From b167818f7dded95a73f83b181fdcd322fc956cbf Mon Sep 17 00:00:00 2001
From: "William A. Rowe Jr" <wrowe@apache.org>
Date: Fri, 29 Jul 2016 15:56:54 +0000
Subject: [PATCH] Introduce ap_scan_http_field_content, ap_scan_http_token and
 ap_get_http_token for more efficient string handling.

git-svn-id: https://svn.apache.org/repos/asf/httpd/httpd/trunk@1754541 13f79535-47bb-0310-9956-ffa450edef68
---
 include/ap_mmn.h |  4 +++-
 include/httpd.h  | 25 +++++++++++++++++++++++++
 server/util.c    | 42 +++++++++++++++++++++++++++++++++++++++++-
 3 files changed, 69 insertions(+), 2 deletions(-)

diff --git a/include/ap_mmn.h b/include/ap_mmn.h
index 1d3b07a72d..b28637a7cf 100644
--- a/include/ap_mmn.h
+++ b/include/ap_mmn.h
@@ -534,6 +534,8 @@
  *                         dav_get_acl_providers.
  * 20160608.5 (2.5.0-dev)  Add ap_proxy_check_backend(), and tmp_bb to struct
  *                         proxy_conn_rec.
+ * 20160508.6 (2.5.0-dev)  Add ap_scan_http_field_content, ap_scan_http_token
+ *                         and ap_get_http_token
  */
 
 #define MODULE_MAGIC_COOKIE 0x41503235UL /* "AP25" */
@@ -541,7 +543,7 @@
 #ifndef MODULE_MAGIC_NUMBER_MAJOR
 #define MODULE_MAGIC_NUMBER_MAJOR 20160608
 #endif
-#define MODULE_MAGIC_NUMBER_MINOR 5                 /* 0...n */
+#define MODULE_MAGIC_NUMBER_MINOR 6                 /* 0...n */
 
 /**
  * Determine if the server's current MODULE_MAGIC_NUMBER is at least a
diff --git a/include/httpd.h b/include/httpd.h
index eba7b53076..931dfac53a 100644
--- a/include/httpd.h
+++ b/include/httpd.h
@@ -1632,6 +1632,31 @@ AP_DECLARE(int) ap_find_etag_weak(apr_pool_t *p, const char *line, const char *t
  */
 AP_DECLARE(int) ap_find_etag_strong(apr_pool_t *p, const char *line, const char *tok);
 
+/* Scan a string for field content chars, as defined by RFC7230 section 3.2
+ * including VCHAR/obs-text, as well as HT and SP
+ * @param ptr The string to scan
+ * @return A pointer to the first (non-HT) ASCII ctrl character.
+ * @note lws and trailing whitespace are scanned, the caller is responsible
+ * for trimming leading and trailing whitespace
+ */
+AP_DECLARE(const char *) ap_scan_http_field_content(const char *ptr);
+
+/* Scan a string for token characters, as defined by RFC7230 section 3.2.6 
+ * @param ptr The string to scan
+ * @return A pointer to the first non-token character.
+ */
+AP_DECLARE(const char *) ap_scan_http_token(const char *ptr);
+
+/* Retrieve a token, advancing the pointer to the first non-token character
+ * and returning a copy of the token string.
+ * @param ptr The string to scan. On return, this points to the first non-token
+ *  character encountered, or NULL if *ptr was not a token character
+ * @return A copy of the token string
+ * @note The caller must handle leading and trailing whitespace as applicable
+ *  and evaluate the terminating character.
+ */
+AP_DECLARE(char *) ap_get_http_token(apr_pool_t *p, const char **ptr);
+
 /**
  * Retrieve an array of tokens in the format "1#token" defined in RFC2616. Only
  * accepts ',' as a delimiter, does not accept quoted strings, and errors on
diff --git a/server/util.c b/server/util.c
index aab8311532..470f6c343b 100644
--- a/server/util.c
+++ b/server/util.c
@@ -80,7 +80,7 @@
  * char in here and get it to work, because if char is signed then it
  * will first be sign extended.
  */
-#define TEST_CHAR(c, f)        (test_char_table[(unsigned)(c)] & (f))
+#define TEST_CHAR(c, f)        (test_char_table[(unsigned char)(c)] & (f))
 
 /* Win32/NetWare/OS2 need to check for both forward and back slashes
  * in ap_getparents() and ap_escape_url.
@@ -1594,6 +1594,46 @@ AP_DECLARE(const char *) ap_parse_token_list_strict(apr_pool_t *p,
     return NULL;
 }
 
+/* Scan a string for HTTP VCHAR/obs-text characters including HT and SP
+ * (as used in header values, for example, in RFC 7230 section 3.2)
+ * returning the pointer to the first non-HT ASCII ctrl character.
+ */
+AP_DECLARE(const char *) ap_scan_http_field_content(const char *ptr)
+{
+    for ( ; !TEST_CHAR(*ptr, T_HTTP_CTRLS); ++ptr) ;
+
+    return ptr;
+}
+
+/* Scan a string for HTTP token characters, returning the pointer to
+ * the first non-token character.
+ */
+AP_DECLARE(const char *) ap_scan_http_token(const char *ptr)
+{
+    for ( ; !TEST_CHAR(*ptr, T_HTTP_TOKEN_STOP); ++ptr) ;
+
+    return ptr;
+}
+
+/* Retrieve a token, advancing the pointer to the first non-token character
+ * and returning a copy of the token string.
+ * The caller must handle whitespace and determine the meaning of the
+ * terminating character. Returns NULL if the character at **ptr is not
+ * a valid token character.
+ */
+AP_DECLARE(char *) ap_get_http_token(apr_pool_t *p, const char **ptr)
+{
+    const char *tok_end = ap_scan_http_token(*ptr);
+    char *tok;
+
+    if (tok_end == *ptr)
+        return NULL;
+
+    tok = apr_pstrmemdup(p, *ptr, tok_end - *ptr);
+    *ptr = tok_end;
+    return tok;
+}
+
 /* Retrieve a token, spacing over it and returning a pointer to
  * the first non-white byte afterwards.  Note that these tokens
  * are delimited by semis and commas; and can also be delimited
-- 
2.50.1