Python で正規表現の構文木を得るには

python

regex

Python で正規表現を構文解析だけした結果（内部表現）を得る方法がないかと思って調べてみたところ、sre_parseを使えばできることが分かったのでメモ。

例えば、メールアドレスにマッチする正規表現（^[a-zA-Z0-9_+-]+(.[a-zA-Z0-9_+-]+)*@([a-zA-Z0-9][a-zA-Z0-9-]*[a-zA-Z0-9]*\.)+[a-zA-Z]{2,}$）を parse すると [(AT, AT_BEGINNING), (MAX_REPEAT, (1, MAXREPEAT, [(IN, [(RANGE, (97, 122)), (RANGE, (65, 90)), (RANGE, (48, 57)), (LITERAL, 95), (LITERAL, 43), (LITERAL, 45)])])), (MAX_REPEAT, (0, MAXREPEAT, [(SUBPATTERN, (1, 0, 0, [(ANY, None), (MAX_REPEAT, (1, MAXREPEAT, [(IN, [(RANGE, (97, 122)), (RANGE, (65, 90)), (RANGE, (48, 57)), (LITERAL, 95), (LITERAL, 43), (LITERAL, 45)])]))]))])), (LITERAL, 64), (MAX_REPEAT, (1, MAXREPEAT, [(SUBPATTERN, (2, 0, 0, [(IN, [(RANGE, (97, 122)), (RANGE, (65, 90)), (RANGE, (48, 57))]), (MAX_REPEAT, (0, MAXREPEAT, [(IN, [(RANGE, (97, 122)), (RANGE, (65, 90)), (RANGE, (48, 57)), (LITERAL, 45)])])), (MAX_REPEAT, (0, MAXREPEAT, [(IN, [(RANGE, (97, 122)), (RANGE, (65, 90)), (RANGE, (48, 57))])])), (LITERAL, 46)]))])), (MAX_REPEAT, (2, MAXREPEAT, [(IN, [(RANGE, (97, 122)), (RANGE, (65, 90))])])), (AT, AT_END)] という表現を得ることができます。

この表現を　sre_compile でコンパイルすると実際に正規表現として利用できるようになるようです。

$ python
Python 3.9.10 (main, Feb 27 2022, 15:02:07)
[GCC 4.8.5 20150623 (Red Hat 4.8.5-44)] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import sre_parse
>>> import sre_compile
>>> sre = sre_parse.parse("^[a-zA-Z0-9_+-]+(.[a-zA-Z0-9_+-]+)*@([a-zA-Z0-9][a-zA-Z0-9-]*[a-zA-Z0-9]*\.)+[a-zA-Z]{2,}$")
>>> sre
[(AT, AT_BEGINNING), (MAX_REPEAT, (1, MAXREPEAT, [(IN, [(RANGE, (97, 122)), (RANGE, (65, 90)), (RANGE, (48, 57)), (LITERAL, 95), (LITERAL, 43), (LITERAL, 45)])])), (MAX_REPEAT, (0, MAXREPEAT, [(SUBPATTERN, (1, 0, 0, [(ANY, None), (MAX_REPEAT, (1, MAXREPEAT, [(IN, [(RANGE, (97, 122)), (RANGE, (65, 90)), (RANGE, (48, 57)), (LITERAL, 95), (LITERAL, 43), (LITERAL, 45)])]))]))])), (LITERAL, 64), (MAX_REPEAT, (1, MAXREPEAT, [(SUBPATTERN, (2, 0, 0, [(IN, [(RANGE, (97, 122)), (RANGE, (65, 90)), (RANGE, (48, 57))]), (MAX_REPEAT, (0, MAXREPEAT, [(IN, [(RANGE, (97, 122)), (RANGE, (65, 90)), (RANGE, (48, 57)), (LITERAL, 45)])])), (MAX_REPEAT, (0, MAXREPEAT, [(IN, [(RANGE, (97, 122)), (RANGE, (65, 90)), (RANGE, (48, 57))])])), (LITERAL, 46)]))])), (MAX_REPEAT, (2, MAXREPEAT, [(IN, [(RANGE, (97, 122)), (RANGE, (65, 90))])])), (AT, AT_END)]
>>> comlied = sre_compile.compile(sre)
>>> comlied.match("hoge@example.com")
<re.Match object; span=(0, 16), match='hoge@example.com'>
>>> comlied.match("hoge..@example.com")
>>>

これを上手く使えば正規表現同士を自動的に合成するようなライブラリを作ったりすることができそうです。

by hsur at 23:13 [5年前][4年前][3年前][2年前][1年前] |

こんな記事もあります「Python で正規表現の構文木を得るには」