runtc: unbreak the protocol decoder output match logic (type vs instance)

Recent libsigrokdecode adjustment changed decoder output. The ID of a
protocol decoder instance need no longer match the decoder type's name
even for simple setups.

Lookup the selected decoder's instance ID during construction of the
decoder stack. Such that the selected decoder's data can get identified
in the output of the stack which is received by runtc(1).

Note that this implementation passes the decoder type's name to the
runtc(1) output file, so that it keeps matching the expected .output
files when pdtest(1) compares those two streams. This approach avoids
the necessity to adjust test.conf or .output files. It will no longer
be appropriate when more complex test configurations shall be supported
which involve multiple instances of the same decoder type or multiple
stacks of decoders.

diff --git a/decoder/runtc.c b/decoder/runtc.c
index 4b534a2564048ae52188c02f2f1d040f9ec1f071..c31b7d3d1ddea03dda8e1c7c965013d6f800b3f6 100644 (file)
--- a/decoder/runtc.c
+++ b/decoder/runtc.c
@@ -60,6 +60,7 @@ struct pd {
 
 struct output {
        const char *pd;
+       const char *pd_id;
        int type;
        const char *class;
        int class_idx;
@@ -168,6 +169,28 @@ static char *py_str_as_str(const PyObject *py_str)
        return outstr;
 }
 
+/*
+ * The following routines are callbacks for libsigrokdecode. They receive
+ * output from protocol decoders, optionally dropping data to only forward
+ * a selected decoder's or class' information. Output is written to either
+ * a specified file or stdout, an external process will compare captured
+ * output against expectations.
+ *
+ * Note that runtc(1) output emits the decoder "class" name instead of the
+ * instance name. So that generated output remains compatible with existing
+ * .output files which hold expected output of test cases. Without this
+ * approach, developers had to "anticipate" instance names from test.conf
+ * setups (and knowledge about internal implementation details of the srd
+ * library), and adjust .output files to reflect those names. Or specify
+ * instance names in each and every test.conf description (-o inst_id=ID).
+ *
+ * It's assumed that runtc(1) is used to check stacked decoders, but not
+ * multiple stacks in parallel and no stacks with multiple instances of
+ * decoders of the same type. When such configurations become desirable,
+ * runtc(1) needs to emit the instance name, and test configurations and
+ * output expectations need adjustment.
+ */
+
 static void srd_cb_py(struct srd_proto_data *pdata, void *cb_data)
 {
        struct output *op;
@@ -175,12 +198,12 @@ static void srd_cb_py(struct srd_proto_data *pdata, void *cb_data)
        GString *out;
        char *s;
 
-       DBG("Python output from %s", pdata->pdo->di->decoder->id);
+       DBG("Python output from %s", pdata->pdo->di->inst_id);
        op = cb_data;
        pydata = pdata->data;
        DBG("ptr %p", pydata);
 
-       if (strcmp(pdata->pdo->di->decoder->id, op->pd))
+       if (strcmp(pdata->pdo->di->inst_id, op->pd_id))
                /* This is not the PD selected for output. */
                return;
 
@@ -191,7 +214,7 @@ static void srd_cb_py(struct srd_proto_data *pdata, void *cb_data)
        s = py_str_as_str(pyrepr);
        Py_DecRef(pyrepr);
 
-       /* Output format for testing is '<ss>-<es> <inst-id>: <repr>\n'. */
+       /* Output format for testing is '<ss>-<es> <decoder-id>: <repr>\n'. */
        out = g_string_sized_new(128);
        g_string_printf(out, "%" PRIu64 "-%" PRIu64 " %s: %s\n",
                        pdata->start_sample, pdata->end_sample,
@@ -211,11 +234,11 @@ static void srd_cb_bin(struct srd_proto_data *pdata, void *cb_data)
        GString *out;
        unsigned int i;
 
-       DBG("Binary output from %s", pdata->pdo->di->decoder->id);
+       DBG("Binary output from %s", pdata->pdo->di->inst_id);
        op = cb_data;
        pdb = pdata->data;
 
