| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
Bug #897 and my own experience caused me to improve the SPDIF decoder.
The following issues were addressed and resolved:
1. The Error: "srd: ValueError: Protocol decoder instance spdif-1:
invalid literal for int() with base 2: '121111012112021111012120'"
This error can happen if the sample rate is marginal. The correct function
of the decoder depends on the position of the data stream. The pulse width
calculation was wrong and the pulse width detection sometimes thought the
same pulse classes to be different.
The new decoder explicitly checks for short pulses and reports an error
with a corresponding message.
2. Bitrates were calculated wrong: The shown results were not usable at all.
The new decoder uses the first ten frames to calculate the bit rates and
uses the correct measurement units.
Possible issue: The bit rate calculation assumes an ongoing data stream.
It uses the time between the first and 10th frame.
They need to be sent without interruption. But this should be no problem
because SPDIF is meant to be a continuous stream.
3. A missing samplerate, e.g. when used in sigrok-cli with binary input, lead
to an error message on the original decoder.
The new decoder just skips the output of the bitrate if the samplerate is
missing. A missing samplerate no longer raises an error but only a message
in the data output.
4. The user was not informed about the integral steps of the decoder.
The new decoder shows the results of the synchronisation process at the
beginning. This can help to understand the behaviour of the decoder.
This fixes bug #897.
|
|
The idea was taken from the vector slicer as suggested in github PR 17.
The code was rewritten to avoid cluttering the decoder set with lots of
tiny implementations. Create a single decoder which does all of the bit
accumulation as well as number format interpretation and also includes
number to text formatting with user selectable presentations. Optional
clock is supported, to avoid too many annotations in glitchy setups. The
IEEE 754 format was added as another interpretation. The enum code path
accepts data files in either the Python or JSON format (the latter feels
backwards to me). Putting enums on separate rows helps visualize state
transitions. Users can disable GUI traces or select CLI rows to silence
the output if it feels noisy.
The most appropriate (useful, and usable) number of supported channels
is yet to get determined. This version accepts up to 16 inputs.
|
|
Support of the now common RGBW type LED strips (uses 4 bytes instead of 3).
Added an option to select RGB or RGBW
|
|
Unclutter the construction of the annotation text for measurements.
Prefer one code block over multiple scatterred lines. Only do string
formatting when an annotation gets emitted. Prefer .format() for its
better control over generated text. Fixup remaining Python idioms in
the unit conversion block while we are here.
|
|
The previous implementation unconditionally waited for up to 1ms, and
optionally handled the user configured timeout period. Use common code
instead to handle the full span of the timeout period. Which somewhat
unclutters the .decode() routine's body by eliminating an unnecessary
step in the sequence. This change also reduces indentation, and prefers
Python's .format() over the old syntax.
|
|
There is only one annotation, rename ss/es for consistency with other
decoders. Move a conversion constant to the top of the source file, and
rename some identifiers to increase readability of math expressions.
There is no point in constructing numbers in MSB first order and then
reversing their bit order. Prefer Python lists for accumulated bits, and
common code for number conversion. Use one code block for the extraction
and conversion of received data instead of a set of scattered lines.
This shall increase readability of the data extraction, and eliminates
redundancy in the implementation.
|
|
Move math expressions with always identical values out of the loop, and
improve readability of the packet timeout condition in the process.
|
|
Merge .reset_data() into .reset() since no other location called it.
Move self.options[] lookup out of the loop, drop state from self when
it's strictly local to .decode(). Use a common annotation emission
routine, and vertically align text variants for zoom levels. This
further reduces the lengths of a few text lines.
|
|
The caliper decoder was written against an older version of the library
and failed to load in recent environments. Fixup the decoder boilerplate
(eliminate ambiguous annotation class/row names, drop the IC tag) and
address other style nits while we are here (rephrase user visible text,
break a few long lines, adjust some of the indentation and whitespace
issues, drop dead code). Remaining issues get addressed later.
|
|
|
|
The IRMP core library is not prepared for threading or interleaved use
by multiple call sites for different data streams, internal state is
kept in global vars (MCU project heritage). Adjust the Python wrapper,
create one usable instance, and several more which fail to execute.
Fail late such that users see error messages.
The approach isn't pretty, but avoids segfaults when re-loaded sessions
assign multiple decoder instances, and raises user's awareness of the
"one instance" limitation by established means: "decoder error" bar, and
log messages, with a description to point out the cause.
