Home / Questions / 32878
Accepted
Jose Hermosilla Rodrigo
Asked: 2020-11-10 17:33:41 +0800 CST 2020-11-10 17:33:41 +0800 CST

Regex to extract certain numeric values

  • 772

I'm trying to parse lines like this with regular expressions in python :

21698213.20307                  -4937213.445 7  -3801759.02548  21698206.56648

These values ​​specifically refer to "observations" of GPS signals. In the line above there would be 5 "remarks". If the observables are L1, L2, C1, C2, P2the values ​​I would like to extract:

L1 : { observación -> 21698213.203, LossOfLockInd -> 0, SignalStrengthInd -> 7}

L2 : { observación -> NOHAY, LossOfLockInd -> NOHAY, SignalStrengthInd -> NOHAY }

C1 : { observación -> -4937213.445, LossOfLockInd -> 0 (NO HAY), SignalStrengthInd -> 7}

C2 : { observación -> -3801759.025, LossOfLockInd -> 4, SignalStrengthInd -> 8}

P2 : { observación -> 21698206.566, LossOfLockInd -> 4, SignalStrengthInd -> 8}

That is, I need to extract each "decimal number from 3 decimal places( observation )" and "each individual number or space ( LossOfLock, SignalStrength )". In the event that there is no value for one of the observables I would like to have 3 empty elements (If there is no observable the separation between each of the observables is 18)

So far I've been able to get the decimals and the integers separately, but I can't also join the empty spaces ( LossOfLock ) or separate the missing observables into 3 empty elements.

This is the expression I'm using at the moment.

([-+]?\d*\.\d{3}|\d)

Example of capture he does so far:

var match = '21698213.20307                  -4937213.445 7  -3801759.02548  21698206.56648'.match(/([-+]?\d*\.\d{3}|\d)/g);
console.log(match);

In the end I used the regular expression : ([-+ \d]{9}[. ][ \d]{3})([\d ])([\d ])proposed by Mariano and a couple of code-based tweaks to fill in the gaps left by the regex at the end :

##Obtenemos la observación
the_obs = re.findall(self.REGEX_PARSE_LINEA_OBS, ''.join(obsArray[obsindex : obsindex + step]) )
## quitamos los espacios de la lista
## El regex devuelve un array de tuplas
## con chain.from_iterable() las tuplas desaparecen 
## y pasan dentro de la lista como strings
the_obs = map(strip_, list(itertools.chain.from_iterable(the_obs)))

## El regex nos puede dejar hiuecos al final si no hay observaciones
## con esto rellenamos los huecos
if(len(the_obs) < len(self.header['OBSERV_TYPES'] * 3)) : 
    ## Cuantos huecos faltan por rellenar ?
    size = (len(self.header['OBSERV_TYPES'] * 3)) - len(the_obs)
    ## rellenamos los huecos DEL FINAL!!!
    the_obs[len(the_obs):] = ['' for x in range(size)]
python

1 Answers

  1. Best Answer

    Seeing as the original text has a fixed width for each element, instead of using a regular expression, I'd recommend retrieving each value based on its position.

    A simplified example for the point raised would be:

    datos = r"""
  21698213.20307                  -4937213.445 7  -3801759.02548  21698206.56648
 121367582.20508  94572134.49208  23095489.677 9  23095481.949 9  23095483.463 7
        42.000          40.000                                                  
 134357446.85408                                                  23095483.463 7
"""

#Los anchos de cada columna
# Separador: 1; Observable: 13; LossOfLock: 1; SignalStrength: 1
columnas = [1,13,1,1]

for linea in datos.splitlines(): #cada línea
    for inicio in range(0, len(linea), sum(columnas)): #cada elemento
        for columna in columnas: #cada valor de la columna dentro del elemento
            print(linea[inicio:inicio+columna])
            inicio += columna

    Result:

    　
 21698213.203
0
7





 -4937213.445

7

 -3801759.025
4
8

    # etc...

    demonstration:

    Ideone.com


    If you still want to keep trying regular expressions, I would use the same logic: always get the element with fixed width. Of course, we can use groups to separate the value of each column .

    r' ([-+ \d]{9}[. ][ \d]{3})([\d ])([\d ])'

    Example:

    datos = r""" 121367582.20508  94572134.49208  23095489.677 9  23095481.949 9  23095483.463 7
        42.000          40.000                                                  
 134357446.85408                                                                """

resultado = re.findall(r' ([-+ \d]{9}[. ][ \d]{3})([\d ])([\d ])', datos)

print(resultado)

    Result

    [('121367582.205', '0', '8'), (' 94572134.492', '0', '8'), (' 23095489.677', ' ', '9'), (' 23095481.949', ' ', '9'), (' 23095483.463', ' ', '7'), ('       42.000', ' ', ' '), ('       40.000', ' ', ' '), ('             ', ' ', ' '), ('             ', ' ', ' '), ('             ', ' ', ' '), ('134357446.854', '0', '8'), ('             ', ' ', ' '), ('             ', ' ', ' '), ('             ', ' ', ' '), ('             ', ' ', ' ')]

    demonstration:

    rextester.com

    • 1