From: Jeeyoung Kim <jeeyoung@contextlogic.com>
Date: Mon, 22 Aug 2011 22:14:39 +0000 (-0700)
Subject: Implemented read_until_regex in tornado. This is similar to read_until(), except... 
X-Git-Tag: v2.1.0~39^2
X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=86404402d4d98e5559346b1533b8566af87978bd;p=thirdparty%2Ftornado.git

Implemented read_until_regex in tornado. This is similar to read_until(), except it looks for a specific delimiter. This function is used to parse the header of the website, where it is delimited by '\n' only (like http://news.ycombinator.com/rss)
---

diff --git a/tornado/iostream.py b/tornado/iostream.py
index f6308d3fd..1a4f4992f 100644
--- a/tornado/iostream.py
+++ b/tornado/iostream.py
@@ -23,6 +23,7 @@ import errno
 import logging
 import socket
 import sys
+import re
 
 from tornado import ioloop
 from tornado import stack_context
@@ -89,6 +90,7 @@ class IOStream(object):
         self._read_buffer_size = 0
         self._write_buffer_frozen = False
         self._read_delimiter = None
+        self._read_regex = None
         self._read_bytes = None
         self._read_until_close = False
         self._read_callback = None
@@ -124,6 +126,20 @@ class IOStream(object):
         self._connect_callback = stack_context.wrap(callback)
         self._add_io_state(self.io_loop.WRITE)
 
+    def read_until_regex(self, regex, callback):
+        """Call callback when we read the given regex pattern."""
+        assert not self._read_callback, "Already reading"
+        self._read_regex = re.compile(regex)
+        self._read_callback = stack_context.wrap(callback)
+        while True:
+            # See if we've already got the data from a previous read
+            if self._read_from_buffer():
+                return
+            self._check_closed()
+            if self._read_to_buffer() == 0:
+                break
+        self._add_io_state(self.io_loop.READ)
+        
     def read_until(self, delimiter, callback):
         """Call callback when we read the given delimiter."""
         assert not self._read_callback, "Already reading"
@@ -374,6 +390,15 @@ class IOStream(object):
                 self._run_callback(callback,
                                    self._consume(loc + delimiter_len))
                 return True
+        elif self._read_regex is not None:
+            _merge_prefix(self._read_buffer, sys.maxint)
+            m = self._read_regex.search(self._read_buffer[0])
+            if m:
+                callback = self._read_callback
+                self._read_callback = None
+                self._read_regex = None
+                self._run_callback(callback, self._consume(m.end()))
+                return True
         return False
 
     def _handle_connect(self):
diff --git a/tornado/simple_httpclient.py b/tornado/simple_httpclient.py
index e1a9ce405..904364fcc 100644
--- a/tornado/simple_httpclient.py
+++ b/tornado/simple_httpclient.py
@@ -268,7 +268,7 @@ class _HTTPConnection(object):
         self.stream.write(b("\r\n").join(request_lines) + b("\r\n\r\n"))
         if self.request.body is not None:
             self.stream.write(self.request.body)
-        self.stream.read_until(b("\r\n\r\n"), self._on_headers)
+        self.stream.read_until_regex(b("\r?\n\r?\n"), self._on_headers)
 
     def _release(self):
         if self.release_callback is not None:
@@ -298,7 +298,7 @@ class _HTTPConnection(object):
 
     def _on_headers(self, data):
         data = native_str(data.decode("latin1"))
-        first_line, _, header_data = data.partition("\r\n")
+        first_line, _, header_data = data.partition("\n")
         match = re.match("HTTP/1.[01] ([0-9]+)", first_line)
         assert match
         self.code = int(match.group(1))