From: Gerhard Sittig Date: Sun, 19 Jul 2020 17:41:18 +0000 (+0200) Subject: numbers_and_state: introduce decoder (based on vector slicer) X-Git-Url: https://sigrok.org/gitaction?a=commitdiff_plain;h=b094fdfc1c1cfe1e34f95c62e73241cff5e4074c;p=libsigrokdecode.git numbers_and_state: introduce decoder (based on vector slicer) 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. --- diff --git a/decoders/numbers_and_state/__init__.py b/decoders/numbers_and_state/__init__.py new file mode 100644 index 0000000..4fe42a3 --- /dev/null +++ b/decoders/numbers_and_state/__init__.py @@ -0,0 +1,41 @@ +## +## This file is part of the libsigrokdecode project. +## +## Copyright (C) 2019 Comlab AG +## Copyright (C) 2020 Gerhard Sittig +## +## This program is free software; you can redistribute it and/or modify +## it under the terms of the GNU General Public License as published by +## the Free Software Foundation; either version 2 of the License, or +## (at your option) any later version. +## +## This program is distributed in the hope that it will be useful, +## but WITHOUT ANY WARRANTY; without even the implied warranty of +## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +## GNU General Public License for more details. +## +## You should have received a copy of the GNU General Public License +## along with this program; if not, see . +## + +''' +This protocol decoder takes a set of logic input signals, and interprets +their bit pattern according to user specifications as different kinds of +numbers, or an enumeration of e.g. machine states. + +Supported formats are: signed and unsigned integers, fixed point numbers, +IEEE754 floating point numbers, and number to text mapping controlled by +external data files. (Support for half precision floats depends on the +Python runtime, and may not universally be available.) + +User provided text mapping files can either use the JSON format: + {"one": 1, "two": 2, "four": 4} +or the Python programming language: + enumtext = { 1: "one", 2: "two", 3: "three", } + +In addition to all enum values on one row (sequential presentation of +the data), a limited number of enum values also are shown in tabular +presentation, which can help visualize state machines or task switches. +''' + +from .pd import Decoder diff --git a/decoders/numbers_and_state/pd.py b/decoders/numbers_and_state/pd.py new file mode 100644 index 0000000..b8ac87e --- /dev/null +++ b/decoders/numbers_and_state/pd.py @@ -0,0 +1,377 @@ +## +## This file is part of the libsigrokdecode project. +## +## Copyright (C) 2019 Comlab AG +## Copyright (C) 2020 Gerhard Sittig +## +## This program is free software; you can redistribute it and/or modify +## it under the terms of the GNU General Public License as published by +## the Free Software Foundation; either version 2 of the License, or +## (at your option) any later version. +## +## This program is distributed in the hope that it will be useful, +## but WITHOUT ANY WARRANTY; without even the implied warranty of +## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +## GNU General Public License for more details. +## +## You should have received a copy of the GNU General Public License +## along with this program; if not, see . +## + +# This implementation started as a "vector slicer", then turned into the +# "numbers and states" decoder, because users always had the freedom to +# connect any logic signal to either of the decoder inputs. That's when +# slicing vectors took second seat, and just was not needed any longer +# in the strict sense. +# +# TODO +# - Find an appropriate number of input channels, and maximum enum slots. +# - Re-check correctness of signed