-       if (strcmp(pdata->pdo->di->decoder->id, op->pd))
+       if (strcmp(pdata->pdo->di->inst_id, op->pd_id))
                /* This is not the PD selected for output. */
                return;
 
@@ -241,6 +264,7 @@ static void srd_cb_bin(struct srd_proto_data *pdata, void *cb_data)
 
 static void srd_cb_ann(struct srd_proto_data *pdata, void *cb_data)
 {
+       struct srd_decoder_inst *di;
        struct srd_decoder *dec;
        struct srd_proto_data_annotation *pda;
        struct output *op;
@@ -248,11 +272,17 @@ static void srd_cb_ann(struct srd_proto_data *pdata, void *cb_data)
        int i;
        char **dec_ann;
 
-       DBG("Annotation output from %s", pdata->pdo->di->decoder->id);
+       /*
+        * Only inspect received annotations when they originate from
+        * the selected protocol decoder, and an optionally specified
+        * annotation class matches the received data.
+        */
        op = cb_data;
        pda = pdata->data;
-       dec = pdata->pdo->di->decoder;
-       if (strcmp(pdata->pdo->di->decoder->id, op->pd))
+       di = pdata->pdo->di;
+       dec = di->decoder;
+       DBG("Annotation output from %s", di->inst_id);
+       if (strcmp(di->inst_id, op->pd_id))
                /* This is not the PD selected for output. */
                return;
 
@@ -263,11 +293,17 @@ static void srd_cb_ann(struct srd_proto_data *pdata, void *cb_data)
                 */
                return;
 
+       /*
+        * Print the annotation information in textual representation
+        * to the specified output file. Prefix the annotation strings
+        * with the start and end sample number, the decoder name, and
+        * the annotation name.
+        */
        dec_ann = g_slist_nth_data(dec->annotations, pda->ann_class);
        line = g_string_sized_new(256);
        g_string_printf(line, "%" PRIu64 "-%" PRIu64 " %s: %s:",
                        pdata->start_sample, pdata->end_sample,
-                       pdata->pdo->di->decoder->id, dec_ann[0]);
+                       dec->id, dec_ann[0]);
        for (i = 0; pda->ann_text[i]; i++)
                g_string_append_printf(line, " \"%s\"", pda->ann_text[i]);
        g_string_append(line, "\n");
@@ -411,6 +447,17 @@ static int run_testcase(const char *infile, GSList *pdlist, struct output *op)
                        return FALSE;
                g_hash_table_destroy(opts);
 
+               /*
+                * Get (a reference to) the decoder instance's ID if we
+                * are about to receive PD output from it. We need to
+                * filter output that carries the decoder instance's name.
+                */
+               if (strcmp(pd->name, op->pd) == 0) {
+                       op->pd_id = di->inst_id;
+                       DBG("Decoder of type \"%s\" has instance ID \"%s\".",
+                           op->pd, op->pd_id);
+               }
+
                /* Map channels. */
                if (pd->channels) {
                        channels = g_hash_table_new_full(g_str_hash, g_str_equal, NULL,
@@ -442,8 +489,14 @@ static int run_testcase(const char *infile, GSList *pdlist, struct output *op)
                }
                prev_di = di;
        }
+       /*
+        * Bail out if we haven't created an instance of the selected
+        * decoder type of which we shall grab output data from.
+        */
+       if (!op->pd_id)
+               return FALSE;
 
-       /* Resolve top decoder's class index, so we can match. */
+       /* Resolve selected decoder's class index, so we can match. */
        dec = srd_decoder_get_by_id(pd->name);
        if (op->class) {
                if (op->type == SRD_OUTPUT_ANN)
@@ -718,6 +771,7 @@ int main(int argc, char **argv)
 
        op = malloc(sizeof(struct output));
        op->pd = NULL;
+       op->pd_id = NULL;
        op->type = -1;
        op->class = NULL;
        op->class_idx = -1;