Tutorial

Here is a demo to showcase the use of PyRegexBuilder.

Problem

Hidden in the text below is a Cyrillic letter that is preceded by exactly 2 Greek or Latin letters (no more, no less).

text = "ウثイا月εЗ人水جИ山γבИЖㄱتאEイЖבИBДE人ججイEㄷجاجㄷアЗاجاㄴACㄴDAエβエЖאג月水水ㄹجAㅁDبエα山הبتウИג月Aاイب山αثㄴاИδㄹ水ア人ㄹتεדβE山ㄴבㅁEאЙエㄷ山אباγجウㄹثㅁEЗИ日山イ日ثاㄷβBAㅁЖ水ㅁ山日水ㅁ人Иオבㄴγב月ت月اβアהγبβㄱبИㄱبオتㅁエ水αEتㄷAアדㄱبדDדㄱエㄹ水ثД山ㄱباイβاイ水δㄹЗㄹ月γB山ЗAアイαEИبДجЖεتИγㅁאاオㄷDЙアEεイㄹAثЖדD日山日δδDИCب月ЗבαتγウЖדبج日גBبהイㄱㅁ月月月ЖجBㄱגエבДجבㄱㅁㄱ人"

Which letter is it? Can you find it using a regular expression?

Solution

Let's see how we would go about coming up with a solution using PyRegexBuilder.

1. Import the required modules.

   from pyregexbuilder import *
   import regex as re
   

2. Create a character set that will match Greek or Latin letters.

   greek_or_latin = PosixClass("IsGreek").union(UnicodeClass("IsLatin"))
   

   See the available character classes.

3. Now let's start constructing the regular expression.

   greek_or_latin = PosixClass("IsGreek").union(UnicodeClass("IsLatin"))
   
   expression = (
       Regex(
           ...
       )
   )
   

4. Create an assertion to check that only the previous 2 characters are Greek or Latin letters.

   greek_or_latin = PosixClass("IsGreek").union(UnicodeClass("IsLatin"))
   
   expression = (
       Regex(
           PositiveLookbehind(
               ChoiceOf(Anchor.WORD_BOUNDARY, greek_or_latin.inverted),
               Repeat(greek_or_latin, count=2)
           )
       )
   )
   

   See the available assertions and quantifiers.

5. After looking behind, capture the current letter if it is part of the Cyrillic alphabet.

   greek_or_latin = PosixClass("IsGreek").union(UnicodeClass("IsLatin"))
   
   expression = (
       Regex(
           PositiveLookbehind(
               ...
           ),
           Capture(PosixClass("IsCyrillic"), name="character"),
       )
   )
   

   Groups can be named or unnamed. See the available groups.

6. Set the flags for the regular expression and compile.

   (Here, these flags are not necessary; they are just an example.)

   greek_or_latin = PosixClass("IsGreek").union(UnicodeClass("IsLatin"))
   
   expression = (
       Regex(
           ...
       )
       .with_flags({"IGNORECASE": True})
       .with_global_flags({"VERSION1": True})
       .compile()
   )
   

   See the available scoped and global flags here.

7. Now you can use the regular expression. Here is the full code:

   from pyregexbuilder import *
   import regex as re
   
   text = "ウثイا月εЗ人水جИ山γבИЖㄱتאEイЖבИBДE人ججイEㄷجاجㄷアЗاجاㄴACㄴDAエβエЖאג月水水ㄹجAㅁDبエα山הبتウИג月Aاイب山αثㄴاИδㄹ水ア人ㄹتεדβE山ㄴבㅁEאЙエㄷ山אباγجウㄹثㅁEЗИ日山イ日ثاㄷβBAㅁЖ水ㅁ山日水ㅁ人Иオבㄴγב月ت月اβアהγبβㄱبИㄱبオتㅁエ水αEتㄷAアדㄱبדDדㄱエㄹ水ثД山ㄱباイβاイ水δㄹЗㄹ月γB山ЗAアイαEИبДجЖεتИγㅁאاオㄷDЙアEεイㄹAثЖדD日山日δδDИCب月ЗבαتγウЖדبج日גBبהイㄱㅁ月月月ЖجBㄱגエבДجבㄱㅁㄱ人"
   
   greek_or_latin = PosixClass("IsGreek").union(UnicodeClass("IsLatin"))
   
   expression = (
       Regex(
           PositiveLookbehind(
               ChoiceOf(Anchor.WORD_BOUNDARY, greek_or_latin.inverted),
               Repeat(
                   greek_or_latin,
                   count=2,
               )
           ),
           Capture(PosixClass("IsCyrillic"), name="character"),
       )
       .with_flags({"IGNORECASE": True})
       .with_global_flags({"VERSION1": True})
       .compile()
   )
   
   match = re.search(expression, text)
   
   print(match)