integers. Signed fixed point is based +# on integers and transparently benefits from fixes and improvements. +# - Local formatting in individual decoders becomes obsolete when common +# support for user selected formatting gets introduced. +# - There is overlap with the 'parallel' decoder. Ideally the numbers +# decoder could stack on top of parallel, but parallel currently is +# severely limited in its number of input channels, and dramatically +# widening the parallel decoder may be undesirable. + +from common.srdhelper import bitpack +import json +import sigrokdecode as srd +import struct + +''' +OUTPUT_PYTHON format: + +Packet: +[, ] + +This is a list of s and their respective values: + - 'RAW': The data is a tuple of bit count and bit pattern (a number, + assuming unsigned integer presentation of the input data bit pattern). + - 'NUMBER': The data is the conversion result of the bit pattern. + - 'ENUM': The data is a tuple of the raw number and its mapped text. +''' + +# TODO Better raise the number of channels to 32. This allows access to +# IEEE754 single precision numbers, and shall cover most busses, _and_ +# remains within most logic analyzers' capabilities, and keeps the UI +# dialog somewhat managable. What's a good default for the number of +# enum slots (which translate to annotation rows)? Notice that 2 to the +# power of the channel count is way out of the question. :) +_max_channels = 16 +_max_enum_slots = 32 + +class ChannelError(Exception): + pass + +class Pin: + CLK, BIT_0 = range(2) + BIT_N = BIT_0 + _max_channels + +class Ann: + RAW, NUM = range(2) + ENUM_0 = NUM + 1 + ENUM_OVR = ENUM_0 + _max_enum_slots + ENUMS = range(ENUM_0, ENUM_OVR) + WARN = ENUM_OVR + 1 + + @staticmethod + def enum_indices(): + return [i for i in range(Ann.ENUMS)] + + @staticmethod + def get_enum_idx(code): + if code in range(_max_enum_slots): + return Ann.ENUM_0 + code + return Ann.ENUM_OVR + +def _channel_decl(count): + return tuple([ + {'id': 'bit{}'.format(i), 'name': 'Bit{}'.format(i), 'desc': 'Bit position {}'.format(i)} + for i in range(count) + ]) + +def _enum_cls_decl(count): + return tuple([ + ('enum{}'.format(i), 'Enumeration slot {}'.format(i)) + for i in range(count) + ] + [('enumovr', 'Enumeration overflow')]) + +def _enum_rows_decl(count): + return tuple([ + ('enums{}'.format(i), 'Enumeration slots {}'.format(i), (Ann.ENUM_0 + i,)) + for i in range(count) + ] + [('enumsovr', 'Enumeration overflows', (Ann.ENUM_OVR,))]) + +class Decoder(srd.Decoder): + api_version = 3 + id = 'numbers_and_state' + name = 'Numbers and State' + longname = 'Interpret bit patters as numbers or state enums' + desc = 'Interpret bit patterns as different kinds of numbers (integer, float, enum).' + license = 'gplv2+' + inputs = ['logic'] + outputs = ['numbers_and_state'] + tags = ['Encoding', 'Util'] + optional_channels = ( + {'id': 'clk', 'name': 'Clock', 'desc': 'Clock'}, + ) + _channel_decl(_max_channels) + options = ( + {'id': 'clkedge', 'desc': 'Clock edge', 'default': 'rising', + 'values': ('rising', 'falling', 'either')}, + {'id': 'count', 'desc': 'Total bits count', 'default': 0}, + {'id': 'interp', 'desc': 'Interpretation', 'default': 'unsigned', + 'values': ('unsigned', 'signed', 'fixpoint', 'fixsigned', 'ieee754', 'enum')}, + {'id': 'fracbits', 'desc': 'Fraction bits count', 'default': 0}, + {'id': 'mapping', 'desc': 'Enum to text map file', + 'default': 'enumtext.json'}, + {'id': 'format', 'desc': 'Number format', 'default': '-', + 'values': ('-', 'bin', 'oct', 'dec', 'hex')}, + ) + annotations = ( + ('raw', 'Raw pattern'), + ('number', 'Number'), + ) + _enum_cls_decl(_max_enum_slots) + ( + ('warning', 