Add script to keep .editorconfig in sync with .gitattributes

Our repo already contained an .editorconfig file, but it was not kept
up to date with .gitattributes.  This adds a script that keeps these
files in sync.  A big advantage of the editorconfig file is that it
many editors/IDEs get automatically configured to trim trailing
newlines and add a final newline on save, while .gitattributes only
complains about these problems instead of automatically fixing them.

This also adds rules to .gitattributes for Python files as well as for
C files in pg_bsd_indent directory (which have a different tab_width
than most C files due to being vendored in).

Author: Jelte Fennema-Nio <github-tech@jeltef.nl>
Discussion: https://www.postgresql.org/message-id/flat/CAGECzQQGzbroAXi+Yicp3HvcCo4=g84kaOgjuvQ5MW9F0ubOGg@mail.gmail.com

.editorconfig

@@ -1,14 +1,176 @@
 root = true
 
-[*.{c,h,l,y,pl,pm}]
-indent_style = tab
+[*]
 indent_size = tab
+
+[*]
+trim_trailing_whitespace = true
+insert_final_newline = true
+indent_style = unset
+tab_width = unset
+
+[*.[chly]]
+trim_trailing_whitespace = true
+insert_final_newline = true
+indent_style = tab
+tab_width = 4
+
+[*.cpp]
+trim_trailing_whitespace = true
+insert_final_newline = true
+indent_style = tab
+tab_width = 4
+
+[*.pl]
+trim_trailing_whitespace = true
+insert_final_newline = true
+indent_style = tab
 tab_width = 4
 
-[*.{sgml,xml}]
+[*.pm]
+trim_trailing_whitespace = true
+insert_final_newline = true
+indent_style = tab
+tab_width = 4
+
+[*.po]
+trim_trailing_whitespace = true
+insert_final_newline = unset
+indent_style = space
+tab_width = unset
+
+[*.py]
+trim_trailing_whitespace = true
+insert_final_newline = true
 indent_style = space
+tab_width = unset
+indent_size = 4
+
+[*.sgml]
+trim_trailing_whitespace = true
+insert_final_newline = true
+indent_style = space
+tab_width = unset
 indent_size = 1
 
-[*.xsl]
+[*.xml]
+trim_trailing_whitespace = true
+insert_final_newline = true
 indent_style = space
+tab_width = unset
 indent_size = 2
+
+[*.xsl]
+trim_trailing_whitespace = true
+insert_final_newline = true
+indent_style = space
+tab_width = unset
+indent_size = 1
+
+[*.data]
+indent_style = unset
+indent_size = unset
+trim_trailing_whitespace = unset
+insert_final_newline = unset
+
+[contrib/pgcrypto/sql/pgp-armor.sql]
+trim_trailing_whitespace = unset
+insert_final_newline = true
+indent_style = unset
+tab_width = unset
+
+[src/backend/catalog/sql_features.txt]
+trim_trailing_whitespace = unset
+insert_final_newline = true
+indent_style = unset
+tab_width = unset
+
+[*.out]
+indent_style = unset
+indent_size = unset
+trim_trailing_whitespace = unset
+insert_final_newline = unset
+
+[src/interfaces/ecpg/test/expected/*]
+indent_style = unset
+indent_size = unset
+trim_trailing_whitespace = unset
+insert_final_newline = unset
+
+[configure]
+indent_style = unset
+indent_size = unset
+trim_trailing_whitespace = unset
+insert_final_newline = unset
+
+[ppport.h]
+indent_style = unset
+indent_size = unset
+trim_trailing_whitespace = unset
+insert_final_newline = unset
+
+[src/backend/jit/llvm/SectionMemoryManager.cpp]
+indent_style = unset
+indent_size = unset
+trim_trailing_whitespace = unset
+insert_final_newline = unset
+
+[src/backend/jit/llvm/SectionMemoryManager.LICENSE]
+indent_style = unset
+indent_size = unset
+trim_trailing_whitespace = unset
+insert_final_newline = unset
+
+[src/backend/regex/COPYRIGHT]
+indent_style = unset
+indent_size = unset
+trim_trailing_whitespace = unset
+insert_final_newline = unset
+
+[src/backend/snowball/libstemmer/*.c]
+indent_style = unset
+indent_size = unset
+trim_trailing_whitespace = unset
+insert_final_newline = unset
+
+[src/backend/utils/mb/Unicode/*-std.txt]
+indent_style = unset
+indent_size = unset
+trim_trailing_whitespace = unset
+insert_final_newline = unset
+
+[src/include/jit/SectionMemoryManager.h]
+indent_style = unset
+indent_size = unset
+trim_trailing_whitespace = unset
+insert_final_newline = unset
+
+[src/include/snowball/libstemmer/*]
+indent_style = unset
+indent_size = unset
+trim_trailing_whitespace = unset
+insert_final_newline = unset
+
+[src/timezone/data/*]
+indent_style = unset
+indent_size = unset
+trim_trailing_whitespace = unset
+insert_final_newline = unset
+
+[src/tools/pg_bsd_indent/*]
+indent_style = unset
+indent_size = unset
+trim_trailing_whitespace = unset
+insert_final_newline = unset
+
+[src/tools/pg_bsd_indent/tests/*]
+indent_style = unset
+indent_size = unset
+trim_trailing_whitespace = unset
+insert_final_newline = unset
+
+[src/tools/pg_bsd_indent/*.[ch]]
+trim_trailing_whitespace = unset
+insert_final_newline = unset
+indent_style = unset
+tab_width = 8