This commit implements a dirty modification of a singleton. It's a pity
that Python appears to lack reliable destruction, hence the whole class
remains blocked even if the instance is released. Move all library use
into pd.py:decode() in the hope that Python's 'with' could help in a
future implementation. Prepare to either present a generic message that
is generated by pd.py, or pass on a text that originates in the Python
wrapper for the C library.
|
|
Recent upstream IRMP core versions introduced a "release" flag in
addition to the "repeat" flag. Prepare the decoder to present these
flags when libraries should pass them in results.
The flags' being orthogonal slightly complicates the logic which
constructs annotation texts. Do provide text variants for all previously
supported zoom levels, yet try to keep the implementation as simple as
possible: Match list lengths for simplified folding. Always print the
flags field even if none of the flags is active (kind of was done before
this change as well, just not visible). This approach easily accepts
more flags as needed in future versions.
|
|
Address Python and sigrok project coding style nits in the IRMP based
decoder for infrared signals. Re-add the attribution which was lost in a
previous copy and forgotten in the recent submission. Eliminate camel
case identifiers and adjust to the simplified Python binding API. Drop
remaining diagnostics from development and dead code (unused carrier
detection). Reword the boilerplate text to match other decoders. Avoid
Python f-strings since they are not portable. Prefer slightly less
cryptic variable names in the construction of annotation texts. Defer
the creation of the library instance until actual decoder use. Start
inspecting the input data at the very first samples in the input stream.
|
|
Commit the decoder as it was provided by Rene Staffen. Which appears to
be a slightly modified version of one of the other IR decoders, though
the boilerplate doesn't say so.
|
|
Detect the MacOS platform by checking for 'Darwin' with the Python
platform(3) module, and use the 'lib<stem>.dylib' filename scheme.
|
|
Rename the Python language binding's source file and identifiers to
eliminate camel case (most of it, stick with camel case in Python class
names as is the convention). Adjust whitespace and arrange tables such
that their indentation will last during maintenance.
Re-add the license text which was missing in the original submission's
copy of another decoder. Add copyright information for this submission.
Don't "import *" from ctypes(3), use explicit references instead. Avoid
double underscores as single leading underscore is already bad enough.
Adjust the Python side to the C library's renamed API routines.
Create a result data structure layout that only has a single level of
nesting, which better represents the C library's interface. Only flags
"get unfolded" in the Python binding, to eliminate magic numbers.
Prepare to support more platforms than Linux (detected) and Windows (the
default when nothing else got detected).
|
|
Commit the IrmpPythonWrap.py file as it was provided by Rene Staffen.
|
|
|
|
The PJDL decoder's previous implementation was incomplete. It assumed
that PAD bits always start with a rising edge. Which made the decoder
miss the next byte when a previous byte's MSB is set, and the last DATA
bit and the next PAD bit kept the signal HIGH between them (no LOW phase
was seen between these symbols).
Keep the check for the LOW level after the byte's last DATA bit within
the bit times' tolerance. But accept when the level remains HIGH, and
check for the HIGH bit's width starting from the end of the last DATA
bit. Also start the PAD bit's annotation from that "virtual" edge.
This patch is based on a fix that was
Submitted-By: Julio Aguirre <jcallano@gmail.com>
|
|
The previous implementation unconditionally assumed a CRC width of
one byte when it calculated the checksum for received frame data.
Do reflect on the CRC8/CRC32 choice instead.
This patch is based on a fix that was
Submitted-By: Julio Aguirre <jcallano@gmail.com>
|
|
|
|
|
|
|
|
|
|
Optionally let users pick the scale for terse timing annotation text.
Which potentially makes numbers show up earlier (at zoom levels of a
further distance). And drops the unit to present mere numbers, which
could speed up navigation during inspection. Keep providing automatic
scaling which then includes the unit text, as it did before.
Extend the automatic scaling to include picoseconds. Which avoids the
fallback to unit-less floating point with uncertain decimals when the
samplerate was 1GHz or higher.
Optionally present distances in terms of sample counts. This supports
decoder development, and can help users spot and judge glitches.
All "terse" presentation so far exclusively affects the 'time' row. It
remains an option for later to migrate averages and deltas as well. For
now it's assumed that high(er) precision and fine grained details are
more important for these rows.
|
|
Break text lines in the options declarations which have become rather
long. Rename 'samples' in the main loop to just 'sa', which better
matches the other 'ss', 'es', 't', etc identifers. Separate the code
for unconditional 'time' classes from optional averaging and deltas.
|
|
Reduce the amount of work which the timing decoder needs to do. Only
keep the deque() filled when averaging is active. Rephrase .put() calls
to reduce text line lengths (and for consistency with a pending change).
