slash : last_bs); + if (value.b == value.e) + continue; + + if (*filename) + append_value_to_filename (filename, &value); + else + *filename = strdupdelim (value.b, value.e); + } + } + + if (*filename) + return true; + else + return false; } + /* Persistent connections. Currently, we cache the most recently used connection as persistent, provided that the HTTP server agrees to @@ -1295,6 +1413,7 @@ struct http_stat char *remote_time; /* remote time-stamp string */ char *error; /* textual HTTP error */ int statcode; /* status code */ + char *message; /* status message */ wgint rd_size; /* amount of data read from socket */ double dltime; /* time it took to download the data */ const char *referer; /* value of the referer header. */ @@ -1303,12 +1422,12 @@ struct http_stat existence after having begun to download (needed in gethttp for when connection is interrupted/restarted. */ - bool timestamp_checked; /* true if pre-download time-stamping checks + bool timestamp_checked; /* true if pre-download time-stamping checks * have already been performed */ char *orig_file_name; /* name of file to compare for time-stamping * (might be != local_file if -K is set) */ wgint orig_file_size; /* size of file to compare for time-stamping */ - time_t orig_file_tstamp; /* time-stamp of file to compare for + time_t orig_file_tstamp; /* time-stamp of file to compare for * time-stamping */ }; @@ -1321,6 +1440,7 @@ free_hstat (struct http_stat *hs) xfree_null (hs->rderrmsg); xfree_null (hs->local_file); xfree_null (hs->orig_file_name); + xfree_null (hs->message); /* Guard against being called twice. */ hs->newloc = NULL; @@ -1328,18 +1448,46 @@ free_hstat (struct http_stat *hs) hs->error = NULL; } +static void +get_file_flags (const char *filename, int *dt) +{ + logprintf (LOG_VERBOSE, _("\ +File %s already there; not retrieving.\n\n"), quote (filename)); + /* If the file is there, we suppose it's retrieved OK. */ + *dt |= RETROKF; + + /* #### Bogusness alert. */ + /* If its suffix is "html" or "htm" or similar, assume text/html. */ + if (has_html_suffix_p (filename)) + *dt |= TEXTHTML; +} + #define BEGINS_WITH(line, string_constant) \ (!strncasecmp (line, string_constant, sizeof (string_constant) - 1) \ && (c_isspace (line[sizeof (string_constant) - 1]) \ || !line[sizeof (string_constant) - 1])) +#ifdef __VMS +#define SET_USER_AGENT(req) do { \ + if (!opt.useragent) \ + request_set_header (req, "User-Agent", \ + aprintf ("Wget/%s (VMS %s %s)", \ + version_string, vms_arch(), vms_vers()), \ + rel_value); \ + else if (*opt.useragent) \ + request_set_header (req, "User-Agent", opt.useragent, rel_none); \ +} while (0) +#else /* def __VMS */ #define SET_USER_AGENT(req) do { \ if (!opt.useragent) \ request_set_header (req, "User-Agent", \ - aprintf ("Wget/%s", version_string), rel_value); \ + aprintf ("Wget/%s (%s)", \ + version_string, OS_TYPE), \ + rel_value); \ else if (*opt.useragent) \ request_set_header (req, "User-Agent", opt.useragent, rel_none); \ } while (0) +#endif /* def __VMS [else] */ /* The flags that allow clobbering the file (opening with "wb"). Defined here to avoid repetition later. #### This will require @@ -1358,7 +1506,8 @@ free_hstat (struct http_stat *hs) If PROXY is non-NULL, the connection will be made to the proxy server, and u->url will be requested. */ static uerr_t -gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) +gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy, + struct iri *iri, int count) { struct request *req; @@ -1402,15 +1551,12 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) is done. */ bool keep_alive; - /* Whether keep-alive should be inhibited. + /* Is the server using the chunked transfer encoding? */ + bool chunked_transfer_encoding = false; - RFC 2068 requests that 1.0 clients not send keep-alive requests - to proxies. This is because many 1.0 proxies do not interpret - the Connection header and transfer it to the remote server, - causing it to not close the connection and leave both the proxy - and the client hanging. */ + /* Whether keep-alive should be inhibited. */ bool inhibit_keep_alive = - !opt.http_keep_alive || opt.ignore_length || proxy != NULL; + !opt.http_keep_alive || opt.ignore_length; /* Headers sent when using POST. */ wgint post_data_size = 0; @@ -1440,6 +1586,7 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) hs->newloc = NULL; hs->remote_time = NULL; hs->error = NULL; + hs->message = NULL; conn = u; @@ -1472,7 +1619,13 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) request_set_header (req, "Referer", (char *) hs->referer, rel_none); if (*dt & SEND_NOCACHE) - request_set_header (req, "Pragma", "no-cache", rel_none); + { + /* Cache-Control MUST be obeyed by all HTTP/1.1 caching mechanisms... */ + request_set_header (req, "Cache-Control", "no-cache, must-revalidate", rel_none); + + /* ... but some HTTP/1.0 caches doesn't implement Cache-Control. */ + request_set_header (req, "Pragma", "no-cache", rel_none); + } if (hs->restval) request_set_header (req, "Range", aprintf ("bytes=%s-", @@ -1488,9 +1641,10 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) user = user ? user : (opt.http_user ? opt.http_user : opt.user); passwd = passwd ? passwd : (opt.http_passwd ? opt.http_passwd : opt.passwd); - if (user && passwd - && !u->user) /* We only do "site-wide" authentication with "global" - user/password values; URL user/password info overrides. */ + /* We only do "site-wide" authentication with "global" user/password + * values unless --auth-no-challange has been requested; URL user/password + * info overrides. */ + if (user && passwd && (!u->user || opt.auth_without_challenge)) { /* If this is a host for which we've already received a Basic * challenge, we'll go ahead and send Basic authentication creds. */ @@ -1518,20 +1672,18 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) rel_value); } - if (!inhibit_keep_alive) - request_set_header (req, "Connection", "Keep-Alive", rel_none); - - if (opt.cookies) - request_set_header (req, "Cookie", - cookie_header (wget_cookie_jar, - u->host, u->port, u->path, -#ifdef HAVE_SSL - u->scheme == SCHEME_HTTPS -#else - 0 -#endif - ), - rel_value); + if (inhibit_keep_alive) + request_set_header (req, "Connection", "Close", rel_none); + else + { + if (proxy == NULL) + request_set_header (req, "Connection", "Keep-Alive", rel_none); + else + { + request_set_header (req, "Connection", "Close", rel_none); + request_set_header (req, "Proxy-Connection", "Keep-Alive", rel_none); + } + } if (opt.post_data || opt.post_file_name) { @@ -1544,8 +1696,8 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) post_data_size = file_size (opt.post_file_name); if (post_data_size == -1) { - logprintf (LOG_NOTQUIET, _("POST data file `%s' missing: %s\n"), - opt.post_file_name, strerror (errno)); + logprintf (LOG_NOTQUIET, _("POST data file %s missing: %s\n"), + quote (opt.post_file_name), strerror (errno)); post_data_size = 0; } } @@ -1554,6 +1706,23 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) rel_value); } + retry_with_auth: + /* We need to come back here when the initial attempt to retrieve + without authorization header fails. (Expected to happen at least + for the Digest authorization scheme.) */ + + if (opt.cookies) + request_set_header (req, "Cookie", + cookie_header (wget_cookie_jar, + u->host, u->port, u->path, +#ifdef HAVE_SSL + u->scheme == SCHEME_HTTPS +#else + 0 +#endif + ), + rel_value); + /* Add the user headers. */ if (opt.user_headers) { @@ -1562,11 +1731,6 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) request_set_user_header (req, opt.user_headers[i]); } - retry_with_auth: - /* We need to come back here when the initial attempt to retrieve - without authorization header fails. (Expected to happen at least - for the Digest authorization scheme.) */ - proxyauth = NULL; if (proxy) { @@ -1602,11 +1766,13 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) request_set_header (req, "Proxy-Authorization", proxyauth, rel_value); } - keep_alive = false; + keep_alive = true; /* Establish the connection. */ - if (!inhibit_keep_alive) + if (inhibit_keep_alive) + keep_alive = false; + else { /* Look for a persistent connection to target host, unless a proxy is used. The exception is when SSL is in use, in which @@ -1629,7 +1795,8 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) sock = pconn.socket; using_ssl = pconn.ssl; logprintf (LOG_VERBOSE, _("Reusing existing connection to %s:%d.\n"), - escnonprint (pconn.host), pconn.port); + quotearg_style (escape_quoting_style, pconn.host), + pconn.port); DEBUGP (("Reusing fd %d.\n", sock)); if (pconn.authorized) /* If the connection is already authorized, the "Basic" @@ -1641,8 +1808,8 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) { request_free (req); logprintf(LOG_NOTQUIET, - _("%s: unable to resolve host address `%s'\n"), - exec_name, relevant->host); + _("%s: unable to resolve host address %s\n"), + exec_name, quote (relevant->host)); return HOSTERR; } } @@ -1711,13 +1878,24 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) resp = resp_new (head); statcode = resp_status (resp, &message); + if (statcode < 0) + { + char *tms = datetime_str (time (NULL)); + logprintf (LOG_VERBOSE, "%d\n", statcode); + logprintf (LOG_NOTQUIET, _("%s ERROR %d: %s.\n"), tms, statcode, + quotearg_style (escape_quoting_style, + _("Malformed status line"))); + xfree (head); + return HERR; + } + hs->message = xstrdup (message); resp_free (resp); xfree (head); if (statcode != 200) { failed_tunnel: logprintf (LOG_NOTQUIET, _("Proxy tunneling failed: %s"), - message ? escnonprint (message) : "?"); + message ? quotearg_style (escape_quoting_style, message) : "?"); xfree_null (message); return CONSSLERR; } @@ -1731,11 +1909,16 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) if (conn->scheme == SCHEME_HTTPS) { - if (!ssl_connect (sock) || !ssl_check_certificate (sock, u->host)) + if (!ssl_connect_wget (sock)) { fd_close (sock); return CONSSLERR; } + else if (!ssl_check_certificate (sock, u->host)) + { + fd_close (sock); + return VERIFCERTERR; + } using_ssl = true; } #endif /* HAVE_SSL */ @@ -1767,6 +1950,7 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) contrange = 0; *dt &= ~RETROKF; +read_header: head = read_http_response_head (sock); if (!head) { @@ -1793,119 +1977,34 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) /* Check for status line. */ message = NULL; statcode = resp_status (resp, &message); - if (!opt.server_response) - logprintf (LOG_VERBOSE, "%2d %s\n", statcode, - message ? escnonprint (message) : ""); - else + if (statcode < 0) { - logprintf (LOG_VERBOSE, "\n"); - print_server_response (resp, " "); + char *tms = datetime_str (time (NULL)); + logprintf (LOG_VERBOSE, "%d\n", statcode); + logprintf (LOG_NOTQUIET, _("%s ERROR %d: %s.\n"), tms, statcode, + quotearg_style (escape_quoting_style, + _("Malformed status line"))); + CLOSE_INVALIDATE (sock); + request_free (req); + xfree (head); + return HERR; } - /* Determine the local filename if needed. Notice that if -O is used - * hstat.local_file is set by http_loop to the argument of -O. */ - if (!hs->local_file) + if (H_10X (statcode)) { - /* Honor Content-Disposition whether possible. */ - if (!opt.content_disposition - || !resp_header_copy (resp, "Content-Disposition", - hdrval, sizeof (hdrval)) - || !parse_content_disposition (hdrval, &hs->local_file)) - { - /* The Content-Disposition header is missing or broken. - * Choose unique file name according to given URL. */ - hs->local_file = url_file_name (u); - } + DEBUGP (("Ignoring response\n")); + xfree (head); + goto read_header; } - - /* TODO: perform this check only once. */ - if (!hs->existence_checked && file_exists_p (hs->local_file)) - { - if (opt.noclobber) - { - /* If opt.noclobber is turned on and file already exists, do not - retrieve the file */ - logprintf (LOG_VERBOSE, _("\ -File `%s' already there; not retrieving.\n\n"), hs->local_file); - /* If the file is there, we suppose it's retrieved OK. */ - *dt |= RETROKF; - - /* #### Bogusness alert. */ - /* If its suffix is "html" or "htm" or similar, assume text/html. */ - if (has_html_suffix_p (hs->local_file)) - *dt |= TEXTHTML; - return RETRUNNEEDED; - } - else if (!ALLOW_CLOBBER) - { - char *unique = unique_name (hs->local_file, true); - if (unique != hs->local_file) - xfree (hs->local_file); - hs->local_file = unique; - } - } - hs->existence_checked = true; - - /* Support timestamping */ - /* TODO: move this code out of gethttp. */ - if (opt.timestamping && !hs->timestamp_checked) + hs->message = xstrdup (message); + if (!opt.server_response) + logprintf (LOG_VERBOSE, "%2d %s\n", statcode, + message ? quotearg_style (escape_quoting_style, message) : ""); + else { - size_t filename_len = strlen (hs->local_file); - char *filename_plus_orig_suffix = alloca (filename_len + sizeof (".orig")); - bool local_dot_orig_file_exists = false; - char *local_filename = NULL; - struct_stat st; - - if (opt.backup_converted) - /* If -K is specified, we'll act on the assumption that it was specified - last time these files were downloaded as well, and instead of just - comparing local file X against server file X, we'll compare local - file X.orig (if extant, else X) against server file X. If -K - _wasn't_ specified last time, or the server contains files called - *.orig, -N will be back to not operating correctly with -k. */ - { - /* Would a single s[n]printf() call be faster? --dan - - Definitely not. sprintf() is horribly slow. It's a - different question whether the difference between the two - affects a program. Usually I'd say "no", but at one - point I profiled Wget, and found that a measurable and - non-negligible amount of time was lost calling sprintf() - in url.c. Replacing sprintf with inline calls to - strcpy() and number_to_string() made a difference. - --hniksic */ - memcpy (filename_plus_orig_suffix, hs->local_file, filename_len); - memcpy (filename_plus_orig_suffix + filename_len, - ".orig", sizeof (".orig")); - - /* Try to stat() the .orig file. */ - if (stat (filename_plus_orig_suffix, &st) == 0) - { - local_dot_orig_file_exists = true; - local_filename = filename_plus_orig_suffix; - } - } - - if (!local_dot_orig_file_exists) - /* Couldn't stat() .orig, so try to stat() . */ - if (stat (hs->local_file, &st) == 0) - local_filename = hs->local_file; - - if (local_filename != NULL) - /* There was a local file, so we'll check later to see if the version - the server has is the same version we already have, allowing us to - skip a download. */ - { - hs->orig_file_name = xstrdup (local_filename); - hs->orig_file_size = st.st_size; - hs->orig_file_tstamp = st.st_mtime; -#ifdef WINDOWS - /* Modification time granularity is 2 seconds for Windows, so - increase local time by 1 second for later comparison. */ - ++hs->orig_file_tstamp; -#endif - } + logprintf (LOG_VERBOSE, "\n"); + print_server_response (resp, " "); } if (!opt.ignore_length @@ -1936,14 +2035,36 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); /* Check for keep-alive related responses. */ if (!inhibit_keep_alive && contlen != -1) { - if (resp_header_copy (resp, "Keep-Alive", NULL, 0)) - keep_alive = true; - else if (resp_header_copy (resp, "Connection", hdrval, sizeof (hdrval))) + if (resp_header_copy (resp, "Connection", hdrval, sizeof (hdrval))) + { + if (0 == strcasecmp (hdrval, "Close")) + keep_alive = false; + } + } + + chunked_transfer_encoding = false; + if (resp_header_copy (resp, "Transfer-Encoding", hdrval, sizeof (hdrval)) + && 0 == strcasecmp (hdrval, "chunked")) + chunked_transfer_encoding = true; + + /* Handle (possibly multiple instances of) the Set-Cookie header. */ + if (opt.cookies) + { + int scpos; + const char *scbeg, *scend; + /* The jar should have been created by now. */ + assert (wget_cookie_jar != NULL); + for (scpos = 0; + (scpos = resp_header_locate (resp, "Set-Cookie", scpos, + &scbeg, &scend)) != -1; + ++scpos) { - if (0 == strcasecmp (hdrval, "Keep-Alive")) - keep_alive = true; + char *set_cookie; BOUNDED_TO_ALLOCA (scbeg, scend, set_cookie); + cookie_handle_set_cookie (wget_cookie_jar, u->host, u->port, + u->path, set_cookie); } } + if (keep_alive) /* The server has promised that it will not close the connection when we're done. This means that we can register it. */ @@ -1952,7 +2073,8 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); if (statcode == HTTP_STATUS_UNAUTHORIZED) { /* Authorization is required. */ - if (keep_alive && !head_only && skip_short_body (sock, contlen)) + if (keep_alive && !head_only + && skip_short_body (sock, contlen, chunked_transfer_encoding)) CLOSE_FINISH (sock); else CLOSE_INVALIDATE (sock); @@ -2002,6 +2124,9 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); register_basic_auth_host (u->host); } xfree (pth); + xfree_null (message); + resp_free (resp); + xfree (head); goto retry_with_auth; } else @@ -2012,6 +2137,9 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); } logputs (LOG_NOTQUIET, _("Authorization failed.\n")); request_free (req); + xfree_null (message); + resp_free (resp); + xfree (head); return AUTHFAILED; } else /* statcode != HTTP_STATUS_UNAUTHORIZED */ @@ -2020,6 +2148,115 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); if (ntlm_seen) pconn.authorized = true; } + + /* Determine the local filename if needed. Notice that if -O is used + * hstat.local_file is set by http_loop to the argument of -O. */ + if (!hs->local_file) + { + char *local_file = NULL; + + /* Honor Content-Disposition whether possible. */ + if (!opt.content_disposition + || !resp_header_copy (resp, "Content-Disposition", + hdrval, sizeof (hdrval)) + || !parse_content_disposition (hdrval, &local_file)) + { + /* The Content-Disposition header is missing or broken. + * Choose unique file name according to given URL. */ + hs->local_file = url_file_name (u, NULL); + } + else + { + DEBUGP (("Parsed filename from Content-Disposition: %s\n", + local_file)); + hs->local_file = url_file_name (u, local_file); + } + } + + /* TODO: perform this check only once. */ + if (!hs->existence_checked && file_exists_p (hs->local_file)) + { + if (opt.noclobber && !opt.output_document) + { + /* If opt.noclobber is turned on and file already exists, do not + retrieve the file. But if the output_document was given, then this + test was already done and the file didn't exist. Hence the !opt.output_document */ + get_file_flags (hs->local_file, dt); + xfree (head); + xfree_null (message); + return RETRUNNEEDED; + } + else if (!ALLOW_CLOBBER) + { + char *unique = unique_name (hs->local_file, true); + if (unique != hs->local_file) + xfree (hs->local_file); + hs->local_file = unique; + } + } + hs->existence_checked = true; + + /* Support timestamping */ + /* TODO: move this code out of gethttp. */ + if (opt.timestamping && !hs->timestamp_checked) + { + size_t filename_len = strlen (hs->local_file); + char *filename_plus_orig_suffix = alloca (filename_len + sizeof (ORIG_SFX)); + bool local_dot_orig_file_exists = false; + char *local_filename = NULL; + struct_stat st; + + if (opt.backup_converted) + /* If -K is specified, we'll act on the assumption that it was specified + last time these files were downloaded as well, and instead of just + comparing local file X against server file X, we'll compare local + file X.orig (if extant, else X) against server file X. If -K + _wasn't_ specified last time, or the server contains files called + *.orig, -N will be back to not operating correctly with -k. */ + { + /* Would a single s[n]printf() call be faster? --dan + + Definitely not. sprintf() is horribly slow. It's a + different question whether the difference between the two + affects a program. Usually I'd say "no", but at one + point I profiled Wget, and found that a measurable and + non-negligible amount of time was lost calling sprintf() + in url.c. Replacing sprintf with inline calls to + strcpy() and number_to_string() made a difference. + --hniksic */ + memcpy (filename_plus_orig_suffix, hs->local_file, filename_len); + memcpy (filename_plus_orig_suffix + filename_len, + ORIG_SFX, sizeof (ORIG_SFX)); + + /* Try to stat() the .orig file. */ + if (stat (filename_plus_orig_suffix, &st) == 0) + { + local_dot_orig_file_exists = true; + local_filename = filename_plus_orig_suffix; + } + } + + if (!local_dot_orig_file_exists) + /* Couldn't stat() .orig, so try to stat() . */ + if (stat (hs->local_file, &st) == 0) + local_filename = hs->local_file; + + if (local_filename != NULL) + /* There was a local file, so we'll check later to see if the version + the server has is the same version we already have, allowing us to + skip a download. */ + { + hs->orig_file_name = xstrdup (local_filename); + hs->orig_file_size = st.st_size; + hs->orig_file_tstamp = st.st_mtime; +#ifdef WINDOWS + /* Modification time granularity is 2 seconds for Windows, so + increase local time by 1 second for later comparison. */ + ++hs->orig_file_tstamp; +#endif + } + } + request_free (req); hs->statcode = statcode; @@ -2037,32 +2274,25 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); char *tmp = strchr (type, ';'); if (tmp) { + /* sXXXav: only needed if IRI support is enabled */ + char *tmp2 = tmp + 1; + while (tmp > type && c_isspace (tmp[-1])) --tmp; *tmp = '\0'; + + /* Try to get remote encoding if needed */ + if (opt.enable_iri && !opt.encoding_remote) + { + tmp = parse_charset (tmp2); + if (tmp) + set_content_encoding (iri, tmp); + } } } hs->newloc = resp_header_strdup (resp, "Location"); hs->remote_time = resp_header_strdup (resp, "Last-Modified"); - /* Handle (possibly multiple instances of) the Set-Cookie header. */ - if (opt.cookies) - { - int scpos; - const char *scbeg, *scend; - /* The jar should have been created by now. */ - assert (wget_cookie_jar != NULL); - for (scpos = 0; - (scpos = resp_header_locate (resp, "Set-Cookie", scpos, - &scbeg, &scend)) != -1; - ++scpos) - { - char *set_cookie; BOUNDED_TO_ALLOCA (scbeg, scend, set_cookie); - cookie_handle_set_cookie (wget_cookie_jar, u->host, u->port, - u->path, set_cookie); - } - } - if (resp_header_copy (resp, "Content-Range", hdrval, sizeof (hdrval))) { wgint first_byte_pos, last_byte_pos, entity_length; @@ -2095,11 +2325,22 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); _("Location: %s%s\n"), hs->newloc ? escnonprint_uri (hs->newloc) : _("unspecified"), hs->newloc ? _(" [following]") : ""); - if (keep_alive && !head_only && skip_short_body (sock, contlen)) + if (keep_alive && !head_only + && skip_short_body (sock, contlen, chunked_transfer_encoding)) CLOSE_FINISH (sock); else CLOSE_INVALIDATE (sock); xfree_null (type); + xfree (head); + /* From RFC2616: The status codes 303 and 307 have + been added for servers that wish to make unambiguously + clear which kind of reaction is expected of the client. + + A 307 should be redirected using the same method, + in other words, a POST should be preserved and not + converted to a GET in that case. */ + if (statcode == HTTP_STATUS_TEMPORARY_REDIRECT) + return NEWLOCATION_KEEP_POST; return NEWLOCATION; } } @@ -2109,47 +2350,41 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); content-type. */ if (!type || 0 == strncasecmp (type, TEXTHTML_S, strlen (TEXTHTML_S)) || - 0 == strncasecmp (type, TEXTXHTML_S, strlen (TEXTXHTML_S))) + 0 == strncasecmp (type, TEXTXHTML_S, strlen (TEXTXHTML_S))) *dt |= TEXTHTML; else *dt &= ~TEXTHTML; - if (opt.html_extension && (*dt & TEXTHTML)) - /* -E / --html-extension / html_extension = on was specified, and this is a - text/html file. If some case-insensitive variation on ".htm[l]" isn't - already the file's suffix, tack on ".html". */ - { - char *last_period_in_local_filename = strrchr (hs->local_file, '.'); + if (type && + 0 == strncasecmp (type, TEXTCSS_S, strlen (TEXTCSS_S))) + *dt |= TEXTCSS; + else + *dt &= ~TEXTCSS; - if (last_period_in_local_filename == NULL - || !(0 == strcasecmp (last_period_in_local_filename, ".htm") - || 0 == strcasecmp (last_period_in_local_filename, ".html"))) + if (opt.adjust_extension) + { + if (*dt & TEXTHTML) + /* -E / --adjust-extension / adjust_extension = on was specified, + and this is a text/html file. If some case-insensitive + variation on ".htm[l]" isn't already the file's suffix, + tack on ".html". */ { - int local_filename_len = strlen (hs->local_file); - /* Resize the local file, allowing for ".html" preceded by - optional ".NUMBER". */ - hs->local_file = xrealloc (hs->local_file, - local_filename_len + 24 + sizeof (".html")); - strcpy(hs->local_file + local_filename_len, ".html"); - /* If clobbering is not allowed and the file, as named, - exists, tack on ".NUMBER.html" instead. */ - if (!ALLOW_CLOBBER && file_exists_p (hs->local_file)) - { - int ext_num = 1; - do - sprintf (hs->local_file + local_filename_len, - ".%d.html", ext_num++); - while (file_exists_p (hs->local_file)); - } - *dt |= ADDED_HTML_EXTENSION; + ensure_extension (hs, ".html", dt); + } + else if (*dt & TEXTCSS) + { + ensure_extension (hs, ".css", dt); } } - if (statcode == HTTP_STATUS_RANGE_NOT_SATISFIABLE) + if (statcode == HTTP_STATUS_RANGE_NOT_SATISFIABLE + || (!opt.timestamping && hs->restval > 0 && statcode == HTTP_STATUS_OK + && contrange == 0 && contlen >= 0 && hs->restval >= contlen)) { /* If `-c' is in use and the file has been fully downloaded (or the remote file has shrunk), Wget effectively requests bytes - after the end of file and the server response with 416. */ + after the end of file and the server response with 416 + (or 200 with a <= Content-Length. */ logputs (LOG_VERBOSE, _("\ \n The file is already fully retrieved; nothing to do.