aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
Diffstat (limited to '')
-rw-r--r--src/gr-fadingui/python/datasource.py18
1 files changed, 9 insertions, 9 deletions
diff --git a/src/gr-fadingui/python/datasource.py b/src/gr-fadingui/python/datasource.py
index 69a2122..12573a8 100644
--- a/src/gr-fadingui/python/datasource.py
+++ b/src/gr-fadingui/python/datasource.py
@@ -35,27 +35,27 @@ class datasource(gr.basic_block):
# a frame has 5 id bits so, there can only be 2 ** 5 chunks per file
# see docstring of frame_obj for more details
nblocks = int(self.fsize / self.frame.payload_length)
- log.debug(f"Loaded {self.fsize} bytes = {nblocks} blocks from name={self.fname}")
+ log.debug(f"Loaded {self.fsize} bytes = {nblocks} blocks from {self.fname}")
assert nblocks < 2 ** 5, "Payload size too small or file too big"
self.fpos = 0
self.blocknr = 0
- # would have been nice to have but does not work
- # self.set_min_noutput_items(frame_obj.length)
+ # Ensure that output buffer is always at least as big a one frame
+ # This probably hits the performance of the simulation quite badly,
+ # it would be better to implement buffering (see below)
+ self.set_output_multiple(self.frame.length)
- # FIXME: implement buffering
- # output buffer
+ # FIXME: Implement buffering if the GR algorithm gives a smaller output
+ # buffer. This is partially solved with the previous line.
self.outbuffer = np.array([])
def general_work(self, input_items, output_items):
out = output_items[0]
-
# FIXME: if there is leftover buffer add that first
- # if self.outbuffer.size > 0:
- # log.debug("Frame did not fit into buffer")
- # out[:len(self.outbuffer)] = self.outbuffer
+ if self.outbuffer.size > 0:
+ log.warning("Frame did not fit into buffer")
if self.fpos + self.frame.payload_length > self.fsize:
# FIXME: implement edge case