imprv: JID: Improve compliance when generating IRIs

2 files changed, 89 insertions, 28 deletions

diff --git a/nbxmpp/xmppiri.py b/nbxmpp/xmppiri.py
index 18af188..bc0035e 100644
--- a/nbxmpp/xmppiri.py
+++ b/nbxmpp/xmppiri.py
@@ -1,57 +1,117 @@
+from __future__ import annotations
 
+from typing import Callable
+
+import functools
 import re
-from gi.repository import GLib
 
 
 # https://www.rfc-editor.org/rfc/rfc3987
-
-ucschar        = r'\xA0-\uD7FF\uF900-\uFDCF\uFDF0-\uFFEF'\
-    r'\U00010000-\U0001FFFD\U00020000-\U0002FFFD\U00030000-\U0003FFFD'\
-    r'\U00040000-\U0004FFFD\U00050000-\U0005FFFD\U00060000-\U0006FFFD'\
-    r'\U00070000-\U0007FFFD\U00080000-\U0008FFFD\U00090000-\U0009FFFD'\
-    r'\U000A0000-\U000AFFFD\U000B0000-\U000BFFFD\U000C0000-\U000CFFFD'\
-    r'\U000D0000-\U000DFFFD\U000E1000-\U000EFFFD'
-unreserved     = r'A-Za-z0-9\-._~'
-iunreserved    = fr'{unreserved}{ucschar}'
-subdelims = r"!$&'()*+,;="
+ucschar = (
+    '\xA0-\uD7FF'
+    '\uF900-\uFDCF'
+    '\uFDF0-\uFFEF'
+    '\U00010000-\U0001FFFD'
+    '\U00020000-\U0002FFFD'
+    '\U00030000-\U0003FFFD'
+    '\U00040000-\U0004FFFD'
+    '\U00050000-\U0005FFFD'
+    '\U00060000-\U0006FFFD'
+    '\U00070000-\U0007FFFD'
+    '\U00080000-\U0008FFFD'
+    '\U00090000-\U0009FFFD'
+    '\U000A0000-\U000AFFFD'
+    '\U000B0000-\U000BFFFD'
+    '\U000C0000-\U000CFFFD'
+    '\U000D0000-\U000DFFFD'
+    '\U000E1000-\U000EFFFD'
+)
+
+ALPHA           = 'A-Za-z'
+DIGIT           = '0-9'
+unreserved      = fr'{ALPHA}{DIGIT}\-\._\~'
+subdelims       = "!$&'()*+,;="
+iunreserved     = f'{unreserved}{ucschar}'
+ipchar          = f'{iunreserved}{re.escape(subdelims)}:@'
+ifragment       = fr'{ipchar}/\?'
 
 # https://www.rfc-editor.org/rfc/rfc5122.html#section-2.2
-nodeallow  = r"!$()*+,;="
-resallow   = r"!$&'()*+,:;="
+nodeallow       = r"!$()*+,;="
+resallow        = r"!$&'()*+,:;="
+inode           = f'{iunreserved}{re.escape(nodeallow)}'
+ires            = f'{iunreserved}{re.escape(resallow)}'
+ivalue          = f'{iunreserved}'
+
+rx_iunreserved  = re.compile(f'[{iunreserved}]*')
+rx_inode        = re.compile(f'[{inode}]')
+rx_ires         = re.compile(f'[{ires}]')
+rx_ikey         = rx_iunreserved
+rx_iquerytype   = rx_iunreserved
+rx_ivalue       = rx_iunreserved
+rx_ifragment     = re.compile(f'[{ifragment}]')
+
+
+class _Quoter(dict[str, str]):
+    """A mapping from a string to its percent encoded form.
+
+    Mapping is only done if string is not in safe range.
 
-# ifragment without iunreserved and pct-encoded
-reserved_chars_allowed_in_ifragment = subdelims + ":@" + "/?"
+    Keeps a cache internally, via __missing__, for efficiency (lookups
+    of cached keys don't call Python code at all).
+    """
+    def __init__(self, safe: re.Pattern[str]) -> None:
+        self._safe = safe
 
-rx_ikey        = f'[{iunreserved}]*'
-rx_iquerytype  = f'[{iunreserved}]*'
+    def __repr__(self):
+        return f"<Quoter {dict(self)!r}>"
+
+    def __missing__(self, b: str):
+        if len(b) != 1:
+            raise ValueError("String must be exactly one character long")
+
+        if self._safe.fullmatch(b) is None:
+            res = "".join(['%{:02X}'.format(i) for i in b.encode()])
+        else:
+            res = b
+        self[b] = res
+        return res
+
+
+@functools.lru_cache
+def _quoter_factory(safe: re.Pattern[str]) -> Callable[[str], str]:
+    return _Quoter(safe).__getitem__
 
 
 def validate_ikey(ikey: str) -> str:
-    res = re.fullmatch(rx_ikey, ikey)
+    res = rx_ikey.fullmatch(ikey)
     if res is None:
         raise ValueError('Not allowed characters in key')
     return ikey
 
 
 def validate_querytype(querytype: str) -> str:
-    res = re.fullmatch(rx_iquerytype, querytype)
+    res = rx_iquerytype.fullmatch(querytype)
     if res is None:
         raise ValueError('Not allowed characters in querytype')
     return querytype
 
 
-def escape_ifragment(ifragment: str) -> str:
-    return GLib.Uri.escape_string(
-        ifragment, reserved_chars_allowed_in_ifragment, True)
+def _escape(string: str, pattern: re.Pattern[str]) -> str:
+    quoter = _quoter_factory(safe=pattern)
+    return ''.join([quoter(c) for c in string])
+
+
+def escape_ifragment(fragment: str) -> str:
+    return _escape(fragment, rx_ifragment)
 
 
-def escape_ivalue(ivalue: str) -> str:
-    return GLib.Uri.escape_string(ivalue, None, True)
+def escape_ivalue(value: str) -> str:
+    return _escape(value, rx_ivalue)
 
 
-def escape_inode(inode: str) -> str:
-    return GLib.Uri.escape_string(inode, nodeallow, True)
+def escape_inode(node: str) -> str:
+    return _escape(node, rx_inode)
 
 
-def escape_ires(ires: str) -> str:
-    return GLib.Uri.escape_string(ires, resallow, True)
+def escape_ires(res: str) -> str:
+    return _escape(res, rx_ires)
diff --git a/test/unit/test_jid_parsing.py b/test/unit/test_jid_parsing.py
index ae59ab4..7199595 100644
--- a/test/unit/test_jid_parsing.py
+++ b/test/unit/test_jid_parsing.py
@@ -194,6 +194,7 @@ class JIDParsing(unittest.TestCase):
         tests = [
             ('nasty!#$%()*+,-.;=?[\\]^_`{|}~node@example.com', 'xmpp:nasty!%23$%25()*+,-.;=%3F%5B%5C%5D%5E_%60%7B%7C%7D~node@example.com'),
             ('node@example.com/repulsive !#"$%&\'()*+,-./:;<=>?@[\\]^_`{|}~resource', 'xmpp:node@example.com/repulsive%20!%23%22$%25&\'()*+,-.%2F:;%3C=%3E%3F%40%5B%5C%5D%5E_%60%7B%7C%7D~resource'),
+            ('jiři@čechy.example/v Praze', 'xmpp:jiři@čechy.example/v%20Praze')
         ]
 
         for jid, iri in tests: