some thing wrong with regex.search

[jxxcr]

i just run the template at readme.md, and regex.search is not work.
Version:
python 3.11
regex 2023.8.8
click 8.1.3
numpy 1.24.3

TypeError                                 Traceback (most recent call last)
Cell In[8], line 2
      1 with open('out.out', 'r') as f:
----> 2     for match in parse_iter(str(f.read())):
      3         print(match.values)
      4     #print(f.read())

File C:\Program Files\Python311\Lib\site-packages\cp2k_output_tools\parser.py:33, in parse_iter(content, matchers, key_mangling)
     31 def parse_iter(content: str, matchers: List[Callable] = builtin_matchers, key_mangling: bool = False) -> Iterator[Dict[str, Any]]:
     32     """Yields the structured data found in the block matches"""
---> 33     for block in parse_iter_blocks(content, matchers, key_mangling):
     34         yield block.data

File C:\Program Files\Python311\Lib\site-packages\cp2k_output_tools\parser.py:22, in parse_iter_blocks(content, matchers, key_mangling)
     20 """Yield BlockMatch objects containing both structured data and metadata for each found match"""
     21 for matcher in matchers:
---> 22     match = matcher(content)
     24     if match:
     25         if key_mangling:

File C:\Program Files\Python311\Lib\site-packages\cp2k_output_tools\blocks\program_info.py:75, in match_program_info(content, start, end, as_tree_obj)
     71 def match_program_info(
     72     content: str, start: int = 0, end: int = sys.maxsize, as_tree_obj: bool = False
     73 ) -> Optional[Union[BlockMatch, ProgramInfo]]:
     74     spans = []
---> 75     match = PROGRAM_INFO_START_RE.search(content, start, end)
     77     if not match:
     78         return None

TypeError: string indices must be integers

[dev-zero]

sorry for the late reply: is there a chance you could attach the file you are trying to parse? The first ~20 lines should be sufficient I think.

[dev-zero]

In case you did also use the file to parse from the README.md, I have the suspicion there is something different for Python on Windows. Can you try to add the following line after line 74 in the C:\Program Files\Python311\Lib\site-packages\cp2k_output_tools\blocks\program_info.py file, try again and send me the output of that line:

    print("end:", end, type(end))

[dev-zero]

Or run this:

import sys
print("maxsize:", sys.maxsize, type(sys.maxsize))
print("hello, there"[0:sys.maxsize])

[jxxcr]

Sorry about my information is not detail, print("end:", end, type(end))end: 9223372036854775807 <class 'int'&gt;--------------------------------- import sys print("maxsize:", sys.maxsize, type(sys.maxsize)) print("hello, there"[0:sys.maxsize])maxsize: 9223372036854775807 <class 'int'&gt; hello, there

…

------------------ 原始邮件 ------------------ 发件人: "cp2k/cp2k-output-tools" ***@***.***&gt;; 发送时间:&nbsp;2023年12月10日(星期天) 凌晨4:42 ***@***.***&gt;; ***@***.******@***.***&gt;; 主题:&nbsp;Re: [cp2k/cp2k-output-tools] some thing wrong with regex.search (Issue #34) sorry for the late reply: is there a chance you could attach the file you are trying to parse? The first ~20 lines should be sufficient I think. — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***&gt;

[dev-zero]

Nevermind, this already helps. My suspicion was that for some reason sys.maxsize is different on Windows, but it does not seem to be the case.

Could you try the following as well?

import sys
import regex
print(regex.compile(r"b+").search("aaaabbbbbbcccc", 0, sys.maxsize))

[jxxcr]

1 import sys 2 import regex ----&gt; 3 print(regex.compile(r"b+").search("aaaabbbbbbcccc", 0, sys.maxsize)) TypeError: string indices must be integers

…

------------------&nbsp;原始邮件&nbsp;------------------ 发件人: "cp2k/cp2k-output-tools" ***@***.***&gt;; 发送时间:&nbsp;2023年12月11日(星期一) 凌晨0:10 ***@***.***&gt;; ***@***.******@***.***&gt;; 主题:&nbsp;Re: [cp2k/cp2k-output-tools] some thing wrong with regex.search (Issue #34) Nevermind, this already helps. My suspicion was that for some reason sys.maxsize is different on Windows, but it does not seem to be the case. Could you try the following as well? import sys import regex print(regex.compile(r"b+").search("aaaabbbbbbcccc", 0, sys.maxsize)) — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***&gt;

[dev-zero]

And does it work with this?

import sys
import re
print(re.compile(r"b+").search("aaaabbbbbbcccc", 0, sys.maxsize))

I am trying to figure out whether it's a general issue with the re package, or a bug in the regex one.

[jxxcr]

I found the template in README.md is working normally in Linux. And the tools is also not commonly in Windows.