\n\n")); /* In case the caller inspects. */ @@ -2160,6 +2395,7 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); xfree_null (type); CLOSE_INVALIDATE (sock); /* would be CLOSE_FINISH, but there might be more bytes in the body. */ + xfree (head); return RETRUNNEEDED; } if ((contrange != 0 && contrange != hs->restval) @@ -2169,6 +2405,7 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); Bail out. */ xfree_null (type); CLOSE_INVALIDATE (sock); + xfree (head); return RANGEERR; } if (contlen == -1) @@ -2205,7 +2442,7 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); logputs (LOG_VERBOSE, opt.ignore_length ? _("ignored") : _("unspecified")); if (type) - logprintf (LOG_VERBOSE, " [%s]\n", escnonprint (type)); + logprintf (LOG_VERBOSE, " [%s]\n", quotearg_style (escape_quoting_style, type)); else logputs (LOG_VERBOSE, "\n"); } @@ -2227,14 +2464,26 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); If not, they can be worked around using `--no-http-keep-alive'. */ CLOSE_FINISH (sock); - else if (keep_alive && skip_short_body (sock, contlen)) + else if (keep_alive + && skip_short_body (sock, contlen, chunked_transfer_encoding)) /* Successfully skipped the body; also keep using the socket. */ CLOSE_FINISH (sock); else CLOSE_INVALIDATE (sock); + xfree (head); return RETRFINISHED; } +/* 2005-06-17 SMS. + For VMS, define common fopen() optional arguments. +*/ +#ifdef __VMS +# define FOPEN_OPT_ARGS "fop=sqo", "acc", acc_cb, &open_id +# define FOPEN_BIN_FLAG 3 +#else /* def __VMS */ +# define FOPEN_BIN_FLAG true +#endif /* def __VMS [else] */ + /* Open the local file. */ if (!output_stream) { @@ -2242,12 +2491,43 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); if (opt.backups) rotate_backups (hs->local_file); if (hs->restval) - fp = fopen (hs->local_file, "ab"); - else if (ALLOW_CLOBBER) - fp = fopen (hs->local_file, "wb"); + { +#ifdef __VMS + int open_id; + + open_id = 21; + fp = fopen (hs->local_file, "ab", FOPEN_OPT_ARGS); +#else /* def __VMS */ + fp = fopen (hs->local_file, "ab"); +#endif /* def __VMS [else] */ + } + else if (ALLOW_CLOBBER || count > 0) + { + if (opt.unlink && file_exists_p (hs->local_file)) + { + int res = unlink (hs->local_file); + if (res < 0) + { + logprintf (LOG_NOTQUIET, "%s: %s\n", hs->local_file, + strerror (errno)); + CLOSE_INVALIDATE (sock); + xfree (head); + return UNLINKERR; + } + } + +#ifdef __VMS + int open_id; + + open_id = 22; + fp = fopen (hs->local_file, "wb", FOPEN_OPT_ARGS); +#else /* def __VMS */ + fp = fopen (hs->local_file, "wb"); +#endif /* def __VMS [else] */ + } else { - fp = fopen_excl (hs->local_file, true); + fp = fopen_excl (hs->local_file, FOPEN_BIN_FLAG); if (!fp && errno == EEXIST) { /* We cannot just invent a new name and use it (which is @@ -2258,6 +2538,7 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); _("%s has sprung into existence.\n"), hs->local_file); CLOSE_INVALIDATE (sock); + xfree (head); return FOPEN_EXCL_ERR; } } @@ -2265,6 +2546,7 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); { logprintf (LOG_NOTQUIET, "%s: %s\n", hs->local_file, strerror (errno)); CLOSE_INVALIDATE (sock); + xfree (head); return FOPENERR; } } @@ -2274,10 +2556,10 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); /* Print fetch message, if opt.verbose. */ if (opt.verbose) { - logprintf (LOG_NOTQUIET, _("Saving to: `%s'\n"), - HYPHENP (hs->local_file) ? "STDOUT" : hs->local_file); + logprintf (LOG_NOTQUIET, _("Saving to: %s\n"), + HYPHENP (hs->local_file) ? quote ("STDOUT") : quote (hs->local_file)); } - + /* This confuses the timestamping code that checks for file size. #### The timestamping code should be smarter about file size. */ if (opt.save_headers && hs->restval == 0) @@ -2297,6 +2579,10 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); /* If the server ignored our range request, instruct fd_read_body to skip the first RESTVAL bytes of body. */ flags |= rb_skip_startpos; + + if (chunked_transfer_encoding) + flags |= rb_chunked_transfer_encoding; + hs->len = hs->restval; hs->rd_size = 0; hs->res = fd_read_body (sock, fp, contlen != -1 ? contlen : 0, @@ -2322,8 +2608,9 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); /* The genuine HTTP loop! This is the part where the retrieval is retried, and retried, and retried, and... */ uerr_t -http_loop (struct url *u, char **newloc, char **local_file, const char *referer, - int *dt, struct url *proxy) +http_loop (struct url *u, struct url *original_url, char **newloc, + char **local_file, const char *referer, int *dt, struct url *proxy, + struct iri *iri) { int count; bool got_head = false; /* used for time-stamping and filename detection */ @@ -2334,16 +2621,18 @@ http_loop (struct url *u, char **newloc, char **local_file, const char *referer, uerr_t err, ret = TRYLIMEXC; time_t tmr = -1; /* remote time-stamp */ struct http_stat hstat; /* HTTP status */ - struct_stat st; + struct_stat st; bool send_head_first = true; + char *file_name; + bool force_full_retrieve = false; /* Assert that no value for *LOCAL_FILE was passed. */ assert (local_file == NULL || *local_file == NULL); - + /* Set LOCAL_FILE parameter. */ if (local_file && opt.output_document) *local_file = HYPHENP (opt.output_document) ? NULL : xstrdup (opt.output_document); - + /* Reset NEWLOC parameter. */ *newloc = NULL; @@ -2351,7 +2640,7 @@ http_loop (struct url *u, char **newloc, char **local_file, const char *referer, here so that we don't go through the hoops if we're just using FTP or whatever. */ if (opt.cookies) - load_cookies(); + load_cookies (); /* Warn on (likely bogus) wildcard usage in HTTP. */ if (opt.ftp_glob && has_wildcards_p (u->path)) @@ -2368,59 +2657,49 @@ http_loop (struct url *u, char **newloc, char **local_file, const char *referer, } else if (!opt.content_disposition) { - hstat.local_file = url_file_name (u); + hstat.local_file = + url_file_name (opt.trustservernames ? u : original_url, NULL); got_name = true; } - /* TODO: Ick! This code is now in both gethttp and http_loop, and is - * screaming for some refactoring. */ - if (got_name && file_exists_p (hstat.local_file) && opt.noclobber) + if (got_name && file_exists_p (hstat.local_file) && opt.noclobber && !opt.output_document) { /* If opt.noclobber is turned on and file already exists, do not - retrieve the file */ - logprintf (LOG_VERBOSE, _("\ -File `%s' already there; not retrieving.