Extracting numbers from sentences

Question

I need to extract some numbers from a text. Text is

x <- "Lorem ipsum dolor sit amet[245], consectetur adipiscing (325). Deinde prima illa, quae in congressu[232]. solemus: Quid tu, inquit, huc? Sequitur disserendi ratio cognitioque 295. naturae;"

The numbers to be extracted are 325 and 232. These are inside brackets and at the end of sentence. Other numbers are excluded. I tried strsplit(text, "[A-Za-z]+"), but is not getting what I needed.

I'm curious on the downvote here? – Tyler Rinker Aug 24 '14 at 18:58 — Tyler Rinker, Aug 24 '14 at 18:58

score 5 · Answer 1 · edited Aug 25 '14 at 00:06

5

Here's a stringi approach

x <- "Lorem ipsum dolor sit amet[245], consectetur adipiscing (325). Deinde prima illa, quae in congressu[232]. solemus: Quid tu, inquit, huc? Sequitur disserendi ratio cognitioque 295. naturae; Claudii libidini, qui tum erat summo ne imperio, dederetur"

library(stringi)
stri_extract_all_regex(x, "(?<=[\\[(])\\d+(?=[\\])][.?!])")

## [[1]]
## [1] "325" "232"

edited Aug 25 '14 at 00:06

hwnd

69,796
4
95
132

answered Aug 24 '14 at 18:32

Tyler Rinker

108,132
65
322
519

2

Curious, doesn't `qdap` have function for getting text between brackets? I though I was you use it a few times before. – Rich Scriven Aug 24 '14 at 19:33
Yeah it has `bracketXtract` but this regex is less general (forces digits between) thus more accurate. And I'm becoming a big fan of the `stringi` package with fast, consistent results. – Tyler Rinker Aug 24 '14 at 20:15

score 4 · Answer 2 · answered Aug 24 '14 at 18:34

4

Another one:

r <- gregexpr("[[(]\\d+[])](?=\\.)", text, perl = TRUE)
(m <- regmatches(text, r)[[1]])
# [1] "(325)" "[232]"

as.integer(gsub("\\D", "", m))
# [1] 325 232

answered Aug 24 '14 at 18:34

lukeA

53,097
5
97
100

hwnd · Answer 3 · 2014-08-24T23:09:26.210

Here is a solution using strsplit....

> x <- 'Lorem ipsum dolor sit amet[245], consectetur adipiscing (325). Deinde prima illa, quae in congressu[232]. solemus: Quid tu, inquit, huc? Sequitur disserendi ratio cognitioque 295. naturae;'
> strsplit(x, '[^0-9]+')[[1]][3:4]
## [1] "325" "232"

Or using base R to extract these values.

> regmatches(x, gregexpr('[[(]\\K\\d+(?=[])](?!,))', x, perl=T))[[1]]
## [1] "325" "232"

score 0 · Answer 4 · answered Oct 30 '14 at 17:45

With re module

import re

string="Lorem ipsum dolor sit amet[245], consectetur adipiscing (325). Deinde prima illa, quae in congressu[232]. solemus: Quid tu, inquit, huc? Sequitur disserendi ratio cognitioque 295. naturae;"

print string

pattern = re.compile(r'(?<=[\[(])\d+(?=[\])]\.)')

result = pattern.findall(string)

print result

Extracting numbers from sentences

4 Answers4