meson: docs: Add xml{lint,proc} wrapper to collect dependencies

meson/ninja do not support specifying dependencies via globs (as those make it
significantly more expensive to check the current build state). Instead
targets should emit dependency information when running that then can be
cheaply re-checked during future builds.

To handle xmllint and xsltproc invocations in the docs, add and use a wrapper
that uses --load-trace to collect dependency information.

Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com>
Author: Nazir Bilal Yavuz <byavuz81@gmail.com>
Author: Andres Freund <andres@anarazel.de>
Discussion: https://postgr.es/m/c5736f70-bb6d-8d25-e35c-e3d886e4e905@enterprisedb.com

doc/src/sgml/xmltools_dep_wrapper

@@ -0,0 +1,54 @@
+#!/usr/bin/env python3
+
+# A small wrapper around xmllint and xsltproc that collects dependency
+# information (in gcc's format) using --load-trace.
+
+import argparse
+import re
+import subprocess
+import sys
+
+parser = argparse.ArgumentParser(
+    description='generate dependency file for docs')
+
+parser.add_argument('--targetname', type=str, required=False, nargs='+')
+parser.add_argument('--depfile', type=str, required=False)
+parser.add_argument('--tool', type=str, required=True)
+parser.add_argument('flags', nargs='*')
+
+args = parser.parse_args()
+
+if args.depfile:
+    command = [args.tool, '--load-trace'] + args.flags
+
+    # list of targets that depend on the loaded files we see via --load-trace
+    line_start = ' '.join(args.targetname) + ': '
+
+    # --load-trace flag displays all the documents loaded during the processing
+    # to stderr
+    res = subprocess.run(command, stderr=subprocess.PIPE,
+                         universal_newlines=True)
+
+    line_re = re.compile('^Loaded URL="([^"]+)"')
+    with open(args.depfile, 'w') as f:
+        for line in res.stderr.splitlines():
+            m = re.match(line_re, line)
+
+            # continue to show errors
+            if m is None:
+                print(line, file=sys.stderr)
+                continue
+            # Absolute paths are printed as file://, relative paths as-is. We
+            # don't care about http://, as a) those will be printed even if
+            # resolved locally b) we couldn't have a dependency anyway.
+            fname = m.group(1)
+            if fname.startswith('http://'):
+                continue
+            if fname.startswith('file://'):
+                fname = fname.split('file://')[1]
+            f.write(line_start + fname + '\n')
+else:
+    command = [args.tool] + args.flags
+    res = subprocess.run(command)
+
+exit(res.returncode)