Optimize string concatenation across multiple files (#288)

Author: mohasarc

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "mkdocs-include-markdown-plugin"
-version = "7.2.0"
+version = "7.2.1"
 description = "Mkdocs Markdown includer plugin."
 readme = "README.md"
 license = "Apache-2.0"

src/mkdocs_include_markdown_plugin/directive.py

@@ -140,7 +140,7 @@ def _maybe_arguments_iter(arguments_string: str) -> Iterable[str]:
     inside_string = False
     escaping = False
     opening_argument = False  # whether we are at the beginning of an argument
-    current_value = ''
+    current_value: list[str] = []
 
     for c in arguments_string:
         if inside_string:
@@ -153,24 +153,25 @@ def _maybe_arguments_iter(arguments_string: str) -> Iterable[str]:
             else:
                 escaping = False
         elif c == '=':
+            current_value_str = ''.join(current_value)
             new_current_value = ''
-            for ch in reversed(current_value):
+            for ch in reversed(current_value_str):
                 if ch in string.whitespace:
-                    current_value = new_current_value[::-1]
+                    current_value_str = new_current_value[::-1]
                     break
                 new_current_value += ch
-            yield current_value
-            current_value = ''
+            yield current_value_str
+            current_value = []
             opening_argument = True
         elif opening_argument:
             opening_argument = False
             if c in ('"', "'"):
                 current_string_opening = c
                 inside_string = True
-                current_value += c
-                current_value += c
+                current_value.append(c)
+                current_value.append(c)
         else:
-            current_value += c
+            current_value.append(c)
 
 
 def warn_invalid_directive_arguments(

src/mkdocs_include_markdown_plugin/event.py

@@ -545,7 +545,7 @@ def found_include_markdown_tag(  # noqa: PLR0912, PLR0915
         # but they have been specified, so the warning(s) must be raised
         expected_but_any_found = [start is not None, end is not None]
 
-        text_to_include = ''
+        text_to_include_parts = []
         for file_path in file_paths_to_include:
             if process.is_url(filename):
                 new_text_to_include = process.read_url(
@@ -621,19 +621,20 @@ def found_include_markdown_tag(  # noqa: PLR0912, PLR0915
                 new_text_to_include
             ):
                 lines = new_text_to_include.splitlines(keepends=True)
-                new_text_to_include = lines[0]
+                indented_lines = [lines[0]]
                 for i in range(1, len(lines)):
-                    new_text_to_include += (
-                        filled_includer_indent + lines[i]
-                    )
+                    indented_lines.append(filled_includer_indent + lines[i])
+                new_text_to_include = ''.join(indented_lines)
 
             if offset:
                 new_text_to_include = process.increase_headings_offset(
                     new_text_to_include,
                     offset=offset + cumulative_heading_offset,
                 )
 
-            text_to_include += new_text_to_include
+            text_to_include_parts.append(new_text_to_include)
+
+        text_to_include = ''.join(text_to_include_parts)
 
         # warn if expected start or ends haven't been found in included content
         for i, delimiter_name in enumerate(['start', 'end']):

src/mkdocs_include_markdown_plugin/process.py

@@ -174,10 +174,12 @@ def transform_p_by_p_skipping_codeblocks(  # noqa: PLR0912, PLR0915
     _maybe_icodeblock_lines: list[str] = []
     _previous_line_was_empty = False
 
-    lines, current_paragraph = ([], '')
+    lines = []
+    current_paragraph_lines: list[str] = []
 
     def process_current_paragraph() -> None:
-        lines.extend(func(current_paragraph).splitlines(keepends=True))
+        lines.extend(func(''.join(current_paragraph_lines),
+                          ).splitlines(keepends=True))
 
     # The next implementation takes into account that indented code
     # blocks must be surrounded by newlines as per the CommonMark
@@ -191,33 +193,32 @@ def process_current_paragraph() -> None:
             if lstripped_line.startswith(('```', '~~~')):
                 _current_fcodeblock_delimiter = lstripped_line[:3]
                 process_current_paragraph()
-                current_paragraph = ''
+                current_paragraph_lines = []
                 lines.append(line)
             elif line.startswith('    '):
                 if not lstripped_line or _maybe_icodeblock_lines:
                     # maybe enter indented codeblock
                     _maybe_icodeblock_lines.append(line)
                 else:
-                    current_paragraph += line
+                    current_paragraph_lines.append(line)
             elif _maybe_icodeblock_lines:
                 process_current_paragraph()
-                current_paragraph = ''
+                current_paragraph_lines = []
                 if not _previous_line_was_empty:
                     # wasn't an indented code block
-                    for line_ in _maybe_icodeblock_lines:
-                        current_paragraph += line_
+                    current_paragraph_lines.extend(_maybe_icodeblock_lines)
                     _maybe_icodeblock_lines = []
-                    current_paragraph += line
+                    current_paragraph_lines.append(line)
                     process_current_paragraph()
-                    current_paragraph = ''
+                    current_paragraph_lines = []
                 else:
                     # exit indented codeblock
                     for line_ in _maybe_icodeblock_lines:
                         lines.append(line_)
                     _maybe_icodeblock_lines = []
                     lines.append(line)
             else:
-                current_paragraph += line
+                current_paragraph_lines.append(line)
             _previous_line_was_empty = not lstripped_line
         else:
             lines.append(line)
@@ -230,14 +231,13 @@ def process_current_paragraph() -> None:
         if not _previous_line_was_empty:
             # at EOF
             process_current_paragraph()
-            current_paragraph = ''
-            for line_ in _maybe_icodeblock_lines:
-                current_paragraph += line_
+            current_paragraph_lines = []
+            current_paragraph_lines.extend(_maybe_icodeblock_lines)
             process_current_paragraph()
-            current_paragraph = ''
+            current_paragraph_lines = []
         else:
             process_current_paragraph()
-            current_paragraph = ''
+            current_paragraph_lines = []
             for line_ in _maybe_icodeblock_lines:
                 lines.append(line_)
     else: