Add new regular expressions for Chunked Encoding

This also moves some regular expressions for QUOTED_PAIR/QUOTED_STRING
into this module from utilities so that they may be reused.

2 files changed, 29 insertions, 26 deletions

diff --git a/src/waitress/rfc7230.py b/src/waitress/rfc7230.py
index 9b25fbd..26e6426 100644
--- a/src/waitress/rfc7230.py
+++ b/src/waitress/rfc7230.py
@@ -5,6 +5,9 @@ needed to properly parse HTTP messages.
 
 import re
 
+HEXDIG = "[0-9a-fA-F]"
+DIGIT = "[0-9]"
+
 WS = "[ \t]"
 OWS = WS + "{0,}?"
 RWS = WS + "{1,}?"
@@ -25,6 +28,12 @@ TOKEN = TCHAR + "{1,}"
 #                  ; visible (printing) characters
 VCHAR = r"\x21-\x7e"
 
+# The '\\' between \x5b and \x5d is needed to escape \x5d (']')
+QDTEXT = "[\t \x21\x23-\x5b\\\x5d-\x7e" + OBS_TEXT + "]"
+
+QUOTED_PAIR = r"\\" + "([\t " + VCHAR + OBS_TEXT + "])"
+QUOTED_STRING = '"(?:(?:' + QDTEXT + ")|(?:" + QUOTED_PAIR + '))*"'
+
 # header-field   = field-name ":" OWS field-value OWS
 # field-name     = token
 # field-value    = *( field-content / obs-fold )
@@ -43,8 +52,24 @@ FIELD_CONTENT = FIELD_VCHAR + "+(?:[ \t]+" + FIELD_VCHAR + "+)*"
 # Which allows the field value here to just see if there is even a value in the first place
 FIELD_VALUE = "(?:" + FIELD_CONTENT + ")?"
 
-HEADER_FIELD = re.compile(
+# chunk-ext      = *( ";" chunk-ext-name [ "=" chunk-ext-val ] )
+# chunk-ext-name = token
+# chunk-ext-val  = token / quoted-string
+
+CHUNK_EXT_NAME = TOKEN
+CHUNK_EXT_VAL = "(?:" + TOKEN + ")|(?:" + QUOTED_STRING + ")"
+CHUNK_EXT = (
+    "(?:;(?P<extension>" + CHUNK_EXT_NAME + ")(?:=(?P<value>" + CHUNK_EXT_VAL + "))?)*"
+)
+
+# Pre-compiled regular expressions for use elsewhere
+ONLY_HEXDIG_RE = re.compile(("^" + HEXDIG + "+$").encode("latin-1"))
+ONLY_DIGIT_RE = re.compile(("^" + DIGIT + "+$").encode("latin-1"))
+HEADER_FIELD_RE = re.compile(
     (
         "^(?P<name>" + TOKEN + "):" + OWS + "(?P<value>" + FIELD_VALUE + ")" + OWS + "$"
     ).encode("latin-1")
 )
+QUOTED_PAIR_RE = re.compile(QUOTED_PAIR)
+QUOTED_STRING_RE = re.compile(QUOTED_STRING)
+CHUNK_EXT_RE = re.compile(("^" + CHUNK_EXT + "$").encode("latin-1"))
diff --git a/src/waitress/utilities.py b/src/waitress/utilities.py
index 3caaa33..6ae4742 100644
--- a/src/waitress/utilities.py
+++ b/src/waitress/utilities.py
@@ -22,7 +22,7 @@ import re
 import stat
 import time
 
-from .rfc7230 import OBS_TEXT, VCHAR
+from .rfc7230 import QUOTED_PAIR_RE, QUOTED_STRING_RE
 
 logger = logging.getLogger("waitress")
 queue_logger = logging.getLogger("waitress.queue")
@@ -216,32 +216,10 @@ def parse_http_date(d):
     return retval
 
 
-# RFC 5234 Appendix B.1 "Core Rules":
-# VCHAR         =  %x21-7E
-#                  ; visible (printing) characters
-vchar_re = VCHAR
-
-# RFC 7230 Section 3.2.6 "Field Value Components":
-# quoted-string = DQUOTE *( qdtext / quoted-pair ) DQUOTE
-# qdtext        = HTAB / SP /%x21 / %x23-5B / %x5D-7E / obs-text
-# obs-text      = %x80-FF
-# quoted-pair   = "\" ( HTAB / SP / VCHAR / obs-text )
-obs_text_re = OBS_TEXT
-
-# The '\\' between \x5b and \x5d is needed to escape \x5d (']')
-qdtext_re = "[\t \x21\x23-\x5b\\\x5d-\x7e" + obs_text_re + "]"
-
-quoted_pair_re = r"\\" + "([\t " + vchar_re + obs_text_re + "])"
-quoted_string_re = '"(?:(?:' + qdtext_re + ")|(?:" + quoted_pair_re + '))*"'
-
-quoted_string = re.compile(quoted_string_re)
-quoted_pair = re.compile(quoted_pair_re)
-
-
 def undquote(value):
     if value.startswith('"') and value.endswith('"'):
         # So it claims to be DQUOTE'ed, let's validate that
-        matches = quoted_string.match(value)
+        matches = QUOTED_STRING_RE.match(value)
 
         if matches and matches.end() == len(value):
             # Remove the DQUOTE's from the value
@@ -249,7 +227,7 @@ def undquote(value):
 
             # Remove all backslashes that are followed by a valid vchar or
             # obs-text
-            value = quoted_pair.sub(r"\1", value)
+            value = QUOTED_PAIR_RE.sub(r"\1", value)
 
             return value
     elif not value.startswith('"') and not value.endswith('"'):