Move another options lookup for deltas out of the main loop.
|
|
In some situations (inspecting a dense run of pulses in a burst of data
communication) it takes a lot of zooming before the 'timing' decoder's
'time' annotations start revealing numbers. Which limits the number of
pulses which can fit in the visible trace area.
This is a stab at improving the usability of the timing decoder for
similar "crowded pulses" scenarios. Try to come up with an annotation
text that is shorter yet communicates the very details which the user
needs in this situation. Drop the frequency, avoid umlauts in the unit
text, don't use decimal places (use all integers within a scale). Even
offer to drop the unit text, assuming that a dense run of pulses results
in all times sharing their scale.
Make the terse presentation optional and off by default, and use a
separate annotation class for maximum backwards compatibility.
|
|
Consistently use ss and es identifiers for annotation emission to match
other decoders, as well as counting distances between sample points to
increase readability. This also dramatically reduces text line length.
|
|
Reduce the number of self members, use local variables instead for data
which is strictly kept within a method and need not remain across calls.
Move options dictionary lookups out of the main loop, as the previous
implementation already did with 'edge'.
|
|
Use symbolic identifiers for pin numbers and annotation classes. Remove
unused variables.
|
|
|
|
The SAE J1850 Variable Pulse Width decoder used to track and annotate
the width of pulses between edges, which duplicates existing features
of the 'timing' decoder. Remove this part from J1850, users can always
connect the input signal to multiple decoders as needed..
Also sort annotation rows while we are here. Top to bottom represents
raw wire bits to highest interpretation layer, as in other decoders.
|
|
Use symbolic identifiers for annotation classes, to improve readability
and maintainability.
|
|
IRC user pman92 reported that this decoder exists, and started migration
to the v3 API. This commit completes the migration, and adds missing
decoder infrastructure which has become mandatory recently.
Adjust the boilerplate: Drop FSF postal address. No Python output, add
category tag, unambiguous annotation class and row names. Add reset()
method. Use common code for edge detection.
This commit also addresses minor style nits. Pass the most recent
pulse's edges as ss and es to the data bit handling routine. Adjust
whitespace to unbreak editor navigation and to improve readability.
Use a more verbose name for the decoder, "vpw" appears a little short
and collision happy, and is not found when users search for "j1850".
[ Indentation changed, see whitespace ignoring diff for the essence. ]
Reported-By: pman92 <dpriestley92@hotmail.com>
|
|
Introduce a protocol decoder for the GM VPW 1x and 4x Vehicle Bus
(SAE J1850, or VPW for variable pulse width).
|
|
Do track the RX and TX information, including their bus IDs. Present bus
numbers as dotted quads. Emit another summary annotation for completed
frames which presents receiver, transmitter, payload, and ACK details at
even higher zoom levels. Rename the last remaining "init CRC" instance
for consistency.
|
|
Since the spec is vague on the subject, and real world captures were
found to occassionally run on odd clocks, internally prepare to inspect
traffic and interpret its content although the input data is invalid in
the strictest sense. Keep this hack internal, don't suggest to users
that invalid traffic would be perfectly acceptable.
|
|
Rename 'pjon-link' to 'pjon_link' for consistency with other decoders.
|
|
Introduce a protocol decoder which accepts 'pjon-link' Python input and
interprets PJON frames. The implementation is assumed to be operational
but most of the protocol's flexibility (optionally present and variable
width fields) has not yet been tested due to lack of example captures.
During development of the PJON decoder only the PJDL link layer decoder
was available, other link layers were not tested.
|
|
Introduce a protocol decoder which generates 'pjon-link' output from
'logic' input by interpreting the PJDL single wire serial communication
link layer of the PJON protocol stack. This decoder extracts frame
markers, data bytes, as well as their pad/sync decoration. Inspection of
data values, or checks for frame validity remain the responsibility of a
stacked decoder which is shared among several link layer types.
This implementation "violates" the PJDL spec in those places where the
spec is incomplete or vague, and real world traffic would not decode at
all when the strict letter of the spec is applied instead of its spirit.
When in doubt, the decoder implementation errs to the usability side.
Carrier sense detection is incomplete in this version. Data extraction
works for all currently available captures. Recovery from synchronization
loss after glitches is acceptable. Glitch filtering is missing (the spec
is silent on this subject).
|
|
Improve processing time by appending bits
instead of inserting them to the lists.
|
|
|
|
Drop the 0x prefix for each byte in annotations (for improved readability).
Also, use 02X instead of 02x (printf-style formats).
|
|
If those are useful for the decoder user, they should be annotations
using the Ann.WARN annotation class.
|
|
|
|
|
|
|
