Regex - Picking text up to a given string

Question

I would like to take the text up to the characters a) and if possible and separate responses also using Regex?

pergunta pergunta pergunta pergunta pergunta pergunta pergunta pergunta pergunta pergunta pergunta pergunta pergunta pergunta pergunta<br /><br />

pergunta pergunta pergunta pergunta pergunta pergunta pergunta pergunta pergunta pergunta pergunta pergunta pergunta pergunta pergunta<br /><br />

  a) resposta a<br />
  b) resposta b<br />
  c) resposta c<br />
  d) resposta d<br />
  e) resposta e<br />

I’m getting the answers with this simple rule http://regexr.com/3hb4p but the question is difficult. Remembering that the question can have several paragraphs.

Answer 1

Marrying question and answers with an expression

If you really want to get married anything up to a letter followed by a letter ), or the end of the string, you can use this Regexp:   _^(regex.)

([\s\S]+?(?=\b[a-z][)]|$))

• [\s\S] is a way to match all characters including line breaks.
  _{Normally, we’d use the flag singleline to change point behavior, but does not exist in ASP.}

• (?=…) is a Lookahead (gives a "peek" ahead), ensuring it was followed by a letter and a ), or at the end of the string. But it has the peculiarity of matching the pattern, while it is not part of the married fragment.

• \b is a word edge.

However, this expression is not very efficient, especially with long texts, and can give a false positive with cases like: "pergunta (veja o segmento b) pergunta".

Bearing in mind that every answer is preceded by (at least) a line break, can use it as an additional condition. This way, we can combine whole lines, provided that one line internal don’t start with [a-z][)].

Regex:   _^(regex101)

[^\r\n]+(?:\r?\n(?!\s*[a-z][)])[^\r\n]*)*

Explanation:

• [^\r\n]+ A whole line.
• (?:\n(?!\s*[a-z][)]).*)* No capture group, to repeat this sub-pattern (0-infinite):
  • \r?\n A break in line.
  • (?!\s*[a-z][)]) Negative Lookahead, ensuring it is not followed by a letter and a ) (with possible spaces from the beginning of the line to the letter).
  • [^\r\n]* A whole line.

Since you are using ASP:

Dim texto
texto = "Os Embargos de Terceiros fazem parte do procedimento especial, previsto no Código de Processo Civil," & _
        " sendo possível sua utilização por quem, não sendo parte no processo, sofre constrição ou sofre" & _
        " ameaça de constrição sobre bens que possua ou sobre os quais tenha direito incompatível com o ato constritivo." & _ 
        " Sobre o ajuizamento dos embargos, assinale a alternativa INCORRETA." & vbNewline & _
        "" & vbNewline & _
        "Considere o segmento “[...] o Estado só percebe o eco enfraquecido.†(2º§). Pode-se afirmar que" & _
        " a partir do recurso de linguagem utilizado pelo enunciador na escolha da palavra “Estadoâ€, identifica-se 7" & vbNewline & _
        "" & vbNewline & _
        "a) o estabelecimento de uma comparação entre “Estado†e “governantesâ€." & vbNewline & _
        "b) o emprego de uma palavra redundante objetivando reforçar a ideia expressa." & vbNewline & _
        " questões que irão ter várias quebra de linhas" & vbNewline & _
        "" & vbNewline & _
        "" & vbNewline & _
        " questões que irão ter várias quebra de linhas" & vbNewline & _
        "c) uma transferência de percepções resultando em uma fusão de impressões sensoriais." & vbNewline & _
        "d) a evocação de um termo em lugar de uma palavra, com a qual se acha relacionada não sendo sinônimos. "

Set re = New RegExp
re.Global = true  'casar todas as coincidências
re.Pattern = "[^\r\n]+(?:\r?\n(?!\s*[a-z][)])[^\r\n]*)*"

'corresponder com a regex
Set matches = re.Execute(texto)
If (matches.Count) Then
    'a primeira é a pergunta <- matches(0)
    Response.Write("Pergunta === " & matches(0))

