From 78c0badc9b3629c5156d4bd0733a3e9a1e4e92c6 Mon Sep 17 00:00:00 2001 From: Aaron Griffin Date: Tue, 31 Oct 2006 06:34:33 +0000 Subject: Added libfetch linux port + misc changes --- lib/libfetch/http.c | 1225 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 1225 insertions(+) create mode 100644 lib/libfetch/http.c (limited to 'lib/libfetch/http.c') diff --git a/lib/libfetch/http.c b/lib/libfetch/http.c new file mode 100644 index 00000000..c173b5d5 --- /dev/null +++ b/lib/libfetch/http.c @@ -0,0 +1,1225 @@ +/*- + * Copyright (c) 2000-2004 Dag-Erling Coïdan Smørgrav + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer + * in this position and unchanged. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. The name of the author may not be used to endorse or promote products + * derived from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR + * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES + * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. + * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, + * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT + * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF + * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#define _GNU_SOURCE + +#include + +/* + * The following copyright applies to the base64 code: + * + *- + * Copyright 1997 Massachusetts Institute of Technology + * + * Permission to use, copy, modify, and distribute this software and + * its documentation for any purpose and without fee is hereby + * granted, provided that both the above copyright notice and this + * permission notice appear in all copies, that both the above + * copyright notice and this permission notice appear in all + * supporting documentation, and that the name of M.I.T. not be used + * in advertising or publicity pertaining to distribution of the + * software without specific, written prior permission. M.I.T. makes + * no representations about the suitability of this software for any + * purpose. It is provided "as is" without express or implied + * warranty. + * + * THIS SOFTWARE IS PROVIDED BY M.I.T. ``AS IS''. M.I.T. DISCLAIMS + * ALL EXPRESS OR IMPLIED WARRANTIES WITH REGARD TO THIS SOFTWARE, + * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. IN NO EVENT + * SHALL M.I.T. BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF + * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND + * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, + * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT + * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +#include +#include + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include + +#include "fetch.h" +#include "common.h" +#include "httperr.h" + +/* Maximum number of redirects to follow */ +#define MAX_REDIRECT 5 + +/* Symbolic names for reply codes we care about */ +#define HTTP_OK 200 +#define HTTP_PARTIAL 206 +#define HTTP_MOVED_PERM 301 +#define HTTP_MOVED_TEMP 302 +#define HTTP_SEE_OTHER 303 +#define HTTP_NEED_AUTH 401 +#define HTTP_NEED_PROXY_AUTH 407 +#define HTTP_BAD_RANGE 416 +#define HTTP_PROTOCOL_ERROR 999 + +#define HTTP_REDIRECT(xyz) ((xyz) == HTTP_MOVED_PERM \ + || (xyz) == HTTP_MOVED_TEMP \ + || (xyz) == HTTP_SEE_OTHER) + +#define HTTP_ERROR(xyz) ((xyz) > 400 && (xyz) < 599) + + +/***************************************************************************** + * I/O functions for decoding chunked streams + */ + +struct httpio +{ + conn_t *conn; /* connection */ + int chunked; /* chunked mode */ + char *buf; /* chunk buffer */ + size_t bufsize; /* size of chunk buffer */ + ssize_t buflen; /* amount of data currently in buffer */ + int bufpos; /* current read offset in buffer */ + int eof; /* end-of-file flag */ + int error; /* error flag */ + size_t chunksize; /* remaining size of current chunk */ +#ifndef NDEBUG + size_t total; +#endif +}; + +/* + * Get next chunk header + */ +static int +_http_new_chunk(struct httpio *io) +{ + char hexdigits[23] = "0123456789abcdefABCDEF\0"; + char *p = NULL; + + if (_fetch_getln(io->conn) == -1) + return (-1); + + if (io->conn->buflen < 2 || !strchr(hexdigits, *p)) + return (-1); + + for (p = io->conn->buf; *p && !isspace(*p); ++p) { + if (*p == ';') + break; + if (!strchr(hexdigits, *p)) + return (-1); + if (isdigit(*p)) { + io->chunksize = io->chunksize * 16 + + *p - '0'; + } else { + io->chunksize = io->chunksize * 16 + + 10 + tolower(*p) - 'a'; + } + } + +#ifndef NDEBUG + if (fetchDebug) { + io->total += io->chunksize; + if (io->chunksize == 0) + fprintf(stderr, "%s(): end of last chunk\n", __func__); + else + fprintf(stderr, "%s(): new chunk: %lu (%lu)\n", + __func__, (unsigned long)io->chunksize, + (unsigned long)io->total); + } +#endif + + return (io->chunksize); +} + +/* + * Grow the input buffer to at least len bytes + */ +static inline int +_http_growbuf(struct httpio *io, size_t len) +{ + char *tmp; + + if (io->bufsize >= len) + return (0); + + if ((tmp = realloc(io->buf, len)) == NULL) + return (-1); + io->buf = tmp; + io->bufsize = len; + return (0); +} + +/* + * Fill the input buffer, do chunk decoding on the fly + */ +static int +_http_fillbuf(struct httpio *io, size_t len) +{ + if (io->error) + return (-1); + if (io->eof) + return (0); + + if (io->chunked == 0) { + if (_http_growbuf(io, len) == -1) + return (-1); + if ((io->buflen = _fetch_read(io->conn, io->buf, len)) == -1) { + io->error = 1; + return (-1); + } + io->bufpos = 0; + return (io->buflen); + } + + if (io->chunksize == 0) { + switch (_http_new_chunk(io)) { + case -1: + io->error = 1; + return (-1); + case 0: + io->eof = 1; + return (0); + } + } + + if (len > io->chunksize) + len = io->chunksize; + if (_http_growbuf(io, len) == -1) + return (-1); + if ((io->buflen = _fetch_read(io->conn, io->buf, len)) == -1) { + io->error = 1; + return (-1); + } + io->chunksize -= io->buflen; + + if (io->chunksize == 0) { + char endl[2]; + + if (_fetch_read(io->conn, endl, 2) != 2 || + endl[0] != '\r' || endl[1] != '\n') + return (-1); + } + + io->bufpos = 0; + + return (io->buflen); +} + +/* + * Read function + */ +static int +_http_readfn(void *v, char *buf, int len) +{ + struct httpio *io = (struct httpio *)v; + int l, pos; + + if (io->error) + return (-1); + if (io->eof) + return (0); + + for (pos = 0; len > 0; pos += l, len -= l) { + /* empty buffer */ + if (!io->buf || io->bufpos == io->buflen) + if (_http_fillbuf(io, len) < 1) + break; + l = io->buflen - io->bufpos; + if (len < l) + l = len; + bcopy(io->buf + io->bufpos, buf + pos, l); + io->bufpos += l; + } + + if (!pos && io->error) + return (-1); + return (pos); +} + +/* + * Write function + */ +static int +_http_writefn(void *v, const char *buf, int len) +{ + struct httpio *io = (struct httpio *)v; + + return (_fetch_write(io->conn, buf, len)); +} + +/* + * Close function + */ +static int +_http_closefn(void *v) +{ + struct httpio *io = (struct httpio *)v; + int r; + + r = _fetch_close(io->conn); + if (io->buf) + free(io->buf); + free(io); + return (r); +} + +/* + * Wrap a file descriptor up + */ +static FILE * +_http_funopen(conn_t *conn, int chunked) +{ + struct httpio *io; + FILE *f; + + if ((io = calloc(1, sizeof(*io))) == NULL) { + _fetch_syserr(); + return (NULL); + } + io->conn = conn; + io->chunked = chunked; + cookie_io_functions_t cookie; + cookie.read = (cookie_read_function_t *)_http_readfn; + cookie.write = (cookie_write_function_t *)_http_writefn; + cookie.seek = (cookie_seek_function_t *)NULL; + cookie.close = (cookie_close_function_t *)_http_closefn; + f = fopencookie(io, "rw", cookie); + if (f == NULL) { + _fetch_syserr(); + free(io); + return (NULL); + } + return (f); +} + + +/***************************************************************************** + * Helper functions for talking to the server and parsing its replies + */ + +/* Header types */ +typedef enum { + hdr_syserror = -2, + hdr_error = -1, + hdr_end = 0, + hdr_unknown = 1, + hdr_content_length, + hdr_content_range, + hdr_last_modified, + hdr_location, + hdr_transfer_encoding, + hdr_www_authenticate +} hdr_t; + +/* Names of interesting headers */ +static struct { + hdr_t num; + const char *name; +} hdr_names[] = { + { hdr_content_length, "Content-Length" }, + { hdr_content_range, "Content-Range" }, + { hdr_last_modified, "Last-Modified" }, + { hdr_location, "Location" }, + { hdr_transfer_encoding, "Transfer-Encoding" }, + { hdr_www_authenticate, "WWW-Authenticate" }, + { hdr_unknown, NULL }, +}; + +/* + * Send a formatted line; optionally echo to terminal + */ +static int +_http_cmd(conn_t *conn, const char *fmt, ...) +{ + va_list ap; + size_t len; + char *msg; + int r; + + va_start(ap, fmt); + len = vasprintf(&msg, fmt, ap); + va_end(ap); + + if (msg == NULL) { + errno = ENOMEM; + _fetch_syserr(); + return (-1); + } + + r = _fetch_putln(conn, msg, len); + free(msg); + + if (r == -1) { + _fetch_syserr(); + return (-1); + } + + return (0); +} + +/* + * Get and parse status line + */ +static int +_http_get_reply(conn_t *conn) +{ + char *p; + + if (_fetch_getln(conn) == -1) + return (-1); + /* + * A valid status line looks like "HTTP/m.n xyz reason" where m + * and n are the major and minor protocol version numbers and xyz + * is the reply code. + * Unfortunately, there are servers out there (NCSA 1.5.1, to name + * just one) that do not send a version number, so we can't rely + * on finding one, but if we do, insist on it being 1.0 or 1.1. + * We don't care about the reason phrase. + */ + if (strncmp(conn->buf, "HTTP", 4) != 0) + return (HTTP_PROTOCOL_ERROR); + p = conn->buf + 4; + if (*p == '/') { + if (p[1] != '1' || p[2] != '.' || (p[3] != '0' && p[3] != '1')) + return (HTTP_PROTOCOL_ERROR); + p += 4; + } + if (*p != ' ' || !isdigit(p[1]) || !isdigit(p[2]) || !isdigit(p[3])) + return (HTTP_PROTOCOL_ERROR); + + conn->err = (p[1] - '0') * 100 + (p[2] - '0') * 10 + (p[3] - '0'); + return (conn->err); +} + +/* + * Check a header; if the type matches the given string, return a pointer + * to the beginning of the value. + */ +static const char * +_http_match(const char *str, const char *hdr) +{ + while (*str && *hdr && tolower(*str++) == tolower(*hdr++)) + /* nothing */; + if (*str || *hdr != ':') + return (NULL); + while (*hdr && isspace(*++hdr)) + /* nothing */; + return (hdr); +} + +/* + * Get the next header and return the appropriate symbolic code. + */ +static hdr_t +_http_next_header(conn_t *conn, const char **p) +{ + int i; + + if (_fetch_getln(conn) == -1) + return (hdr_syserror); + while (conn->buflen && isspace(conn->buf[conn->buflen - 