Text processing/2
The following task concerns data that came from a pollution monitoring station with twenty-four instruments monitoring twenty-four aspects of pollution in the air. Periodically a record is added to the file, each record being a line of 49 fields separated by white-space, which can be one or more space or tab characters.
You are encouraged to solve this task according to the task description, using any language you may know.
The fields (from the left) are:
DATESTAMP [ VALUEn FLAGn ] * 24
i.e. a datestamp followed by twenty-four repetitions of a floating-point instrument value and that instrument's associated integer flag. Flag values are >= 1 if the instrument is working and < 1 if there is some problem with it, in which case that instrument's value should be ignored.
A sample from the full data file readings.txt, which is also used in the Text processing/1 task, follows:
Data is no longer available at that link. Zipped mirror available here
1991-03-30 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 1991-03-31 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 10.000 1 20.000 1 20.000 1 20.000 1 35.000 1 50.000 1 60.000 1 40.000 1 30.000 1 30.000 1 30.000 1 25.000 1 20.000 1 20.000 1 20.000 1 20.000 1 20.000 1 35.000 1 1991-03-31 40.000 1 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 0.000 -2 1991-04-01 0.000 -2 13.000 1 16.000 1 21.000 1 24.000 1 22.000 1 20.000 1 18.000 1 29.000 1 44.000 1 50.000 1 43.000 1 38.000 1 27.000 1 27.000 1 24.000 1 23.000 1 18.000 1 12.000 1 13.000 1 14.000 1 15.000 1 13.000 1 10.000 1 1991-04-02 8.000 1 9.000 1 11.000 1 12.000 1 12.000 1 12.000 1 27.000 1 26.000 1 27.000 1 33.000 1 32.000 1 31.000 1 29.000 1 31.000 1 25.000 1 25.000 1 24.000 1 21.000 1 17.000 1 14.000 1 15.000 1 12.000 1 12.000 1 10.000 1 1991-04-03 10.000 1 9.000 1 10.000 1 10.000 1 9.000 1 10.000 1 15.000 1 24.000 1 28.000 1 24.000 1 18.000 1 14.000 1 12.000 1 13.000 1 14.000 1 15.000 1 14.000 1 15.000 1 13.000 1 13.000 1 13.000 1 12.000 1 10.000 1 10.000 1
- Task
- Confirm the general field format of the file.
- Identify any DATESTAMPs that are duplicated.
- Report the number of records that have good readings for all instruments.
11l
V debug = 0B
V datePat = re:‘\d{4}-\d{2}-\d{2}’
V valuPat = re:‘[-+]?\d+\.\d+’
V statPat = re:‘-?\d+’
V totalLines = 0
Set[String] dupdate
Set[String] badform
Set[String] badlen
V badreading = 0
Set[String] datestamps
L(line) File(‘readings.txt’).read().rtrim("\n").split("\n")
totalLines++
V fields = line.split("\t")
V date = fields[0]
V pairs = (1 .< fields.len).step(2).map(i -> (@fields[i], @fields[i + 1]))
V lineFormatOk = datePat.match(date)
& all(pairs.map(p -> :valuPat.match(p[0])))
& all(pairs.map(p -> :statPat.match(p[1])))
I !lineFormatOk
I debug
print(‘Bad formatting ’line)
badform.add(date)
I pairs.len != 24 | any(pairs.map(p -> Int(p[1]) < 1))
I debug
print(‘Missing values ’line)
I pairs.len != 24
badlen.add(date)
I any(pairs.map(p -> Int(p[1]) < 1))
badreading++
I date C datestamps
I debug
print(‘Duplicate datestamp ’line)
dupdate.add(date)
datestamps.add(date)
print("Duplicate dates:\n "sorted(Array(dupdate)).join("\n "))
print("Bad format:\n "sorted(Array(badform)).join("\n "))
print("Bad number of fields:\n "sorted(Array(badlen)).join("\n "))
print("Records with good readings: #. = #2.2%\n".format(
totalLines - badreading, (totalLines - badreading) / Float(totalLines) * 100))
print(‘Total records: ’totalLines)
- Output:
Duplicate dates: 1990-03-25 1991-03-31 1992-03-29 1993-03-28 1995-03-26 Bad format: Bad number of fields: Records with good readings: 5017 = 91.70% Total records: 5471
Ada
with Ada.Calendar; use Ada.Calendar;
with Ada.Text_IO; use Ada.Text_IO;
with Strings_Edit; use Strings_Edit;
with Strings_Edit.Floats; use Strings_Edit.Floats;
with Strings_Edit.Integers; use Strings_Edit.Integers;
with Generic_Map;
procedure Data_Munging_2 is
package Time_To_Line is new Generic_Map (Time, Natural);
use Time_To_Line;
File : File_Type;
Line_No : Natural := 0;
Count : Natural := 0;
Stamps : Map;
begin
Open (File, In_File, "readings.txt");
loop
declare
Line : constant String := Get_Line (File);
Pointer : Integer := Line'First;
Flag : Integer;
Year, Month, Day : Integer;
Data : Float;
Stamp : Time;
Valid : Boolean := True;
begin
Line_No := Line_No + 1;
Get (Line, Pointer, SpaceAndTab);
Get (Line, Pointer, Year);
Get (Line, Pointer, Month);
Get (Line, Pointer, Day);
Stamp := Time_Of (Year_Number (Year), Month_Number (-Month), Day_Number (-Day));
begin
Add (Stamps, Stamp, Line_No);
exception
when Constraint_Error =>
Put (Image (Year) & Image (Month) & Image (Day) & ": record at " & Image (Line_No));
Put_Line (" duplicates record at " & Image (Get (Stamps, Stamp)));
end;
Get (Line, Pointer, SpaceAndTab);
for Reading in 1..24 loop
Get (Line, Pointer, Data);
Get (Line, Pointer, SpaceAndTab);
Get (Line, Pointer, Flag);
Get (Line, Pointer, SpaceAndTab);
Valid := Valid and then Flag >= 1;
end loop;
if Pointer <= Line'Last then
Put_Line ("Unrecognized tail at " & Image (Line_No) & ':' & Image (Pointer));
elsif Valid then
Count := Count + 1;
end if;
exception
when End_Error | Data_Error | Constraint_Error | Time_Error =>
Put_Line ("Syntax error at " & Image (Line_No) & ':' & Image (Pointer));
end;
end loop;
exception
when End_Error =>
Close (File);
Put_Line ("Valid records " & Image (Count) & " of " & Image (Line_No) & " total");
end Data_Munging_2;
Sample output
1990-3-25: record at 85 duplicates record at 84 1991-3-31: record at 456 duplicates record at 455 1992-3-29: record at 820 duplicates record at 819 1993-3-28: record at 1184 duplicates record at 1183 1995-3-26: record at 1911 duplicates record at 1910 Valid records 5017 of 5471 total
Aime
check_format(list l)
{
integer i;
text s;
if (~l != 49) {
error("bad field count");
}
s = l[0];
if (match("????-??-??", s)) {
error("bad date format");
}
l[0] = s.delete(7).delete(4).atoi;
i = 1;
while (i < 49) {
atof(l[i]);
i += 1;
l[i >> 1] = atoi(l[i]);
i += 1;
}
l.erase(25, -1);
}
main(void)
{
integer goods, i, v;
file f;
list l;
index x;
goods = 0;
f.affix("readings.txt");
while (f.list(l, 0) != -1) {
if (!trap(check_format, l)) {
if ((x[v = lf_x_integer(l)] += 1) != 1) {
v_form("duplicate ~ line\n", v);
}
i = 1;
l.ucall(min_i, 1, i);
goods += iclip(0, i, 1);
}
}
o_(goods, " good lines\n");
0;
}
- Output:
(the "reading.txt" needs to be converted to UNIX end-of-line)
duplicate 19900325 line duplicate 19910331 line duplicate 19920329 line duplicate 19930328 line duplicate 19950326 line 5017 good lines
Amazing Hopper
#include <basico.h>
algoritmo
número de campos correcto = `awk 'NF != 49' basica/readings.txt`
fechas repetidas = `awk '++count[$1] >= 2{print $1, "(",count[$1],")"}' basica/readings.txt`
resultados buenos = `awk '{rec++;ok=1; for(i=0;i<24;i++){if($(2*i+3)<1){ok=0}}; recordok += ok} END {print "Total records",rec,"OK records", recordok, "or", recordok/rec*100,"%"}' basica/readings.txt`
"Check field number by line: ", #( !(number(número de campos correcto)) ? "Ok\n" : "Nok\n";),\
"\nCheck duplicated dates:\n", fechas repetidas,NL, \
"Number of records have good readings for all instruments:\n",resultados buenos,\
"(including "
fijar separador( NL )
contar tokens en 'fechas repetidas'
" duplicated records)\n", luego imprime todo
terminar
- Output:
Check field number by line: Ok Check duplicated dates: 1990-03-25 ( 2 ) 1991-03-31 ( 2 ) 1992-03-29 ( 2 ) 1993-03-28 ( 2 ) 1995-03-26 ( 2 ) Number of records have good readings for all instruments: Total records 5471 OK records 5017 or 91,7017 % (including 5 duplicated records)
AutoHotkey
; Author: AlephX Aug 17 2011
data = %A_scriptdir%\readings.txt
Loop, Read, %data%
{
Lines := A_Index
StringReplace, dummy, A_LoopReadLine, %A_Tab%,, All UseErrorLevel
Loop, parse, A_LoopReadLine, %A_Tab%
{
wrong := 0
if A_index = 1
{
Date := A_LoopField
if (Date == OldDate)
{
WrongDates = %WrongDates%%OldDate% at %Lines%`n
TotwrongDates++
Wrong := 1
break
}
}
else
{
if (A_loopfield/1 < 0)
{
Wrong := 1
break
}
}
}
if (wrong == 1)
totwrong++
else
valid++
if (errorlevel <> 48)
{
if (wrong == 0)
{
totwrong++
valid--
}
unvalidformat++
}
olddate := date
}
msgbox, Duplicate Dates:`n%wrongDates%`nRead Lines: %lines%`nValid Lines: %valid%`nwrong lines: %totwrong%`nDuplicates: %TotWrongDates%`nWrong Formatted: %unvalidformat%`n
Sample Output:
Duplicate Dates: 1990-03-25 at 85 1991-03-31 at 456 1992-03-29 at 820 1993-03-28 at 1184 1995-03-26 at 1911 Read Lines: 5471 Valid Lines: 5129 wrong lines: 342 Duplicates: 5 Wrong Formatted: 0
AWK
A series of AWK one-liners are shown as this is often what is done. If this information were needed repeatedly, (and this is not known), a more permanent shell script might be created that combined multi-line versions of the scripts below.
Gradually tie down the format.
(In each case offending lines will be printed)
If their are any scientific notation fields then their will be an e in the file:
bash$ awk '/[eE]/' readings.txt
bash$
Quick check on the number of fields:
bash$ awk 'NF != 49' readings.txt
bash$
Full check on the file format using a regular expression:
bash$ awk '!(/^[0-9][0-9][0-9][0-9]-[0-9][0-9]-[0-9][0-9]([ \t]+[-]?[0-9]+\.[0-9]+[\t ]+[-]?[0-9]+)+$/ && NF==49)' readings.txt
bash$
Full check on the file format as above but using regular expressions allowing intervals (gnu awk):
bash$ awk --re-interval '!(/^[0-9]{4}-[0-9]{2}-[0-9]{2}([ \t]+[-]?[0-9]+\.[0-9]+[\t ]+[-]?[0-9]+){24}+$/ )' readings.txt
bash$
Identify any DATESTAMPs that are duplicated.
Accomplished by counting how many times the first field occurs and noting any second occurrences.
bash$ awk '++count[$1]==2{print $1}' readings.txt
1990-03-25
1991-03-31
1992-03-29
1993-03-28
1995-03-26
bash$
What number of records have good readings for all instruments.