'Warning'), + ) + annotation_rows = ( + ('raws', 'Raw bits', (Ann.RAW,)), + ('numbers', 'Numbers', (Ann.NUM,)), + ) + _enum_rows_decl(_max_enum_slots) + ( + ('warnings', 'Warnings', (Ann.WARN,)), + ) + + def __init__(self): + self.reset() + + def reset(self): + pass + + def start(self): + self.out_ann = self.register(srd.OUTPUT_ANN) + self.out_python = self.register(srd.OUTPUT_PYTHON) + + def putg(self, ss, es, cls, data): + self.put(ss, es, self.out_ann, [cls, data]) + + def putpy(self, ss, es, ptype, pdata): + self.put(ss, es, self.out_python, (ptype, pdata)) + + def grab_pattern(self, pins): + '''Get a bit pattern from potentially incomplete probes' values.''' + + # Pad and trim the input data, to achieve the user specified + # total number of bits. Map all unassigned signals to 0 (low). + # Return raw number (unsigned integer interpreation). + bits = pins + (None,) * self.bitcount + bits = bits[:self.bitcount] + bits = [b if b in (0, 1) else 0 for b in bits] + pattern = bitpack(bits) + return pattern + + def handle_pattern(self, ss, es, pattern): + fmt = '{{:0{}b}}'.format(self.bitcount) + txt = fmt.format(pattern) + self.putg(ss, es, Ann.RAW, [txt]) + self.putpy(ss, es, 'RAW', (self.bitcount, pattern)) + + try: + value = self.interpreter(ss, es, pattern) + except: + value = None + if value is None: + return + self.putpy(ss, es, 'NUMBER', value) + try: + formatted = self.formatter(ss, es, value) + except: + formatted = None + if formatted: + self.putg(ss, es, Ann.NUM, formatted) + if self.interpreter == self.interp_enum: + cls = Ann.get_enum_idx(pattern) + self.putg(ss, es, cls, formatted) + self.putpy(ss, es, 'ENUM', (value, formatted)) + + def interp_unsigned(self, ss, es, pattern): + value = pattern + return value + + def interp_signed(self, ss, es, pattern): + if not 'signmask' in self.interp_state: + self.interp_state.update({ + 'signmask': 1 << (self.bitcount - 1), + 'signfull': 1 << self.bitcount, + }) + is_neg = pattern & self.interp_state['signmask'] + if is_neg: + value = -(self.interp_state['signfull'] - pattern) + else: + value = pattern + return value + + def interp_fixpoint(self, ss, es, pattern): + if not 'fixdiv' in self.interp_state: + self.interp_state.update({ + 'fixsign': self.options['interp'] == 'fixsigned', + 'fixdiv': 2 ** self.options['fracbits'], + }) + if self.interp_state['fixsign']: + value = self.interp_signed(ss, es, pattern) + else: + value = self.interp_unsigned(ss, es, pattern) + value /= self.interp_state['fixdiv'] + return value + + def interp_ieee754(self, ss, es, pattern): + if not 'ieee_has_16bit' in self.interp_state: + self.interp_state.update({ + 'ieee_fmt_int_16': '=H', + 'ieee_fmt_flt_16': '=e', + 'ieee_fmt_int_32': '=L', + 'ieee_fmt_flt_32': '=f', + 'ieee_fmt_int_64': '=Q', + 'ieee_fmt_flt_64': '=d', + }) + try: + fmt = self.interp_state.update['ieee_fmt_flt_16'] + has_16bit_support = 8 * struct.calcsize(fmt) == 16 + except: + has_16bit_support = False + self.interp_state['ieee_has_16bit'] = has_16bit_support + if self.bitcount == 16: + if not self.interp_state['ieee_has_16bit']: + return None + buff = struct.pack(self.interp_state['ieee_fmt_int_16'], pattern) + value, = struct.unpack(self.interp_state['ieee_fmt_flt_16'], buff) + return value + if self.bitcount == 32: + buff = struct.pack(self.interp_state['ieee_fmt_int_32'], pattern) + value, = struct.unpack(self.interp_state['ieee_fmt_flt_32'], buff) + return value + if self.bitcount == 64: + buff = struct.pack(self.interp_state['ieee_fmt_int_64'], pattern) + value, = struct.unpack(self.interp_state['ieee_fmt_flt_64'], buff) + return value + return None + + def interp_enum(self, ss, es, pattern): + if not 