1])) + conn->buflen--; + conn->buf[conn->buflen] = '\0'; + if (conn->buflen == 0) + return (hdr_end); + /* + * We could check for malformed headers but we don't really care. + * A valid header starts with a token immediately followed by a + * colon; a token is any sequence of non-control, non-whitespace + * characters except "()<>@,;:\\\"{}". + */ + for (i = 0; hdr_names[i].num != hdr_unknown; i++) + if ((*p = _http_match(hdr_names[i].name, conn->buf)) != NULL) + return (hdr_names[i].num); + return (hdr_unknown); +} + +/* + * Parse a last-modified header + */ +static int +_http_parse_mtime(const char *p, time_t *mtime) +{ + char locale[64], *r; + struct tm tm; + + strncpy(locale, setlocale(LC_TIME, NULL), sizeof(locale)); + setlocale(LC_TIME, "C"); + r = (char *)strptime(p, "%a, %d %b %Y %H:%M:%S GMT", &tm); + /* XXX should add support for date-2 and date-3 */ + setlocale(LC_TIME, locale); + if (r == NULL) + return (-1); + DEBUG(fprintf(stderr, "last modified: [%04d-%02d-%02d " + "%02d:%02d:%02d]\n", + tm.tm_year + 1900, tm.tm_mon + 1, tm.tm_mday, + tm.tm_hour, tm.tm_min, tm.tm_sec)); + *mtime = timegm(&tm); + return (0); +} + +/* + * Parse a content-length header + */ +static int +_http_parse_length(const char *p, off_t *length) +{ + off_t len; + + for (len = 0; *p && isdigit(*p); ++p) + len = len * 10 + (*p - '0'); + if (*p) + return (-1); + DEBUG(fprintf(stderr, "content length: [%lld]\n", + (long long)len)); + *length = len; + return (0); +} + +/* + * Parse a content-range header + */ +static int +_http_parse_range(const char *p, off_t *offset, off_t *length, off_t *size) +{ + off_t first, last, len; + + if (strncasecmp(p, "bytes ", 6) != 0) + return (-1); + p += 6; + if (*p == '*') { + first = last = -1; + ++p; + } else { + for (first = 0; *p && isdigit(*p); ++p) + first = first * 10 + *p - '0'; + if (*p != '-') + return (-1); + for (last = 0, ++p; *p && isdigit(*p); ++p) + last = last * 10 + *p - '0'; + } + if (first > last || *p != '/') + return (-1); + for (len = 0, ++p; *p && isdigit(*p); ++p) + len = len * 10 + *p - '0'; + if (*p || len < last - first + 1) + return (-1); + if (first == -1) { + DEBUG(fprintf(stderr, "content range: [*/%lld]\n", + (long long)len)); + *length = 0; + } else { + DEBUG(fprintf(stderr, "content range: [%lld-%lld/%lld]\n", + (long long)first, (long long)last, (long long)len)); + *length = last - first + 1; + } + *offset = first; + *size = len; + return (0); +} + + +/***************************************************************************** + * Helper functions for authorization + */ + +/* + * Base64 encoding + */ +static char * +_http_base64(const char *src) +{ + static const char base64[] = + "ABCDEFGHIJKLMNOPQRSTUVWXYZ" + "abcdefghijklmnopqrstuvwxyz" + "0123456789+/"; + char *str, *dst; + size_t l; + int t, r; + + l = strlen(src); + if ((str = malloc(((l + 2) / 3) * 4 + 1)) == NULL) + return (NULL); + dst = str; + r = 0; + + while (l >= 3) { + t = (src[0] << 16) | (src[1] << 8) | src[2]; + dst[0] = base64[(t >> 18) & 0x3f]; + dst[1] = base64[(t >> 12) & 0x3f]; + dst[2] = base64[(t >> 6) & 0x3f]; + dst[3] = base64[(t >> 0) & 0x3f]; + src += 3; l -= 3; + dst += 4; r += 4; + } + + switch (l) { + case 2: + t = (src[0] << 16) | (src[1] << 8); + dst[0] = base64[(t >> 18) & 0x3f]; + dst[1] = base64[(t >> 12) & 0x3f]; + dst[2] = base64[(t >> 6) & 0x3f]; + dst[3] = '='; + dst += 4; + r += 4; + break; + case 1: + t = src[0] << 16; + dst[0] = base64[(t >> 18) & 0x3f]; + dst[1] = base64[(t >> 12) & 0x3f]; + dst[2] = dst[3] = '='; + dst += 4; + r += 