.gitattributes

@@ -1,11 +1,15 @@
+# IMPORTANT: After updating this file, also run src/tools/generate_editorconfig.py
+
 *		whitespace=space-before-tab,trailing-space
 *.[chly]	whitespace=space-before-tab,trailing-space,indent-with-non-tab,tabwidth=4
 *.cpp		whitespace=space-before-tab,trailing-space,indent-with-non-tab,tabwidth=4
 *.pl		whitespace=space-before-tab,trailing-space,tabwidth=4
 *.pm		whitespace=space-before-tab,trailing-space,tabwidth=4
 *.po		whitespace=space-before-tab,trailing-space,tab-in-indent,-blank-at-eof
+*.py		whitespace=space-before-tab,trailing-space,tab-in-indent
 *.sgml		whitespace=space-before-tab,trailing-space,tab-in-indent
-*.x[ms]l	whitespace=space-before-tab,trailing-space,tab-in-indent
+*.xml		whitespace=space-before-tab,trailing-space,tab-in-indent
+*.xsl		whitespace=space-before-tab,trailing-space,tab-in-indent
 
 # Avoid confusing ASCII underlines with leftover merge conflict markers
 README		conflict-marker-size=32
@@ -33,3 +37,4 @@ src/include/snowball/libstemmer/*	-whitespace
 src/timezone/data/*			-whitespace
 src/tools/pg_bsd_indent/*		-whitespace
 src/tools/pg_bsd_indent/tests/*		-whitespace
+src/tools/pg_bsd_indent/*.[ch]		whitespace=-blank-at-eol,-blank-at-eof,tabwidth=8

src/tools/generate_editorconfig.py

@@ -0,0 +1,94 @@
+#!/usr/bin/env python3
+
+import os
+
+
+def cd_to_repo_root():
+    abspath = os.path.abspath(__file__)
+    dname = os.path.join(os.path.dirname(abspath), "..", "..")
+    os.chdir(dname)
+
+
+# Space based indentation levels are not tracked in .gitattributes, so
+# we hardcode them here for the relevant filetypes.
+space_based_indent_sizes = {
+    "*.py": 4,
+    "*.sgml": 1,
+    "*.xsl": 1,
+    "*.xml": 2,
+}
+
+
+def main():
+    cd_to_repo_root()
+
+    with open(".gitattributes", "r") as f:
+        lines = f.read().splitlines()
+
+    new_contents = """root = true
+
+[*]
+indent_size = tab
+"""
+
+    for line in lines:
+        if line.startswith("#") or len(line) == 0:
+            continue
+        name, git_rules = line.split()
+        if git_rules == "-whitespace":
+            rules = [
+                "indent_style = unset",
+                "indent_size = unset",
+                "trim_trailing_whitespace = unset",
+                "insert_final_newline = unset",
+            ]
+        elif git_rules.startswith("whitespace="):
+            git_whitespace_rules = git_rules.replace("whitespace=", "").split(",")
+            rules = []
+            if "-blank-at-eol" in git_whitespace_rules:
+                rules += ["trim_trailing_whitespace = unset"]
+            else:
+                rules += ["trim_trailing_whitespace = true"]
+
+            if "-blank-at-eof" in git_whitespace_rules:
+                rules += ["insert_final_newline = unset"]
+            else:
+                rules += ["insert_final_newline = true"]
+
+            if "tab-in-indent" in git_whitespace_rules:
+                rules += ["indent_style = space"]
+            elif "indent-with-non-tab" in git_whitespace_rules:
+                rules += ["indent_style = tab"]
+            elif name in ["*.pl", "*.pm"]:
+                # We want editors to use tabs for indenting Perl
+                # files, but we cannot add it such a rule to
+                # .gitattributes, because certain lines are still
+                # indented with spaces (e.g. SYNOPSIS blocks).  So we
+                # hardcode the rule here.
+
+                rules += ["indent_style = tab"]
+            else:
+                rules += ["indent_style = unset"]
+
+            tab_width = "unset"
+            for rule in git_whitespace_rules:
+                if rule.startswith("tabwidth="):
+                    tab_width = rule.replace("tabwidth=", "")
+            rules += [f"tab_width = {tab_width}"]
+
+            if name in space_based_indent_sizes:
+                indent_size = space_based_indent_sizes[name]
+                rules += [f"indent_size = {indent_size}"]
+
+        else:
+            continue
+
+        rules = "\n".join(rules)
+        new_contents += f"\n[{name}]\n{rules}\n"
+
+    with open(".editorconfig", "w") as f:
+        f.write(new_contents)
+
+
+if __name__ == "__main__":
+    main()