add *.pyd to gitignore, imc_channel.hpp: fix int size_t conversions

* makefile: versioning and tags: properly deal with multiple digits version tags
* setup.py: support both sdist and bdist pip wheels

.gitignore

@@ -15,6 +15,7 @@ cython/*.cpp
 
 *.log
 *.so
+*.pyd
 *.o
 *.csv
 *.parquet

cython/setup.py

@@ -1,6 +1,28 @@
 from distutils.core import setup
 from distutils.extension import Extension
 from Cython.Build import cythonize
+import os
+import sys
+
+print("building on platform: "+sys.platform)
+os.system("git tag -l --sort=version:refname > gittags.log")
+with open ("gittags.log","r") as gt:
+    taglst = gt.readlines()
+os.remove("gittags.log")
+if len(taglst) > 0 :
+    version = taglst[-1].replace('\n','').replace('v','')
+else:
+    version = 'unkown'
+print("building version: "+version)
+
+if sys.platform == "linux" or sys.platform == "darwin" :
+	cmpargs = ['-std=c++17','-Wno-unused-variable']
+	lnkargs = ['-std=c++17','-Wno-unused-variable']
+elif sys.platform == "win32" :
+	cmpargs = ['/EHsc','/std:c++17']
+	lnkargs = []
+else :
+	raise RuntimeError("unknown platform")
 
 extensions = Extension(
     name="imc_termite",
@@ -9,13 +31,13 @@ extensions = Extension(
     # library_dirs=["lib"],
     include_dirs=["lib"],
     language='c++',
-    extra_compile_args=['-std=c++17','-Wno-unused-variable'],
-    extra_link_args=['-std=c++17'],
+    extra_compile_args=cmpargs,
+    extra_link_args=lnkargs,
 )
 
 setup(
-    name="imc_termite",
-    version='1.2.8',
+    name='imc_termite',
+    version='1.2.10',
     description='IMCtermite cython extension',
     author='Record Evolution GmbH',
     author_email='mario.fink@record-evolution.de',

lib/imc_channel.hpp

@@ -3,10 +3,12 @@
 #ifndef IMCCHANNEL
 #define IMCCHANNEL
 
-#include <sstream>
 #include "imc_datatype.hpp"
 #include "imc_conversion.hpp"
+#include <sstream>
 #include <math.h>
+#include <chrono>
+#include <ctime>
 
 //---------------------------------------------------------------------------//
 
@@ -106,7 +108,8 @@ namespace imc
   std::string joinvec(std::vector<dt> myvec, unsigned long int limit = 10, int prec = 10, bool fixed = true)
   {
     // include entire list for limit = 0
-    limit = (limit == 0) ? myvec.size() : limit;
+    unsigned long int myvecsize = (unsigned long int)myvec.size();
+    limit = (limit == 0) ? myvecsize : limit;
 
     std::stringstream ss;
     ss<<"[";
@@ -120,14 +123,14 @@ namespace imc
     }
     else
     {
-      unsigned long int heals = (unsigned long int)(limit/2.);
+      unsigned long int heals = limit/2;
       for ( unsigned long int i = 0; i < heals; i++ )
       {
         customize_stream(ss,prec,fixed);
         ss<<myvec[i]<<",";
       }
       ss<<"...";
-      for ( unsigned long int i = myvec.size()-heals; i < myvec.size(); i++ )
+      for ( unsigned long int i = myvecsize-heals; i < myvecsize; i++ )
       {
         customize_stream(ss,prec,fixed);
         ss<<myvec[i]<<",";
@@ -152,6 +155,8 @@ namespace imc
     std::string uuid_;
     std::string name_, comment_;
     std::string origin_, origin_comment_, text_;
+    std::chrono::system_clock::time_point trigger_time_, absolute_trigger_time_;
+    double trigger_time_frac_secs_;
     std::string language_code_, codepage_;
     std::string yname_, yunit_;
     std::string xname_, xunit_;
@@ -162,6 +167,7 @@ namespace imc
     int signbits_, num_bytes_;
     // unsigned long int byte_offset_;
     unsigned long int buffer_offset_, buffer_size_;
+    long int addtime_;
     int datatp_;
     imc::datatype dattyp_;
     std::vector<imc::datatype> ydata_;
@@ -235,6 +241,7 @@ namespace imc
         buffer_offset_ = std::stoul(blocks_->at(chnenv_.Cbuuid_).get_parameter(prms[6]));
         buffer_size_ = std::stoul(blocks_->at(chnenv_.Cbuuid_).get_parameter(prms[7]));
         xoffset_ = std::stod(blocks_->at(chnenv_.Cbuuid_).get_parameter(prms[11]));
+        addtime_ = (long int)std::stod(blocks_->at(chnenv_.Cbuuid_).get_parameter(prms[12]));
       }
 
       // extract associated CR data
@@ -277,8 +284,35 @@ namespace imc
         language_code_ = blocks_->at(chnenv_.NLuuid_).get_parameter(prms[3]);
       }
 
+      // obtain NT data
+      // - https://en.cppreference.com/w/cpp/chrono/c/tm
+      // - https://en.cppreference.com/w/cpp/io/manip/put_time
+      if ( blocks_->count(chnenv_.NTuuid_) == 1 )
+      {
+        prms = blocks_->at(chnenv_.NTuuid_).get_parameters();
+        //std::tm tm{};
+        std::tm tms = std::tm();
+        tms.tm_mday = std::stoi(blocks_->at(chnenv_.NTuuid_).get_parameter(prms[2]));
+        tms.tm_mon = std::stoi(blocks_->at(chnenv_.NTuuid_).get_parameter(prms[3])) - 1;
+        tms.tm_year = std::stoi(blocks_->at(chnenv_.NTuuid_).get_parameter(prms[4])) - 1900;
+        tms.tm_hour = std::stoi(blocks_->at(chnenv_.NTuuid_).get_parameter(prms[5]));
+        tms.tm_min = std::stoi(blocks_->at(chnenv_.NTuuid_).get_parameter(prms[6]));
+        long double secs = std::stold(blocks_->at(chnenv_.NTuuid_).get_parameter(prms[7]));
+        double secs_int;
+        trigger_time_frac_secs_ = modf((double)secs,&secs_int);
+        tms.tm_sec = (int)secs_int;
+
+        // generate std::chrono::system_clock::time_point type
+        std::time_t ts = std::mktime(&tms);
+        trigger_time_ = std::chrono::system_clock::from_time_t(ts);
+      }
+
       // start converting binary buffer to imc::datatype
       if ( !chnenv_.CSuuid_.empty() ) convert_buffer();
+
+      // calculate absolute trigger-time
+      absolute_trigger_time_ = trigger_time_ + std::chrono::seconds(addtime_);
+      //                                       + std::chrono::nanoseconds((long int)(trigger_time_frac_secs_*1.e9));
     }
 
     // convert buffer to actual datatype
@@ -351,7 +385,7 @@ namespace imc
       // fill xdata_
       for ( unsigned long int i = 0; i < num_values; i++ )
       {
-        xdata_.push_back(xoffset_+i*xstepwidth_);
+        xdata_.push_back(xoffset_+(double)i*xstepwidth_);
       }
 
       // employ data transformation
@@ -368,12 +402,18 @@ namespace imc
     // get info string
     std::string get_info(int width = 20)
     {
+      // prepare printable trigger-time
+      std::time_t tt = std::chrono::system_clock::to_time_t(trigger_time_);
+      std::time_t att = std::chrono::system_clock::to_time_t(absolute_trigger_time_);
+
       std::stringstream ss;
       ss<<std::setw(width)<<std::left<<"uuid:"<<uuid_<<"\n"
         <<std::setw(width)<<std::left<<"name:"<<name_<<"\n"
         <<std::setw(width)<<std::left<<"comment:"<<comment_<<"\n"
         <<std::setw(width)<<std::left<<"origin:"<<origin_<<"\n"
         <<std::setw(width)<<std::left<<"description:"<<text_<<"\n"
+        <<std::setw(width)<<std::left<<"trigger-time-nt:"<<std::put_time(std::localtime(&tt),"%FT%T")<<"\n"
+        <<std::setw(width)<<std::left<<"trigger-time:"<<std::put_time(std::localtime(&att),"%FT%T")<<"\n"
         <<std::setw(width)<<std::left<<"language-code:"<<language_code_<<"\n"
         <<std::setw(width)<<std::left<<"codepage:"<<codepage_<<"\n"
         <<std::setw(width)<<std::left<<"yname:"<<yname_<<"\n"
@@ -382,6 +422,7 @@ namespace imc
         <<std::setw(width)<<std::left<<"significant bits:"<<signbits_<<"\n"
         <<std::setw(width)<<std::left<<"buffer-offset:"<<buffer_offset_<<"\n"
         <<std::setw(width)<<std::left<<"buffer-size:"<<buffer_size_<<"\n"
+        <<std::setw(width)<<std::left<<"add-time:"<<addtime_<<"\n"
         <<std::setw(width)<<std::left<<"xname:"<<xname_<<"\n"
         <<std::setw(width)<<std::left<<"xunit:"<<xunit_<<"\n"
         <<std::setw(width)<<std::left<<"xstepwidth:"<<xstepwidth_<<"\n"
@@ -399,17 +440,24 @@ namespace imc
     // provide JSON string of metadata
     std::string get_json(bool include_data = false)
     {
+      // prepare printable trigger-time
+      std::time_t tt = std::chrono::system_clock::to_time_t(trigger_time_);
+      std::time_t att = std::chrono::system_clock::to_time_t(absolute_trigger_time_);
+
       std::stringstream ss;
       ss<<"{"<<"\"uuid\":\""<<uuid_
              <<"\",\"name\":\""<<name_
              <<"\",\"comment\":\""<<comment_
              <<"\",\"origin\":\""<<origin_
              <<"\",\"description\":\""<<text_
+             <<"\",\"trigger-time-nt\":\""<<std::put_time(std::localtime(&tt),"%FT%T")
+             <<"\",\"trigger-time\":\""<<std::put_time(std::localtime(&att),"%FT%T")
              <<"\",\"language-code\":\""<<language_code_
              <<"\",\"codepage\":\""<<codepage_
              <<"\",\"yname\":\""<<yname_
              <<"\",\"yunit\":\""<<yunit_
              <<"\",\"significantbits\":\""<<signbits_
+             <<"\",\"addtime\":\""<<addtime_
              <<"\",\"xname\":\""<<xname_
              <<"\",\"xunit\":\""<<xunit_
              <<"\",\"xstepwidth\":\""<<xstepwidth_

makefile

@@ -17,13 +17,13 @@ HPP = $(wildcard $(LIB)/*.hpp)
 # choose compiler and its options
 CC = g++ -std=c++17
 #OPT = -O3 -Wall -mavx -mno-tbm -mf16c -mno-f16c
-OPT = -O3 -Wall -Werror -Wunused-variable -Wsign-compare
+OPT = -O3 -Wall -Wconversion -Wpedantic -Werror -Wunused-variable -Wsign-compare
 
 # determine git version/commit and release tag
-GTAG := $(shell git tag | tail -n1)
+GTAG := $(shell git tag -l --sort=version:refname | tail -n1)
 GHSH := $(shell git rev-parse HEAD | head -c8)
-RTAG := v$(shell cat pip/setup.py | grep version | grep -oP "([0-9]\.){2}[0-9]")
-CTAG := v$(shell cat cython/setup.py | grep version | grep -oP "([0-9]\.){2}[0-9]")
+RTAG := v$(shell cat pip/setup.py | grep version | grep -oP "([0-9]\.){2}[0-9]{1,2}")
+CTAG := v$(shell cat cython/setup.py | grep version | grep -oP "([0-9]\.){2}[0-9]{1,2}")
 
 # define install location
 INST := /usr/local/bin
@@ -99,7 +99,7 @@ cython-clean :
 # pip
 
 pip-release: check-vtag $(RTAG) cython-build
-	cd ./pip/ && make publish
+	cd ./pip/ && make publish-source
 
 #-----------------------------------------------------------------------------#
 # clean

pip/makefile

@@ -2,15 +2,23 @@
 
 SHELL := /bin/bash
 
-publish: sdist upload
+publish-source: sdist upload
+publish-binary: bdist upload
 
-sdist: ../cython/py_imc_termite.pyx ../cython/imc_termite.pxd ../cython/py_imc_termite.cpp
+prepdist: ../cython/py_imc_termite.pyx ../cython/imc_termite.pxd ../cython/py_imc_termite.cpp
 	cp -v $? ./
 	cp -v $(shell ls ../lib/imc_*.hpp) ./
 	tail -n 212 ../README.md > ./README.md
 	cp -v ../LICENSE ./
+
+sdist: prepdist
 	python3 setup.py sdist
 
+# TODO use manylinux wheel to avoid ERROR "unsupported platform tag 'linux_x86_64'" 
+# see: - https://github.com/pypa/manylinux
+bdist: prepdist
+	python3 setup.py bdist_wheel
+
 # authentication:
 # - username: __token__
 # - password: <token value including pypi-prefix>

pip/setup.py

@@ -18,7 +18,7 @@ with open("README.md", "r", encoding="utf-8") as fh:
 
 setup(
     name="IMCtermite",
-    version="1.2.8",
+    version="1.2.10",
     author="Record Evolution GmbH",
     author_email="mario.fink@record-evolution.de",
     maintainer="Record Evolution GmbH",

