Member Avatar for romes87

Basically, I am trying to extract text between two strings within a loop as one of the two words changes after the information is extracted.

so for example, the string is:

string = alpha 111 bravo 222 alpha somethingA end, 333 bravo somethingB end 444 alpha 555 bravo

So I want to extract the text between alpha and end and then bravo and end. I have quite a few of these unique words in my file so I have a list and a counter to go through them. See the code below:

string = 'alpha 111 bravo 222 alpha somethingA end, 333 bravo somethingB end 444 alpha 555 bravo'

words = ['alpha', 'bravo'] #there will be more words here

counter = 0

stringOut = ''

#going through the list of words

while counter < len(words):

    firstWord = words[counter]

    lastWord = 'end'     

    data = string[string.find(firstWord)+len(firstWord):string.find(lastWord)].strip() 

    #this will give the text between the first ocurrance of "alpha" and "end"
    #since I want just the smallest string between "alpha" and "end", I use another while loop
    #to see if firstWord occurs again

    while firstWord in data:
            ignore,ignore2,data = data.partition(str(firstWord))

    counter = counter + 1

    stringOut += str(data) + str('\n')

print('output string is \n' + str(stringOut))

#this code gives the correct output for the text between the first word ("alpha") and "end".
#but when the list moves to the next string "bravo", it takes the text between the first "bravo" 
#and the "end" that was associated with the information required for "alpha" ("somethingA")

Can anyone help me with this please? Any suggestions are welcome.

Many thanks.

Edited by romes87
  • 3 Contributors
  • 6 Replies
  • 269 Views
  • 4 Days Discussion Span
  • Latest Post Latest Post by TrustyTony

Dani AI

Generated

A concise recap and a robust, practical option.

showed the situation: many different keywords in a list, each followed somewhere later by an end marker, and the goal is to extract the shortest text between a matched keyword and the next end. demonstrated that a regex can pull the correct blocks, and sketched a split-on-end approach that avoids some regex pitfalls. The main gaps to cover are safe, dynamic construction of the keyword alternation (handle regex metacharacters and partial matches), ensuring whole-word matches (so “bend” or “send” do not trigger), allowing multi-word payloads, and keeping it reasonably fast for moderate file sizes.

A straightforward, safe pattern (build alternation with escaping, prefer longer keywords first, use non-greedy capture and dot-all) extracts key + payload in order:

import re

# words is the list of keywords (provided)
alts = '|'.join(sorted((re.escape(w) for w in words), key=len, reverse=True))
pattern = re.compile(r'\b(' + alts + r')\b\s*(.*?)\s*\bend\b', re.I | re.S)

results = []
for m in pattern.finditer(text):
    results.append((m.group(1), m.group(2).strip()))

Notes and troubleshooting

  • re.escape prevents keywords with regex specials from breaking the alternation.
  • Sorting by length (longer first) avoids shorter keywords stealing matches from longer ones that contain them.
  • \b prevents matches inside words like “bend”; if punctuation rules differ, tune the boundary (or match end with (?<!\w)end(?!\w)).
  • If end can legitimately appear inside payloads, a simple pattern won’t be sufficient — either pick a more reliable delimiter or parse by scanning each end and choosing the nearest preceding keyword (the split/rfind technique from ), or use an Aho–Corasick scan for very large keyword sets.

Precompile the regex once for repeated work; for thousands of keywords or gigabyte-sized inputs, consider a linear scanner (Aho–Corasick) or the split-on-end approach for memory and speed.

Recommended Answers

All 6 Replies

Member Avatar for snippsat

For fun one with regex,but i guess this is a school task?

import re

string = 'alpha 111 bravo 222 alpha somethingA end, 333 bravo somethingB end 444 alpha 555 bravo'
pattern = re.compile(r'(alpha|bravo)(\s\w+\s)(end)')
for match in pattern.finditer(string):
    print match.group(2).strip()

"""Output-->
somethingA
somethingB
"""
Edited by snippsat
Member Avatar for TrustyTony

This code would fail with words like 'bend' or 'send' inside the data, but could give you idea.

t = "alpha 111 bravo 222 alpha somethingA end, 333 bravo somethingB end 444 alpha 555 bravo"
keywords = 'alpha', 'bravo'
if 'end' in t:
    for part in t.split('end')[:-1]:
        last, key = max((part.rfind(key)+len(key), key) for key in keywords)
        print key,':',part[last:]
Edited by TrustyTony
Member Avatar for romes87

Thank you both for your reply. snippsat. You answer works perfectly.

I managed to get around the problem by marking the index of the string where word appears

word = ['alpha', 'bravo'] #...
counter = 0

marker0 = fileString.index(word)
marker0 = marker0 + len(word)
marker0 = fileString.index(word,marker0)

But your solution looks more robust as it doesnt matter how many times the word appears in the string before the required information!

Thanks! :)

Member Avatar for romes87

But would using this line of code :

pattern = re.compile(r'(alpha|bravo)(\s\w+\s)(end)')

mean that the whole list of words need to written within this manually?
The reasom I am asking is that I have many of these unique words that I need to extract info for. So :

word = ['alpha', 'bravo', '....', '.....' , 'etc'] #could you quite a few here

Is there a way to use variable within the re.compile statement?

Member Avatar for TrustyTony 
'|'.join(words)
Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.