4; + break; + case 0: + break; + } + + *dst = 0; + return (str); +} + +/* + * Encode username and password + */ +static int +_http_basic_auth(conn_t *conn, const char *hdr, const char *usr, const char *pwd) +{ + char *upw, *auth; + int r; + + DEBUG(fprintf(stderr, "usr: [%s]\n", usr)); + DEBUG(fprintf(stderr, "pwd: [%s]\n", pwd)); + if (asprintf(&upw, "%s:%s", usr, pwd) == -1) + return (-1); + auth = _http_base64(upw); + free(upw); + if (auth == NULL) + return (-1); + r = _http_cmd(conn, "%s: Basic %s", hdr, auth); + free(auth); + return (r); +} + +/* + * Send an authorization header + */ +static int +_http_authorize(conn_t *conn, const char *hdr, const char *p) +{ + /* basic authorization */ + if (strncasecmp(p, "basic:", 6) == 0) { + char *user, *pwd, *str; + int r; + + /* skip realm */ + for (p += 6; *p && *p != ':'; ++p) + /* nothing */ ; + if (!*p || strchr(++p, ':') == NULL) + return (-1); + if ((str = strdup(p)) == NULL) + return (-1); /* XXX */ + user = str; + pwd = strchr(str, ':'); + *pwd++ = '\0'; + r = _http_basic_auth(conn, hdr, user, pwd); + free(str); + return (r); + } + return (-1); +} + + +/***************************************************************************** + * Helper functions for connecting to a server or proxy + */ + +/* + * Connect to the correct HTTP server or proxy. + */ +static conn_t * +_http_connect(struct url *URL, struct url *purl, const char *flags) +{ + conn_t *conn; + int verbose; + int af; + +#ifdef INET6 + af = AF_UNSPEC; +#else + af = AF_INET; +#endif + + verbose = CHECK_FLAG('v'); + if (CHECK_FLAG('4')) + af = AF_INET; +#ifdef INET6 + else if (CHECK_FLAG('6')) + af = AF_INET6; +#endif + + if (purl && strcasecmp(URL->scheme, SCHEME_HTTPS) != 0) { + URL = purl; + } else if (strcasecmp(URL->scheme, SCHEME_FTP) == 0) { + /* can't talk http to an ftp server */ + /* XXX should set an error code */ + return (NULL); + } + + if ((conn = _fetch_connect(URL->host, URL->port, af, verbose)) == NULL) + /* _fetch_connect() has already set an error code */ + return (NULL); + if (strcasecmp(URL->scheme, SCHEME_HTTPS) == 0 && + _fetch_ssl(conn, verbose) == -1) { + _fetch_close(conn); + /* grrr */ + _fetch_syserr(); + return (NULL); + } + +#ifdef TCP_NOPUSH + int val = 1; + setsockopt(conn->sd, IPPROTO_TCP, TCP_NOPUSH, &val, sizeof(val)); +#endif + + return (conn); +} + +static struct url * +_http_get_proxy(const char *flags) +{ + struct url *purl; + char *p; + + if (flags != NULL && strchr(flags, 'd') != NULL) + return (NULL); + if (((p = getenv("HTTP_PROXY")) || (p = getenv("http_proxy"))) && + *p && (purl = fetchParseURL(p))) { + if (!*purl->scheme) + strcpy(purl->scheme, SCHEME_HTTP); + if (!purl->port) + purl->port = _fetch_default_proxy_port(purl->scheme); + if (strcasecmp(purl->scheme, SCHEME_HTTP) == 0) + return (purl); + fetchFreeURL(purl); + } + return (NULL); +} + +static void +_http_print_html(FILE *out, FILE *in) +{ + size_t len; + unsigned int sz = 1024; + char *line, *p, *q; + int comment, tag; + + line = (char *)calloc(1024, sizeof(char)); + comment = tag = 0; + while ((len = getline(&line, &sz, in)) > 0) { + while (len && isspace(line[len - 1])) + --len; + for (p = q = line; q < line + len; ++q) { + if (comment && *q == '-') { + if (q + 2 < line + len && + strcmp(q, "-->") == 0) { + tag = comment = 0; + q += 2; + } + } else if (tag && !comment && *q == '>') { + p = q + 1; + tag = 0; + } else if (!tag && *q == '<') { + if (q > p) + fwrite(p, q - p, 1, out); + tag = 1; + if (q + 3 < line + len && + strcmp(q, "