Wide a dataframe and insert missing columns

If there is a precio for every item, we could use the code below, which count how many precio entries have occurred in the prior rows. lag(column, default = "") is used to avoid having "NA" for the first item, which would break the cumulative count.

df %>%
  mutate(row = cumsum(lag(column, default = "") == "precio")) %>% 
  pivot_wider(names_from = column, values_from = text)


# A tibble: 4 x 6
    row promo                    marca    descripción                                       cantidad precio 
  <int> <chr>                    <chr>    <chr>                                             <chr>    <chr>  
1     0 2 X$39.990Ahorro:$14.990 Pampers  Pañales Desechables Premium Care XXG 112 Unidades 112 Un   $27.490
2     1 2 X$39.990Ahorro:$14.990 Pampers  Pañales Desechables Premium Care XG 112 Unidades  112 Un   $27.490
3     2 NA                       Babysec  Toalla Húmeda Premium X 140/6                     1 Un     $2.590 
4     3 NA                       Emuwipes Toallitas Húmedas Premium 85 Un c/u Bolsa         2 Un     $2.650 

EDIT: Alternate approach using marca, but where promo is assigned to the next grouping (since it is assumed to always precede marca whenever it appears). Same output for the sample data.

df %>%
  mutate(row = cumsum(column == "marca") + if_else(column == "promo", 1, 0)) %>%
  pivot_wider(names_from = column, values_from = text)

Source data

df <- data.frame(
  stringsAsFactors = FALSE,
                    text = c("2 X$39.990Ahorro:$14.990","Pampers",
                             "Pañales Desechables Premium Care XXG 112 Unidades","112 Un","$27.490",
                             "2 X$39.990Ahorro:$14.990","Pampers",
                             "Pañales Desechables Premium Care XG 112 Unidades","112 Un","$27.490",
                             "Babysec","Toalla Húmeda Premium X 140/6",
                             "1 Un","$2.590","Emuwipes",
                             "Toallitas Húmedas Premium 85 Un c/u Bolsa","2 Un","$2.650"),
                  column = c("promo","marca",
                             "descripción","cantidad","precio","promo",
                             "marca","descripción","cantidad","precio","marca",
                             "descripción","cantidad","precio","marca",
                             "descripción","cantidad","precio")
      )

Using dcast from data.table. Convert the 'data.frame' to 'data.table' (setDT), create a formula with cumulative sum (based on taking the lag of 'column' (shift), check if the values are equal to 'precio', get the cumulative sum (cumsum)) with the 'column', and specify the value.var as 'text' in dcast to reshape from 'long' to 'wide' format

library(data.table)
 dcast(setDT(df), cumsum(shift(column, fill = "") == "precio" )~ column, value.var = 'text')

Update: to concerns raised by OP and others with new provided data and with the knowledge gained here: Recognize a given pattern in a vector and add the lacking elements to get the repitition of the given pattern

library(tidyverse)
df1 %>% 
    group_by(column) %>% 
    mutate(row = row_number()) %>%
    pivot_wider(
        names_from = column,
        values_from = text
    )

Gives:

     row promo                    marca           descripción                                                                    cantidad precio 
   <int> <chr>                    <chr>           <chr>                                                                          <chr>    <chr>  
 1     1 2 X$39.990Ahorro:$14.990 Pampers         Pañales Desechables Premium Care XXG 112 Unidades                              112 Un   $27.490
 2     2 2 X$39.990Ahorro:$14.990 Pampers         Pañales Desechables Premium Care XG 112 Unidades                               112 Un   $27.490
 3     3 3 X$45.990Ahorro:$13.980 Babysec         Toalla Húmeda Premium X 140/6                                                  1 Un     $2.590 
 4     4 2 X$39.990Ahorro:$14.990 Emuwipes        Toallitas Húmedas Premium 85 Un c/u Bolsa                                      2 Un     $2.650 
 5     5 NA                       Parent's Choice Toallitas Húmedas Ultra Soft con Aceite de Emu 160 Un                          160 Un   $2.550 
 6     6 NA                       Emuwipes        Toallitas Húmedas sin Alcohol (2 Bolsas de 80 Un c/u) Bolsa 2 Un               1 Un     $1.990 
 7     7 NA                       Babysec         Pañal Super Premium XXG68                                                      160 Un   $19.990
 8     8 NA                       Parent's Choice Toallitas Húmedas Ultra Soft con Aceite de Emu y sin Aroma (2 Un de 80 Un c/u) 124 Un   $2.550 
 9     9 NA                       Pampers         Pañales Desechables Premium Care G 124 Unidades                                NA       $27.490
10    10 NA                       Huggies         NA                                                                             NA       NA     

First answer:

We could create a unique identifier row for each column and then use pivot_wider

library(tidyverse)
df %>% 
    group_by(column) %>% 
    mutate(row = row_number()) %>%
    pivot_wider(
        names_from = column,
        values_from = text
    ) 

    row promo                    marca    descripción                                       cantidad precio 
  <int> <chr>                    <chr>    <chr>                                             <chr>    <chr>  
1     1 2 X$39.990Ahorro:$14.990 Pampers  Pañales Desechables Premium Care XXG 112 Unidades 112 Un   $27.490
2     2 2 X$39.990Ahorro:$14.990 Pampers  Pañales Desechables Premium Care XG 112 Unidades  112 Un   $27.490
3     3 NA                       Babysec  Toalla Húmeda Premium X 140/6                     1 Un     $2.590 
4     4 NA                       Emuwipes Toallitas Húmedas Premium 85 Un c/u Bolsa         2 Un     $2.650