python/example.py

@@ -1,137 +0,0 @@
-
-#-----------------------------------------------------------------------------#
-
-import raw_eater
-import raw_meat
-import pyarrow as pa
-import pyarrow.parquet as pq
-from pathlib import Path
-
-fileobj1 = Path("samples/datasetA/").rglob("*.raw")
-rawlist1 = [str(fl) for fl in fileobj1]
-
-fileobj2 = Path("samples/datasetB/").rglob("*.raw")
-rawlist2 = [str(fl) for fl in fileobj2]
-
-rawlist = rawlist1 #[rawlist1[0],rawlist1[4],rawlist2[0],rawlist2[6]]
-for fil in rawlist2 :
-    rawlist.append(fil)
-rawlist.append("./README.md")
-
-print("")
-print(rawlist)
-print()
-
-#-----------------------------------------------------------------------------#
-
-# alternatively create "empty" instance of "raw_eater" and set file names
-eatraw = raw_eater.raweater()
-# eatraw.set_file("../smp/pressure_Vacuum.raw".encode())
-
-# convert every single listed file
-for rf in rawlist :
-
-    print("converting " + str(rf) + "...\n" + 90*("-") + "\n")
-
-    # setup instance of "raw_eater" and trigger conversion
-    # eatraw = raw_eater.raweater(rf.encode())
-    # eatraw = raw_meat.rawmerger(rf.encode())
-
-    # use global instance of "raw_eater" to set file and perform decoding
-    eatraw.set_file(rf.encode())
-    try :
-        eatraw.do_conversion()
-    except RuntimeError as e :
-        print("conversion failed: " + str(e))
-
-    # check validity of file format
-    if eatraw.validity() :
-
-        # show channel name and its unit
-        entity = eatraw.channel_name().decode(encoding='UTF-8',errors='ignore')
-        unit = eatraw.unit().decode(encoding='UTF-8',errors='ignore')
-        print("\nentity: " + str(entity))
-        print("unit:   " + str(unit) + "\n")
-
-        # obtain extracted data
-        xt = eatraw.get_time()
-        yt = eatraw.get_channel()
-
-        # show excerpt of data
-        print("time (length: " + str(len(xt)) + ") \n"
-             + str(xt[:10]) + "\n...\n" + str(xt[-10:]) + "\n")
-        yttrunc = [round(y,4) for y in yt]
-        print(str(entity) + " (length: " + str(len(yttrunc)) + ") \n"
-            + str(yttrunc[:10]) + "\n...\n" + str(yttrunc[-10:]) + "\n")
-
-        outname = rf.split('/')[-1].replace('raw','csv')
-
-        print("write output to : " + outname)
-        eatraw.write_table(("output/"+outname).encode(),ord(' '))
-
-    else :
-
-        print("\nerror: invalid/corrupt .raw file")
-
-    print("\n")
-
-#-----------------------------------------------------------------------------#
-
-print("convert and merge channels " + "\n" + 90*("-") + "\n")
-
-# setup new instance to merge channels
-eatmea = raw_meat.rawmerger(''.encode()) #rawlist[0].encode())
-
-# add every single channel/file in list
-for rf in rawlist :
-    print("\nadding channel " + str(rf))
-    try :
-        succ = eatmea.add_channel(rf.encode())
-        print("\nrecent time series: length: " + str(len(eatmea.get_time_series())) + "\n")
-    except RuntimeError as e :
-        print("failed to add channel: " + str(e))
-
-# show summary of successfully merged channels
-print("\nmerged channels:\n")
-
-# write merged table to .csv output
-eatmea.write_table_all('output/allchannels.csv'.encode(),ord(','))
-
-# get number of successfully merged channels and their names (+units)
-numch = eatmea.get_num_channels()
-chnames = [chnm.decode(encoding='UTF-8',errors='ignore') for chnm in eatmea.get_channel_names()]
-print("number of channels: " + str(numch))
-print("channel names: " + str(chnames))
-
-# obtain final time series
-timse = eatmea.get_time_series()
-print("\nfinal time series:\nlength:" + str(len(timse)) + "\n")
-
-# get time unit and prepend column name
-chnames.insert(0,"Time ["+str(eatmea.time_unit().decode(encoding='UTF-8',errors='ignore'))+"]")
-
-# prepare list of pyarrow arrays
-pyarrs = []
-pyarrs.append(pa.array(timse))
-
-for i in range(0,numch) :
-    print("\n" + str(i) + "  " + str(chnames[i]))
-    dat = eatmea.get_channel_by_index(i)
-    print("length: " + str(len(dat)))
-    pyarrs.append(pa.array(dat))
-print("")
-# print("\npyarrow arrays\n" + str(pyarrs))
-
-# create pyarrow table from data
-pyarwtab = pa.Table.from_arrays(pyarrs,chnames)
-print("\n" + 60*"-" + "\n" + str(pyarwtab) + "\n")
-
-# write pyarrow table to .parquet file with compression
-pq.write_table(pyarwtab,'output/allchannels.parquet',compression='BROTLI')  # compression='BROTLI', 'SNAPPY')
-
-# try to read and decode the .parquet file
-df = pq.read_table('output/allchannels.parquet')
-print(df.to_pandas())
-# df.to_pandas().to_csv('allchannels.csv',index=False,encoding='utf-8',sep=",")
-
-#-----------------------------------------------------------------------------#

python/numpy_to_pyarrow.py

@@ -1,24 +0,0 @@
-
-import pyarrow as pa
-import numpy as np
-import pyarrow.parquet as pq
-
-db = pa.array(np.linspace(10,50,6))
-print(db)
-da = pa.array(np.linspace(0,5,6))
-print(db)
-
-filenam = 'pyarrow_testtab.parquet'
-
-patab = pa.Table.from_arrays([da,db],['entity A [unitA]','entity B [unitB]'])
-print(patab)
-
-# pq.write_table(patab,filenam,compression='BROTLI')
-pq.write_table(patab,filenam,compression='SNAPPY')
-
-df = pq.read_table(filenam)
-print(df)
-print(df.to_pandas())
-
-#import readline
-#readline.write_history_file('generate_pyarrow_table_and_write_parquet.py')

