X-Git-Url: http://sjero.net/git/?a=blobdiff_plain;f=src%2Fhttp.c;h=52f65fed9215cd8381a427c3eeb480648fef0405;hb=123f5c39669abc055987d69a311785c861494c87;hp=e6f48aa2b2b8c1832d902461ae3069f567d0dd66;hpb=68740f10dd55cb272bcad0bd0c5199bbdef5b26e;p=wget diff --git a/src/http.c b/src/http.c index e6f48aa2..52f65fed 100644 --- a/src/http.c +++ b/src/http.c @@ -1,6 +1,6 @@ /* HTTP support. Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, - 2004, 2005, 2006, 2007 Free Software Foundation, Inc. + 2004, 2005, 2006, 2007, 2008 Free Software Foundation, Inc. This file is part of GNU Wget. @@ -17,17 +17,18 @@ GNU General Public License for more details. You should have received a copy of the GNU General Public License along with Wget. If not, see . -In addition, as a special exception, the Free Software Foundation -gives permission to link the code of its release of Wget with the -OpenSSL project's "OpenSSL" library (or with modified versions of it -that use the same license as the "OpenSSL" library), and distribute -the linked executables. You must obey the GNU General Public License -in all respects for all of the code used other than "OpenSSL". If you -modify this file, you may extend this exception to your version of the -file, but you are not obligated to do so. If you do not wish to do -so, delete this exception statement from your version. */ +Additional permission under GNU GPL version 3 section 7 -#include +If you modify this program, or any covered work, by linking or +combining it with the OpenSSL project's OpenSSL library (or a +modified version of that library), containing parts covered by the +terms of the OpenSSL or SSLeay licenses, the Free Software Foundation +grants you additional permission to convey the resulting work. +Corresponding Source for a non-source form of such a combination +shall include the source code for the parts of OpenSSL used as well +as that of the covered work. */ + +#include "wget.h" #include #include @@ -40,7 +41,6 @@ so, delete this exception statement from your version. */ #include #include -#include "wget.h" #include "hash.h" #include "http.h" #include "utils.h" @@ -69,11 +69,13 @@ so, delete this exception statement from your version. */ 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 @@ -86,6 +88,7 @@ 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_20X(x) (((x) >= 200) && ((x) < 300)) @@ -389,27 +392,35 @@ static struct hash_table *basic_authed_hosts; * it the username, password. A temporary measure until we can get * proper authentication in place. */ -static int +static bool maybe_send_basic_creds (const char *hostname, const char *user, const char *passwd, struct request *req) { - int did_challenge = 0; + bool do_challenge = false; - if (basic_authed_hosts + if (opt.auth_without_challenge) + { + 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)); - request_set_header (req, "Authorization", - basic_authentication_encode (user, passwd), - rel_value); - did_challenge = 1; + 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) + { + request_set_header (req, "Authorization", + basic_authentication_encode (user, passwd), + rel_value); } - return did_challenge; + return do_challenge; } static void @@ -422,7 +433,7 @@ 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))); } } @@ -802,7 +813,8 @@ print_response_line(const char *prefix, const char *b, const char *e) { char *copy; BOUNDED_TO_ALLOCA(b, e, copy); - logprintf (LOG_VERBOSE, "%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 @@ -864,8 +876,11 @@ parse_content_range (const char *hdr, wgint *first_byte_ptr, return false; *last_byte_ptr = num; ++hdr; - for (num = 0; c_isdigit (*hdr); hdr++) - num = 10 * num + (*hdr - '0'); + if (*hdr == '*') + num = -1; + else + for (num = 0; c_isdigit (*hdr); hdr++) + num = 10 * num + (*hdr - '0'); *entity_length_ptr = num; return true; } @@ -1284,10 +1299,15 @@ 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. */ char *local_file; /* local file name. */ + bool existence_checked; /* true if we already checked for a file's + 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 * have already been performed */ char *orig_file_name; /* name