bpo-29979: Rewrite cgi.parse_multipart to make it consistent with FieldStorage (#991)

Doc/library/cgi.rst

@@ -294,19 +294,20 @@ algorithms implemented in this module in other circumstances.
    This function is deprecated in this module. Use :func:`urllib.parse.parse_qsl`
    instead. It is maintained here only for backward compatibility.
 
-.. function:: parse_multipart(fp, pdict)
+.. function:: parse_multipart(fp, pdict, encoding="utf-8")
 
    Parse input of type :mimetype:`multipart/form-data` (for  file uploads).
-   Arguments are *fp* for the input file and *pdict* for a dictionary containing
+   Arguments are *fp* for the input file, *pdict* for a dictionary containing
-   other parameters in the :mailheader:`Content-Type` header.
+   other parameters in the :mailheader:`Content-Type` header, and *encoding*,
+   the request encoding.
 
-   Returns a dictionary just like :func:`urllib.parse.parse_qs` keys are the field names, each
+   Returns a dictionary just like :func:`urllib.parse.parse_qs`: keys are the
-   value is a list of values for that field.  This is easy to use but not much good
+   field names, each value is a list of values for that field. For non-file
-   if you are expecting megabytes to be uploaded --- in that case, use the
+   fields, the value is a list of strings.
-   :class:`FieldStorage` class instead which is much more flexible.
 
-   Note that this does not parse nested multipart parts --- use
+   This is easy to use but not much good if you are expecting megabytes to be
-   :class:`FieldStorage` for that.
+   uploaded --- in that case, use the :class:`FieldStorage` class instead
+   which is much more flexible.
 
 
 .. function:: parse_header(string)

Doc/whatsnew/3.7.rst

@@ -95,6 +95,14 @@ New Modules
 Improved Modules
 ================
 
+cgi
+---
+
+:func:`~cgi.parse_multipart` returns the same results as
+:class:`~FieldStorage` : for non-file fields, the value associated to a key
+is a list of strings, not bytes.
+(Contributed by Pierre Quentel in :issue:`29979`.)
+
 binascii
 --------
 

Lib/cgi.py

@@ -198,105 +198,28 @@ def parse_qsl(qs, keep_blank_values=0, strict_parsing=0):
          DeprecationWarning, 2)
     return urllib.parse.parse_qsl(qs, keep_blank_values, strict_parsing)
 
-def parse_multipart(fp, pdict):
+def parse_multipart(fp, pdict, encoding="utf-8"):
     """Parse multipart input.
 
     Arguments:
     fp   : input file
     pdict: dictionary containing other parameters of content-type header
+    encoding: request encoding
 
     Returns a dictionary just like parse_qs(): keys are the field names, each
-    value is a list of values for that field.  This is easy to use but not
+    value is a list of values for that field. For non-file fields, the value
-    much good if you are expecting megabytes to be uploaded -- in that case,
+    is a list of strings.
-    use the FieldStorage class instead which is much more flexible.  Note
-    that content-type is the raw, unparsed contents of the content-type
-    header.
-
-    XXX This does not parse nested multipart parts -- use FieldStorage for
-    that.
-
-    XXX This should really be subsumed by FieldStorage altogether -- no
-    point in having two implementations of the same parsing algorithm.
-    Also, FieldStorage protects itself better against certain DoS attacks
-    by limiting the size of the data read in one chunk.  The API here
-    does not support that kind of protection.  This also affects parse()
-    since it can call parse_multipart().
-
     """
-    import http.client
+    # RFC 2026, Section 5.1 : The "multipart" boundary delimiters are always
-
+    # represented as 7bit US-ASCII.
-    boundary = b""
+    boundary = pdict['boundary'].decode('ascii')
-    if 'boundary' in pdict:
+    ctype = "multipart/form-data; boundary={}".format(boundary)
-        boundary = pdict['boundary']
+    headers = Message()
-    if not valid_boundary(boundary):
+    headers.set_type(ctype)
-        raise ValueError('Invalid boundary in multipart form: %r'
+    headers['Content-Length'] = pdict['CONTENT-LENGTH']
-                            % (boundary,))
+    fs = FieldStorage(fp, headers=headers, encoding=encoding,
-
+        environ={'REQUEST_METHOD': 'POST'})
-    nextpart = b"--" + boundary
+    return {k: fs.getlist(k) for k in fs}
-    lastpart = b"--" + boundary + b"--"
-    partdict = {}
-    terminator = b""
-
-    while terminator != lastpart:
-        bytes = -1
-        data = None
-        if terminator:
-            # At start of next part.  Read headers first.
-            headers = http.client.parse_headers(fp)
-            clength = headers.get('content-length')
-            if clength:
-                try:
-                    bytes = int(clength)
-                except ValueError:
-                    pass
-            if bytes > 0:
-                if maxlen and bytes > maxlen:
-                    raise ValueError('Maximum content length exceeded')
-                data = fp.read(bytes)
-            else:
-                data = b""
-        # Read lines until end of part.
-        lines = []
-        while 1:
-            line = fp.readline()
-            if not line:
-                terminator = lastpart # End outer loop
-                break
-            if line.startswith(b"--"):
-                terminator = line.rstrip()
-                if terminator in (nextpart, lastpart):
-                    break
-            lines.append(line)
-        # Done with part.
-        if data is None:
-            continue
-        if bytes < 0:
-            if lines:
-                # Strip final line terminator
-                line = lines[-1]
-                if line[-2:] == b"\r\n":
-                    line = line[:-2]
-                elif line[-1:] == b"\n":
-                    line = line[:-1]
-                lines[-1] = line
-                data = b"".join(lines)
-        line = headers['content-disposition']
-        if not line:
-            continue
-        key, params = parse_header(line)
-        if key != 'form-data':
-            continue
-        if 'name' in params:
-            name = params['name']
-        else:
-            continue
-        if name in partdict:
-            partdict[name].append(data)
-        else:
-            partdict[name] = [data]
-
-    return partdict
-
 
 def _parseparam(s):
     while s[:1] == ';':

Lib/test/test_cgi.py

@@ -126,8 +126,8 @@ def test_parse_multipart(self):
         env = {'boundary': BOUNDARY.encode('latin1'),
                'CONTENT-LENGTH': '558'}
         result = cgi.parse_multipart(fp, env)
-        expected = {'submit': [b' Add '], 'id': [b'1234'],
+        expected = {'submit': [' Add '], 'id': ['1234'],
-                    'file': [b'Testing 123.\n'], 'title': [b'']}
+                    'file': [b'Testing 123.\n'], 'title': ['']}
         self.assertEqual(result, expected)
 
     def test_fieldstorage_properties(self):

Misc/NEWS

@@ -317,6 +317,10 @@ Extension Modules
 Library
 -------
 
+- bpo-29979: rewrite cgi.parse_multipart, reusing the FieldStorage class and
+  making its results consistent with those of FieldStorage for
+  multipart/form-data requests. Patch by Pierre Quentel.
+
 - bpo-30243: Removed the __init__ methods of _json's scanner and encoder.
   Misusing them could cause memory leaks or crashes.  Now scanner and encoder
   objects are completely initialized in the __new__ methods.