bash$ awk '{rec++;ok=1; for(i=0;i<24;i++){if($(2*i+3)<1){ok=0}}; recordok += ok} END {print "Total records",rec,"OK records", recordok, "or", recordok/rec*100,"%"}' readings.txt
Total records 5471 OK records 5017 or 91.7017 %
bash$
C
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#include <unistd.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
typedef struct { const char *s; int ln, bad; } rec_t;
int cmp_rec(const void *aa, const void *bb)
{
const rec_t *a = aa, *b = bb;
return a->s == b->s ? 0 : !a->s ? 1 : !b->s ? -1 : strncmp(a->s, b->s, 10);
}
int read_file(const char *fn)
{
int fd = open(fn, O_RDONLY);
if (fd == -1) return 0;
struct stat s;
fstat(fd, &s);
char *txt = malloc(s.st_size);
read(fd, txt, s.st_size);
close(fd);
int i, j, lines = 0, k, di, bad;
for (i = lines = 0; i < s.st_size; i++)
if (txt[i] == '\n') {
txt[i] = '\0';
lines++;
}
rec_t *rec = calloc(sizeof(rec_t), lines);
const char *ptr, *end;
rec[0].s = txt;
rec[0].ln = 1;
for (i = 0; i < lines; i++) {
if (i + 1 < lines) {
rec[i + 1].s = rec[i].s + strlen(rec[i].s) + 1;
rec[i + 1].ln = i + 2;
}
if (sscanf(rec[i].s, "%4d-%2d-%2d", &di, &di, &di) != 3) {
printf("bad line %d: %s\n", i, rec[i].s);
rec[i].s = 0;
continue;
}
ptr = rec[i].s + 10;
for (j = k = 0; j < 25; j++) {
if (!strtod(ptr, (char**)&end) && end == ptr) break;
k++, ptr = end;
if (!(di = strtol(ptr, (char**)&end, 10)) && end == ptr) break;
k++, ptr = end;
if (di < 1) rec[i].bad = 1;
}
if (k != 48) {
printf("bad format at line %d: %s\n", i, rec[i].s);
rec[i].s = 0;
}
}
qsort(rec, lines, sizeof(rec_t), cmp_rec);
for (i = 1, bad = rec[0].bad, j = 0; i < lines && rec[i].s; i++) {
if (rec[i].bad) bad++;
if (strncmp(rec[i].s, rec[j].s, 10)) {
j = i;
} else
printf("dup line %d: %.10s\n", rec[i].ln, rec[i].s);
}
free(rec);
free(txt);
printf("\n%d out %d lines good\n", lines - bad, lines);
return 0;
}
int main()
{
read_file("readings.txt");
return 0;
}
- Output:
dup line 85: 1990-03-25 dup line 456: 1991-03-31 dup line 820: 1992-03-29 dup line 1184: 1993-03-28 dup line 1911: 1995-03-26 5017 out 5471 lines good
C#
using System;
using System.Collections.Generic;
using System.Text.RegularExpressions;
using System.IO;
namespace TextProc2
{
class Program
{
static void Main(string[] args)
{
Regex multiWhite = new Regex(@"\s+");
Regex dateEx = new Regex(@"^\d{4}-\d{2}-\d{2}$");
Regex valEx = new Regex(@"^\d+\.{1}\d{3}$");
Regex flagEx = new Regex(@"^[1-9]{1}$");
int missformcount = 0, totalcount = 0;
Dictionary<int, string> dates = new Dictionary<int, string>();
using (StreamReader sr = new StreamReader("readings.txt"))
{
string line = sr.ReadLine();
while (line != null)
{
line = multiWhite.Replace(line, @" ");
string[] splitLine = line.Split(' ');
if (splitLine.Length != 49)
missformcount++;
if (!dateEx.IsMatch(splitLine[0]))
missformcount++;
else
dates.Add(totalcount + 1, dateEx.Match(splitLine[0]).ToString());
int err = 0;
for (int i = 1; i < splitLine.Length; i++)
{
if (i%2 != 0)
{
if (!valEx.IsMatch(splitLine[i]))
err++;
}
else
{
if (!flagEx.IsMatch(splitLine[i]))
err++;
}
}
if (err != 0) missformcount++;
line = sr.ReadLine();
totalcount++;
}
}
int goodEntries = totalcount - missformcount;
Dictionary<string,List<int>> dateReverse = new Dictionary<string,List<int>>();
foreach (KeyValuePair<int, string> kvp in dates)
{
if (!dateReverse.ContainsKey(kvp.Value))
dateReverse[kvp.Value] = new List<int>();
dateReverse[kvp.Value].Add(kvp.Key);
}
Console.WriteLine(goodEntries + " valid Records out of " + totalcount);
foreach (KeyValuePair<string, List<int>> kvp in dateReverse)
{
if (kvp.Value.Count > 1)
Console.WriteLine("{0} is duplicated at Lines : {1}", kvp.Key, string.Join(",", kvp.Value));
}
}
}
}
5017 valid Records out of 5471 1990-03-25 is duplicated at Lines : 84,85 1991-03-31 is duplicated at Lines : 455,456 1992-03-29 is duplicated at Lines : 819,820 1993-03-28 is duplicated at Lines : 1183,1184 1995-03-26 is duplicated at Lines : 1910,1911
C++
#include <boost/regex.hpp>
#include <fstream>
#include <iostream>
#include <vector>
#include <string>
#include <set>
#include <cstdlib>
#include <algorithm>
using namespace std ;
boost::regex e ( "\\s+" ) ;
int main( int argc , char *argv[ ] ) {
ifstream infile( argv[ 1 ] ) ;
vector<string> duplicates ;
set<string> datestamps ; //for the datestamps
if ( ! infile.is_open( ) ) {
cerr << "Can't open file " << argv[ 1 ] << '\n' ;
return 1 ;
}
int all_ok = 0 ;//all_ok for lines in the given pattern e
int pattern_ok = 0 ; //overall field pattern of record is ok
while ( infile ) {
string eingabe ;
getline( infile , eingabe ) ;
boost::sregex_token_iterator i ( eingabe.begin( ), eingabe.end( ) , e , -1 ), j ;//we tokenize on empty fields
vector<string> fields( i, j ) ;
if ( fields.size( ) == 49 ) //we expect 49 fields in a record
pattern_ok++ ;
else
cout << "Format not ok!\n" ;
if ( datestamps.insert( fields[ 0 ] ).second ) { //not duplicated
int howoften = ( fields.size( ) - 1 ) / 2 ;//number of measurement
//devices and values
for ( int n = 1 ; atoi( fields[ 2 * n ].c_str( ) ) >= 1 ; n++ ) {
if ( n == howoften ) {
all_ok++ ;
break ;
}
}
}
else {
duplicates.push_back( fields[ 0 ] ) ;//first field holds datestamp
}
}
infile.close( ) ;
cout << "The following " << duplicates.size() << " datestamps were duplicated:\n" ;
copy( duplicates.begin( ) , duplicates.end( ) ,
ostream_iterator<string>( cout , "\n" ) ) ;
cout << all_ok << " records were complete and ok!\n" ;
return 0 ;
}
- Output:
Format not ok! The following 6 datestamps were duplicated: 1990-03-25 1991-03-31 1992-03-29 1993-03-28 1995-03-26 2004-12-31
Clojure
(defn parse-line [s]
(let [[date & data-toks] (str/split s #"\s+")
data-fields (map read-string data-toks)
valid-date? (fn [s] (re-find #"\d{4}-\d{2}-\d{2}" s))
valid-line? (and (valid-date? date)
(= 48 (count data-toks))
(every? number? data-fields))
readings (for [[v flag] (partition 2 data-fields)]
{:val v :flag flag})]
(when (not valid-line?)
(println "Malformed Line: " s))
{:date date
:no-missing-readings? (and (= 48 (count data-toks))
(every? pos? (map :flag readings)))}))
(defn analyze-file [path]
(reduce (fn [m line]
(let [{:keys [all-dates dupl-dates n-full-recs invalid-lines]} m
this-date (:date line)
dupl? (contains? all-dates this-date)
full? (:no-missing-readings? line)]
(cond-> m
dupl? (update-in [:dupl-dates] conj this-date)
full? (update-in [:n-full-recs] inc)
true (update-in [:all-dates] conj this-date))))
{:dupl-dates #{} :all-dates #{} :n-full-recs 0}
(->> (slurp path)
clojure.string/split-lines
(map parse-line))))
(defn report-summary [path]
(let [m (analyze-file path)]
(println (format "%d unique dates" (count (:all-dates m))))
(println (format "%d duplicated dates [%s]"
(count (:dupl-dates m))
(clojure.string/join " " (sort (:dupl-dates m)))))
(println (format "%d lines with no missing data" (:n-full-recs m)))))
- Output:
5466 unique dates 5 duplicated dates [1990-03-25 1991-03-31 1992-03-29 1993-03-28 1995-03-26] 5017 lines with no missing data
COBOL
IDENTIFICATION DIVISION.
PROGRAM-ID. text-processing-2.
ENVIRONMENT DIVISION.
INPUT-OUTPUT SECTION.
FILE-CONTROL.
SELECT readings ASSIGN Input-File-Path
ORGANIZATION LINE SEQUENTIAL
FILE STATUS file-status.
DATA DIVISION.
FILE SECTION.
FD readings.
01 reading-record.
03 date-stamp PIC X(10).
03 FILLER PIC X.
03 input-data PIC X(300).
LOCAL-STORAGE SECTION.
78 Input-File-Path VALUE "readings.txt".
78 Num-Data-Points VALUE 48.
01 file-status PIC XX.
01 current-line PIC 9(5).
01 num-date-stamps-read PIC 9(5).
01 read-date-stamps-area.
03 read-date-stamps PIC X(10) OCCURS 1 TO 10000 TIMES
DEPENDING ON num-date-stamps-read
INDEXED BY date-stamp-idx.
01 offset PIC 999.
01 data-len PIC 999.
01 data-flag PIC X.
88 data-not-found VALUE "N".
01 data-field PIC X(25).
01 i PIC 99.
01 num-good-readings PIC 9(5).
01 reading-flag PIC X.
88 bad-reading VALUE "B".
01 delim PIC X.
PROCEDURE DIVISION.
DECLARATIVES.
readings-error SECTION.
USE AFTER ERROR ON readings
DISPLAY "An error occurred while using " Input-File-Path
DISPLAY "Error code " file-status
DISPLAY "The program will terminate."
CLOSE readings
GOBACK
.
END DECLARATIVES.
main-line.
OPEN INPUT readings
*> Process each line of the file.
PERFORM FOREVER
READ readings
AT END
EXIT PERFORM
END-READ
ADD 1 TO current-line
IF reading-record = SPACES
DISPLAY "Line " current-line " is blank."
EXIT PERFORM CYCLE
END-IF
PERFORM check-duplicate-date-stamp
*> Check there are 24 data pairs and see if all the
*> readings are ok.
INITIALIZE offset, reading-flag, data-flag
PERFORM VARYING i FROM 1 BY 1 UNTIL Num-Data-Points < i
PERFORM get-next-field
IF data-not-found
DISPLAY "Line " current-line " has missing "
"fields."
SET bad-reading TO TRUE
EXIT PERFORM
END-IF
*> Every other data field is the instrument flag.
IF FUNCTION MOD(i, 2) = 0 AND NOT bad-reading
IF FUNCTION NUMVAL(data-field) <= 0
SET bad-reading TO TRUE
END-IF
END-IF
ADD data-len TO offset
END-PERFORM
IF NOT bad-reading
ADD 1 TO num-good-readings
END-IF
END-PERFORM
CLOSE readings
*> Display results.
DISPLAY SPACE
DISPLAY current-line " lines read."
DISPLAY num-good-readings " have good readings for all "
"instruments."
GOBACK
.
check-duplicate-date-stamp.
SEARCH read-date-stamps
AT END
ADD 1 TO num-date-stamps-read
MOVE date-stamp
TO read-date-stamps (num-date-stamps-read)
WHEN read-date-stamps (date-stamp-idx) = date-stamp
DISPLAY "Date " date-stamp " is duplicated at "
"line " current-line "."
END-SEARCH
.
get-next-field.
INSPECT input-data (offset:) TALLYING offset
FOR LEADING X"09"
*> The fields are normally delimited by a tab.
MOVE X"09" TO delim
PERFORM find-num-chars-before-delim
*> If the delimiter was not found...
IF FUNCTION SUM(data-len, offset) > 300
*> The data may be delimited by a space if it is at the
*> end of the line.
MOVE SPACE TO delim
PERFORM find-num-chars-before-delim
IF FUNCTION SUM(data-len, offset) > 300
SET data-not-found TO TRUE
EXIT PARAGRAPH
END-IF
END-IF
IF data-len = 0
SET data-not-found TO TRUE
EXIT PARAGRAPH
END-IF
MOVE input-data (offset:data-len) TO data-field
.
find-num-chars-before-delim.
INITIALIZE data-len
INSPECT input-data (offset:) TALLYING data-len
FOR CHARACTERS BEFORE delim
.
- Output:
Date 1990-03-25 is duplicated at line 00084. Date 1991-03-31 is duplicated at line 00455. Date 1992-03-29 is duplicated at line 00819. Date 1993-03-28 is duplicated at line 01183. Date 1995-03-26 is duplicated at line 01910. 05470 lines read. 05016 have good readings for all instruments.
D
void main() {
import std.stdio, std.array, std.string, std.regex, std.conv,
std.algorithm;
auto rxDate = `^\d\d\d\d-\d\d-\d\d$`.regex;
// Works but eats lot of RAM in DMD 2.064.
// auto rxDate = ctRegex!(`^\d\d\d\d-\d\d-\d\d$`);
int[string] repeatedDates;
int goodReadings;
foreach (string line; "readings.txt".File.lines) {
try {
auto parts = line.split;
if (parts.length != 49)
throw new Exception("Wrong column count");
if (parts[0].match(rxDate).empty)
throw new Exception("Date is wrong");
repeatedDates[parts[0]]++;
bool noProblem = true;
for (int i = 1; i < 48; i += 2) {
if (parts[i + 1].to!int < 1)
// don't break loop because it's validation too.
noProblem = false;
if (!parts[i].isNumeric)
throw new Exception("Reading is wrong: "~parts[i]);
}
if (noProblem)
goodReadings++;
} catch(Exception ex) {
writefln(`Problem in line "%s": %s`, line, ex);
}
}
writefln("Duplicated timestamps: %-(%s, %)",
repeatedDates.byKey.filter!(k => repeatedDates[k] > 1));
writeln("Good reading records: ", goodReadings);
}
- Output:
Duplicated timestamps: 1990-03-25, 1991-03-31, 1992-03-29, 1993-03-28, 1995-03-26 Good reading records: 5017
DuckDB
The read_csv_auto() function can read readings.txt (a TSV file), with a minimum of fuss, but since there are no headers in the file, it will be useful to rename them. Here we do so using DuckDB's support for "dynamic SQL", after the data has been ingested.
