* all of robustness and flexibility and correctness.
* - The current implementation splits on "any run of CR and LF". Which
* translates to: Line numbers are wrong in the presence of empty
- * lines in the input stream.
- * - The current implementation insists in the presence of end-of-line
- * markers on _every_ line in the input stream. "Incomplete" text
- * files that are so typical on the Windows platform get rejected as
- * invalid.
+ * lines in the input stream. See below for an (expensive) fix.
* - Dropping support for CR style end-of-line markers could improve
* the situation a lot. Code could search for and split on LF, and
* trim optional trailing CR. This would result in proper support
* input stream is scanned for the first occurance of either of the
* supported termination styles (which is good). For the remaining
* session a consistent encoding of the text lines is assumed (which
- * is acceptable). Potential absence of the terminator for the last
- * line is orthogonal, and can get handled by a "force" flag when
- * the end() routine calls the process_buffer() routine.
+ * is acceptable).
* - When line numbers need to be correct and reliable, _and_ the full
* set of previously supported line termination sequences are required,
* and potentially more are to get added for improved compatibility
return ret;
}
+/*
+ * Gets called from initial_receive(), which runs until the end-of-line
+ * encoding of the input stream could get determined. Assumes that this
+ * routine receives enough buffered initial input data to either see the
+ * BOM when there is one, or that no BOM will follow when a text line
+ * termination sequence was seen. Silently drops the UTF-8 BOM sequence
+ * from the input buffer if one was seen. Does not care to protect
+ * against multiple execution or dropping the BOM multiple times --
+ * there should be at most one in the input stream.
+ */
+static void initial_bom_check(const struct sr_input *in)
+{
+ static const char *utf8_bom = "\xef\xbb\xbf";
+
+ if (in->buf->len < strlen(utf8_bom))
+ return;
+ if (strncmp(in->buf->str, utf8_bom, strlen(utf8_bom)) != 0)
+ return;
+ g_string_erase(in->buf, 0, strlen(utf8_bom));
+}
+
static int initial_receive(const struct sr_input *in)
{
struct context *inc;
char *p;
const char *termination;
+ initial_bom_check(in);
+
inc = in->priv;
termination = get_line_termination(in->buf);
return ret;
}
-static int process_buffer(struct sr_input *in)
+static int process_buffer(struct sr_input *in, gboolean is_eof)
{
struct sr_datafeed_packet packet;
struct sr_datafeed_meta meta;
* maximum amount of accumulated data that consists of full text
* lines, and process what has been received so far, leaving not
* yet complete lines for the next invocation.
+ *
+ * Enforce that all previously buffered data gets processed in
+ * the "EOF" condition. Do not insist in the presence of the
+ * termination sequence for the last line (may often be missing
+ * on Windows). A present termination sequence will just result
+ * in the "execution of an empty line", and does not harm.
*/
if (!in->buf->len)
return SR_OK;
- p = g_strrstr_len(in->buf->str, in->buf->len, inc->termination);
- if (!p)
- return SR_ERR;
- *p = '\0';
- p += strlen(inc->termination);
+ if (is_eof) {
+ p = in->buf->str + in->buf->len;
+ } else {
+ p = g_strrstr_len(in->buf->str, in->buf->len, inc->termination);
+ if (!p)
+ return SR_ERR;
+ *p = '\0';
+ p += strlen(inc->termination);
+ }
g_strstrip(in->buf->str);
ret = SR_OK;
return SR_OK;
}
- ret = process_buffer(in);
+ ret = process_buffer(in, FALSE);
return ret;
}
int ret;
if (in->sdi_ready)
- ret = process_buffer(in);
+ ret = process_buffer(in, TRUE);
else
ret = SR_OK;