Extract a dictionary of substring in bash? - regex

Question

I have files of the kind:
(1), (2), (3), (4), (5), (6), (10), (11), (12), (13), (14), (15), (16), (17), (18), (24), (25), (26), (27), (28), (29), (30), (31), (32), (33), (34), (35), (36), (37), (38), (39), (40), (41), (42), (43), (51), (52), (53), (54), (55), (56), (57), (58), (62), (63), (64), (65), (66), (67), (68), (69), (70), (71), (72), (73), (74) Use method number 1. (7), (8), (9), (19), (20), (21), (22), (23), (59), (60), (61) Use method number 2. (44), (45), (46), (47), (48), (49), (50) Use method number 3.
I would like to build a dictionary containing the numbers between parentheses and link them to the sentences of the type: "Use method number #". So, in this case:
1,2,3,4,5...74 --> Use method number 1.
7,8,9,19....61 --> Use method number 2.
Currently I am building a complex while that reads regexs (^ *\([0-9]+\)), extracts each number, deletes the coincidence and starts again until regex is not found and then extracts the sentence. But this is quite poor in performance and tedious to maintain.
Have you got any suggestions on how to improve this through more compact methods other than the while do?
I am not bothered by the dictionary structure, do not consider it right now if it does not imply modifying the method.
EDIT. ADDING REAL DATA STRING:
(12), (13), (14), (15) P.S.: 3 días en cultivo de invernadero. Efectuar un máximo de 6 aplicaciones por
campaña a intervalos de 7 días utilizando un volumen máximo de caldo de 600 l/Ha. y un máximo de
7,5 Kg de cobre inorgánico por campaña.
(28) Tratamiento en otoño, pulverizando hasta una altura de 1,5 m.
(44), (45), (46), (47), (48), (49), (50), (51) Efectuar sólo tratamientos desde la cosecha hasta la
floración, limitando la aplicación a 1200 l. de caldo/Ha. y un máximo de 3 aplicaciones por campaña
(con un intervalo de tratamientos de 14 días) y un máximo de 7,5 Kg. de cobre inorgánico/Ha.por
campaña.

Answer 1

You can use sed:
sed -r 's/( *\(|\))//g;s/\./\n/g' input.txt
This assumes that your input file does not contain line breaks. If it contains line breaks the command needs to get modified a bit.
Explanation:
The first command s/( *\(|\))//g removes the parentheses and additional whitespace. The second command s/\./\n/g adds a newline after a dot.
Oh I missed that you want to add an additional -->. If you really need that, the second sed commands needs to get modified:
sed -r 's/( *\(|\))//g;s/U[^.]+\./--> \0\n/g' input.txt
Now the second command searches for the sequence U --> until a dot and prepends a --> plus adds the newline after the dot.
Output:
1,2,3,4,5,6,10,...,74 --> Use method number 1.
7,8,9,19,20,21,22,23,59,60,61 --> Use method number 2.
44,45,46,47,48,49,50 --> Use method number 3.
One another thing: The above commands adds an additional newline at the end of output. You can suppress that by adding a third sed command s/\n$// which removes the additional new line before the end of the output:
sed -r 's/( *\(|\))//g;s/U[^.]+\./--> \0\n/g;s/\n$//' input.txt

Answer 2

Quite an idiomatic gnu awk solution:
awk -v RS="Use method number [0-9]."
-v OFS=" --> "
'NF{gsub(/\s*|\(|\)/, ""); print $0, RT}' file
Test
$ awk -v RS="Use method number [0-9]." -v OFS=" --> " 'NF{gsub(/\s*|\(|\)/, ""); print $0, RT}' a
1,2,3,4,5,6,10,11,12,13,14,15,16,17,18,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,51,52,53,54,55,56,57,58,62,63,64,65,66,67,68,69,70,71,72,73,74 --> Use method number 1.
7,8,9,19,20,21,22,23,59,60,61 --> Use method number 2.
44,45,46,47,48,49,50 --> Use method number 3.
Explanation
-v RS="Use method number [0-9]." set the record separator to the string "Use method number X.`, X being a digit.
-v OFS=" --> " set the print separator.
NF{gsub(/\s*|\(|\)/, ""); print $0, RT} main code
-- NF {} if there is at least one field, proceed.
-- gsub(/\s*|\(|\)/, "") remove all spaces, ( and ) from the string.
-- print $0, RT print the replaced string together with the record separator that was used ("Use method number X."). Using RT instead of RS so that we catch the value of the specific X used in the string.
From man awk:
RT
The record terminator. Gawk sets RT to the input text that matched
the character or regular expression specified by RS.

Answer 3

you can very intuitively do it with an ed script,
:: ed.script ::
# first you split your data in multiple lines
,s/\(\(([0-9]*), \)*([0-9]*)\)/\
\1\
/g
# then for each matching line with numbers, you remove unwanted chars
# and append " --> " to the next line
,g/\(\(([0-9]*), \)*([0-9]*)\)/\
s/[)( ]//g\
a\
-->\
.
# and finally you join lines
,g/^ -->/-1,+1j
# save if you want
w
Then you launch it with the following command:
cat ed.script | ed -s file.txt
that was the part intuitive... and it works with your sample data.