of file to compare for time-stamping @@ -1306,6 +1326,7 @@ 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; @@ -1425,6 +1446,7 @@ 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; @@ -1482,41 +1504,6 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) basic_auth_finished = maybe_send_basic_creds(u->host, user, passwd, req); } - proxyauth = NULL; - if (proxy) - { - char *proxy_user, *proxy_passwd; - /* For normal username and password, URL components override - command-line/wgetrc parameters. With proxy - authentication, it's the reverse, because proxy URLs are - normally the "permanent" ones, so command-line args - should take precedence. */ - if (opt.proxy_user && opt.proxy_passwd) - { - proxy_user = opt.proxy_user; - proxy_passwd = opt.proxy_passwd; - } - else - { - proxy_user = proxy->user; - proxy_passwd = proxy->passwd; - } - /* #### This does not appear right. Can't the proxy request, - say, `Digest' authentication? */ - if (proxy_user && proxy_passwd) - proxyauth = basic_authentication_encode (proxy_user, proxy_passwd); - - /* If we're using a proxy, we will be connecting to the proxy - server. */ - conn = proxy; - - /* Proxy authorization over SSL is handled below. */ -#ifdef HAVE_SSL - if (u->scheme != SCHEME_HTTPS) -#endif - request_set_header (req, "Proxy-Authorization", proxyauth, rel_value); - } - /* Generate the Host header, HOST:PORT. Take into account that: - Broken server-side software often doesn't recognize the PORT @@ -1564,8 +1551,8 @@ 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; } } @@ -1587,6 +1574,41 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) without authorization header fails. (Expected to happen at least for the Digest authorization scheme.) */ + proxyauth = NULL; + if (proxy) + { + char *proxy_user, *proxy_passwd; + /* For normal username and password, URL components override + command-line/wgetrc parameters. With proxy + authentication, it's the reverse, because proxy URLs are + normally the "permanent" ones, so command-line args + should take precedence. */ + if (opt.proxy_user && opt.proxy_passwd) + { + proxy_user = opt.proxy_user; + proxy_passwd = opt.proxy_passwd; + } + else + { + proxy_user = proxy->user; + proxy_passwd = proxy->passwd; + } + /* #### This does not appear right. Can't the proxy request, + say, `Digest' authentication? */ + if (proxy_user && proxy_passwd) + proxyauth = basic_authentication_encode (proxy_user, proxy_passwd); + + /* If we're using a proxy, we will be connecting to the proxy + server. */ + conn = proxy; + + /* Proxy authorization over SSL is handled below. */ +#ifdef HAVE_SSL + if (u->scheme != SCHEME_HTTPS) +#endif + request_set_header (req, "Proxy-Authorization", proxyauth, rel_value); + } + keep_alive = false; /* Establish the connection. */ @@ -1614,7 +1636,8 @@ 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" @@ -1626,8 +1649,8 @@ 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; } } @@ -1696,13 +1719,14 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) resp = resp_new (head); statcode = resp_status (resp, &message); + 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; } @@ -1778,9 +1802,10 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) /* Check for status line. */ message = NULL; statcode = resp_status (resp, &message); + hs->message = xstrdup (message); if (!opt.server_response) logprintf (LOG_VERBOSE, "%2d %s\n", statcode, - message ? escnonprint (message) : ""); + message ? quotearg_style (escape_quoting_style, message) : ""); else { logprintf (LOG_VERBOSE, "\n"); @@ -1804,14 +1829,15 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy) } /* TODO: perform this check only once. */ - if (file_exists_p (hs->local_file)) + if (!hs->existence_checked && file_exists_p (hs->local_file)) { - if (opt.noclobber) + if (opt.noclobber && !opt.output_document) { /* If opt.noclobber is turned on and file already exists, do not - retrieve the file */ + 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 */ logprintf (LOG_VERBOSE, _("\ -File `%s' already there; not retrieving.\n\n"), hs->local_file); +File %s already there; not retrieving.