\n\n"), - hstat.local_file); - /* If the file is there, we suppose it's retrieved OK. */ - *dt |= RETROKF; - - /* #### Bogusness alert. */ - /* If its suffix is "html" or "htm" or similar, assume text/html. */ - if (has_html_suffix_p (hstat.local_file)) - *dt |= TEXTHTML; - - return RETRUNNEEDED; + retrieve the file. But if the output_document was given, then this + test was already done and the file didn't exist. Hence the !opt.output_document */ + get_file_flags (hstat.local_file, dt); + ret = RETROK; + goto exit; } /* Reset the counter. */ count = 0; - + /* Reset the document type. */ *dt = 0; - - /* Skip preliminary HEAD request if we're not in spider mode AND - * if -O was given or HTTP Content-Disposition support is disabled. */ - if (!opt.spider - && (got_name || !opt.content_disposition)) + + /* Skip preliminary HEAD request if we're not in spider mode. */ + if (!opt.spider) send_head_first = false; - /* Send preliminary HEAD request if -N is given and we have an existing + /* Send preliminary HEAD request if -N is given and we have an existing * destination file. */ - if (opt.timestamping - && !opt.content_disposition - && file_exists_p (url_file_name (u))) + file_name = url_file_name (opt.trustservernames ? u : original_url, NULL); + if (opt.timestamping && (file_exists_p (file_name) + || opt.content_disposition)) send_head_first = true; - + xfree (file_name); + /* THE loop */ do { /* Increment the pass counter. */ ++count; sleep_between_retrievals (count); - + /* Get the current time string. */ tms = datetime_str (time (NULL)); - + if (opt.spider && !got_head) logprintf (LOG_VERBOSE, _("\ Spider mode enabled. Check if remote file exists.\n")); @@ -2429,20 +2708,20 @@ Spider mode enabled. Check if remote file exists.\n")); if (opt.verbose) { char *hurl = url_string (u, URL_AUTH_HIDE_PASSWD); - - if (count > 1) + + if (count > 1) { char tmp[256]; sprintf (tmp, _("(try:%2d)"), count); logprintf (LOG_NOTQUIET, "--%s-- %s %s\n", tms, tmp, hurl); } - else + else { logprintf (LOG_NOTQUIET, "--%s-- %s\n", tms, hurl); } - + #ifdef WINDOWS ws_changetitle (hurl); #endif @@ -2452,13 +2731,15 @@ Spider mode enabled. Check if remote file exists.\n")); /* Default document type is empty. However, if spider mode is on or time-stamping is employed, HEAD_ONLY commands is encoded within *dt. */ - if (send_head_first && !got_head) + if (send_head_first && !got_head) *dt |= HEAD_ONLY; else *dt &= ~HEAD_ONLY; /* Decide whether or not to restart. */ - if (opt.always_rest + if (force_full_retrieve) + hstat.restval = hstat.len; + else if (opt.always_rest && got_name && stat (hstat.local_file, &st) == 0 && S_ISREG (st.st_mode)) @@ -2485,11 +2766,11 @@ Spider mode enabled. Check if remote file exists.\n")); *dt &= ~SEND_NOCACHE; /* Try fetching the document, or at least its head. */ - err = gethttp (u, &hstat, dt, proxy); + err = gethttp (u, &hstat, dt, proxy, iri, count); /* Time? */ tms = datetime_str (time (NULL)); - + /* Get the new location (with or without the redirection). */ if (hstat.newloc) *newloc = xstrdup (hstat.newloc); @@ -2507,10 +2788,10 @@ Spider mode enabled. Check if remote file exists.\n")); case FWRITEERR: case FOPENERR: /* Another fatal error. */ logputs (LOG_VERBOSE, "\n"); - logprintf (LOG_NOTQUIET, _("Cannot write to `%s' (%s).\n"), - hstat.local_file, strerror (errno)); - case HOSTERR: case CONIMPOSSIBLE: case PROXERR: case AUTHFAILED: - case SSLINITFAILED: case CONTNOTSUPPORTED: + logprintf (LOG_NOTQUIET, _("Cannot write to %s (%s).\n"), + quote (hstat.local_file), strerror (errno)); + case HOSTERR: case CONIMPOSSIBLE: case PROXERR: case AUTHFAILED: + case SSLINITFAILED: case CONTNOTSUPPORTED: case VERIFCERTERR: /* Fatal errors just return from the function. */ ret = err; goto exit; @@ -2519,7 +2800,15 @@ Spider mode enabled. Check if remote file exists.\n")); logprintf (LOG_NOTQUIET, _("Unable to establish SSL connection.\n")); ret = err; goto exit; + case UNLINKERR: + /* Another fatal error. */ + logputs (LOG_VERBOSE, "\n"); + logprintf (LOG_NOTQUIET, _("Cannot unlink %s (%s).\n"), + quote (hstat.local_file), strerror (errno)); + ret = err; + goto exit; case NEWLOCATION: + case NEWLOCATION_KEEP_POST: /* Return the new location to the caller. */ if (!*newloc) { @@ -2528,9 +2817,9 @@ Spider mode enabled. Check if remote file exists.\n")); hstat.statcode); ret = WRONGCODE; } - else + else { - ret = NEWLOCATION; + ret = err; } goto exit; case RETRUNNEEDED: @@ -2544,7 +2833,7 @@ Spider mode enabled. Check if remote file exists.\n")); /* All possibilities should have been exhausted. */ abort (); } - + if (!(*dt & RETROKF)) { char *hurl = NULL; @@ -2563,11 +2852,13 @@ Spider mode enabled. Check if remote file exists.\n")); continue; } /* Maybe we should always keep track of broken links, not just in - * spider mode. */ - else if (opt.spider) + * spider mode. + * Don't log error if it was UTF-8 encoded because we will try + * once unencoded. */ + else if (opt.spider && !iri->utf8_encode) { /* #### Again: ugly ugly ugly! */ - if (!hurl) + if (!hurl) hurl = url_string (u, URL_AUTH_HIDE_PASSWD); nonexisting_url (hurl); logprintf (LOG_NOTQUIET, _("\ @@ -2576,7 +2867,8 @@ Remote file does not exist -- broken link!!!\n")); else { logprintf (LOG_NOTQUIET, _("%s ERROR %d: %s.\n"), - tms, hstat.statcode, escnonprint (hstat.error)); + tms, hstat.statcode, + quotearg_style (escape_quoting_style, hstat.error)); } logputs (LOG_VERBOSE, "\n"); ret = WRONGCODE; @@ -2604,7 +2896,7 @@ Last-modified header invalid -- time-stamp ignored.\n")); if (*dt & HEAD_ONLY) time_came_from_head = true; } - + if (send_head_first) { /* The time-stamping section. */ @@ -2615,7 +2907,7 @@ Last-modified header invalid -- time-stamp ignored.\n")); we're supposed to download already exists. */ { - if (hstat.remote_time && + if (hstat.remote_time && tmr != (time_t) (-1)) { /* Now time-stamping can be used validly. @@ -2626,12 +2918,12 @@ Last-modified header invalid -- time-stamp ignored.\n")); download procedure is resumed. */ if (hstat.orig_file_tstamp >= tmr) { - if (hstat.contlen == -1 + if (hstat.contlen == -1 || hstat.orig_file_size == hstat.contlen) { logprintf (LOG_VERBOSE, _("\ -Server file no newer than local file `%s' -- not retrieving.\n\n"), - hstat.orig_file_name); +Server file no newer than local file %s -- not retrieving.\n\n"), + quote (hstat.orig_file_name)); ret = RETROK; goto exit; } @@ -2643,32 +2935,36 @@ The sizes do not match (local %s) -- retrieving.\n"), } } else - logputs (LOG_VERBOSE, - _("Remote file is newer, retrieving.\n")); + { + force_full_retrieve = true; + logputs (LOG_VERBOSE, + _("Remote file is newer, retrieving.