python/usage_ext.py

@@ -0,0 +1,50 @@
+
+import imc_termite
+import json
+import os
+import datetime
+
+# declare and initialize instance of "imctermite" by passing a raw-file
+try :
+    imcraw = imc_termite.imctermite(b"samples/sampleB.raw")
+except RuntimeError as e :
+    raise Exception("failed to load/parse raw-file: " + str(e))
+
+# obtain list of channels as list of dictionaries (without data)
+channels = imcraw.get_channels(False)
+print(json.dumps(channels,indent=4, sort_keys=False))
+
+# obtain all channels (including full data)
+channelsdata = imcraw.get_channels(True)
+
+# everything that follows is an example that specifically makes use only of
+# the first (index = 0) channel ...
+idx = 0
+
+if len(channelsdata) > 0 :
+
+    # get first channel's data
+    chnydata = channelsdata[idx]['ydata']
+    chnxdata = channelsdata[idx]['xdata']
+    print("xdata: " + str(len(chnxdata)))
+    print("ydata: " + str(len(chnydata)))
+
+    # extract trigger-time
+    trigtim = datetime.datetime.fromisoformat(channels[idx]["trigger-time"])
+    print(trigtim)
+
+    # file output of data with absolute timestamp in 1st column
+    filname = os.path.join("./",channelsdata[idx]['name']+".csv")
+    print("writing output into " + filname)
+    with open(filname,'w') as fout :
+        # include column header
+        fout.write( str(channelsdata[idx]['xname']) + '[' + str(channelsdata[idx]['xunit']) + "]"
+                  + ","
+                  + str(channelsdata[idx]['yname']) + '[' + str(channelsdata[idx]['yunit']) + "]"
+                  + "\n" )
+        # add data (introduce time shift according to trigger-time)
+        for row in range(0,len(chnxdata)) :
+            fout.write( str( (trigtim + datetime.timedelta(seconds=chnxdata[row])).isoformat() )
+                      + ","
+                      + str( chnydata[row])
+                      + "\n" )

shell/check_markers.sh

@@ -1,12 +1,50 @@
 ##!/bin/bash/
 
-dir=$1
+fildir=$1
 
-#ls ${dir} | while read fn; do echo $fn; cat ${dir}$fn | grep -a "|[A-Z][A-Z]," -o | wc -l; done;
+if [ -z "${fildir}" ]; then
+  echo "CLI argument missing: provide file or directory" >&2
+  exit 1
+fi
 
-#ls ${dir} | while read fn; do echo $fn; cat ${dir}$fn | grep -a "|[A-Z][A-Z]," -o; done;
+list-markers()
+{
+  arg="$1"
+  if [ -z "${arg}" ]; then
+    echo "list-markers: missing file argument" >&2
+    exit 1
+  else
+    if [ -d "${arg}" ]; then
+      echo "list-markers: file argument is a directory" >&2
+      exit 1
+    fi
+  fi
   
-#ls ${dir} | while read fn; do echo $fn; cat ${dir}$fn | xxd | head -n10 | tail -n3; done;
+  echo -e "\n${arg}\n"
+  mrks=$(cat ${arg} | grep -a "|[A-Z][a-zA-Z]," -o)
+  mrksnum=$(echo "${mrks}" | wc -l)
+  echo -e "(${mrksnum})\n${mrks}"
+}
+
+if [ -f "${fildir}" ]; then
+
+  echo "analyzing single file ${fildir} ..."
+  list-markers "${fildir}"
+
+elif [ -d "${fildir}" ]; then
+
+  echo "analyzing all *.raw files in directory ${fildir} ..."
+  lsfls=$(ls ${fildir}/*.raw | sed 's/\/\//\//g')
+  echo -e "\n${lsfls}"
+
+  for fl in ${lsfls}; do
+    list-markers "${fl}"
+  done
+
+else
+
+  echo "does not exist: ${fildir}" >&2
+  exit 1
+
+fi
 
-ls ${dir} | while read fn; do echo $fn; cat ${dir}$fn | grep -a "|[A-Z][a-zA-Z]," -o | wc -l; done;
-ls ${dir} | while read fn; do echo $fn; cat ${dir}$fn | grep -a "|[A-Z][a-zA-Z]," -o; done;