'enum_map' in self.interp_state: + self.interp_state.update({ + 'enum_fn': self.options['mapping'], + 'enum_map': {}, + 'enum_have_map': False, + }) + try: + fn = self.interp_state['enum_fn'] + # TODO Optionally try in several locations? Next to the + # decoder implementation? Where else? Expect users to + # enter absolute paths? + with open(fn, 'r') as f: + maptext = f.read() + maptable = {} + if fn.endswith('.js') or fn.endswith('.json'): + # JSON requires string literals on the LHS, so the + # table is written "in reverse order". + js_table = json.loads(maptext) + for k, v in js_table.items(): + maptable[v] = k + elif fn.endswith('.py'): + # Expect a specific identifier at the Python module + # level, and assume that it's a dictionary. + py_table = {} + exec(maptext, py_table) + maptable.update(py_table['enumtext']) + self.interp_state['enum_map'].update(maptable) + self.interp_state['enum_have_map'] = True + except: + # Silently ignore failure. This happens while the user + # is typing the filename, and is non-fatal. If the file + # exists and is not readable or not valid or of unknown + # format, the worst thing that can happen is that the + # decoder implementation keeps using "anonymous" phrases + # until a mapping has become available. No harm is done. + # This decoder cannot tell intermediate from final file + # read attempts, so we cannot raise severity here. + pass + value = self.interp_state['enum_map'].get(pattern, None) + if value is None: + value = pattern + return value + + def format_native(self, ss, es, value): + return ['{}'.format(value),] + + def format_bin(self, ss, es, value): + if not self.format_string: + self.format_string = '{{:0{}b}}'.format(self.bitcount) + return [self.format_string.format(value)] + + def format_oct(self, ss, es, value): + if not self.format_string: + self.format_string = '{{:0{}o}}'.format((self.bitcount + 3 - 1) // 3) + return [self.format_string.format(value)] + + def format_dec(self, ss, es, value): + if not self.format_string: + self.format_string = '{:d}' + return [self.format_string.format(value)] + + def format_hex(self, ss, es, value): + if not self.format_string: + self.format_string = '{{:0{}x}}'.format((self.bitcount + 4 - 1) // 4) + return [self.format_string.format(value)] + + def decode(self): + channels = [ch for ch in range(_max_channels) if self.has_channel(ch)] + have_clk = Pin.CLK in channels + if have_clk: + channels.remove(Pin.CLK) + if not channels: + raise ChannelError("Need at least one bit channel.") + if have_clk: + clkedge = { + 'rising': 'r', + 'falling': 'f', + 'either': 'e', + }.get(self.options['clkedge']) + wait_cond = {Pin.CLK: clkedge} + else: + wait_cond = [{ch: 'e'} for ch in channels] + + bitcount = self.options['count'] + if not bitcount: + bitcount = channels[-1] - Pin.BIT_0 + 1 + self.bitcount = bitcount + + self.interpreter = { + 'unsigned': self.interp_unsigned, + 'signed': self.interp_signed, + 'fixpoint': self.interp_fixpoint, + 'fixsigned': self.interp_fixpoint, + 'ieee754': self.interp_ieee754, + 'enum': self.interp_enum, + }.get(self.options['interp']) + self.interp_state = {} + self.formatter = { + '-': self.format_native, + 'bin': self.format_bin, + 'oct': self.format_oct, + 'dec': self.format_dec, + 'hex': self.format_hex, + }.get(self.options['format']) + self.format_string = None + + pins = self.wait() + ss = self.samplenum + prev_pattern = self.grab_pattern(pins[Pin.BIT_0:]) + while True: + pins = self.wait(wait_cond) + es = self.samplenum + pattern = self.grab_pattern(pins[Pin.BIT_0:]) + if pattern == prev_pattern: + continue + self.handle_pattern(ss, es, prev_pattern) + ss = es + prev_pattern = pattern