    'o resto são as respostas <- matches(m)
    For m = 1 To matches.Count - 1
        Response.Write(vbNewline & "Resposta " & m & " === ")
        Response.Write(matches(m))
    Next
End If

Set matches = Nothing
Set re = Nothing

Upshot:

Pergunta === Os Embargos de Terceiros fazem parte do procedimento especial, previsto no Código de Processo Civil, sendo possível sua utilização por quem, não sendo parte no processo, sofre constrição ou sofre ameaça de constrição sobre bens que possua ou sobre os quais tenha direito incompatível com o ato constritivo. Sobre o ajuizamento dos embargos, assinale a alternativa INCORRETA.

Considere o segmento “[...] o Estado só percebe o eco enfraquecido.†(2º§). Pode-se afirmar que a partir do recurso de linguagem utilizado pelo enunciador na escolha da palavra “Estadoâ€, identifica-se 7
Resposta 1 === a) o estabelecimento de uma comparação entre “Estado†e “governantesâ€.
Resposta 2 === b) o emprego de uma palavra redundante objetivando reforçar a ideia expressa.
 questões que irão ter várias quebra de linhas


 questões que irão ter várias quebra de linhas
Resposta 3 === c) uma transferência de percepções resultando em uma fusão de impressões sensoriais.
Resposta 4 === d) a evocação de um termo em lugar de uma palavra, com a qual se acha relacionada não sendo sinônimos. 

I clicked on a free lodging if you want to test it:
http://mariano.somee.com/258904/index.asp

Answer 2

You can use /(.*?a\))/s to obtain all characters including to)

You can use /(.*?)a\)/s for all characters prior to to)

Explanation:

.*? Identifies all characters

a Identifies a a literally (differentiates capital from lowercase)

\( Identifies the first parentheses

Answer 3

Use that regex:

((.|\n)*?)(a\).*?)\n*?(b\).*?)\n*?(c\).*?)\n*?(d\).*?)\n*?(e\).*?)$|\n*?

It will separate the text into groups where:

• Group 1 - Contains text before option a).
• Group 2 - Captures nothing, but encapsulates the group 1 capture options.
• Group 3 - Contains the contents of the option a up to the line break (where you would start option b in your example).
• Group 4 - Contains the contents of the option b until the line breaks.
• Group 5 - Contains the contents of the option c until the line breaks.
• Group 6 - Contains the contents of the option d until the line breaks.
• Group 7 - Contains the contents of the option e up to the line break or end of the text.

You can see how this regex works here

Explanation of the regex

((.|\n)*?) - It will capture any character and line break of the content until the first delimiter arrives.

(a\).*?)\n*? - a\) is equal to a) and it will be used as a delimiter, so that the first capture group stops capturing in the first occurrence of the sequence a), after that the regex will capture all the content until the first line break.

(b\).*?)\n*? - The operation of that catch group shall be equal to that of group 3, only that it is taken from the b).

(c\).*?)\n*? - The operation of that catch group shall be equal to that of group 3, only that it is taken from the c).

(d\).*?)\n*? - The operation of that catch group shall be equal to that of group 3, only that it is taken from the d).

(e\).*?)\$|\n*? - The operation of that catch group shall be equal to that of group 3, only that it is taken from the e)and stops at the end of the text or in a line break, in case you use this regex in a file that has many questions and texts.

Answer 4

If you are using a library with more resources like PCRE.

Could use :

/(?(?=\s+)|(?(?=\w\))(?<a>\w\).*)|(?<q>.*?\n)))/g

See working on REGEX101

Here I am using "conditional" and "nominal groups"

Explanation

• (?(?=\s+)|...) - This parole basically says you should ignore "spaces", because if you find the first condition is "do nothing".
• (?(?=\w\)) - Here in fact we have the condition for questions and answers, because if you find [a-z0-9_] followed by ) is answer, if not, is a question.

Note

• I used \w to facilitate, the correct I believe would be [[:alpha:]], or if you use the modifier i simplifies to [a-z]

Problems

• This Regex ends up generating a lot of garbage by not capturing the first condition.