2 * Copyright (C) 2009-2010 Andy Spencer <andy753421@gmail.com>
4 * This program is free software: you can redistribute it and/or modify
5 * it under the terms of the GNU General Public License as published by
6 * the Free Software Foundation, either version 3 of the License, or
7 * (at your option) any later version.
9 * This program is distributed in the hope that it will be useful,
10 * but WITHOUT ANY WARRANTY; without even the implied warranty of
11 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12 * GNU General Public License for more details.
14 * You should have received a copy of the GNU General Public License
15 * along with this program. If not, see <http://www.gnu.org/licenses/>.
20 * @short_description: Hyper Text Transfer Protocol
22 * #GritsHttp is a small wrapper around libsoup to provide data access using
23 * the Hyper Text Transfer Protocol. Each #GritsHttp should be associated with
24 * a particular server or dataset, all the files downloaded for this dataset
25 * will be cached together in $HOME/.cache/grits/
30 #include <glib/gstdio.h>
31 #include <libsoup/soup.h>
33 #include "grits-http.h"
35 gchar *_get_cache_path(GritsHttp *http, const gchar *local)
37 return g_build_filename(g_get_user_cache_dir(), PACKAGE,
38 http->prefix, local, NULL);
43 * @prefix: The prefix in the cache to store the downloaded files.
44 * For example: * "/nexrad/level2/".
46 * Create a new #GritsHttp for the given prefix
48 * Returns: the new #GritsHttp
50 GritsHttp *grits_http_new(const gchar *prefix)
52 g_debug("GritsHttp: new - %s", prefix);
53 GritsHttp *http = g_new0(GritsHttp, 1);
54 http->soup = soup_session_sync_new();
55 http->prefix = g_strdup(prefix);
56 g_object_set(http->soup, "user-agent", PACKAGE_STRING, NULL);
62 * @http: the #GritsHttp to free
64 * Frees resources used by @http and cancels any pending requests.
66 void grits_http_free(GritsHttp *http)
68 g_debug("GritsHttp: free - %s", http->prefix);
69 soup_session_abort(http->soup);
70 g_object_unref(http->soup);
75 /* For passing data to the chunck callback */
79 GritsChunkCallback callback;
82 struct _CacheInfoMain {
84 GritsChunkCallback callback;
89 /* call the user callback from the main thread,
90 * since it's usually UI updates */
91 static gboolean _chunk_main_cb(gpointer _infomain)
93 struct _CacheInfoMain *infomain = _infomain;
94 infomain->callback(infomain->path,
95 infomain->cur, infomain->total,
102 * Append data to the file and call the users callback if they supplied one.
104 static void _chunk_cb(SoupMessage *message, SoupBuffer *chunk, gpointer _info)
106 struct _CacheInfo *info = _info;
108 if (!SOUP_STATUS_IS_SUCCESSFUL(message->status_code)) {
109 g_warning("GritsHttp: _chunk_cb - soup failed with %d",
110 message->status_code);
114 if (!fwrite(chunk->data, chunk->length, 1, info->fp))
115 g_error("GritsHttp: _chunk_cb - Unable to write data");
117 if (info->callback) {
118 struct _CacheInfoMain *infomain = g_new0(struct _CacheInfoMain, 1);
119 infomain->path = info->path;
120 infomain->callback = info->callback;
121 infomain->user_data = info->user_data;
122 infomain->cur = ftell(info->fp);
124 soup_message_headers_get_content_range(message->response_headers,
125 &st, &end, &infomain->total);
126 g_idle_add(_chunk_main_cb, infomain);
133 * @http: the #GritsHttp connection to use
134 * @uri: the URI to fetch
135 * @local: the local name to give to the file
136 * @mode: the update type to use when fetching data
137 * @callback: callback to call when a chunk of data is received
138 * @user_data: user data to pass to the callback
140 * Fetch a file from the cache. Whether the file is actually loaded from the
141 * remote server depends on the value of @mode.
