|author||James Bursa <email@example.com>||2003-04-17 21:35:02 +0000|
|committer||James Bursa <firstname.lastname@example.org>||2003-04-17 21:35:02 +0000|
[project @ 2003-04-17 21:35:02 by bursa]
Max one fetch from each host at once, fix multiple fetches of same url. svn path=/import/netsurf/; revision=127
Diffstat (limited to 'content/fetch.h')
1 files changed, 24 insertions, 1 deletions
diff --git a/content/fetch.h b/content/fetch.h
index ab3029b23..92c269a02 100644
@@ -1,5 +1,28 @@
- * $Id: fetch.h,v 1.2 2003/03/15 15:53:20 bursa Exp $
+ * $Id: fetch.h,v 1.3 2003/04/17 21:35:02 bursa Exp $
+ * This module handles fetching of data from any url.
+ * Usage:
+ * fetch_init() must be called once before any other function. fetch_quit()
+ * must be called before exiting.
+ * fetch_start() will begin fetching a url. The function returns immediately.
+ * A pointer to an opaque struct fetch is returned, which can be passed to
+ * fetch_abort() to abort the fetch at any time. The caller must supply a
+ * callback function which is called when anything interesting happens. The
+ * callback function is first called with msg = FETCH_TYPE, with the
+ * Content-Type header in data, then one or more times with FETCH_DATA with
+ * some data for the url, and finally with FETCH_FINISHED. Alternatively,
+ * FETCH_ERROR indicates an error occurred: data contains an error message.
+ * Some private data can be passed as the last parameter to fetch_start, and
+ * callbacks will contain this.
+ * fetch_poll() must be called regularly to make progress on fetches.
+ * fetch_filetype() is used internally to determine the mime type of local
+ * files. It is platform specific, and implemented elsewhere.