\n")); + } logputs (LOG_VERBOSE, "\n"); } } - + /* free_hstat (&hstat); */ hstat.timestamp_checked = true; } - + if (opt.spider) { + bool finished = true; if (opt.recursive) { if (*dt & TEXTHTML) { logputs (LOG_VERBOSE, _("\ Remote file exists and could contain links to other resources -- retrieving.\n\n")); + finished = false; } - else + else { logprintf (LOG_VERBOSE, _("\ Remote file exists but does not contain any link -- not retrieving.\n\n")); ret = RETROK; /* RETRUNNEEDED is not for caller. */ - goto exit; } } else @@ -2679,12 +2975,20 @@ Remote file exists but does not contain any link -- not retrieving.\n\n")); Remote file exists and could contain further links,\n\ but recursion is disabled -- not retrieving.\n\n")); } - else + else { logprintf (LOG_VERBOSE, _("\ Remote file exists.\n\n")); } ret = RETROK; /* RETRUNNEEDED is not for caller. */ + } + + if (finished) + { + logprintf (LOG_NONVERBOSE, + _("%s URL: %s %2d %s\n"), + tms, u->url, hstat.statcode, + hstat.message ? quotearg_style (escape_quoting_style, hstat.message) : ""); goto exit; } } @@ -2695,21 +2999,14 @@ Remote file exists.\n\n")); continue; } /* send_head_first */ } /* !got_head */ - - if ((tmr != (time_t) (-1)) + + if (opt.useservertimestamps + && (tmr != (time_t) (-1)) && ((hstat.len == hstat.contlen) || ((hstat.res == 0) && (hstat.contlen == -1)))) { - /* #### This code repeats in http.c and ftp.c. Move it to a - function! */ const char *fl = NULL; - if (opt.output_document) - { - if (output_stream_regular) - fl = opt.output_document; - } - else - fl = hstat.local_file; + set_local_file (&fl, hstat.local_file); if (fl) { time_t newtmr = -1; @@ -2718,7 +3015,7 @@ Remote file exists.\n\n")); && hstat.remote_time && hstat.remote_time[0]) { newtmr = http_atotm (hstat.remote_time); - if (newtmr != -1) + if (newtmr != (time_t)-1) tmr = newtmr; } touch (fl, tmr); @@ -2733,9 +3030,14 @@ Remote file exists.\n\n")); { if (*dt & RETROKF) { + bool write_to_stdout = (opt.output_document && HYPHENP (opt.output_document)); + logprintf (LOG_VERBOSE, - _("%s (%s) - `%s' saved [%s/%s]\n\n"), - tms, tmrate, hstat.local_file, + write_to_stdout + ? _("%s (%s) - written to stdout %s[%s/%s]\n\n") + : _("%s (%s) - %s saved [%s/%s]\n\n"), + tms, tmrate, + write_to_stdout ? "" : quote (hstat.local_file), number_to_static_string (hstat.len), number_to_static_string (hstat.contlen)); logprintf (LOG_NONVERBOSE, @@ -2745,14 +3047,14 @@ Remote file exists.\n\n")); number_to_static_string (hstat.contlen), hstat.local_file, count); } - ++opt.numurls; - total_downloaded_bytes += hstat.len; + ++numurls; + total_downloaded_bytes += hstat.rd_size; /* Remember that we downloaded the file for later ".orig" code. */ if (*dt & ADDED_HTML_EXTENSION) - downloaded_file(FILE_DOWNLOADED_AND_HTML_EXTENSION_ADDED, hstat.local_file); + downloaded_file (FILE_DOWNLOADED_AND_HTML_EXTENSION_ADDED, hstat.local_file); else - downloaded_file(FILE_DOWNLOADED_NORMALLY, hstat.local_file); + downloaded_file (FILE_DOWNLOADED_NORMALLY, hstat.local_file); ret = RETROK; goto exit; @@ -2760,28 +3062,33 @@ Remote file exists.\n\n")); else if (hstat.res == 0) /* No read error */ { if (hstat.contlen == -1) /* We don't know how much we were supposed - to get, so assume we succeeded. */ + to get, so assume we succeeded. */ { if (*dt & RETROKF) { + bool write_to_stdout = (opt.output_document && HYPHENP (opt.output_document)); + logprintf (LOG_VERBOSE, - _("%s (%s) - `%s' saved [%s]\n\n"), - tms, tmrate, hstat.local_file, + write_to_stdout + ? _("%s (%s) - written to stdout %s[%s]\n\n") + : _("%s (%s) - %s saved [%s]\n\n"), + tms, tmrate, + write_to_stdout ? "" : quote (hstat.local_file), number_to_static_string (hstat.len)); logprintf (LOG_NONVERBOSE, "%s URL:%s [%s] -> \"%s\" [%d]\n", tms, u->url, number_to_static_string (hstat.len), hstat.local_file, count); } - ++opt.numurls; - total_downloaded_bytes += hstat.len; + ++numurls; + total_downloaded_bytes += hstat.rd_size; /* Remember that we downloaded the file for later ".orig" code. */ if (*dt & ADDED_HTML_EXTENSION) - downloaded_file(FILE_DOWNLOADED_AND_HTML_EXTENSION_ADDED, hstat.local_file); + downloaded_file (FILE_DOWNLOADED_AND_HTML_EXTENSION_ADDED, hstat.local_file); else - downloaded_file(FILE_DOWNLOADED_NORMALLY, hstat.local_file); - + downloaded_file (FILE_DOWNLOADED_NORMALLY, hstat.local_file); + ret = RETROK; goto exit; } @@ -2794,10 +3101,18 @@ Remote file exists.\n\n")); printwhat (count, opt.ntry); continue; } - else + else if (hstat.len != hstat.restval) /* Getting here would mean reading more data than requested with content-length, which we never do. */ abort (); + else + { + /* Getting here probably means that the content-length was + * _less_ than the original, local size. We should probably + * truncate or re-read, or something. FIXME */ + ret = RETROK; + goto exit; + } } else /* from now on hstat.res can only be -1 */ { @@ -2827,10 +3142,10 @@ Remote file exists.\n\n")); while (!opt.ntry || (count < opt.ntry)); exit: - if (ret == RETROK) + if (ret == RETROK && local_file) *local_file = xstrdup (hstat.local_file); free_hstat (&hstat); - + return ret; } @@ -2902,13 +3217,24 @@ http_atotm (const char *time_string) Netscape cookie specification.) */ }; const char *oldlocale; - int i; + char savedlocale[256]; + size_t i; time_t ret = (time_t) -1; /* Solaris strptime fails to recognize English month names in non-English locales, which we work around by temporarily setting locale to C before invoking strptime. */ oldlocale = setlocale (LC_TIME, NULL); + if (oldlocale) + { + size_t l = strlen (oldlocale) + 1; + if (l >= sizeof savedlocale) + savedlocale[0] = '\0'; + else + memcpy (savedlocale, oldlocale, l); + } + else savedlocale[0] = '\0'; + setlocale (LC_TIME, "C"); for (i = 0; i < countof (time_formats); i++) @@ -2928,7 +3254,8 @@ http_atotm (const char *time_string) } /* Restore the previous locale. */ - setlocale (LC_TIME, oldlocale); + if (savedlocale[0]) + setlocale (LC_TIME, savedlocale); return ret; } @@ -2981,7 +3308,7 @@ dump_hash (char *buf, const unsigned char *hash) { int i; - for (i = 0; i < MD5_HASHLEN; i++, hash++) + for (i = 0; i < MD5_DIGEST_SIZE; i++, hash++) { *buf++ = XNUM_TO_digit (*hash >> 4); *buf++ = XNUM_TO_digit (*hash & 0xf); @@ -3013,10 +3340,12 @@ digest_authentication_encode (const char *au, const char *user, au += 6; /* skip over `Digest' */ while (extract_param (&au, &name, &value, ',')) { - int i; + size_t i; + size_t namelen = name.e - name.b; for (i = 0; i < countof (options); i++) - if (name.e - name.b == strlen (options[i].name) - && 0 == strncmp (name.b, options[i].name, name.e - name.b)) + if (namelen == strlen (options[i].name) + && 0 == strncmp (name.b, options[i].name, + namelen)) { *options[i].variable = strdupdelim (value.b, value.e); break; @@ -3032,37 +3361,37 @@ digest_authentication_encode (const char *au, const char *user, /* Calculate the digest value. */ { - ALLOCA_MD5_CONTEXT (ctx); - unsigned char hash[MD5_HASHLEN]; - char a1buf[MD5_HASHLEN * 2 + 1], a2buf[MD5_HASHLEN * 2 + 1]; - char response_digest[MD5_HASHLEN * 2 + 1]; + struct md5_ctx ctx; + unsigned char hash[MD5_DIGEST_SIZE]; + char a1buf[MD5_DIGEST_SIZE * 2 + 1], a2buf[MD5_DIGEST_SIZE * 2 + 1]; + char response_digest[MD5_DIGEST_SIZE * 2 + 1]; /* A1BUF = H(user ":" realm ":" password) */ - gen_md5_init (ctx); - gen_md5_update ((unsigned char *)user, strlen (user), ctx); - gen_md5_update ((unsigned char *)":", 1, ctx); - gen_md5_update ((unsigned char *)realm, strlen (realm), ctx); - gen_md5_update ((unsigned char *)":", 1, ctx); - gen_md5_update ((unsigned char *)passwd, strlen (passwd), ctx); - gen_md5_finish (ctx, hash); + md5_init_ctx (&ctx); + md5_process_bytes ((unsigned char *)user, strlen (user), &ctx); + md5_process_bytes ((unsigned char *)":", 1, &ctx); + md5_process_bytes ((unsigned char *)realm, strlen (realm), &ctx); + md5_process_bytes ((unsigned char *)":", 1, &ctx); + md5_process_bytes ((unsigned char *)passwd, strlen (passwd), &ctx); + md5_finish_ctx (&ctx, hash); dump_hash (a1buf, hash); /* A2BUF = H(method ":" path) */ - gen_md5_init (ctx); - gen_md5_update ((unsigned char *)method, strlen (method), ctx); - gen_md5_update ((unsigned char *)":", 1, ctx); - gen_md5_update ((unsigned char *)path, strlen (path), ctx); - gen_md5_finish (ctx, hash); + md5_init_ctx (&ctx); + md5_process_bytes ((unsigned char *)method, strlen (method), &ctx); + md5_process_bytes ((unsigned char *)":", 1, &ctx); + md5_process_bytes ((unsigned char *)path, strlen (path), &ctx); + md5_finish_ctx (&ctx, hash); dump_hash (a2buf, hash); /* RESPONSE_DIGEST = H(A1BUF ":" nonce ":" A2BUF) */ - gen_md5_init (ctx); - gen_md5_update ((unsigned char *)a1buf, MD5_HASHLEN * 2, ctx); - gen_md5_update ((unsigned char *)":", 1, ctx); - gen_md5_update ((unsigned char *)nonce, strlen (nonce), ctx); - gen_md5_update ((unsigned char *)":", 1, ctx); - gen_md5_update ((unsigned char *)a2buf, MD5_HASHLEN * 2, ctx); - gen_md5_finish (ctx, hash); + md5_init_ctx (&ctx); + md5_process_bytes ((unsigned char *)a1buf, MD5_DIGEST_SIZE * 2, &ctx); + md5_process_bytes ((unsigned char *)":", 1, &ctx); + md5_process_bytes ((unsigned char *)nonce, strlen (nonce), &ctx); + md5_process_bytes ((unsigned char *)":", 1, &ctx); + md5_process_bytes ((unsigned char *)a2buf, MD5_DIGEST_SIZE * 2, &ctx); + md5_finish_ctx (&ctx, hash); dump_hash (response_digest, hash); res = xmalloc (strlen (user) @@ -3070,7 +3399,7 @@ digest_authentication_encode (const char *au, const char *user, + strlen (realm) + strlen (nonce) + strlen (path) - + 2 * MD5_HASHLEN /*strlen (response_digest)*/ + + 2 * MD5_DIGEST_SIZE /*strlen (response_digest)*/ + (opaque ? strlen (opaque) : 0) + 128); sprintf (res, "Digest \ @@ -3096,9 +3425,10 @@ username=\"%s\", realm=\"%s\", nonce=\"%s\", uri=\"%s\", response=\"%s\"", first argument and are followed by whitespace or terminating \0. The comparison is case-insensitive. */ #define STARTS(literal, b, e) \ - ((e) - (b) >= STRSIZE (literal) \ + ((e > b) \ + && ((size_t) ((e) - (b))) >= STRSIZE (literal) \ && 0 == strncasecmp (b, literal, STRSIZE (literal)) \ - && ((e) - (b) == STRSIZE (literal) \ + && ((size_t) ((e) - (b)) == STRSIZE (literal) \ || c_isspace (b[STRSIZE (literal)]))) static bool @@ -3181,6 +3511,42 @@ http_cleanup (void) cookie_jar_delete (wget_cookie_jar); } +void +ensure_extension (struct http_stat *hs, const char *ext, int *dt) +{ + char *last_period_in_local_filename = strrchr (hs->local_file, '.'); + char shortext[8]; + int len = strlen (ext); + if (len == 5) + { + strncpy (shortext, ext, len - 1); + shortext[len - 1] = '\0'; + } + + if (last_period_in_local_filename == NULL + || !(0 == strcasecmp (last_period_in_local_filename, shortext) + || 0 == strcasecmp (last_period_in_local_filename, ext))) + { + int local_filename_len = strlen (hs->local_file); + /* Resize the local file, allowing for ".html" preceded by + optional ".NUMBER". */ + hs->local_file = xrealloc (hs->local_file, + local_filename_len + 24 + len); + strcpy (hs->local_file + local_filename_len, ext); + /* If clobbering is not allowed and the file, as named, + exists, tack on ".NUMBER.html" instead. */ + if (!ALLOW_CLOBBER && file_exists_p (hs->local_file)) + { + int ext_num = 1; + do + sprintf (hs->local_file + local_filename_len, + ".%d%s", ext_num++, ext); + while (file_exists_p (hs->local_file)); + } + *dt |= ADDED_HTML_EXTENSION; + } +} + #ifdef TESTING @@ -3189,32 +3555,28 @@ test_parse_content_disposition() { int i; struct { - char *hdrval; - char *opt_dir_prefix; + char *hdrval; char *filename; bool result; } test_array[] = { - { "filename=\"file.ext\"", NULL, "file.ext", true }, - { "filename=\"file.ext\"", "somedir", "somedir/file.ext", true }, - { "attachment; filename=\"file.ext\"", NULL, "file.ext", true }, - { "attachment; filename=\"file.ext\"", "somedir", "somedir/file.ext", true }, - { "attachment; filename=\"file.ext\"; dummy", NULL, "file.ext", true }, - { "attachment; filename=\"file.ext\"; dummy", "somedir", "somedir/file.ext", true }, - { "attachment", NULL, NULL, false }, - { "attachment", "somedir", NULL, false }, + { "filename=\"file.ext\"", "file.ext", true }, + { "attachment; filename=\"file.ext\"", "file.ext", true }, + { "attachment; filename=\"file.ext\"; dummy", "file.ext", true }, + { "attachment", NULL, false }, + { "attachement; filename*=UTF-8'en-US'hello.txt", "hello.txt", true }, + { "attachement; filename*0=\"hello\"; filename*1=\"world.txt\"", "helloworld.txt", true }, }; - - for (i = 0; i < sizeof(test_array)/sizeof(test_array[0]); ++i) + + for (i = 0; i < sizeof(test_array)/sizeof(test_array[0]); ++i) { char *filename; bool res; - opt.dir_prefix = test_array[i].opt_dir_prefix; res = parse_content_disposition (test_array[i].hdrval, &filename); - mu_assert ("test_parse_content_disposition: wrong result", + mu_assert ("test_parse_content_disposition: wrong result", res == test_array[i].result - && (res == false + && (res == false || 0 == strcmp (test_array[i].filename, filename))); }