143 * Returns: The local path to the complete file
145 /* TODO: use .part extentions and continue even when using GRITS_ONCE */
146 gchar *grits_http_fetch(GritsHttp *http, const gchar *uri, const char *local,
147 GritsCacheType mode, GritsChunkCallback callback, gpointer user_data)
149 g_debug("GritsHttp: fetch - %s mode=%d", local, mode);
150 gchar *path = _get_cache_path(http, local);
152 /* Unlink the file if we're refreshing it */
153 if (mode == GRITS_REFRESH)
156 /* Do the cache if necessasairy */
157 if (!(mode == GRITS_ONCE && g_file_test(path, G_FILE_TEST_EXISTS)) &&
158 mode != GRITS_LOCAL) {
159 g_debug("GritsHttp: fetch - Caching file %s", local);
161 /* Open the file for writting */
163 if (!g_file_test(path, G_FILE_TEST_EXISTS))
164 part = g_strdup_printf("%s.part", path);
165 FILE *fp = fopen_p(part, "ab");
166 fseek(fp, 0, SEEK_END); // "a" is broken on Windows, twice
169 struct _CacheInfo info = {
172 .callback = callback,
173 .user_data = user_data,
176 /* Download the file */
177 SoupMessage *message = soup_message_new("GET", uri);
179 g_error("message is null, cannot parse uri");
180 g_signal_connect(message, "got-chunk", G_CALLBACK(_chunk_cb), &info);
182 soup_message_headers_set_range(message->request_headers, ftell(fp), -1);
183 if (mode == GRITS_REFRESH)
184 soup_message_headers_replace(message->request_headers,
185 "Cache-Control", "max-age=0");
186 soup_session_send_message(http->soup, message);
190 if (path != part && SOUP_STATUS_IS_SUCCESSFUL(message->status_code)) {
191 g_rename(part, path);
196 if (message->status_code == SOUP_STATUS_CANCELLED) {
198 } else if (message->status_code == SOUP_STATUS_REQUESTED_RANGE_NOT_SATISFIABLE) {
199 /* Range unsatisfiable, file already complete */
200 } else if (!SOUP_STATUS_IS_SUCCESSFUL(message->status_code)) {
201 g_warning("GritsHttp: done_cb - error copying file, status=%d\n"
204 message->status_code, uri, path);
210 /* TODO: free everything.. */
215 * grits_http_available:
216 * @http: the #GritsHttp connection to use
217 * @filter: filter used to extract files from the index, or NULL
218 * For example: "href=\"([^"]*)\""
219 * @cache: path to the local cache, or NULL to not search the cache
220 * @extract: regex used to extract filenames from the page, should match the
221 * filename as $1, or NULL to use /http="([^"])"/
222 * @index: path to the index page, or NULL to not search online
224 * Look through the cache and an HTTP index page for a list of available files.
225 * The name of each file that matches the filter is added to the returned list.
227 * The list as well as the strings contained in it should be freed afterwards.
229 * Returns the list of matching filenames
231 GList *grits_http_available(GritsHttp *http,
232 gchar *filter, gchar *cache,
233 gchar *extract, gchar *index)
235 g_debug("GritsHttp: available - %s~=%s %s~=%s",
236 filter, cache, extract, index);
237 GRegex *filter_re = g_regex_new(filter, 0, 0, NULL);
240 /* Add cached files */
243 gchar *path = _get_cache_path(http, cache);
244 GDir *dir = g_dir_open(path, 0, NULL);
245 while ((file = g_dir_read_name(dir)))
246 if (g_regex_match(filter_re, file, 0, NULL))
247 files = g_list_prepend(files, g_strdup(file));
252 /* Add online files if online */
255 g_snprintf(tmp, sizeof(tmp), ".index.%x", g_random_int());
256 gchar *path = grits_http_fetch(http, index, tmp,
257 GRITS_REFRESH, NULL, NULL);
261 g_file_get_contents(path, &html, NULL, NULL);
265 /* Match hrefs by default, this regex is not very accurate */
266 GRegex *extract_re = g_regex_new(
267 extract ?: "href=\"([^\"]*)\"", 0, 0, NULL);
269 g_regex_match(extract_re, html, 0, &info);
270 while (g_match_info_matches(info)) {
271 gchar *file = g_match_info_fetch(info, 1);
273 if (g_regex_match(filter_re, file, 0, NULL))
274 files = g_list_prepend(files, file);
278 g_match_info_next(info, NULL);
281 g_regex_unref(extract_re);
282 g_match_info_free(info);
288 g_regex_unref(filter_re);