added is_python field for code blocks; added comment explaining regex duplication

Author: SamuelRoettgermann

bot/exts/info/codeblock/_instructions.py

@@ -44,7 +44,7 @@ def _get_bad_ticks_message(code_block: _parsing.CodeBlock) -> str | None:
     log.trace("Check if the bad ticks code block also has issues with the language specifier.")
     addition_msg = _get_bad_lang_message(code_block.content)
     if not addition_msg and not code_block.language:
-        addition_msg = _get_no_lang_message(code_block.content)
+        addition_msg = _get_no_lang_message(code_block)
 
     # Combine the back ticks message with the language specifier message. The latter will
     # already have an example code block.
@@ -112,15 +112,15 @@ def _get_bad_lang_message(content: str) -> str | None:
     return None
 
 
-def _get_no_lang_message(content: str) -> str | None:
+def _get_no_lang_message(code_block: _parsing.CodeBlock) -> str | None:
     """
     Return instructions on specifying a language for a code block.
 
     If `content` is not valid Python or Python REPL code, return None.
     """
     log.trace("Creating instructions for a missing language.")
 
-    if _parsing.is_python_code(content):
+    if code_block.is_python:
         example_blocks = _get_example("py")
 
         # Note that _get_bad_ticks_message expects the first line to have two newlines.
@@ -160,6 +160,6 @@ def get_instructions(content: str) -> str | None:
             # Check for a bad language first to avoid parsing content into an AST.
             instructions = _get_bad_lang_message(block.content)
             if not instructions:
-                instructions = _get_no_lang_message(block.content)
+                instructions = _get_no_lang_message(block)
 
     return instructions

bot/exts/info/codeblock/_parsing.py

@@ -47,7 +47,8 @@
     """,
     re.DOTALL | re.VERBOSE
 )
-_RE_CODE_BLOCK_REGEX = regex.compile(_RE_CODE_BLOCK.pattern, regex.DOTALL | regex.VERBOSE)
+# copy of _RE_CODE_BLOCK. Done like this for highlighting reasons (regex.compile doesn't properly highlight)
+_REGEX_CODE_BLOCK = regex.compile(_RE_CODE_BLOCK.pattern, regex.DOTALL | regex.VERBOSE)
 
 _RE_LANGUAGE = re.compile(
     fr"""
@@ -66,6 +67,7 @@ class CodeBlock(NamedTuple):
     language: str
     ticks: str
     tick: str
+    is_python: bool
 
 
 class BadLanguage(NamedTuple):
@@ -89,7 +91,7 @@ def find_faulty_code_blocks(message: str) -> Sequence[CodeBlock] | None:
     log.trace("Finding all code blocks in a message.")
 
     code_blocks = []
-    for match in _RE_CODE_BLOCK_REGEX.finditer(message, overlapped=True):
+    for match in _REGEX_CODE_BLOCK.finditer(message, overlapped=True):
         # Used to ensure non-matched groups have an empty string as the default value.
         groups = match.groupdict("")
         language = groups["lang"].strip()  # Strip the whitespace cause it's included in the group.
@@ -102,11 +104,12 @@ def find_faulty_code_blocks(message: str) -> Sequence[CodeBlock] | None:
             log.trace("Skipped a code block shorter than 4 lines.")
             continue
 
+        is_python = is_python_code(groups["code"])
         if (groups["tick"] == BACKTICK
-                or (language in PY_LANG_CODES and is_python_code(groups["code"]))
+                or (language in PY_LANG_CODES and is_python)
                 or len(groups["ticks"]) >= 2):
             log.trace("Message has an invalid code block.")
-            code_block = CodeBlock(groups["code"], language, groups["ticks"], groups["tick"])
+            code_block = CodeBlock(groups["code"], language, groups["ticks"], groups["tick"], is_python)
             code_blocks.append(code_block)
         else:
             log.trace("Skipped invalid code block due to uncertainty if it is supposed to be a code block.")