\n\n"), quote (hs->local_file)); /* If the file is there, we suppose it's retrieved OK. */ *dt |= RETROKF; @@ -1830,6 +1856,7 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); hs->local_file = unique; } } + hs->existence_checked = true; /* Support timestamping */ /* TODO: move this code out of gethttp. */ @@ -2052,7 +2079,10 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); wgint first_byte_pos, last_byte_pos, entity_length; if (parse_content_range (hdrval, &first_byte_pos, &last_byte_pos, &entity_length)) - contrange = first_byte_pos; + { + contrange = first_byte_pos; + contlen = last_byte_pos - first_byte_pos + 1; + } } resp_free (resp); @@ -2095,34 +2125,25 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); 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.html_extension) + { + if (*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". */ { - 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); } } @@ -2152,7 +2173,10 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file); CLOSE_INVALIDATE (sock); return RANGEERR; } - hs->contlen = contlen + contrange; + if (contlen == -1) + hs->contlen = -1; + else + hs->contlen = contlen + contrange; if (opt.verbose) { @@ -2183,7 +2207,7 @@ 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"); } @@ -2252,8 +2276,8 @@ 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. @@ -2350,6 +2374,28 @@ http_loop (struct url *u, char **newloc, char **local_file, const char *referer, 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 && !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 */ + logprintf (LOG_VERBOSE, _("\ +File %s already there; not retrieving.\n\n"), + quote (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; + + ret = RETROK; + goto exit; + } + /* Reset the counter. */ count = 0; @@ -2465,8 +2511,8 @@ 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)); + 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: /* Fatal errors just return from the function. */ @@ -2534,7 +2580,8 @@ 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; @@ -2588,8 +2635,8 @@ Last-modified header invalid -- time-stamp ignored.\n")); || 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; } @@ -2614,26 +2661,44 @@ The sizes do not match (local %s) -- retrieving.\n"), 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 { 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 { - logprintf (LOG_VERBOSE, _("\ -Remote file exists but recursion is disabled -- not retrieving.\n\n")); + if (*dt & TEXTHTML) + { + logprintf (LOG_VERBOSE, _("\ +Remote file exists and could contain further links,\n\ +but recursion is disabled -- not retrieving.\n\n")); + } + 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; } } @@ -2683,8 +2748,8 @@ Remote file exists but recursion is disabled -- not retrieving.\n\n")); if (*dt & RETROKF) { logprintf (LOG_VERBOSE, - _("%s (%s) - `%s' saved [%s/%s]\n\n"), - tms, tmrate, hstat.local_file, + _("%s (%s) - %s saved [%s/%s]\n\n"), + tms, tmrate, quote (hstat.local_file), number_to_static_string (hstat.len), number_to_static_string (hstat.contlen)); logprintf (LOG_NONVERBOSE, @@ -2714,8 +2779,8 @@ Remote file exists but recursion is disabled -- not retrieving.\n\n")); if (*dt & RETROKF) { logprintf (LOG_VERBOSE, - _("%s (%s) - `%s' saved [%s]\n\n"), - tms, tmrate, hstat.local_file, + _("%s (%s) - %s saved [%s]\n\n"), + tms, tmrate, quote (hstat.local_file), number_to_static_string (hstat.len)); logprintf (LOG_NONVERBOSE, "%s URL:%s [%s] -> \"%s\" [%d]\n", @@ -2743,10 +2808,18 @@ Remote file exists but recursion is disabled -- not retrieving.\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 */ { @@ -2851,7 +2924,7 @@ http_atotm (const char *time_string) Netscape cookie specification.) */ }; const char *oldlocale; - int i; + size_t i; time_t ret = (time_t) -1; /* Solaris strptime fails to recognize English month names in @@ -2962,10 +3035,12 @@ 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; @@ -3045,9 +3120,10 @@ 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 @@ -3130,6 +3206,42 @@ 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 - 2] = '\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