Reading the data
create or replace table t as
from read_csv_auto('readings.txt', header=false);
Renaming the columns
To rename the 'date' column:
alter table t rename column column00 to date;
To make it easy to distinguish the "value" columns from the "flag" columns, we shall name the latter, but because there are so many of them, it would be tedious to write out the necessary SQL. Instead, we'll generate and execute the SQL dynamically using DuckDB's "dot command" language:
.mode list
.headers off
.once dynamic.tmp
select 'alter table t rename column ''column' || format('{:02d}', 2 * n) || ''' to ''flag' || format('{:02d}'';', n) from range(1,25) t(n);
.read dynamic.tmp
.mode duckdb
.headers on
To preview the results:
D from t limit 1; ┌────────────┬──────────┬────────┬──────────┬────────┬──────────┬────────┬──────────┬────────┬───┬──────────┬────────┬──────────┬────────┬──────────┬────────┬──────────┬────────┬──────────┬────────┐ │ date │ column01 │ flag01 │ column03 │ flag02 │ column05 │ flag03 │ column07 │ flag04 │ … │ column39 │ flag20 │ column41 │ flag21 │ column43 │ flag22 │ column45 │ flag23 │ column47 │ flag24 │ │ date │ double │ int64 │ double │ int64 │ double │ int64 │ double │ int64 │ │ double │ int64 │ double │ int64 │ double │ int64 │ double │ int64 │ double │ int64 │ ├────────────┼──────────┼────────┼──────────┼────────┼──────────┼────────┼──────────┼────────┼───┼──────────┼────────┼──────────┼────────┼──────────┼────────┼──────────┼────────┼──────────┼────────┤ │ 1990-01-01 │ 0.0 │ 0 │ 0.0 │ 0 │ 30.0 │ 1 │ 30.0 │ 1 │ … │ 30.0 │ 1 │ 20.0 │ 1 │ 20.0 │ 1 │ 20.0 │ 1 │ 20.0 │ 1 │ ├────────────┴──────────┴────────┴──────────┴────────┴──────────┴────────┴──────────┴────────┴───┴──────────┴────────┴──────────┴────────┴──────────┴────────┴──────────┴────────┴──────────┴────────┤ │ 1 rows 49 columns (19 shown) │ └────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┘
Duplicate dates
D select * from (select date, count(*) as n from t group by date) where n > 1; ┌────────────┬───────┐ │ date │ n │ │ date │ int64 │ ├────────────┼───────┤ │ 1990-03-25 │ 2 │ │ 1991-03-31 │ 2 │ │ 1995-03-26 │ 2 │ │ 1992-03-29 │ 2 │ │ 1993-03-28 │ 2 │ └────────────┴───────┘
Number of NULLs
# The values in a struct as JSON values
create or replace function struct_values(s) as (
with j as (select s::JSON as j)
SELECT list_transform( json_keys(j), k -> j[k]) from j
);
# Number of nulls in an entire table:
select sum(length(list_filter(struct_values(t), x -> x = 'null'))) as "|null|"
from (select t from t);
- Output:
┌────────┐ │ |null| │ │ int128 │ ├────────┤ │ 0 │ └────────┘
Rows in which at least one flag is less than 1
DuckDB has a syntax for gathering all the columns starting with `flag`: namely: COLUMNS('^flag')
This makes it easy to gather all the flag-value pairs into a JSON object, and thus to identify the rows which have a non-working instrument.
# j is a JSON object ; return a list of the keys for which the value is less than 1
create or replace function report(j) as (
list_filter(json_keys(j), k -> (j -> k) < 1 )
);
select count(*) from
from (select report(struct_pack(*COLUMNS('flag'))::JSON) as flags
from t
where flags = []);
- Output:
┌──────────────┐ │ count_star() │ │ int64 │ ├──────────────┤ │ 5017 │ └──────────────┘
Instruments that never worked
with allflags as (
select array_agg(flags) as flags
from (select report(struct_pack(*COLUMNS('flag'))::JSON) as flags from t)
)
select list_reduce( flags, (acc,x) -> list_intersect(acc,x) ) as "never worked"
from allflags;
- Output:
┌──────────────┐ │ never worked │ │ varchar[] │ ├──────────────┤ │ [] │ └──────────────┘
The number of instruments that were not working at any one time
Here's the frequency count for the number of instruments that were not working at any one time:
D select histogram(length(flags)) as nflags from (select report_flags(struct_pack(*COLUMNS('flag'))::JSON) as flags from t); ┌─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┐ │ nflags │ │ map(bigint, ubigint) │ ├─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┤ │ {0=5017, 1=262, 2=89, 3=19, 4=8, 5=3, 6=2, 7=6, 8=1, 9=5, 10=4, 11=2, 12=1, 13=1, 14=2, 15=1, 16=2, 20=1, 23=12, 24=33} │ └─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┘ D
Eiffel
class
APPLICATION
create
make
feature
make
-- Finds double date stamps and wrong formats.
local
found: INTEGER
double: STRING
do
read_wordlist
fill_hash_table
across
hash as h
loop
if h.key.has_substring ("_double") then
io.put_string ("Double date stamp: %N")
double := h.key
double.remove_tail (7)
io.put_string (double)
io.new_line
end
if h.item.count /= 24 then
io.put_string (h.key.out + " has the wrong format. %N")
found := found + 1
end
end
io.put_string (found.out + " records have not 24 readings.%N")
good_records
end
good_records
-- Number of records that have flag values > 0 for all readings.
local
count, total: INTEGER
end_date: STRING
do
create end_date.make_empty
across
hash as h
loop
count := 0
across
h.item as d
loop
if d.item.flag > 0 then
count := count + 1
end
end
if count = 24 then
total := total + 1
end
end
io.put_string ("%NGood records: " + total.out + ". %N")
end
original_list: STRING = "readings.txt"
read_wordlist
--Preprocesses data in 'data'.
local
l_file: PLAIN_TEXT_FILE
do
create l_file.make_open_read_write (original_list)
l_file.read_stream (l_file.count)
data := l_file.last_string.split ('%N')
l_file.close
end
data: LIST [STRING]
fill_hash_table
--Fills 'hash' using the date as key.
local
by_dates: LIST [STRING]
date: STRING
data_tup: TUPLE [val: REAL; flag: INTEGER]
data_arr: ARRAY [TUPLE [val: REAL; flag: INTEGER]]
i: INTEGER
do
create hash.make (data.count)
across
data as d
loop
if not d.item.is_empty then
by_dates := d.item.split ('%T')
date := by_dates [1]
by_dates.prune (date)
create data_tup
create data_arr.make_empty
from
i := 1
until
i > by_dates.count - 1
loop
data_tup := [by_dates [i].to_real, by_dates [i + 1].to_integer]
data_arr.force (data_tup, data_arr.count + 1)
i := i + 2
end
hash.put (data_arr, date)
if not hash.inserted then
date.append ("_double")
hash.put (data_arr, date)
end
end
end
end
hash: HASH_TABLE [ARRAY [TUPLE [val: REAL; flag: INTEGER]], STRING]
end
- Output:
Double date stamp: 1990-03-25 Double date stamp: 1991-03-31 Double date stamp: 1992-03-29 Double date stamp: 1993-03-28 Double date stamp: 1995-03-26 0 records have not 24 readings. Good records: 5017.
Erlang
Uses function from Text_processing/1. It does some correctness checks for us.
-module( text_processing2 ).
-export( [task/0] ).
task() ->
Name = "priv/readings.txt",
try
File_contents = text_processing:file_contents( Name ),
[correct_field_format(X) || X<- File_contents],
{_Previous, Duplicates} = lists:foldl( fun date_duplicates/2, {"", []}, File_contents ),
io:fwrite( "Duplicates: ~p~n", [Duplicates] ),
Good = [X || X <- File_contents, is_all_good_readings(X)],
io:fwrite( "Good readings: ~p~n", [erlang:length(Good)] )
catch
_:Error ->
io:fwrite( "Error: Failed when checking ~s: ~p~n", [Name, Error] )
end.
correct_field_format( {_Date, Value_flags} ) ->
Corret_number = value_flag_records(),
{correct_field_format, Corret_number} = {correct_field_format, erlang:length(Value_flags)}.
date_duplicates( {Date, _Value_flags}, {Date, Acc} ) -> {Date, [Date | Acc]};
date_duplicates( {Date, _Value_flags}, {_Other, Acc} ) -> {Date, Acc}.
is_all_good_readings( {_Date, Value_flags} ) -> value_flag_records() =:= erlang:length( [ok || {_Value, ok} <- Value_flags] ).
value_flag_records() -> 24.
- Output:
12> text_processing2:task(). Duplicates: ["1995-03-26","1993-03-28","1992-03-29","1991-03-31","1990-03-25"] Good readings: 5017
F#
let file = @"readings.txt"
let dates = HashSet(HashIdentity.Structural)
let mutable ok = 0
do
for line in System.IO.File.ReadAllLines file do
match String.split [' '; '\t'] line with
| [] -> ()
| date::xys ->
if dates.Contains date then
printf "Date %s is duplicated\n" date
else
dates.Add date
let f (b, t) h = not b, if b then int h::t else t
let _, states = Seq.fold f (false, []) xys
if Seq.forall (fun s -> s >= 1) states then
ok <- ok + 1
printf "%d records were ok\n" ok
Prints:
Date 1990-03-25 is duplicated
Date 1991-03-31 is duplicated
Date 1992-03-29 is duplicated
Date 1993-03-28 is duplicated
Date 1995-03-26 is duplicated
5017 records were ok
Factor
USING: io io.encodings.ascii io.files kernel math math.parser
prettyprint sequences sequences.extras sets splitting ;
: check-format ( seq -- )
[ " \t" split length 49 = ] all?
"Format okay." "Format not okay." ? print ;
"readings.txt" ascii file-lines [ check-format ] keep
[ "Duplicates:" print [ "\t" split1 drop ] map duplicates . ]
[ [ " \t" split rest <odds> [ string>number 0 <= ] none? ] count ]
bi pprint " records were good." print
- Output:
Format okay. Duplicates: { "1990-03-25" "1991-03-31" "1992-03-29" "1993-03-28" "1995-03-26" } 5017 records were good.
Fortran
The trouble with the dates rather suggests that they should be checked for correctness in themselves, and that the sequence check should be that each new record advances the date by one day. Daynumber calculations were long ago presented by H. F. Fliegel and T.C. van Flandern, in Communications of the ACM, Vol. 11, No. 10 (October, 1968).
Rather than copy today's data to a PDATA holder so that on the next read the new data may be compared to the old, a two-row array is used, with IT flip-flopping 1,2,1,2,1,2,... Comparison of the data as numerical values rather than text strings means that different texts that evoke the same value will not be regarded as different. If the data format were invalid, there would be horrible messages. There aren't, so ... the values should be read and plotted...
Crunches a set of hourly data. Starts with a date, then 24 pairs of value,indicator for that day, on one line.
INTEGER Y,M,D !Year, month, and day.
INTEGER GOOD(24,2) !The indicators.
REAL*8 V(24,2) !The grist.
CHARACTER*10 DATE(2) !Along with the starting date.
INTEGER IT,TI !A flipper and its antiflipper.
INTEGER NV !Number of entirely good records.
INTEGER I,NREC,HIC !Some counters.
LOGICAL INGOOD !State flipper for the runs of data.
INTEGER IN,MSG !I/O mnemonics.
CHARACTER*666 ACARD !Scratchpad, of sufficient length for all expectation.
IN = 10 !Unit number for the input file.
MSG = 6 !Output.
OPEN (IN,FILE="Readings1.txt", FORM="FORMATTED", !This should be a function.
1 STATUS ="OLD",ACTION="READ") !Returning success, or failure.
NV = 0 !No pure records seen.
NREC = 0 !No records read.
HIC = 0 !Provoking no complaints.
DATE = "snargle" !No date should look like this!
IT = 2 !Syncopation for the 1-2 flip flop.
Chew into the file.
10 READ (IN,11,END=100,ERR=666) L,ACARD(1:MIN(L,LEN(ACARD))) !With some protection.
NREC = NREC + 1 !So, a record has been read.
11 FORMAT (Q,A) !Obviously, Q ascertains the length of the record being read.
READ (ACARD,12,END=600,ERR=601) Y,M,D !The date part is trouble, as always.
12 FORMAT (I4,2(1X,I2)) !Because there are no delimiters between the parts.
TI = IT !Thus finger the previous value.
IT = 3 - IT !Flip between 1 and 2.
DATE(IT) = ACARD(1:10) !Save the date field.
READ (ACARD(11:L),*,END=600,ERR=601) (V(I,IT),GOOD(I,IT),I = 1,24) !But after the date, delimiters abound.
Comparisons. Should really convert the date to a daynumber, check it by reversion, and then check for + 1 day only.
20 IF (DATE(IT).EQ.DATE(TI)) THEN !Same date?
IF (ALL(V(:,IT) .EQ.V(:,TI)) .AND. !Yes. What about the data?
1 ALL(GOOD(:,IT).EQ.GOOD(:,TI))) THEN !This disregards details of the spacing of the data.
WRITE (MSG,21) NREC,DATE(IT),"same." !Also trailing zeroes, spurious + signs, blah blah.
21 FORMAT ("Record",I8," Duplicate date field (",A,"), data ",A) !Say it.
ELSE !But if they're not all equal,
WRITE (MSG,21) NREC,DATE(IT),"different!" !They're different!
END IF !So much for comparing the data.
END IF !So much for just comparing the date's text.
IF (ALL(GOOD(:,IT).GT.0)) NV = NV + 1 !A fully healthy record, either way?
GO TO 10 !More! More! I want more!!
Complaints. Should really distinguish between trouble in the date part and in the data part.
600 WRITE (MSG,*) '"END" declared - insufficient data?' !Not enough numbers, presumably.
GO TO 602 !Reveal the record.
601 WRITE (MSG,*) '"ERR" declared - improper number format?' !Ah, but which number?
602 WRITE (MSG,603) NREC,L,ACARD(1:L) !Anyway, reveal the uninterpreted record.
603 FORMAT("Record",I8,", length ",I0," reads ",A) !Just so.
HIC = HIC + 1 !This may grow into a habit.
IF (HIC.LE.12) GO TO 10 !But if not yet, try the next record.
STOP "Enough distaste." !Or, give up.
666 WRITE (MSG,101) NREC,"format error!" !For A-style data? Should never happen!
GO TO 900 !But if it does, give up!
Closedown.
100 WRITE (MSG,101) NREC,"then end-of-file" !Discovered on the next attempt.
101 FORMAT ("Record",I8,": ",A) !A record number plus a remark.
WRITE (MSG,102) NV !The overall results.
102 FORMAT (" with",I8," having all values good.") !This should do.
900 CLOSE(IN) !Done.
END !Spaghetti rules.
Output:
Record 85 Duplicate date field (1990-03-25), data different! Record 456 Duplicate date field (1991-03-31), data different! Record 820 Duplicate date field (1992-03-29), data different! Record 1184 Duplicate date field (1993-03-28), data different! Record 1911 Duplicate date field (1995-03-26), data different! Record 5471: then end-of-file with 5017 having all values good.
Go
package main
import (
"bufio"
"fmt"
"log"
"os"
"strconv"
"strings"
"time"
)
const (
filename = "readings.txt"
readings = 24 // per line
fields = readings*2 + 1 // per line
dateFormat = "2006-01-02"
)
func main() {
file, err := os.Open(filename)
if err != nil {
log.Fatal(err)
}
defer file.Close()
var allGood, uniqueGood int
// map records not only dates seen, but also if an all-good record was
// seen for the key date.
m := make(map[time.Time]bool)
s := bufio.NewScanner(file)
for s.Scan() {
f := strings.Fields(s.Text())
if len(f) != fields {
log.Fatal("unexpected format,", len(f), "fields.")
}
ts, err := time.Parse(dateFormat, f[0])
if err != nil {
log.Fatal(err)
}
good := true
for i := 1; i < fields; i += 2 {
flag, err := strconv.Atoi(f[i+1])
if err != nil {
log.Fatal(err)
}
if flag > 0 { // value is good
_, err := strconv.ParseFloat(f[i], 64)
if err != nil {
log.Fatal(err)
}
} else { // value is bad
good = false
}
}
if good {
allGood++
}
previouslyGood, seen := m[ts]
if seen {
fmt.Println("Duplicate datestamp:", f[0])
}
m[ts] = previouslyGood || good
if !previouslyGood && good {
uniqueGood++
}
}
if err := s.Err(); err != nil {
log.Fatal(err)
}
fmt.Println("\nData format valid.")
fmt.Println(allGood, "records with good readings for all instruments.")
fmt.Println(uniqueGood,
"unique dates with good readings for all instruments.")
}
- Output:
Duplicate datestamp: 1990-03-25 Duplicate datestamp: 1991-03-31 Duplicate datestamp: 1992-03-29 Duplicate datestamp: 1993-03-28 Duplicate datestamp: 1995-03-26 Data format valid. 5017 records with good readings for all instruments. 5013 unique dates with good readings for all instruments.
Haskell
import Data.List (nub, (\\))
data Record = Record {date :: String, recs :: [(Double, Int)]}
duplicatedDates rs = rs \\ nub rs
goodRecords = filter ((== 24) . length . filter ((>= 1) . snd) . recs)
parseLine l = let ws = words l in Record (head ws) (mapRecords (tail ws))
mapRecords [] = []
mapRecords [_] = error "invalid data"
mapRecords (value:flag:tail) = (read value, read flag) : mapRecords tail
main = do
inputs <- (map parseLine . lines) `fmap` readFile "readings.txt"
putStr (unlines ("duplicated dates:": duplicatedDates (map date inputs)))
putStrLn ("number of good records: " ++ show (length $ goodRecords inputs))
this script outputs:
duplicated dates: 1990-03-25 1991-03-31 1992-03-29 1993-03-28 1995-03-26 number of good records: 5017
Icon and Unicon
The following works in both languages. It assumes there is nothing wrong with duplicated timestamps that are on well-formed records.
procedure main(A)
dups := set()
goodRecords := 0
lastDate := badFile := &null
f := A[1] | "readings.txt"
fin := open(f) | stop("Cannot open file '",f,"'")
while (fields := 0, badReading := &null, line := read(fin)) do {
line ? {
ldate := tab(many(&digits ++ '-')) | (badFile := "yes", next)
if \lastDate == ldate then insert(dups, ldate)
lastDate := ldate
while tab(many(' \t')) do {
(value := real(tab(many(&digits++'-.'))),
tab(many(' \t')),
flag := integer(tab(many(&digits++'-'))),
fields +:= 1) | (badFile := "yes")
if flag < 1 then badReading := "yes"
}
}
if fields = 24 then goodRecords +:= (/badReading, 1)
else badFile := "yes"
}
if (\badFile) then write(f," has field format issues.")
write("There are ",goodRecords," records with all good readings.")
if *dups > 0 then {
write("The following dates have multiple records:")
every writes(" ",!sort(dups))
write()
}
end
Sample run:
->tp2 There are 5017 records with all good readings. The following dates have multiple records: 1990-03-25 1991-03-31 1992-03-29 1993-03-28 1995-03-26 ->
J
require 'tables/dsv dates'
dat=: TAB readdsv jpath '~temp/readings.txt'
Dates=: getdate"1 >{."1 dat
Vals=: _99 ". >(1 + +: i.24){"1 dat
Flags=: _99 ". >(2 + +: i.24){"1 dat
# Dates NB. Total # lines
5471
+/ *./"1 ] 0 = Dates NB. # lines with invalid date formats
0
+/ _99 e."1 Vals,.Flags NB. # lines with invalid value or flag formats
0
+/ *./"1 [0 < Flags NB. # lines with only valid flags
5017
~. (#~ (i.~ ~: i:~)) Dates NB. Duplicate dates
1990 3 25
1991 3 31
1992 3 29
1993 3 28
1995 3 26
Java
import java.util.*;
import java.util.regex.*;
import java.io.*;
public class DataMunging2 {
public static final Pattern e = Pattern.compile("\\s+");
public static void main(String[] args) {
try {
BufferedReader infile = new BufferedReader(new FileReader(args[0]));
List<String> duplicates = new ArrayList<String>();
Set<String> datestamps = new HashSet<String>(); //for the datestamps
String eingabe;
int all_ok = 0;//all_ok for lines in the given pattern e
while ((eingabe = infile.readLine()) != null) {
String[] fields = e.split(eingabe); //we tokenize on empty fields
if (fields.length != 49) //we expect 49 fields in a record
System.out.println("Format not ok!");
if (datestamps.add(fields[0])) { //not duplicated
int howoften = (fields.length - 1) / 2 ; //number of measurement
//devices and values
for (int n = 1; Integer.parseInt(fields[2*n]) >= 1; n++) {
if (n == howoften) {
all_ok++ ;
break ;
}
}
} else {
duplicates.add(fields[0]); //first field holds datestamp
}
}
infile.close();
System.out.println("The following " + duplicates.size() + " datestamps were duplicated:");
for (String x : duplicates)
System.out.println(x);
System.out.println(all_ok + " records were complete and ok!");
} catch (IOException e) {
System.err.println("Can't open file " + args[0]);
System.exit(1);
}
}
}
The program produces the following output:
The following 5 datestamps were duplicated: 1990-03-25 1991-03-31 1992-03-29 1993-03-28 1995-03-26 5013 records were complete and ok!
JavaScript
// wrap up the counter variables in a closure.
function analyze_func(filename) {
var dates_seen = {};
var format_bad = 0;
var records_all = 0;
var records_good = 0;
return function() {
var fh = new ActiveXObject("Scripting.FileSystemObject").openTextFile(filename, 1); // 1 = for reading
while ( ! fh.atEndOfStream) {
records_all ++;
var allOK = true;
var line = fh.ReadLine();
var fields = line.split('\t');
if (fields.length != 49) {
format_bad ++;
continue;
}
var date = fields.shift();
if (has_property(dates_seen, date))
WScript.echo("duplicate date: " + date);
else
dates_seen[date] = 1;
while (fields.length > 0) {
var value = parseFloat(fields.shift());
var flag = parseInt(fields.shift(), 10);
if (isNaN(value) || isNaN(flag)) {
format_bad ++;
}
else if (flag <= 0) {
allOK = false;
}
}
if (allOK)
records_good ++;
}
fh.close();
WScript.echo("total records: " + records_all);
WScript.echo("Wrong format: " + format_bad);
WScript.echo("records with no bad readings: " + records_good);
}
}
function has_property(obj, propname) {
return typeof(obj[propname]) == "undefined" ? false : true;
}
var analyze = analyze_func('readings.txt');
analyze();
jq
For this problem, it is convenient to use jq in a pipeline: the first invocation of jq will convert the text file into a stream of JSON arrays (one array per line):
$ jq -R '[splits("[ \t]+")]' Text_processing_2.txt
The second part of the pipeline performs the task requirements. The following program is used in the second invocation of jq.
Generic Utilities
# Given any array, produce an array of [item, count] pairs for each run.
def runs:
reduce .[] as $item
( [];
if . == [] then [ [ $item, 1] ]
else .[length-1] as $last
| if $last[0] == $item then (.[0:length-1] + [ [$item, $last[1] + 1] ] )
else . + [[$item, 1]]
end
end ) ;
def is_float: test("^[-+]?[0-9]*[.][0-9]*([eE][-+]?[0-9]+)?$");
def is_integral: test("^[-+]?[0-9]+$");
def is_date: test("[12][0-9]{3}-[0-9][0-9]-[0-9][0-9]");
Validation:
# Report line and column numbers using conventional numbering (IO=1).
def validate_line(nr):
def validate_date:
if is_date then empty else "field 1 in line \(nr) has an invalid date: \(.)" end;
def validate_length(n):
if length == n then empty else "line \(nr) has \(length) fields" end;
def validate_pair(i):
( .[2*i + 1] as $n
| if ($n | is_float) then empty else "field \(2*i + 2) in line \(nr) is not a float: \($n)" end),
( .[2*i + 2] as $n
| if ($n | is_integral) then empty else "field \(2*i + 3) in line \(nr) is not an integer: \($n)" end);
(.[0] | validate_date),
(validate_length(49)),
(range(0; (length-1) / 2) as $i | validate_pair($i)) ;
def validate_lines:
. as $in
| range(0; length) as $i | ($in[$i] | validate_line($i + 1));
Check for duplicate timestamps
def duplicate_timestamps:
[.[][0]] | sort | runs | map( select(.[1]>1) );
Number of valid readings for all instruments:
# The following ignores any issues with respect to duplicate dates,
# but does check the validity of the record, including the date format:
def number_of_valid_readings:
def check:
. as $in
| (.[0] | is_date)
and length == 49
and all(range(0; 24) | $in[2*. + 1] | is_float)
and all(range(0; 24) | $in[2*. + 2] | (is_integral and tonumber >= 1) );
map(select(check)) | length ;
Generate Report
validate_lines,
"\nChecking for duplicate timestamps:",
duplicate_timestamps,
"\nThere are \(number_of_valid_readings) valid rows altogether."
- Output:
Part 1: Simple demonstration
To illustrate that the program does report invalid lines, we first use the six lines at the top but mangle the last line.
$ jq -R '[splits("[ \t]+")]' Text_processing_2.txt | jq -s -r -f Text_processing_2.jq
field 1 in line 6 has an invalid date: 991-04-03
line 6 has 47 fields
field 2 in line 6 is not a float: 10000
field 3 in line 6 is not an integer: 1.0
field 47 in line 6 is not an integer: x
Checking for duplicate timestamps:
[
[
"1991-03-31",
2
]
]
There are 5 valid rows altogether.
Part 2: readings.txt
$ jq -R '[splits("[ \t]+")]' readings.txt | jq -s -r -f Text_processing_2.jq
Checking for duplicate timestamps:
[
[
"1990-03-25",
2
],
[
"1991-03-31",
2
],
[
"1992-03-29",
2
],
[
"1993-03-28",
2
],
[
"1995-03-26",
2
]
]
There are 5017 valid rows altogether.
Julia
Refer to the code at https://rosettacode.org/wiki/Text_processing/1#Julia. Add at the end of that code the following:
dupdate = df[nonunique(df[:,[:Date]]),:][:Date]
println("The following rows have duplicate DATESTAMP:")
println(df[df[:Date] .== dupdate,:])
println("All values good in these rows:")
println(df[df[:ValidValues] .== 24,:])
- Output:
The following rows have duplicate DATESTAMP: 2×29 DataFrames.DataFrame │ Row │ Date │ Mean │ ValidValues │ MaximumGap │ GapPosition │ 0:00 │ 1:00 │ 2:00 │ 3:00 │ 4:00 │ ├─────┼─────────────────────┼─────────┼─────────────┼────────────┼─────────────┼──────┼──────┼──────┼──────┼──────┤ │ 1 │ 1991-03-31T00:00:00 │ 23.5417 │ 24 │ 0 │ 0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ │ 2 │ 1991-03-31T00:00:00 │ 40.0 │ 1 │ 23 │ 2 │ 40.0 │ NaN │ NaN │ NaN │ NaN │ │ Row │ 5:00 │ 6:00 │ 7:00 │ 8:00 │ 9:00 │ 10:00 │ 11:00 │ 12:00 │ 13:00 │ 14:00 │ 15:00 │ 16:00 │ 17:00 │ 18:00 │ ├─────┼──────┼──────┼──────┼──────┼──────┼───────┼───────┼───────┼───────┼───────┼───────┼───────┼───────┼───────┤ │ 1 │ 10.0 │ 10.0 │ 20.0 │ 20.0 │ 20.0 │ 35.0 │ 50.0 │ 60.0 │ 40.0 │ 30.0 │ 30.0 │ 30.0 │ 25.0 │ 20.0 │ │ 2 │ NaN │ NaN │ NaN │ NaN │ NaN │ NaN │ NaN │ NaN │ NaN │ NaN │ NaN │ NaN │ NaN │ NaN │ │ Row │ 19:00 │ 20:00 │ 21:00 │ 22:00 │ 23:00 │ ├─────┼───────┼───────┼───────┼───────┼───────┤ │ 1 │ 20.0 │ 20.0 │ 20.0 │ 20.0 │ 35.0 │ │ 2 │ NaN │ NaN │ NaN │ NaN │ NaN │ All values good in these rows: 4×29 DataFrames.DataFrame │ Row │ Date │ Mean │ ValidValues │ MaximumGap │ GapPosition │ 0:00 │ 1:00 │ 2:00 │ 3:00 │ 4:00 │ ├─────┼─────────────────────┼─────────┼─────────────┼────────────┼─────────────┼──────┼──────┼──────┼──────┼──────┤ │ 1 │ 1991-03-30T00:00:00 │ 10.0 │ 24 │ 0 │ 0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ │ 2 │ 1991-03-31T00:00:00 │ 23.5417 │ 24 │ 0 │ 0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ │ 3 │ 1991-04-02T00:00:00 │ 19.7917 │ 24 │ 0 │ 0 │ 8.0 │ 9.0 │ 11.0 │ 12.0 │ 12.0 │ │ 4 │ 1991-04-03T00:00:00 │ 13.9583 │ 24 │ 0 │ 0 │ 10.0 │ 9.0 │ 10.0 │ 10.0 │ 9.0 │ │ Row │ 5:00 │ 6:00 │ 7:00 │ 8:00 │ 9:00 │ 10:00 │ 11:00 │ 12:00 │ 13:00 │ 14:00 │ 15:00 │ 16:00 │ 17:00 │ 18:00 │ ├─────┼──────┼──────┼──────┼──────┼──────┼───────┼───────┼───────┼───────┼───────┼───────┼───────┼───────┼───────┤ │ 1 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ │ 2 │ 10.0 │ 10.0 │ 20.0 │ 20.0 │ 20.0 │ 35.0 │ 50.0 │ 60.0 │ 40.0 │ 30.0 │ 30.0 │ 30.0 │ 25.0 │ 20.0 │ │ 3 │ 12.0 │ 27.0 │ 26.0 │ 27.0 │ 33.0 │ 32.0 │ 31.0 │ 29.0 │ 31.0 │ 25.0 │ 25.0 │ 24.0 │ 21.0 │ 17.0 │ │ 4 │ 10.0 │ 15.0 │ 24.0 │ 28.0 │ 24.0 │ 18.0 │ 14.0 │ 12.0 │ 13.0 │ 14.0 │ 15.0 │ 14.0 │ 15.0 │ 13.0 │ │ Row │ 19:00 │ 20:00 │ 21:00 │ 22:00 │ 23:00 │ ├─────┼───────┼───────┼───────┼───────┼───────┤ │ 1 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ 10.0 │ │ 2 │ 20.0 │ 20.0 │ 20.0 │ 20.0 │ 35.0 │ │ 3 │ 14.0 │ 15.0 │ 12.0 │ 12.0 │ 10.0 │ │ 4 │ 13.0 │ 13.0 │ 12.0 │ 10.0 │ 10.0 │
Kotlin
// version 1.2.31
import java.io.File
fun main(args: Array<String>) {
val rx = Regex("""\s+""")
val file = File("readings.txt")
var count = 0
var invalid = 0
var allGood = 0
var map = mutableMapOf<String, Int>()
file.forEachLine { line ->
count++
val fields = line.split(rx)
val date = fields[0]
if (fields.size == 49) {
if (map.containsKey(date))
map[date] = map[date]!! + 1
else
map.put(date, 1)
var good = 0
for (i in 2 until fields.size step 2) {
if (fields[i].toInt() >= 1) {
good++
}
}
if (good == 24) allGood++
}
else invalid++
}
println("File = ${file.name}")
println("\nDuplicated dates:")
for ((k,v) in map) {
if (v > 1) println(" $k ($v times)")
}
println("\nTotal number of records : $count")
var percent = invalid.toDouble() / count * 100.0
println("Number of invalid records : $invalid (${"%5.2f".format(percent)}%)")
percent = allGood.toDouble() / count * 100.0
println("Number which are all good : $allGood (${"%5.2f".format(percent)}%)")
}
- Output:
File = readings.txt Duplicated dates: 1990-03-25 (2 times) 1991-03-31 (2 times) 1992-03-29 (2 times) 1993-03-28 (2 times) 1995-03-26 (2 times) Total number of records : 5471 Number of invalid records : 0 ( 0.00%) Number which are all good : 5017 (91.70%)
Lua
filename = "readings.txt"
io.input( filename )
dates = {}
duplicated, bad_format = {}, {}
num_good_records, lines_total = 0, 0
while true do
line = io.read( "*line" )
if line == nil then break end
lines_total = lines_total + 1
date = string.match( line, "%d+%-%d+%-%d+" )
if dates[date] ~= nil then
duplicated[#duplicated+1] = date
end
dates[date] = 1
count_pairs, bad_values = 0, false
for v, w in string.gmatch( line, "%s(%d+[%.%d+]*)%s(%-?%d)" ) do
count_pairs = count_pairs + 1
if tonumber(w) <= 0 then
bad_values = true
end
end
if count_pairs ~= 24 then
bad_format[#bad_format+1] = date
end
if not bad_values then
num_good_records = num_good_records + 1
end
end
print( "Lines read:", lines_total )
print( "Valid records: ", num_good_records )
print( "Duplicate dates:" )
for i = 1, #duplicated do
print( " ", duplicated[i] )
end
print( "Bad format:" )
for i = 1, #bad_format do
print( " ", bad_format[i] )
end
Output:
Lines read: 5471 Valid records: 5017 Duplicate dates: 1990-03-25 1991-03-31 1992-03-29 1993-03-28 1995-03-26 Bad format:
M2000 Interpreter
File is in user dir. Use Win Dir$ to open the explorer window and copy there the readings.txt
Module TestThis {
Document a$, exp$
\\ automatic find the enconding and the line break
Load.doc a$, "readings.txt"
m=0
n=doc.par(a$)
k=list
nl$={
}
l=0
exp$=format$("Records: {0}", n)+nl$
For i=1 to n
b$=paragraph$(a$, i)
If exist(k,Left$(b$, 10)) then
m++ : where=eval(k)
exp$=format$("Duplicate for {0} at {1}",where, i)+nl$
Else
Append k, Left$(b$, 10):=i
End if
Stack New {
Stack Mid$(Replace$(chr$(9)," ", b$), 11)
while not empty {
Read a, b
if b<=0 then l++ : exit
}
}
Next
exp$= format$("Duplicates {0}",m)+nl$
exp$= format$("Valid Records {0}",n-l)+nl$
clipboard exp$
report exp$
}
TestThis
- Output:
Records: 5471 Duplicate for 84 at 85 Duplicate for 455 at 456 Duplicate for 819 at 820 Duplicate for 1183 at 1184 Duplicate for 1910 at 1911 Duplicates 5 Valid Records 5017
Mathematica /Wolfram Language
data = Import["Readings.txt","TSV"]; Print["duplicated dates: "];
Select[Tally@data[[;;,1]], #[[2]]>1&][[;;,1]]//Column
Print["number of good records: ", Count[(Times@@#[[3;;All;;2]])& /@ data, 1],
" (out of a total of ", Length[data], ")"]
- Output:
duplicated dates: 1990-03-25 1991-03-31 1992-03-29 1993-03-28 1995-03-26 number of good records: 5017 (out of a total of 5471)
MATLAB / Octave
function [val,count] = readdat(configfile)
% READDAT reads readings.txt file
%
% The value of boolean parameters can be tested with
% exist(parameter,'var')
if nargin<1,
filename = 'readings.txt';
end;
fid = fopen(filename);
if fid<0, error('cannot open file %s\n',a); end;
[val,count] = fscanf(fid,'%04d-%02d-%02d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d %f %d \n');
fclose(fid);
count = count/51;
if (count<1) || count~=floor(count),
error('file has incorrect format\n')
end;
val = reshape(val,51,count)'; % make matrix with 51 rows and count columns, then transpose it.
d = datenum(val(:,1:3)); % compute timestamps
printf('The following records are followed by a duplicate:');
dix = find(diff(d)==0) % check for to consequtive timestamps with zero difference
printf('number of valid records: %i\n ', sum( all( val(:,5:2:end) >= 1, 2) ) );
>> [val,count]=readdat; The following records are followed by a duplicate:dix = 84 455 819 1183 1910 number of valid records: 5017
Nim
import strutils, tables
const NumFields = 49
const DateField = 0
const FlagGoodValue = 1
var badRecords: int # Number of records that have invalid formatted values.
var totalRecords: int # Total number of records in the file.
var badInstruments: int # Total number of records that have at least one instrument showing error.
var seenDates: Table[string, bool] # Table to keep track of what dates we have seen.
proc checkFloats(floats: seq[string]): bool =
## Ensure we can parse all records as floats (except the date stamp).
for index in 1..<NumFields:
try:
# We're assuming all instrument flags are floats not integers.
discard parseFloat(floats[index])
except ValueError:
return false
true
proc areAllFlagsOk(instruments: seq[string]): bool =
## Ensure that all sensor flags are ok.
# Flags start at index 2, and occur every 2 fields.
for index in countup(2, NumFields, 2):
# We're assuming all instrument flags are floats not integers
var flag = parseFloat(instruments[index])
if flag < FlagGoodValue: return false
true
# Note: we're not checking the format of the date stamp.
# Main.
var currentLine = 0
for line in "readings.txt".lines:
currentLine.inc
if line.len == 0: continue # Empty lines don't count as records.
var tokens = line.split({' ', '\t'})
totalRecords.inc
if tokens.len != NumFields:
badRecords.inc
continue
if not checkFloats(tokens):
badRecords.inc
continue
if not areAllFlagsOk(tokens):
badInstruments.inc
if seenDates.hasKeyOrPut(tokens[DateField], true):
echo tokens[DateField], " duplicated on line ", currentLine
let goodRecords = totalRecords - badRecords
let goodInstruments = goodRecords - badInstruments
echo "Total Records: ", totalRecords
echo "Records with wrong format: ", badRecords
echo "Records where all instruments were OK: ", goodInstruments
- Output:
1990-03-25 duplicated on line 85 1991-03-31 duplicated on line 456 1992-03-29 duplicated on line 820 1993-03-28 duplicated on line 1184 1995-03-26 duplicated on line 1911 Total Records: 5471 Records with wrong format: 0 Records where all instruments were OK: 5017
OCaml
#load "str.cma"
open Str
let strip_cr str =
let last = pred (String.length str) in
if str.[last] <> '\r' then str else String.sub str 0 last
let map_records =
let rec aux acc = function
| value::flag::tail ->
let e = (float_of_string value, int_of_string flag) in
aux (e::acc) tail
| [_] -> invalid_arg "invalid data"
| [] -> List.rev acc
in
aux [] ;;
let duplicated_dates =
let same_date (d1,_) (d2,_) = (d1 = d2) in
let date (d,_) = d in
let rec aux acc = function
| a::b::tl when same_date a b ->
aux (date a::acc) tl
| _::tl ->
aux acc tl
| [] ->
List.rev acc
in
aux [] ;;
let record_ok (_,record) =
let is_ok (_,v) = v >= 1 in
let sum_ok =
List.fold_left (fun sum this ->
if is_ok this then succ sum else sum) 0 record
in
sum_ok = 24
let num_good_records =
List.fold_left (fun sum record ->
if record_ok record then succ sum else sum) 0 ;;
let parse_line line =
let li = split (regexp "[ \t]+") line in
let records = map_records (List.tl li)
and date = List.hd li in
(date, records)
let () =
let ic = open_in "readings.txt" in
let rec read_loop acc =
let line_opt = try Some (strip_cr (input_line ic))
with End_of_file -> None
in
match line_opt with
None -> close_in ic; List.rev acc
| Some line -> read_loop (parse_line line :: acc)
in
let inputs = read_loop [] in
Printf.printf "%d total lines\n" (List.length inputs);
Printf.printf "duplicated dates:\n";
let dups = duplicated_dates inputs in
List.iter print_endline dups;
Printf.printf "number of good records: %d\n" (num_good_records inputs);
;;
this script outputs:
5471 total lines duplicated dates: 1990-03-25 1991-03-31 1992-03-29 1993-03-28 1995-03-26 number of good records: 5017
Perl
use List::MoreUtils 'natatime';
use constant FIELDS => 49;
binmode STDIN, ':crlf';
# Read the newlines properly even if we're not running on
# Windows.
my ($line, $good_records, %dates) = (0, 0);
while (<>)
{++$line;
my @fs = split /\s+/;
@fs == FIELDS or die "$line: Bad number of fields.\n";
for (shift @fs)
{/\d{4}-\d{2}-\d{2}/ or die "$line: Bad date format.\n";
++$dates{$_};}
my $iterator = natatime 2, @fs;
my $all_flags_okay = 1;
while ( my ($val, $flag) = $iterator->() )
{$val =~ /\d+\.\d+/ or die "$line: Bad value format.\n";
$flag =~ /\A-?\d+/ or die "$line: Bad flag format.\n";
$flag < 1 and $all_flags_okay = 0;}
$all_flags_okay and ++$good_records;}
print "Good records: $good_records\n",
"Repeated timestamps:\n",
map {" $_\n"}
grep {$dates{$_} > 1}
sort keys %dates;
Output:
Good records: 5017 Repeated timestamps: 1990-03-25 1991-03-31 1992-03-29 1993-03-28 1995-03-26
Phix
-- demo\rosetta\TextProcessing2.exw with javascript_semantics -- (include version/first of next three lines only) include readings.e -- global constant lines, or: --assert(write_lines("readings.txt",lines)!=-1) -- first run, then: --constant lines = read_lines("readings.txt") include builtins\timedate.e integer all_good = 0 string fmt = "%d-%d-%d\t"&join(repeat("%f",48),'\t') sequence extset = sq_mul(tagset(24),2), -- {2,4,6,..48} curr, last for i=1 to length(lines) do string li = lines[i] sequence r = scanf(li,fmt) if length(r)!=1 then printf(1,"bad line [%d]:%s\n",{i,li}) else curr = r[1][1..3] if i>1 and curr=last then printf(1,"duplicate line for %04d/%02d/%02d\n",last) end if last = curr all_good += sum(sq_le(extract(r[1][4..$],extset),0))=0 end if end for printf(1,"Valid records %d of %d total\n",{all_good, length(lines)}) ?"done" {} = wait_key()
- Output:
duplicate line for 1990/03/25 duplicate line for 1991/03/31 duplicate line for 1992/03/29 duplicate line for 1993/03/28 duplicate line for 1995/03/26 Valid records 5017 of 5471 total
PHP
$handle = fopen("readings.txt", "rb");
$missformcount = 0;
$totalcount = 0;
$dates = array();
while (!feof($handle)) {
$buffer = fgets($handle);
$line = preg_replace('/\s+/',' ',$buffer);
$line = explode(' ',trim($line));
$datepattern = '/^\d{4}-\d{2}-\d{2}$/';
$valpattern = '/^\d+\.{1}\d{3}$/';
$flagpattern = '/^[1-9]{1}$/';
if(count($line) != 49) $missformcount++;
if(!preg_match($datepattern,$line[0],$check)) $missformcount++;
else $dates[$totalcount+1] = $check[0];
$errcount = 0;
for($i=1;$i<count($line);$i++){
if($i%2!=0){
if(!preg_match($valpattern,$line[$i],$check)) $errcount++;
}else{
if(!preg_match($flagpattern,$line[$i],$check)) $errcount++;
}
}
if($errcount != 0) $missformcount++;
$totalcount++;
}
fclose ($handle);
$good = $totalcount - $missformcount;
$duplicates = array_diff_key( $dates , array_unique( $dates ));
echo 'Valid records ' . $good . ' of ' . $totalcount . ' total<br>';
echo 'Duplicates : <br>';
foreach ($duplicates as $key => $val){
echo $val . ' at Line : ' . $key . '<br>';
}
Valid records 5017 of 5471 total Duplicates : 1990-03-25 at Line : 85 1991-03-31 at Line : 456 1992-03-29 at Line : 820 1993-03-28 at Line : 1184 1995-03-26 at Line : 1911
Picat
import util.
go =>
Readings = [split(Record) : Record in read_file_lines("readings.txt")],
DateStamps = new_map(),
GoodReadings = 0,
foreach({Rec,Id} in zip(Readings,1..Readings.length))
if Rec.length != 49 then printf("Entry %d has bad_length %d\n", Id, Rec.length) end,
Date = Rec[1],
if DateStamps.has_key(Date) then
printf("Entry %d (date %w) is a duplicate of entry %w\n", Id, Date, DateStamps.get(Date))
else
if sum([1: I in 3..2..49, check_field(Rec[I])]) == 0 then
GoodReadings := GoodReadings + 1
end
end,
DateStamps.put(Date, Id)
end,
nl,
printf("Total readings: %d\n",Readings.len),
printf("Good readings: %d\n",GoodReadings),
nl.
check_field(Field) =>
Field == "-2" ; Field == "-1" ; Field == "0".
- Output:
Entry 85 (date 1990-03-25) is a duplicate of entry 84 Entry 456 (date 1991-03-31) is a duplicate of entry 455 Entry 820 (date 1992-03-29) is a duplicate of entry 819 Entry 1184 (date 1993-03-28) is a duplicate of entry 1183 Entry 1911 (date 1995-03-26) is a duplicate of entry 1910 Total readings: 5471 Good readings: 5013
PicoLisp
Put the following into an executable file "checkReadings":
#!/usr/bin/picolisp /usr/lib/picolisp/lib.l
(load "@lib/misc.l")
(in (opt)
(until (eof)
(let Lst (split (line) "^I")
(unless
(and
(= 49 (length Lst)) # Check total length
($dat (car Lst) "-") # Check for valid date
(fully # Check data format
'((L F)
(if F # Alternating:
(format L 3) # Number
(>= 9 (format L) -9) ) ) # or flag
(cdr Lst)
'(T NIL .) ) )
(prinl "Bad line format: " (glue " " Lst))
(bye 1) ) ) ) )
(bye)
Then it can be called as
$ ./checkReadings readings.txt
PL/I
/* To process readings produced by automatic reading stations. */
check: procedure options (main);
declare 1 date, 2 (yy, mm, dd) character (2),
(j1, j2) character (1);
declare old_date character (6);
declare line character (330) varying;
declare R(24) fixed decimal, Machine(24) fixed binary;
declare (i, k, n, faulty static initial (0)) fixed binary;
declare input file;
open file (input) title ('/READINGS.TXT,TYPE(CRLF),RECSIZE(300)');
on endfile (input) go to done;
old_date = '';
k = 0;
do forever;
k = k + 1;
get file (input) edit (line) (L);
get string(line) edit (yy, j1, mm, j2, dd) (a(4), a(1), a(2), a(1), a(2));
line = substr(line, 11);
do i = 1 to length(line);
if substr(line, i, 1) = '09'x then substr(line, i, 1) = ' ';
end;
line = trim(line);
n = tally(line, ' ') - tally (line, ' ') + 1;
if n ^= 48 then
do;
put skip list ('There are ' || n || ' readings in line ' || k);
end;
n = n/2;
line = line || ' ';
get string(line) list ((R(i), Machine(i) do i = 1 to n));
if any(Machine < 1) ^= '0'B then
faulty = faulty + 1;
if old_date ^= ' ' then if old_date = string(date) then
put skip list ('Dates are the same at line' || k);
old_date = string(date);
end;
done:
put skip list ('There were ' || k || ' sets of readings');
put skip list ('There were ' || faulty || ' faulty readings' );
put skip list ('There were ' || k-faulty || ' good readings' );
end check;
PowerShell
$dateHash = @{}
$goodLineCount = 0
get-content c:\temp\readings.txt |
ForEach-Object {
$line = $_.split(" |`t",2)
if ($dateHash.containskey($line[0])) {
$line[0] + " is duplicated"
} else {
$dateHash.add($line[0], $line[1])
}
$readings = $line[1].split()
$goodLine = $true
if ($readings.count -ne 48) { $goodLine = $false; "incorrect line length : $line[0]" }
for ($i=0; $i -lt $readings.count; $i++) {
if ($i % 2 -ne 0) {
if ([int]$readings[$i] -lt 1) {
$goodLine = $false
}
}
}
if ($goodLine) { $goodLineCount++ }
}
[string]$goodLineCount + " good lines"
Output:
1990-03-25 is duplicated 1991-03-31 is duplicated 1992-03-29 is duplicated 1993-03-28 is duplicated 1995-03-26 is duplicated 5017
An alternative using regular expression syntax:
$dateHash = @{}
$goodLineCount = 0
ForEach ($rawLine in ( get-content c:\temp\readings.txt) ){
$line = $rawLine.split(" |`t",2)
if ($dateHash.containskey($line[0])) {
$line[0] + " is duplicated"
} else {
$dateHash.add($line[0], $line[1])
}
$readings = [regex]::matches($line[1],"\d+\.\d+\s-?\d")
if ($readings.count -ne 24) { "incorrect number of readings for date " + $line[0] }
$goodLine = $true
foreach ($flagMatch in [regex]::matches($line[1],"\d\.\d*\s(?<flag>-?\d)")) {
if ([int][string]$flagMatch.groups["flag"].value -lt 1) {
$goodLine = $false
}
}
if ($goodLine) { $goodLineCount++}
}
[string]$goodLineCount + " good lines"
Output:
1990-03-25 is duplicated 1991-03-31 is duplicated 1992-03-29 is duplicated 1993-03-28 is duplicated 1995-03-26 is duplicated 5017 good lines
PureBasic
Using regular expressions.
Define filename.s = "readings.txt"
#instrumentCount = 24
Enumeration
#exp_date
#exp_instruments
#exp_instrumentStatus
EndEnumeration
Structure duplicate
date.s
firstLine.i
line.i
EndStructure
NewMap dates() ;records line date occurs first
NewList duplicated.duplicate()
NewList syntaxError()
Define goodRecordCount, totalLines, line.s, i
Dim inputDate.s(0)
Dim instruments.s(0)
If ReadFile(0, filename)
CreateRegularExpression(#exp_date, "\d+-\d+-\d+")
CreateRegularExpression(#exp_instruments, "(\t|\x20)+(\d+\.\d+)(\t|\x20)+\-?\d")
CreateRegularExpression(#exp_instrumentStatus, "(\t|\x20)+(\d+\.\d+)(\t|\x20)+")
Repeat
line = ReadString(0, #PB_Ascii)
If line = "": Break: EndIf
totalLines + 1
ExtractRegularExpression(#exp_date, line, inputDate())
If FindMapElement(dates(), inputDate(0))
AddElement(duplicated())
duplicated()\date = inputDate(0)
duplicated()\firstLine = dates()
duplicated()\line = totalLines
Else
dates(inputDate(0)) = totalLines
EndIf
ExtractRegularExpression(#exp_instruments, Mid(line, Len(inputDate(0)) + 1), instruments())
Define pairsCount = ArraySize(instruments()), containsBadValues = #False
For i = 0 To pairsCount
If Val(ReplaceRegularExpression(#exp_instrumentStatus, instruments(i), "")) < 1
containsBadValues = #True
Break
EndIf
Next
If pairsCount <> #instrumentCount - 1
AddElement(syntaxError()): syntaxError() = totalLines
EndIf
If Not containsBadValues
goodRecordCount + 1
EndIf
ForEver
CloseFile(0)
If OpenConsole()
ForEach duplicated()
PrintN("Duplicate date: " + duplicated()\date + " occurs on lines " + Str(duplicated()\line) + " and " + Str(duplicated()\firstLine) + ".")
Next
ForEach syntaxError()
PrintN( "Syntax error in line " + Str(syntaxError()))
Next
PrintN(#CRLF$ + Str(goodRecordCount) + " of " + Str(totalLines) + " lines read were valid records.")
Print(#CRLF$ + #CRLF$ + "Press ENTER to exit"): Input()
CloseConsole()
EndIf
EndIf
Sample output:
Duplicate date: 1990-03-25 occurs on lines 85 and 84. Duplicate date: 1991-03-31 occurs on lines 456 and 455. Duplicate date: 1992-03-29 occurs on lines 820 and 819. Duplicate date: 1993-03-28 occurs on lines 1184 and 1183. Duplicate date: 1995-03-26 occurs on lines 1911 and 1910. 5017 of 5471 lines read were valid records.
Python
import re
import zipfile
import StringIO
def munge2(readings):
datePat = re.compile(r'\d{4}-\d{2}-\d{2}')
valuPat = re.compile(r'[-+]?\d+\.\d+')
statPat = re.compile(r'-?\d+')
allOk, totalLines = 0, 0
datestamps = set([])
for line in readings:
totalLines += 1
fields = line.split('\t')
date = fields[0]
pairs = [(fields[i],fields[i+1]) for i in range(1,len(fields),2)]
lineFormatOk = datePat.match(date) and \
all( valuPat.match(p[0]) for p in pairs ) and \
all( statPat.match(p[1]) for p in pairs )
if not lineFormatOk:
print 'Bad formatting', line
continue
if len(pairs)!=24 or any( int(p[1]) < 1 for p in pairs ):
print 'Missing values', line
continue
if date in datestamps:
print 'Duplicate datestamp', line
continue
datestamps.add(date)
allOk += 1
print 'Lines with all readings: ', allOk
print 'Total records: ', totalLines
#zfs = zipfile.ZipFile('readings.zip','r')
#readings = StringIO.StringIO(zfs.read('readings.txt'))
readings = open('readings.txt','r')
munge2(readings)
The results indicate 5013 good records, which differs from the Awk implementation. The final few lines of the output are as follows
Missing values 2004-12-29 2.900 1 2.700 1 2.800 1 3.300 1 2.900 1 2.300 1 0.000 0 1.700 1 1.900 1 2.300 1 2.600 1 2.900 1 2.600 1 2.600 1 2.600 1 2.700 1 2.300 1 2.200 1 2.100 1 2.000 1 2.100 1 2.100 1 2.300 1 2.400 1 Missing values 2004-12-30 2.400 1 2.600 1 2.600 1 2.600 1 3.000 1 0.000 0 3.300 1 2.600 1 2.900 1 2.400 1 2.300 1 2.900 1 3.500 1 3.700 1 3.600 1 4.000 1 3.400 1 2.400 1 2.500 1 2.600 1 2.600 1 2.800 1 2.400 1 2.200 1 Missing values 2004-12-31 2.400 1 2.500 1 2.500 1 2.400 1 0.000 0 2.400 1 2.400 1 2.400 1 2.200 1 2.400 1 2.500 1 2.000 1 1.700 1 1.400 1 1.500 1 1.900 1 1.700 1 2.000 1 2.000 1 2.200 1 1.700 1 1.500 1 1.800 1 1.800 1 Lines with all readings: 5013 Total records: 5471
Second Version
Modification of the version above to:
- Remove continue statements so it counts as the AWK example does.
- Generate mostly summary information that is easier to compare to other solutions.
import re
import zipfile
import StringIO
def munge2(readings, debug=False):
datePat = re.compile(r'\d{4}-\d{2}-\d{2}')
valuPat = re.compile(r'[-+]?\d+\.\d+')
statPat = re.compile(r'-?\d+')
totalLines = 0
dupdate, badform, badlen, badreading = set(), set(), set(), 0
datestamps = set([])
for line in readings:
totalLines += 1
fields = line.split('\t')
date = fields[0]
pairs = [(fields[i],fields[i+1]) for i in range(1,len(fields),2)]
lineFormatOk = datePat.match(date) and \
all( valuPat.match(p[0]) for p in pairs ) and \
all( statPat.match(p[1]) for p in pairs )
if not lineFormatOk:
if debug: print 'Bad formatting', line
badform.add(date)
if len(pairs)!=24 or any( int(p[1]) < 1 for p in pairs ):
if debug: print 'Missing values', line
if len(pairs)!=24: badlen.add(date)
if any( int(p[1]) < 1 for p in pairs ): badreading += 1
if date in datestamps:
if debug: print 'Duplicate datestamp', line
dupdate.add(date)
datestamps.add(date)
print 'Duplicate dates:\n ', '\n '.join(sorted(dupdate))
print 'Bad format:\n ', '\n '.join(sorted(badform))
print 'Bad number of fields:\n ', '\n '.join(sorted(badlen))
print 'Records with good readings: %i = %5.2f%%\n' % (
totalLines-badreading, (totalLines-badreading)/float(totalLines)*100 )
print 'Total records: ', totalLines
readings = open('readings.txt','r')
munge2(readings)
bash$ /cygdrive/c/Python26/python munge2.py Duplicate dates: 1990-03-25 1991-03-31 1992-03-29 1993-03-28 1995-03-26 Bad format: Bad number of fields: Records with good readings: 5017 = 91.70% Total records: 5471 bash$
R
# Read in data from file
dfr <- read.delim("d:/readings.txt", colClasses=c("character", rep(c("numeric", "integer"), 24)))
dates <- strptime(dfr[,1], "%Y-%m-%d")
# Any bad values?
dfr[which(is.na(dfr))]
# Any duplicated dates
dates[duplicated(dates)]
# Number of rows with no bad values
flags <- as.matrix(dfr[,seq(3,49,2)])>0
sum(apply(flags, 1, all))
Racket
#lang racket
(read-decimal-as-inexact #f)
;; files to read is a sequence, so it could be either a list or vector of files
(define (text-processing/2 files-to-read)
(define seen-datestamps (make-hash))
(define (datestamp-seen? ds) (hash-ref seen-datestamps ds #f))
(define (datestamp-seen! ds pos) (hash-set! seen-datestamps ds pos))
(define (fold-into-pairs l (acc null))
(match l ['() (reverse acc)]
[(list _) (reverse (cons l acc))]
[(list-rest a b tl) (fold-into-pairs tl (cons (list a b) acc))]))
(define (match-valid-field line pos)
(match (string-split line)
;; if we don't hit an error, then the file is valid
((list-rest (not (pregexp #px"[[:digit:]]{4}-[[:digit:]]{2}-[[:digit:]]{2}")) _)
(error 'match-valid-field "invalid format non-datestamp at head: ~a~%" line))
;; check for duplicates
((list-rest (? datestamp-seen? ds) _)
(printf "duplicate datestamp: ~a at line: ~a (first seen at: ~a)~%"
ds pos (datestamp-seen? ds))
#f)
;; register the datestamp as seen, then move on to rest of match
((list-rest ds _) (=> next-match-rule) (datestamp-seen! ds pos) (next-match-rule))
((list-rest
_
(app fold-into-pairs
(list (list (app string->number (and (? number?) vs))
(app string->number (and (? integer?) statuss)))
...)))
