Merge pull request #77 from DiamondLightSource/incorrect_organization_replacement_in_general_README

Changed the ignore functionality so individual lines in files can be ignored

Author: evvaaaa

src/python3_pip_skeleton/__main__.py

@@ -5,7 +5,7 @@
 from pathlib import Path
 from subprocess import STDOUT, CalledProcessError, call, check_output
 from tempfile import TemporaryDirectory
-from typing import List
+from typing import Callable, Dict, List, Tuple
 
 import tomli
 
@@ -20,12 +20,18 @@
 MERGE_BRANCH = "skeleton-merge-branch"
 # Extensions to change
 CHANGE_SUFFIXES = [".py", ".rst", ".cfg", "", ".toml"]
-# Files not to change
-IGNORE_FILES = [
-    "test_boilerplate_removed.py",
-    "pin-requirements.rst",
-    "update-tools.rst",
-]
+# Files not to change where IGNORE_FILES[x] is a list of tuples where substitutions
+# will be ignored in that file in any substring between the two strings.
+# An empty list will ignore the whole file.
+IGNORE_FILES: Dict[str, List[Tuple[str, str]]] = {
+    "update-tools.rst": [],
+    "test_boilerplate_removed.py": [],
+    "pin-requirements.rst": [],
+    "0002-switched-to-pip-skeleton.rst:": [],
+    "README.rst": [
+        ("adopt this skeleton project see", "that describes what your module does")
+    ],
+}
 
 SKELETON_ROOT_COMMIT = "ededf00035e6ccfac78946213009c1ecd7c110a9"
 
@@ -54,6 +60,48 @@ def __truediv__(self, other) -> Path:
         return Path(self.name) / other
 
 
+def find_ignore_sections(
+    file_name: str, file_text: str, ignore_sections: List[Tuple[str, str]]
+) -> List[re.Match]:
+    ignore_section_matches = []
+    for sub_strings in ignore_sections:
+        pre_sub_string, post_sub_string = sub_strings
+        regex = rf"(?s){pre_sub_string}(.*?){post_sub_string}"
+        # finditer so we can throw an error if the used ignore strings ignore
+        # more than once in the file
+        original_substrings = list(re.finditer(regex, file_text))
+        assert original_substrings, (
+            f"could not find substrings {pre_sub_string} or "
+            f"{post_sub_string} in {file_name}."
+        )
+        assert len(original_substrings) == 1, (
+            f"multiple substrings found between {pre_sub_string} and "
+            f"{post_sub_string} in {file_name}."
+        )
+        ignore_section_matches.append(original_substrings[0])
+
+    ignore_section_matches.sort(key=lambda x: x.start())
+    return ignore_section_matches
+
+
+def replace_text_ignoring_sections(
+    text: str,
+    ignore_section_matches: List[re.Match],
+    text_replacement_method: Callable,
+) -> str:
+    replacement_text = ""
+    next_start = 0
+    for ignore_section in ignore_section_matches:
+        replacement_text += text_replacement_method(
+            text[next_start : ignore_section.start()]
+        )
+        replacement_text += text[ignore_section.start() : ignore_section.end()]
+        next_start = ignore_section.end()
+
+    replacement_text += text_replacement_method(text[next_start : len(text)])
+    return replacement_text
+
+
 def merge_skeleton(
     path: Path,
     org: str,
@@ -106,6 +154,21 @@ def replace_in_file(file_path: Path, text_from: str, text_to: str):
             if child.suffix in CHANGE_SUFFIXES and child.name not in IGNORE_FILES:
                 text = replace_text(child.read_text())
                 child.write_text(text)
+            # Replace the file, ignoring text between specified substrings
+            elif (
+                child.suffix in CHANGE_SUFFIXES
+                and child.name in IGNORE_FILES
+                and IGNORE_FILES[child.name]
+            ):
+                original_text = child.read_text()
+                ignore_sections = find_ignore_sections(
+                    child.name, original_text, IGNORE_FILES[child.name]
+                )
+                child.write_text(
+                    replace_text_ignoring_sections(
+                        original_text, ignore_sections, replace_text
+                    )
+                )
 
         # Change instructions in the docs to reflect which pip skeleton is in use
         replace_in_file(