(=> next-match-rule)
(unless (= (length vs) 24) (next-match-rule))
(not (for/first ((s statuss) #:unless (positive? s)) #t)))
;; if we don't hit an error, then the file is valid
(else (error 'match-valid-field "bad field format: ~a~%" line))))
(define (sub-t-p/1)
(for/sum ((line (in-lines))
(line-number (in-naturals 1)))
(if (match-valid-field line line-number) 1 0)))
(for/sum ((file-name files-to-read))
(with-input-from-file file-name sub-t-p/1)))
(printf "~a records have good readings for all instruments~%"
(text-processing/2 (current-command-line-arguments)))
Example session:
$ racket 2.rkt readings/readings.txt duplicate datestamp: 1990-03-25 at line: 85 (first seen at: 84) duplicate datestamp: 1991-03-31 at line: 456 (first seen at: 455) duplicate datestamp: 1992-03-29 at line: 820 (first seen at: 819) duplicate datestamp: 1993-03-28 at line: 1184 (first seen at: 1183) duplicate datestamp: 1995-03-26 at line: 1911 (first seen at: 1910) 5013 records have good readings for all instruments
Raku
(formerly Perl 6)
This version does validation with a single Raku regex that is much more readable than the typical regex, and arguably expresses the data structure more straightforwardly. Here we use normal quotes for literals, and \h for horizontal whitespace.
Variables like $good-record that are going to be autoincremented do not need to be initialized.
The .push method on a hash is magical and loses no information; if a duplicate key is found in the pushed pair, an array of values is automatically created of the old value and the new value pushed. Hence we can easily track all the lines that a particular duplicate occurred at.
The .all method does "junctional" logic: it autothreads through comparators as any English speaker would expect. Junctions can also short-circuit as soon as they find a value that doesn't match, and the evaluation order is up to the computer, so it can be optimized or parallelized.
The final line simply greps out the pairs from the hash whose value is an array with more than 1 element. (Those values that are not arrays nevertheless have a .elems method that always reports 1.) The .pairs is merely there for clarity; grepping a hash directly has the same effect. Note that we sort the pairs after we've grepped them, not before; this works fine in Raku, sorting on the key and value as primary and secondary keys. Finally, pairs and arrays provide a default print format that is sufficient without additional formatting in this case.
my $good-records;
my $line;
my %dates;
for lines() {
$line++;
/ ^
(\d ** 4 '-' \d\d '-' \d\d)
[ \h+ \d+'.'\d+ \h+ ('-'?\d+) ] ** 24
$ /
or note "Bad format at line $line" and next;
%dates.push: $0 => $line;
$good-records++ if $1.all >= 1;
}
say "$good-records good records out of $line total";
say 'Repeated timestamps (with line numbers):';
.say for sort %dates.pairs.grep: *.value.elems > 1;
Output:
5017 good records out of 5471 total Repeated timestamps (with line numbers): 1990-03-25 => [84 85] 1991-03-31 => [455 456] 1992-03-29 => [819 820] 1993-03-28 => [1183 1184] 1995-03-26 => [1910 1911]
REXX
This REXX program process the file mentioned in "text processing 1" and does further validate on the dates, flags, and data.
Some of the checks performed are:
- checks for duplicated date records.
- checks for a bad date (YYYY-MM-DD) format, among:
- wrong length
- year > current year
- year < 1970 (to allow for posthumous data)
- mm < 1 or mm > 12
- dd < 1 or dd > days for the month
- yyyy, dd, mm isn't numeric
- missing data (or flags)
- flag isn't an integer
- flag contains a decimal point
- data isn't numeric
In addition, all of the presented numbers may have commas inserted.
The program has (negated) code to write the report to a file in addition to the console.
/*REXX program to process instrument data from a data file. */
numeric digits 20 /*allow for bigger numbers. */
ifid='READINGS.TXT' /*name of the input file. */
ofid='READINGS.OUT' /* " " " output " */
grandSum=0 /*grand sum of the whole file. */
grandFlg=0 /*grand number of flagged data. */
grandOKs=0
Lflag=0 /*longest period of flagged data. */
Cflag=0 /*longest continuous flagged data. */
oldDate =0 /*placeholder of penultimate date. */
w =16 /*width of fields when displayed. */
dupDates=0 /*count of duplicated timestamps. */
badFlags=0 /*count of bad flags (not integer). */
badDates=0 /*count of bad dates (bad format). */
badData =0 /*count of bad data (not numeric). */
ignoredR=0 /*count of ignored records, bad records*/
maxInstruments=24 /*maximum number of instruments. */
yyyyCurr=right(date(),4) /*get the current year (today). */
monDD. =31 /*number of days in every month. */
/*# days in Feb. is figured on the fly.*/
monDD.4 =30
monDD.6 =30
monDD.9 =30
monDD.11=30
do records=1 while lines(ifid)\==0 /*read until finished. */
rec=linein(ifid) /*read the next record (line). */
parse var rec datestamp Idata /*pick off the the dateStamp and data. */
if datestamp==oldDate then do /*found a duplicate timestamp. */
dupDates=dupDates+1 /*bump the dupDate counter*/
call sy datestamp copies('~',30),
'is a duplicate of the',
"previous datestamp."
ignoredR=ignoredR+1 /*bump # of ignoredRecs.*/
iterate /*ignore this duplicate record. */
end
parse var datestamp yyyy '-' mm '-' dd /*obtain YYYY, MM, and the DD. */
monDD.2=28+leapyear(yyyy) /*how long is February in year YYYY ? */
/*check for various bad formats. */
if verify(yyyy||mm||dd,1234567890)\==0 |,
length(datestamp)\==10 |,
length(yyyy)\==4 |,
length(mm )\==2 |,
length(dd )\==2 |,
yyyy<1970 |,
yyyy>yyyyCurr |,
mm=0 | dd=0 |,
mm>12 | dd>monDD.mm then do
badDates=badDates+1
call sy datestamp copies('~'),
'has an illegal format.'
ignoredR=ignoredR+1 /*bump number ignoredRecs.*/
iterate /*ignore this bad record. */
end
oldDate=datestamp /*save datestamp for the next read. */
sum=0
flg=0
OKs=0
do j=1 until Idata='' /*process the instrument data. */
parse var Idata data.j flag.j Idata
if pos('.',flag.j)\==0 |, /*does flag have a decimal point -or- */
\datatype(flag.j,'W') then do /* ··· is the flag not a whole number? */
badFlags=badFlags+1 /*bump badFlags counter*/
call sy datestamp copies('~'),
'instrument' j "has a bad flag:",
flag.j
iterate /*ignore it and it's data. */
end
if \datatype(data.j,'N') then do /*is the flag not a whole number?*/
badData=badData+1 /*bump counter.*/
call sy datestamp copies('~'),
'instrument' j "has bad data:",
data.j
iterate /*ignore it & it's flag.*/
end
if flag.j>0 then do /*if good data, ~~~ */
OKs=OKs+1
sum=sum+data.j
if Cflag>Lflag then do
Ldate=datestamp
Lflag=Cflag
end
Cflag=0
end
else do /*flagged data ~~~ */
flg=flg+1
Cflag=Cflag+1
end
end /*j*/
if j>maxInstruments then do
badData=badData+1 /*bump the badData counter.*/
call sy datestamp copies('~'),
'too many instrument datum'
end
if OKs\==0 then avg=format(sum/OKs,,3)
else avg='[n/a]'
grandOKs=grandOKs+OKs
_=right(commas(avg),w)
grandSum=grandSum+sum
grandFlg=grandFlg+flg
if flg==0 then call sy datestamp ' average='_
else call sy datestamp ' average='_ ' flagged='right(flg,2)
end /*records*/
records=records-1 /*adjust for reading the end─of─file. */
if grandOKs\==0 then grandAvg=format(grandsum/grandOKs,,3)
else grandAvg='[n/a]'
call sy
call sy copies('=',60)
call sy ' records read:' right(commas(records ),w)
call sy ' records ignored:' right(commas(ignoredR),w)
call sy ' grand sum:' right(commas(grandSum),w+4)
call sy ' grand average:' right(commas(grandAvg),w+4)
call sy ' grand OK data:' right(commas(grandOKs),w)
call sy ' grand flagged:' right(commas(grandFlg),w)
call sy ' duplicate dates:' right(commas(dupDates),w)
call sy ' bad dates:' right(commas(badDates),w)
call sy ' bad data:' right(commas(badData ),w)
call sy ' bad flags:' right(commas(badFlags),w)
if Lflag\==0 then call sy ' longest flagged:' right(commas(LFlag),w) " ending at " Ldate
call sy copies('=',60)
exit /*stick a fork in it, we're all done.*/
/*────────────────────────────────────────────────────────────────────────────*/
commas: procedure; parse arg _; n=_'.9'; #=123456789; b=verify(n,#,"M")
e=verify(n,#'0',,verify(n,#"0.",'M'))-4
do j=e to b by -3; _=insert(',',_,j); end /*j*/; return _
/*────────────────────────────────────────────────────────────────────────────*/
leapyear: procedure; arg y /*year could be: Y, YY, YYY, or YYYY*/
if length(y)==2 then y=left(right(date(),4),2)y /*adjust for YY year.*/
if y//4\==0 then return 0 /* not divisible by 4? Not a leapyear*/
return y//100\==0 | y//400==0 /*apply the 100 and the 400 year rule.*/
/*────────────────────────────────────────────────────────────────────────────*/
sy: say arg(1); call lineout ofid,arg(1); return
output when using the default input file:
∙ ∙ ∙ 1991-03-31 average= 23.542 1991-03-31 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ is a duplicate of the previous datestamp. 1991-04-01 average= 23.217 flagged= 1 1991-04-02 average= 19.792 1991-04-03 average= 13.958 ∙ ∙ ∙ ============================================================ records read: 5,471 records ignored: 5 grand sum: 1,357,152.400 grand average: 10.496 grand OK data: 129,306 grand flagged: 1,878 duplicate dates: 5 bad dates: 0 bad data: 0 bad flags: 0 longest flagged: 589 ending at 1993-03-05 ============================================================
Ruby
require 'set'
def munge2(readings, debug=false)
datePat = /^\d{4}-\d{2}-\d{2}/
valuPat = /^[-+]?\d+\.\d+/
statPat = /^-?\d+/
totalLines = 0
dupdate, badform, badlen, badreading = Set[], Set[], Set[], 0
datestamps = Set[[]]
for line in readings
totalLines += 1
fields = line.split(/\t/)
date = fields.shift
pairs = fields.enum_slice(2).to_a
lineFormatOk = date =~ datePat &&
pairs.all? { |x,y| x =~ valuPat && y =~ statPat }
if !lineFormatOk
puts 'Bad formatting ' + line if debug
badform << date
end
if pairs.length != 24 ||
pairs.any? { |x,y| y.to_i < 1 }
puts 'Missing values ' + line if debug
end
if pairs.length != 24
badlen << date
end
if pairs.any? { |x,y| y.to_i < 1 }
badreading += 1
end
if datestamps.include?(date)
puts 'Duplicate datestamp ' + line if debug
dupdate << date
end
datestamps << date
end
puts 'Duplicate dates:', dupdate.sort.map { |x| ' ' + x }
puts 'Bad format:', badform.sort.map { |x| ' ' + x }
puts 'Bad number of fields:', badlen.sort.map { |x| ' ' + x }
puts 'Records with good readings: %i = %5.2f%%' % [
totalLines-badreading, (totalLines-badreading)/totalLines.to_f*100 ]
puts
puts 'Total records: %d' % totalLines
end
open('readings.txt','r') do |readings|
munge2(readings)
end
Scala
object DataMunging2 {
import scala.io.Source
import scala.collection.immutable.{TreeMap => Map}
val pattern = """^(\d+-\d+-\d+)""" + """\s+(\d+\.\d+)\s+(-?\d+)""" * 24 + "$" r;
def main(args: Array[String]) {
val files = args map (new java.io.File(_)) filter (file => file.isFile && file.canRead)
val (numFormatErrors, numValidRecords, dateMap) =
files.iterator.flatMap(file => Source fromFile file getLines ()).
foldLeft((0, 0, new Map[String, Int] withDefaultValue 0)) {
case ((nFE, nVR, dM), line) => pattern findFirstMatchIn line map (_.subgroups) match {
case Some(List(date, rawData @ _*)) =>
val allValid = (rawData map (_ toDouble) iterator) grouped 2 forall (_.last > 0)
(nFE, nVR + (if (allValid) 1 else 0), dM(date) += 1)
case None => (nFE + 1, nVR, dM)
}
}
dateMap foreach {
case (date, repetitions) if repetitions > 1 => println(date+": "+repetitions+" repetitions")
case _ =>
}
println("""|
|Valid records: %d
|Duplicated dates: %d
|Duplicated records: %d
|Data format errors: %d
|Invalid data records: %d
|Total records: %d""".stripMargin format (
numValidRecords,
dateMap filter { case (_, repetitions) => repetitions > 1 } size,
dateMap.valuesIterable filter (_ > 1) map (_ - 1) sum,
numFormatErrors,
dateMap.valuesIterable.sum - numValidRecords,
dateMap.valuesIterable.sum))
}
}
Sample output:
1990-03-25: 2 repetitions 1991-03-31: 2 repetitions 1992-03-29: 2 repetitions 1993-03-28: 2 repetitions 1995-03-26: 2 repetitions Valid records: 5017 Duplicated dates: 5 Duplicated records: 5 Data format errors: 0 Invalid data records: 454 Total records: 5471
Sidef
var good_records = 0;
var dates = Hash();
ARGF.each { |line|
var m = /^(\d\d\d\d-\d\d-\d\d)((?:\h+\d+\.\d+\h+-?\d+){24})\s*$/.match(line);
m || (warn "Bad format at line #{$.}"; next);
dates{m[0]} := 0 ++;
var i = 0;
m[1].words.all{|n| i++.is_even || (n.to_num >= 1) } && ++good_records;
}
say "#{good_records} good records out of #{$.} total";
say 'Repeated timestamps:';
say dates.to_a.grep{ .value > 1 }.map { .key }.sort.join("\n");
- Output:
$ sidef script.sf < readings.txt 5017 good records out of 5471 total Repeated timestamps: 1990-03-25 1991-03-31 1992-03-29 1993-03-28 1995-03-26
Snobol4
Developed using the Snobol4 dialect Spitbol for Linux, version 4.0
* Read text/2
v = array(24)
f = array(24)
tos = char(9) " " ;* break characters are both tab and space
pat1 = break(tos) . dstamp
pat2 = span(tos) break(tos) . *v[i] span(tos) (break(tos) | (len(1) rem)) . *f[i]
rowcount = 0
hold_dstamp = ""
num_bad_rows = 0
num_invalid_rows = 0
in0
row = input :f(endinput)
rowcount = rowcount + 1
row ? pat1 = :f(invalid_row)
* duplicated datestamp?
* if dstamp = hold_dstamp then duplicated
hold_dstamp = differ(hold_dstamp,dstamp) dstamp :s(nodup)
output = dstamp ": datestamp at row " rowcount " duplicates datestamp at " rowcount - 1
nodup
i = 1
in1
row ? pat2 = :f(invalid_row)
i = lt(i,24) i + 1 :s(in1)
* Is this a goodrow?
* if any flag is < 1 then row has bad data
c = 0
goodrow
c = lt(c,24) c + 1 :f(goodrow2)
num_bad_rows = lt(f[c],1) num_bad_rows + 1 :s(goodrow2)f(goodrow)
goodrow2
:(in0)
invalid_row
num_invalid_rows = num_invalid_rows + 1
:(in0)
endinput
output =
output = "Total number of rows : " rowcount
output = "Total number of rows with invalid format: " num_invalid_rows
output = "Total number of rows with bad data : " num_bad_rows
output = "Total number of good rows : " rowcount - num_invalid_rows - num_bad_rows
end
- Output:
1990-03-25: datestamp at row 85 duplicates datestamp at 84 1991-03-31: datestamp at row 456 duplicates datestamp at 455 1992-03-29: datestamp at row 820 duplicates datestamp at 819 1993-03-28: datestamp at row 1184 duplicates datestamp at 1183 1995-03-26: datestamp at row 1911 duplicates datestamp at 1910 Total number of rows : 5471 Total number of rows with invalid format: 0 Total number of rows with bad data : 454 Total number of good rows : 5017
Tcl
set data [lrange [split [read [open "readings.txt" "r"]] "\n"] 0 end-1]
set total [llength $data]
set correct $total
set datestamps {}
foreach line $data {
set formatOk true
set hasAllMeasurements true
set date [lindex $line 0]
if {[llength $line] != 49} { set formatOk false }
if {![regexp {\d{4}-\d{2}-\d{2}} $date]} { set formatOk false }
if {[lsearch $datestamps $date] != -1} { puts "Duplicate datestamp: $date" } {lappend datestamps $date}
foreach {value flag} [lrange $line 1 end] {
if {$flag < 1} { set hasAllMeasurements false }
if {![regexp -- {[-+]?\d+\.\d+} $value] || ![regexp -- {-?\d+} $flag]} {set formatOk false}
}
if {!$hasAllMeasurements} { incr correct -1 }
if {!$formatOk} { puts "line \"$line\" has wrong format" }
}
puts "$correct records with good readings = [expr $correct * 100.0 / $total]%"
puts "Total records: $total"
$ tclsh munge2.tcl Duplicate datestamp: 1990-03-25 Duplicate datestamp: 1991-03-31 Duplicate datestamp: 1992-03-29 Duplicate datestamp: 1993-03-28 Duplicate datestamp: 1995-03-26 5017 records with good readings = 91.7016998721% Total records: 5471
Second version
To demonstate a different method to iterate over the file, and different ways to verify data types:
set total [set good 0]
array set seen {}
set fh [open readings.txt]
while {[gets $fh line] != -1} {
incr total
set fields [regexp -inline -all {[^ \t\r\n]+} $line]
if {[llength $fields] != 49} {
puts "bad format: not 49 fields on line $total"
continue
}
if { ! [regexp {^(\d{4}-\d\d-\d\d)$} [lindex $fields 0] -> date]} {
puts "bad format: invalid date on line $total: '$date'"
continue
}
if {[info exists seen($date)]} {
puts "duplicate date on line $total: $date"
}
incr seen($date)
set line_format_ok true
set readings_ignored 0
foreach {value flag} [lrange $fields 1 end] {
if { ! [string is double -strict $value]} {
puts "bad format: value not a float on line $total: '$value'"
set line_format_ok false
}
if { ! [string is int -strict $flag]} {
puts "bad format: flag not an integer on line $total: '$flag'"
set line_format_ok false
}
if {$flag < 1} {incr readings_ignored}
}
if {$line_format_ok && $readings_ignored == 0} {incr good}
}
close $fh
puts "total: $total"
puts [format "good: %d = %5.2f%%" $good [expr {100.0 * $good / $total}]]
Results:
duplicate date on line 85: 1990-03-25 duplicate date on line 456: 1991-03-31 duplicate date on line 820: 1992-03-29 duplicate date on line 1184: 1993-03-28 duplicate date on line 1911: 1995-03-26 total: 5471 good: 5017 = 91.70%
Ursala
compiled and run in a single step, with the input file accessed as a list of strings pre-declared in readings_dot_txt
#import std
#import nat
readings = (*F ~&c/;digits+ rlc ==+ ~~ -={` ,9%cOi&,13%cOi&}) readings_dot_txt
valid_format = all -&length==49,@tK27 all ~&w/`.&& ~&jZ\digits--'-.',@tK28 all ~&jZ\digits--'-'&-
duplicate_dates = :/'duplicated dates:'+ ~&hK2tFhhPS|| -[(none)]-!
good_readings = --' good readings'@h+ %nP+ length+ *~ @tK28 all ~='0'&& ~&wZ/`-
#show+
main = valid_format?(^C/good_readings duplicate_dates,-[invalid format]-!) readings
output:
5017 good readings duplicated dates: 1995-03-26 1993-03-28 1992-03-29 1991-03-31 1990-03-25
VBScript
Set objFSO = CreateObject("Scripting.FileSystemObject")
Set objFile = objFSO.OpenTextFile(objFSO.GetParentFolderName(WScript.ScriptFullName) &_
"\readings.txt",1)
Set objDateStamp = CreateObject("Scripting.Dictionary")
Total_Records = 0
Valid_Records = 0
Duplicate_TimeStamps = ""
Do Until objFile.AtEndOfStream
line = objFile.ReadLine
If line <> "" Then
token = Split(line,vbTab)
If objDateStamp.Exists(token(0)) = False Then
objDateStamp.Add token(0),""
Total_Records = Total_Records + 1
If IsValid(token) Then
Valid_Records = Valid_Records + 1
End If
Else
Duplicate_TimeStamps = Duplicate_TimeStamps & token(0) & vbCrLf
Total_Records = Total_Records + 1
End If
End If
Loop
Function IsValid(arr)
IsValid = True
Bad_Readings = 0
n = 1
Do While n <= UBound(arr)
If n + 1 <= UBound(arr) Then
If CInt(arr(n+1)) < 1 Then
Bad_Readings = Bad_Readings + 1
End If
End If
n = n + 2
Loop
If Bad_Readings > 0 Then
IsValid = False
End If
End Function
WScript.StdOut.Write "Total Number of Records = " & Total_Records
WScript.StdOut.WriteLine
WScript.StdOut.Write "Total Valid Records = " & Valid_Records
WScript.StdOut.WriteLine
WScript.StdOut.Write "Duplicate Timestamps:"
WScript.StdOut.WriteLine
WScript.StdOut.Write Duplicate_TimeStamps
WScript.StdOut.WriteLine
objFile.Close
Set objFSO = Nothing
- Output:
Total Number of Records = 5471 Total Valid Records = 5013 Duplicate Timestamps: 1990-03-25 1991-03-31 1992-03-29 1993-03-28 1995-03-26
Vedit macro language
This implementation does the following checks:
- Checks for duplicate date fields. Note: duplicates can still be counted as valid records, as in other implementations.
- Checks date format.
- Checks that value fields have 1 or more digits followed by decimal point followed by 3 digits
- Reads flag value and checks if it is positive
- Requires 24 value/flag pairs on each line
#50 = Buf_Num // Current edit buffer (source data)
File_Open("|(PATH_ONLY)\output.txt")
#51 = Buf_Num // Edit buffer for output file
Buf_Switch(#50)
#11 = #12 = #13 = #14 = #15 = 0
Reg_Set(15, "xxx")
While(!At_EOF) {
#10 = 0
#12++
// Check for repeated date field
if (Match(@15) == 0) {
#20 = Cur_Line
Buf_Switch(#51) // Output file
Reg_ins(15) IT(": duplicate record at ") Num_Ins(#20)
Buf_Switch(#50) // Input file
#13++
}
// Check format of date field
if (Match("|d|d|d|d-|d|d-|d|d|w", ADVANCE) != 0) {
#10 = 1
#14++
}
Reg_Copy_Block(15, BOL_pos, Cur_Pos-1)
// Check data fields and flags:
Repeat(24) {
if ( Match("|d|*.|d|d|d|w", ADVANCE) != 0 || Num_Eval(ADVANCE) < 1) {
#10 = 1
#15++
Break
}
Match("|W", ADVANCE)
}
if (#10 == 0) { #11++ } // record was OK
Line(1, ERRBREAK)
}
Buf_Switch(#51) // buffer for output data
IN
IT("Valid records: ") Num_Ins(#11)
IT("Duplicates: ") Num_Ins(#13)
IT("Date format errors: ") Num_Ins(#14)
IT("Invalid data records:") Num_Ins(#15)
IT("Total records: ") Num_Ins(#12)
Sample output:
1990-03-25: duplicate record at 85
1991-03-31: duplicate record at 456
1992-03-29: duplicate record at 820
1993-03-28: duplicate record at 1184
1995-03-26: duplicate record at 1911
Valid records: 5017
Duplicates: 5
Date format errors: 0
Invalid data records: 454
Total records: 5471
Wren
import "io" for File
import "./pattern" for Pattern
import "./fmt" for Fmt
import "./sort" for Sort
var p = Pattern.new("+1/s")
var fileName = "readings.txt"
var lines = File.read(fileName).trimEnd().split("\r\n")
var count = 0
var invalid = 0
var allGood = 0
var map = {}
for (line in lines) {
count = count + 1
var fields = p.splitAll(line)
var date = fields[0]
if (fields.count == 49) {
map[date] = map.containsKey(date) ? map[date] + 1 : 1
var good = 0
var i = 2
while (i < fields.count) {
if (Num.fromString(fields[i]) >= 1) good = good + 1
i = i + 2
}
if (good == 24) allGood = allGood + 1
} else {
invalid = invalid + 1
}
}
Fmt.print("File = $s", fileName)
System.print("\nDuplicated dates:")
var keys = map.keys.toList
Sort.quick(keys)
for (k in keys) {
var v = map[k]
if (v > 1) Fmt.print(" $s ($d times)", k, v)
}
Fmt.print("\nTotal number of records : $d", count)
var percent = invalid/count * 100
Fmt.print("Number of invalid records : $d ($5.2f)\%", invalid, percent)
percent = allGood/count * 100
Fmt.print("Number which are all good : $d ($5.2f)\%", allGood, percent)
- Output:
File = readings.txt Duplicated dates: 1990-03-25 (2 times) 1991-03-31 (2 times) 1992-03-29 (2 times) 1993-03-28 (2 times) 1995-03-26 (2 times) Total number of records : 5471 Number of invalid records : 0 ( 0.00)% Number which are all good : 5017 (91.70)%
zkl
// the RegExp engine has a low limit on groups so
// I can't use it to select all fields, only verify them
re:=RegExp(0'|^(\d+-\d+-\d+)| + 0'|\s+\d+\.\d+\s+-*\d+| * 24 + ".+$");
w:=[1..].zip(File("readings.txt")); //-->lazy (line #,line)
reg datep,N, good=0, dd=0;
foreach n,line in (w){
N=n; // since n is local to this scope
if (not re.search(line)){ println("Line %d: malformed".fmt(n)); continue; }
date:=line[re.matchedNs[1].xplode()]; // I can group the date field
if (datep==date){ dd+=1; println("Line %4d: dup date: %s".fmt(n,date)); }
datep=date;
if (line.replace("\t"," ").split(" ").filter()[1,*] // blow fields apart, drop date
.pump(Void,Void.Read, // get (reading,status)
fcn(_,s){ // stop on first problem status and return True
if(s.strip().toInt()<1) T(Void.Stop,True) else False
})) continue;
good+=1;
}
println("%d records read, %d duplicate dates, %d valid".fmt(N,dd,good));
- Output:
Line 85: dup date: 1990-03-25 Line 456: dup date: 1991-03-31 Line 820: dup date: 1992-03-29 Line 1184: dup date: 1993-03-28 Line 1911: dup date: 1995-03-26 5471 records read, 5 duplicate dates, 5017 valid